Climate Change Adaptation: a Green Infrastructure Planning Framework for Resilient Urban Regions Yaser F

University of Massachusetts Amherst ScholarWorks@UMass Amherst

Open Access Dissertations

9-2013 Climate Change Adaptation: A Green Infrastructure Planning Framework for Resilient Urban Regions Yaser F. Abunnasr University of Massachusetts Amherst, [email protected]

Follow this and additional works at: https://scholarworks.umass.edu/open_access_dissertations Part of the Landscape Architecture Commons, and the Urban, Community and Regional Planning Commons

Recommended Citation Abunnasr, Yaser F., "Climate Change Adaptation: A Green Infrastructure Planning Framework for Resilient Urban Regions" (2013). Open Access Dissertations. 775. https://doi.org/10.7275/yg71-qy42 https://scholarworks.umass.edu/open_access_dissertations/775

This Open Access Dissertation is brought to you for free and open access by ScholarWorks@UMass Amherst. It has been accepted for inclusion in Open Access Dissertations by an authorized administrator of ScholarWorks@UMass Amherst. For more information, please contact [email protected].

CLIMATE CHANGE ADAPTATION: A GREEN INFRASTRUCTURE PLANNING FRAMEWORK FOR RESILIENT URBAN REGIONS

A Dissertation Presented

YASER ABUNNASR

Submitted to the Graduate School of the University of Massachusetts Amherst in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILISOPHY

September 2013

Regional Planning Department of Landscape Architecture and Regional Planning

CLIMATE CHANGE ADAPTATION: A GREEN INFRASTRUCTURE PLANNING FRAMEWORK FOR RESILIENT URBAN REGIONS

A Dissertation Presented By

YASER ABUNNASR

Approved as to style and content by:

Elisabeth Hamin, Chair

Elizabeth Brabec, Member

Qian Yu, Member

Elisabeth Hamin, Department Head Department of Landscape Architecture and Regional Planning

DEDICATION

Maria, Zayn & Jad

Muna & Fadeel

ACKNOWLEDGEMENTS

I would like to thank all my committee members, Professor Elisabeth Hamin,

Professor Elizabeth Brabec, and Professor Qian Yu for their support in a long and arduous journey.

Elisabeth, I highly appreciate your unequivocal support, efficiency in handling matters and confidence in my capabilities to make the leap from design to research. You did not stop believing in me when I doubted myself. Thank you!

Elizabeth, I appreciate the time you took to follow-up on my progress during your travel in different time zones and for invigorating discussions.

Qian, for inspiring me to be immersed in geographic information systems and remote sensing. These topics are dear to my heart as they come closest to fulfill a childhood dream of mine.

Thank you to Ray La Raja and Scott Auerbach for their informal teaching through great discussions.

Thank you to Maria for your love and patience in sharing these absolutely challenging years.

v ABSTRACT

CLIMATE CHANGE ADPATATION: A GREEN INFRASTRCTURE PLANNING FRAMEWORK FOR RESILIENT URBAN REGIONS

SEPTEMBER 2013

YASER ABUNNASR, B.ARCH., AMERICAN UNIVERSITY OF BEIRUT

M.L.A., UNIVERSITY OF MASSACHUSETTS AMHERST

Ph.D., UNIVERSITY OF MASSACHUSETTS AMHERST

Directed by: Elisabeth Hamin, Associate Professor

The research explores multiple facets of a green infrastructure planning

framework for climate change adaptation in urban regions. The research is organized in

three distinct, but related parts. The first develops an adaptation implementation model based on triggering conditions rather than time. The approach responds to policy makers’

reluctance to engage in adaptation planning due to uncertain future conditions. The model

is based on planning and adaptation literature and applied to two case studies.

Uncertainty during implementation may be reduced by incremental and flexible policy

implementation, disbursing investments as needs arise, monitoring conditions, and

organizing adaptation measures along no-regrets to transformational measures. The

second part develops the green infrastructure transect as an organizational framework for

mainstreaming adaptation planning policies. The framework integrates multi-scalar and

context aspects of green infrastructure for vertical and horizontal integration of policy.

The framework integrates literature from urban and landscape planning and tested on

Boston. Prioritization of adaptation measures depends on location. Results suggest that

green infrastructure adaptation policies should respond to configuration of zones. Cross

vi jurisdiction coordination at regional and parcel scales supports mainstreaming. A secondary conclusion suggests that green infrastructure is space intensive and becomes the basis of the empirical study in part three. A spatial assessment method is introduced

to formulate opportunities for green infrastructure network implementation within land-

uses and across an urban-rural gradient. Spatial data in GIS for Boston is utilized to

develop a percent pervious metric allowing the characterization of the study area into six

zones of varying perviousness. Opportunities across land uses were assessed then

maximum space opportunities were defined based on conservation, intensification,

transformation and expansion. The opportunities for transformation of impervious

surfaces to vegetal surfaces are highest in the urban center and its surrounding.

Intensification of vegetation on pervious surfaces along all land uses is high across the

gradient. Conservation of existing forested land is significant for future climate proofing.

The concluding section argues for a green infrastructure planning framework for

adaptation based on integration into existing infrastructural bodies, regional vision,

incremental implementation, ecosystem benefits accounting, and conditions based

planning rather than time based.

vii TABLE OF CONTENTS

Pages

ACKNOWLEDGEMENTS ...... v

ABSTRACT ...... vi

LIST OF TABLES ...... xii

LIST OF FIGURES ...... xiii

CHAPTERS

1 INTRODUCTION ...... 1

1.1.Green Infrastructure as an adaptation suite of measures ...... 1 1.2.Dissertation propositions ...... 7 1.3.Dissertation Organization ...... 11

2 GREEN INFRASTRUCTURE, URBAN FORM AND ADAPTIVE CAPACITY: COMPLEMENTARITY, SYNERGY AND CONTRADICTION ...... 16

2.1 Introduction ...... 16 2.2 Resilience: An Interpretive framework ...... 18 2.3 Adaptation to Climate Change ...... 21 2.4 Green Infrastructure ...... 25

2.4.1 Network and Connectivity ...... 27 2.4.2 Multi-scalar spatial elements ...... 29 2.4.3 Multi-functionality as Ecosystem Benefits ...... 32 2.4.4 Implementation ...... 34 2.4.5 Complementarity Between Green Infrastructure and Adaptation Measures ...... 36

2.5 Urban Form ...... 37 2.6 Green Infrastructure, Space and Adaptation ...... 40

3 WINDOWS OF OPPORTUNITY: ADDRESSING CLIMATE UNCERTAINTY THROUGH ADAPTATION PLAN IMPLEMENTATION ...... 43

3.1 Introduction ...... 43 3.2 Climate Uncertainty and Urban Planning Policy ...... 44

3.2.1 Climate Change Timing Uncertainty ...... 45 3.2.2 Climate Variability and Climate Change ...... 46

viii 3.2.3 Uncertainty in Impacts at the Urban Micro-Climate Scale ...... 47 3.2.4 The Challenge of Uncertainty to Municipal Planning ...... 48

3.3 Municipal Adaptation Planning Processes ...... 49

3.3.1 Maladaptation in the Planning Process ...... 51 3.3.2 Mainstreaming and Incremental Change ...... 52

3.4 Windows of Opportunity Model...... 52

3.4.1 Phasing Policies ...... 54 3.4.2 ATPs and Types of Indicators ...... 55 3.4.3 Example: London and the Thames Barrier ...... 57

3.5 Application in Case Study Communities ...... 58

3.5.1 Managing Rising Sea Level Impacts: The City of Clarence, Tasmania, Australia...... 60 3.5.2 Managing flood impacts from Extreme Precipitation Events: The City of Copenhagen, Denmark ...... 65 3.5.3 Case Study Discussion ...... 69

3.6 Concluding Remarks ...... 70

4 THE GREEN INFRASTRUCTURE TRANSECT: AN ORGANIZATIONAL FRAMEWORK FOR MAINSTREAMING ADAPTATION PLANNING POLICIES ...... 73

4.1 Introduction ...... 73 4.2 The Green Infrastructure Transect ...... 75 4.3 Boston Metropolitan Region ...... 81 4.4 Conclusion ...... 88

5 PERCENT PERVIOUS: AN ASSESSMENT METHOD OF GREEN INFRASTRCTURE SPACE OPPORTUNITIES - AN APPLICATION TO THE BOSTON METROPOLITAN GRADIENT FOR URBAN HEAT ISLAND ADPATATION ...... 90

5.1 Introduction ...... 90 5.2 The Urban Heat Island and Urban Form ...... 96 5.3 Effective Vegetated Green Infrastructure Measures for the Urban Heat Island ...... 106 5.4 Framework of Strategies to Increase Vegetated Green Infrastructure Surfaces ...... 112 5.5 Method ...... 117

5.5.1 Study Area ...... 118 5.5.2 Percent Pervious ...... 120 5.5.3 Pervious Data Set...... 121 5.5.4 Characterizing the Study Area ...... 124 5.5.5 Assess Pervious Surface Opportunities ...... 125

5.5.5.1 Percent Pervious Across Land-use Classes ...... 126 5.5.5.2 Opportunity in Pervious Land-use Classes: Forest and Agriculture ...... 128 5.5.5.3 Opportunity in Pervious and Impervious Land-use Classes ... 128 5.5.6 Assess Maximum Opportunities of Landover for the Urban Heat Island ...... 131

5.5.6.1 Patches of Vegetated Forest Surfaces ...... 131 5.5.6.2 Green Roofs across land-use classes ...... 132 5.5.6.3 Pervious Surfaces Across Road Categories ...... 132 5.5.6.4 Parking Surface Opportunities Across Land-use Classes ...... 133 5.5.7 Gradient of Green Infrastructure Strategies ...... 134

5.6 Results ...... 135

5.6.1 Gradient Zones of Percent Pervious Across Study Area ...... 135 5.6.2 Pervious and Impervious Surfaces within All Land-use classes and Across Gradient Zones ...... 138 5.6.3 Pervious Surface Opportunities Across Land-uses ...... 141 5.6.4 Maximum Opportunities of Pervious Surfaces for the Urban Heat Island ...... 143

5.6.4.1 Preserve and Intensify: Large Patches of Contiguous Pervious land Surfaces ...... 143 5.6.4.2 Intensify: Patches of Pervious Land Surfaces Across Land-use classes ...... 145 5.6.4.3 Transforming Building Roofs to Green Roofs ...... 146 5.6.4.4 Intensify and Shade for Trees: Pervious Surface Opportunities in Road Easements ...... 148 5.6.4.5 Transform Roads to Vegetated Surfaces ...... 149 5.6.4.6 Transform for Tree Shading: Pervious Surface Opportunities in Parking Surfaces ...... 150

5.6.5 Gradient of Green Infrastructure Policy Opportunities ...... 152

5.7 Discussion and Conclusion ...... 156

6 INTEGRATING GREEN INFRASTRUCTURE AND ADAPTATION PLANNING ...... 159

6.1 Prologue ...... 159 6.2 Green Infrastructure and Adaptive Capacity ...... 160 6.3 Green Infrastructure Framework for adaptation at the metropolitan scale ... 169 6.4 Future Research Work ...... 174

6.4.1 Adaptation Planning ...... 174 6.4.2 Green Infrastructure Planning ...... 176 6.4.2 Metropolitan Planning and Urban Form ...... 177

6.5 Epilogue ...... 178

APPENDICES

A TYPES OF URBAN HEAT ISLAND ...... 181

B THE TABLES ...... 184

BIBLIOGRAPHY ...... 202

xi LIST OF TABLES Table Page

2.1 Climate change impacts and their indicators ...... 25

2.2 Typology of Green infrastructure spatial elements ...... 31

2.3: Typology of Green Infrastructure based on spatial configuration ...... 32

2.4 Complementarity of GI benefits and adaptation needs ...... 37

3.1 Components of the implementation model used to assess case study examples...... 59

3.2 Example of integrating adaptation measures across geographic scales and reduction of risk for flooding from extreme rain events and rising sea level ...... 67

5.1 Derived data sets used in the study from raw spatial data downloaded from Mass GIS...... 124

5.2 Categorization of perviousness of land-use classes ...... 127

5.3 Classification scheme of coefficient of use (CU) and coefficient of tree planting (CT) for each land use class...... 130

5.4 Relationship between planning strategies, green infrastructure strategies, and contribution to temperature reduction ...... 135

5.5 Results of the Boston Metropolitan Area characterization and comparison to similar studies...... 137

5.6 Highest provision of opportunity across land-use classes by green infrastructure policy...... 155

5.7 Green infrastructure prioritization scheme across gradient zones...... 156

5.8 List of towns included in the study area with basic population, area, and land cover data...... 193

5.9 Table of land-use classes used in the study (MassGIS)...... 197

5.10 Original data set used to develop study data layers (adapted from MassGIS web site, http://www.mass.gov/anf/research-and-tech/it-serv-and-support/application- serv/office-of-geographic-information-massgis/datalayers/layerlist.html) ...... 199

5.11 Detailed information on calculation of coefficient of use by land use class ...... 201

xii LIST OF FIGURES

Figure Page

2.1 Three steps to achieve resiliency by managers and planners (Salt and Walker, 2006)...... 20

2.2 Four types of regional urban form: Chicago, sprawl/constellation city (top, left); Buenos Aires, Stellar/Linear (top, right); Washington DC, Green City (bottom, left); and Calgary, Nucleus (bottom, right)...... 42

3.1 Timing of regional and global climate change impacts under high-emissions scenarios...... 46

3.2 The Adaptation Planning process as recommended by the U.S. National Research Council...... 49

3.3 Local adaptation planning process advised in the State of California Adaptation Planning Guide...... 50

3.4 Windows of Opportunity Climate Adaptation Model is based on phasing of policy to match new conditions...... 53

3.5 An example of phased engineering adaptation for the Thames Barrier, responding to potential climate change trigger points ...... 58

3.6 Phasing Model for Lauderdale Area, City of Clarence, Australia...... 61

3.7 Phasing Model applied to the Copenhagen Climate Adaptation Plan ...... 68

4.1 The Green Infrastructure Transect: Concept and Organization ...... 78

4.2 Multi-tiered GI adaptation planning framework: horizontal and vertical integration...... 80

4.3 Metropolitan region of Boston: spatial distribution of pervious and impervious surfaces ...... 82

4.4 Metropolitan region of Boston: green infrastructure across town boundaries ...... 83

4.5 Green infrastructure transect application to Boston region, step one: vulnerability assessment ...... 84

4.6 Green infrastructure transect application to Boston region, step two: existing green infrastructure patterns ...... 85

xiii 4.7 Green infrastructure transect application to Boston region, step three: identiﬁcation of GI policies...... 87

5.1 Primary variable for the provision of ecosystem services is the area of hydrologically active surfaces or vegetated surfaces...... 91

5.2 Impact of land use classes on space availability and morphology of pervious surfaces...... 93

5.3 Typical Surface and air temperature differences during the day and at night ...... 97

5.4 Process of UHI formation in relation to urbanization process...... 100

5.5 Diagram showing how increasing physical and biological properties of vegetated green infrastructure decreases the urban island heat...... 109

5.6 Strategies to increase pervious surfaces...... 116

5.7 Steps included in the method...... 118

5.8 Study Area boundaries ...... 119

5.9 Reduction of variables of interest to two primary surrogate indicators, pervious and impervious surfaces...... 120

5.10 Six Gradient Zones by percent pervious and pervious-to-impervious ratio ...... 136

5.11 Metropolitan gradient zones compared to population density an major road network ...... 138

5. 12Distribution of pervious, impervious and water surfaces across the urban-rural gradient...... 139

5.13 Distribution of percent pervious across land uses within the MB...... 140

5.14 Percent pervious surfaces across land-use types within each gradient zone and ranking of the highest three values ...... 143

5.15 Percent pervious values for four hectare patches across land uses and their ranking cross the six zones...... 145

5.16 Comparative table and ranking of PP as defined by use-coefficient across land-use classes...... 146

5.17 Qualifying potential building roofs for green roof implementation ...... 148

5.18 Copmartive duagram and table of PP potential within road easements categorized by road class ...... 149

xiv 5.19 Class five roads (local arterial) that could potentially be transformed to green streets ...... 150

5.20 Opportunities to transform impervious surfaces to vegetated surfaces within qualifying land-use classes ...... 151

5.21 Spatial distribution of each green infrastructure measure type across the gradient zones compared to the urban heat island distribution for MB...... 153

5.22 Opportunities for green infrastructure measures across gradient zones...... 154

5.23 Opportunities for green infrastructure measures across gradient zones including forest and agricultural land use classes...... 155

5.25 Urban Boundary Layer (Source: Kuttler, 2008) ...... 182

xv CHAPTER 1

INTRODUCTION

1.1. Green Infrastructure as an adaptation suite of measures

Green infrastructure is a strategic and spatial approach to landscape and urban

planning. It is spatial because it is a network of patches and corridors of land that are

planned and managed for biodiversity conservation, “nature” protection, water resources,

land protection, recreation, cultural uses, urban development control, and more recently,

for climate change mitigation and adaptation. A green infrastructure network or system

may include networks of greenways, open spaces, greenbelts, urban greening, cultural

landscapes, urban open spaces, ecological networks, agricultural land, and natural

systems (and all un-built land that could support vegetation) that form the necessary

support system for living. Green infrastructure can be conceived and understood by stakeholders, planners and policy makers (Benedict and McMahon, 2002; Ahern, 2007;

Forman, 2008; Mell, 2009).

It is strategic because it fulfills multiple sustainable planning objectives. Green infrastructure is conceived in tandem with development rather than an antithesis. It is conceived as a network that exists and compliments other networks/systems within a landscape or urban system (Benedict and McMahon, 2002, 2006). Green infrastructure combines principles from landscape ecology and landscape planning (Forman, 2008) combined with principles of growth management and sustainable development. The aim of the multidisciplinary nature of green infrastructure planning is to achieve sustainable development in response to humanity’s needs and to maintain “nature” for the ecosystem services it provides for the benefit of humanity and other species. Ecosystem services are

1 derived from geological, biological and ecological processes and generalized under four

categories: provisioning services (food, water, timber, and fiber), regulating services

(regulation of climate, floods, disease, wastes, and water), and cultural (recreational,

aesthetic, and spiritual benefits), and supporting services (soil formation, photosynthesis,

and nutrient cycling) (MEA, 2005).Within metropolitan contexts, regulating, cultural, and

supporting services are derived from green infrastructure to improve wellbeing of

communities and environmental conditions. Yet, the explicit connection between green

infrastructure and climate change adaptation planning (as an ecosystem benefit) remains

less prominent.

The main premise to adapt to climate change is that the climate is already

changing due to historic emissions originating in the mid-nineteenth century (IPCC,

2007). These emissions result in increased temperatures across the globe with varying

impacts on humans and other species. Direct impacts of climate change are

environmental and physical with clear and strong repercussions on social and economic

dimensions. Some of these impacts include increased heat temperatures (resulting in increased intense heat human mortalities and species migration), flooding due to extreme precipitation events (resulting in flooding with building and infrastructural damage as well as related mortalities), rising sea level (inundation of large tracts of developed coastal urban land resulting in lost property) and extreme variability of local and micro climates (McEvoy et al. 2006). Accordingly, the magnitude of exposure and risk to these events vary the extent of vulnerability of communities (Brooks et al., 2005;

Gallopin,2006; Kazmierczak and Cavan,2011).Through understanding the interdependencies between vulnerability, exposure and risk of specific communities and

2 contexts, adaptation planning aims to adjust human and natural systems to the impacts of

increased temperatures (IPCC, 2007).These adjustments are and will be vast and costly

and cover institutional, administrative, environmental, physical and cultural changes.

Urban and metropolitan populations will incur the financial, moral and temporal costs of these changes because, currently, more than fifty percent of the world population is

residing in cities and mega-cities (UN, 2007).

More recently and in tandem with mitigation, adaptation planning is gaining more

attention with a strong focus on urban contexts. It is a developing field for planning.

Frameworks have and are being established to advance adaptation plans. The process

begins with climate science to identify current and future climate impacts to be able to

assess vulnerabilities and risks. Current adaptive capacities are then assessed followed by

adaptation strategies implementation and monitoring. Adaptation plans face many

limitations and include uncertainty of scientific information and projections, future

climatic projections, extent of materialization of projected future impacts, and cost of

adaptation. These impediments render many policy makers reluctant to pursue adaptation

plans and policies in an aggressive and explicit manner.

Emerging adaptation plans are beginning to set forth a series of multi-tiered and

multi-objective measures to address adaptation, including green infrastructure. Many of

these measures are conventional in approach and high cost engineering projects. Green

infrastructure has much to contribute to adapt to climate change impacts and adaptation

planning in a cost effective way, without explicitly advancing the climate change

adaptation agenda, and through gradual policies that are more acceptable to communities.

3 Yet, the extent and explicit contribution of green infrastructure has not been researched

enough.

The dissertation contributes to this discussion and furthers the understanding of

green infrastructure as an adaptation measure. The overarching argument is that the

theory and application of green infrastructure planning and practice can be an integral

and effective part of adaptation planning and implementation in metropolitan areas

(urban regions) because green infrastructure is: 1) multi-disciplinary 2) multi-objective,

3) spatial and hierarchal, and 4) meaningful to communities. Green infrastructure planning is supported by theories from landscape ecology, landscape architecture, growth management, and regional planning; operationalized by quantification of ecosystem benefits from natural sciences and environmental engineering; and advanced by practical applications from the fields of ecology, forestry and landscape architecture/planning. This multi-disciplinary characteristic, coupled with hazard and risk management strategies, provides a theoretical and practical basis to contribute and advance integrative approaches to adaptation planning1 . The convergence of all these fields prompt common

1 Two approaches for determining adaptation needs dominate the literature: The first is a linear cause–effect chain in which climate scenarios are the basis for estimating future climate impacts, which then deﬁne adaptation needs (Fussel, 2007). Adaptation to climate change is seen as largely separate from other social processes and activities, and adaptation needs are largely determined by scientiﬁc analysis, such as the IPCC report (1994-2007). The second approach is more complex and is characteristic of recent adaptation assessments. This approach arrives at a more comprehensive description of climate-related risks now and in the future by considering future climate change together with current climate variability and non-climate factors. The risk assessment is further informed by experience from management of past climatic hazards. Recommendations for adaption are determined by their potential to reduce current and future climate risks but also by their synergy with other policy objectives, for example sustainable development goals. The second approach considers the wider policy context of adaptation that usually lead to suggestions for mainstreaming climate adaptation into existing management activity and development plans (e.g. Huq et al. 2003).

4 multiple objectives for green infrastructure and adaptation by making urban and

landscape sustainability, climate change, and the wellbeing of communities as common

objectives. The fact that green infrastructure is all around us and can be seen and touched, and when used for adaptation, becomes a tangible adaptation measure that is meaningful to communities. Open space planning, urban greening, urban forest management and other vegetation policies are already underway in many US cities. The task upon us, as planners concerned with climate change, is to support and increase these initiatives and make explicit the connection to climate change

The dissertation includes reviews of literature on green infrastructure and adaptation and makes original contribution on how to integrate and make effective the use of green infrastructure for adaptation planning and measures. The dissertation argues that green infrastructure is an essential tool for addressing climate change adaptation, but

to be successful it needs to be: 1) conceptualized and organized, and managed based on

the green infrastructure transect (GIT); 2) implemented incrementally and in phases with

attention to where communities choose to live; and 3) with a focus on maximizing

ecosystem adaptation benefits by estimating potential space across land uses to

accommodate multi-scale green infrastructure measures.

Two key concepts provide the basis for the synergy between green infrastructure

and adaptation: the integrative/adaptive approach to adaptation and complementarity of

purpose. The integrative approach to adaptation arrives at a more comprehensive

description of climate-related risks by considering future climate change together with

past and current climate variability and non-climate factors. Recommendations for

adaption are determined by their potential to reduce current and future climate risks but

5 also by their synergy with other policy objectives, for example sustainable development

goals. Adaptation is often conducted under uncertainty of scientific information. The

field of adaptive management provides procedural approaches that provide a framework to advance adaptation by experimenting using piecemeal approaches with monitoring controls that allow flexibility in plan improvement and implementation.

Complementarity offers a clear and explicit way to connect green infrastructure

and adaptation at the operational level. Direct climate change impacts can be ameliorated

to some extent by the provision of green infrastructure and the ecosystem services it

provides. Complementarity is also at the conceptual level where green infrastructure

provides a multi-scalar framework to move adaptation from a solely local approach to a

regional one where local adaptation planning and implementation is integrated into a

regional vision.

The dissertation employs a mixed method approach that utilizes literature reviews,

case study analysis, policy analysis, and geographic information system (GIS) mapping

to address the propositions. Literature is reviewed to formulate the propositions and

develop theoretical frameworks and approaches. An incremental and phasing approach is

developed to address uncertainty in adaptation implementation that is based on adaptive

management. Policy analysis is used to understand the state of practice and to identify

synergies with developed theories. Case studies are used for application and to test the

approaches and frameworks. Lessons are then integrated into a proposed theoretical

framework to mainstream adaptation based on green infrastructure. Case studies are used

to test the adaptation implementation approach. Taking the urban island heat as a

surrogate to climate change, spatial mapping and analysis of land cover and land uses is

6 carried to test the theoretical green infrastructure framework in the Boston Metropolitan

area.

1.2. Dissertation propositions

The individual chapters are unified by four general propositions. These propositions

were developed to guide the research and act as a basis for the relationships between the chapters. Figure 1.1 provides a hierarchal schema of the interrelations between the main argument, propositions and the aspects addressed for each proposition.

Figure 1.1 was developed to ensure that each proposition is addressed and contributes to advancing the main argument of this dissertation. Different aspects of each proposition are addressed in more than one chapter and from different perspectives. The concluding chapter summarizes the propositions and evaluates the extent each was addressed.

1.2.1. Proposition 1: Green infrastructure planning provides an effective analogy to plan and manage climate change adaptation at the local and regional scales.

Green infrastructure is a strategic approach to urban and landscape planning that offers practical solutions for adaptation and an analogy to mainstream adaptation at multiple scales. Three arguments support this proposition: 1) the multi-scalar and multi- contextual nature of green infrastructure, 2) the nested hierarchy of green infrastructure physical typologies, and 3) green infrastructure is not bounded by administrative boundaries.

When impacts from a changing climate occur, they will occur regardless of administrative boundaries. Adaptation is conceptualized as a planning initiative conducted at the local scale. Adapting at the town or city scale is important because it allows communities to be involved in issues directly impacting them. Yet, coordination

7 across boundaries will be necessary to curb impacts that transcend these boundaries, such

as large flooding events or extreme heat waves. Green infrastructure provides an analogy

and practical road map to conceptualize and organize adaptation efforts into a regional

initiative. A regional vision for adaptation provides the basis for coordination of local

plans and resources. In such an approach, local initiatives will contribute to the regional

adaptation effort and a regional vision would guide local policy makers towards priorities within their jurisdiction. This becomes especially crucial when using green infrastructure as an adaption strategy. Small and large scale green infrastructure elements transcend administrative boundaries by the mere fact that they depend on biological and ecological processes for the delivery of benefits. These processes are interconnected through air, soil, water and species migration. Therefore, a regional vision that directs adaptation

planning and green infrastructure implementation serves the purpose of mainstreaming

adaptation.

Chapter 4 proposes a regional framework that includes scale and context

integration for adaptation planning using green infrastructure. Chapter five applies this framework to define the physical element and spatial distribution across a metropolitan area.

1.2.2. Proposition 2: The effectiveness of green infrastructure as an adaptation strategy is contingent on the provision of maximum pervious surface.

Green infrastructure depends on the biological and ecological processes to provide ecosystem services. This depends on the presence of soil, vegetation, and water as a harbor for these processes to occur. In urban contexts, pervious surface is the primary indicator of the extent and potential of the delivery of ecosystem services. This

8 proposition is supported with the following arguments: 1) pervious surface exists and

should be accounted for across all land-use types, 2) the provision of ecosystem services

is irrelevant of land ownership type and use, and 3) there are opportunities to intensify

existing pervious surfaces and transform impervious surfaces.

Chapter two discusses through a literature review the relationship of green infrastructure and the delivery of ecosystem surfaces within urban and surrounding contexts. Chapter four suggests a framework for accounting pervious surfaces across varying metropolitan contexts. Chapter five estimates the current pervious surface area and potential of intensification of existing pervious surfaces and transformation of impervious surfaces to hydrological active surfaces.

1.2.3. Proposition 3: Adaptive management offers practical advantages to implement climate adaptation plans.

Adaptive management posits an incremental approach to managing systems with a central component focused on monitoring and experimentation. This is particularly relevant to adaptation when much of adaptation planning is conducted under conditions

of uncertain future climatic projections and impacts. Uncertainty is not new to planning.

Planning, in general, projects future outcomes based on historical precedence. In the

context of climate change, actions described in adaptation plans are best determined

based on downscaled global or regional climatic projections. Until climatic projections

and impacts could be advanced to higher levels of precision, adaptive management

provides a robust analogy to define methodologies to advance adaptation planning, while

recognizing uncertainty. The precursor here is that we cannot wait until impacts begin to

materialize to begin action. In more specific, incrementalism, mainstreaming, and

9 experimentation are crucial theories to advance in adaptation planning while recognizing uncertainty.

Chapter two, surveys adaptive management and resiliency theory and its relevance to adaptation planning. Chapter three integrates adaptive management principles into the implementation phase of recognized adaptation procedural frameworks and proposes an incremental approach to implementing adaptation measures. Chapter four applies the incremental approach by proposing to estimate current existing area of pervious surfaces prior to implementing large scale green infrastructure policies.

1.2.4. Proposition 4: Green infrastructure contributes and strengthens the adaptive capacity within metropolitan regions.

Adaptive capacity is the inherent capacity within a system to recover from a disturbance. In the context of adaptation planning, exposure and risk determine the extent of vulnerability to climate change impacts that may change or affect the system (i.e.in this dissertation the metropolitan system). In the context of this dissertation, risk, exposure and vulnerability are related to the location of physical assets and communities.

Green infrastructure reduces risk by decreasing the intensity of impacts from a changing climate (i.e. reduce overall temperatures within a metropolitan area). Green infrastructure reduces exposure by protecting from hazards resulting from extreme events. By decreasing intensity of impacts and protection from hazards, it follows that vulnerability of communities may be reduced. While the proposition is focused on enhancing physical and environmental dimension of adaptive capacity, it is understood that the combination of physical measures and social, cultural and economic policies will result in a total

enhancement of the capacity to curb climate change impacts.

10 Chapter two explores the relationship between risk, exposure, and vulnerability and the contribution of green infrastructure. Chapter four proposes a framework to relate and prioritize climate change impacts, green infrastructure policies, and location of communities across a metropolitan gradient. Chapter five measures the potential of contribution of green infrastructure to the adaptive capacity through the estimate of vegetated cover to mitigate the impacts of the urban heat island.

Lastly, while this is not a core proposition of the dissertation, the idea that green

infrastructure as an adaptation strategy should be explicitly and aggressively pursued runs

across all chapters. The fact that vegetated surfaces deliver ecosystem services that

contribute to adaptation and wellbeing of communities seems very intuitive to be ignored.

The no-regrets nature and the cost: benefit of green infrastructure strategies in the short and long terms makes evident that green infrastructure is an opportunity that may prove to be very effective in curbing the pace of climate change, at least within metropolitan contexts, and not to be missed.

1.3. Dissertation Organization

The dissertation is comprised four main chapters and a conclusion. The first chapter is a general literature review and the remaining three are written in the form of publishable articles. One was already published as a refereed conference proceeding, one submitted for review and the third completed for this dissertation. All three articles were conceived to be included in this dissertation. Chapter five is a summary and conclusions of the articles as well as an assessment of the extent the propositions have been addressed. The references from each article were merged into a master reference section.

11 1.3.1. Chapter Two: Green Infrastructure and Adaptive capacity: Complementarity and Synergy

This chapter surveys the literature on green infrastructure and adaptation

planning. The aim is to identify the overlaps and contradictions between both fields. The

survey of the literature discusses theory, practice and applied case studies in both fields.

The purpose is to understand green infrastructure planning, its components and

procedures. Similarly, adaptation frameworks are surveyed to understand where green

infrastructure could contribute to adaptation planning. The main findings indicate that the provision of ecosystem services complimentary with requirements for adaptation, at least within the environmental dimension. The primary variable for the efficacy of using green infrastructure is: 1) space and 2) extent of pervious surfaces across the metropolitan area.

1.3.2. Chapter Three: Windows of Opportunity: Addressing Climate Uncertainty through Adaptation Plan Implementation

There is a pressing need for municipalities and regions to undertake adaptation planning that will create urban systems suited to current as well as future climates, but uptake of adaptation has been slow. This is particularly unfortunate in that patterns of

urban built form interact with climate change in ways that can reduce, or intensify, the

impacts of overall global change. For policy-makers, uncertainty regarding the timing

and magnitude of climate change is a significant barrier to implementing adaptation

planning. Resiliency theory suggests an approach to evaluating adaptation options for

cities that can bring these factors together. The method we propose focuses on

implementation of adaptation, and phasing of policy. It removes time as a decision

marker, instead arguing for an initial comprehensive plan to prevent maladaptive policy

choices, implemented incrementally after testing the micro-climate outcomes of previous

12 interventions. Policies begin with no-regrets decisions that reduce the long-term need for more intensive adaptive actions and generate immediate policy benefits, while gradually enabling transformative infrastructure and design responses to increased climate impacts.

Global and local indicators assume a larger role in the process, to evaluate when tipping points are in sight. We use case studies from two exemplary municipal plans to demonstrate this method's usefulness. While framed for urban planning, the approach is applicable to natural resource areas and other planning that must manage uncertainty.

The model is applied to two case studies in Australia and Copenhagen with new insights on an adaptive approach to adaptation implementation.

1.3.3. Chapter Four: The Green Infrastructure Transect: An Organizational Framework for Mainstreaming Adaptation Planning Policies

When considering the range of spatial planning actions that cities can take to adapt to climate change, many of them fall under the conceptual umbrella of green infrastructure (GI). GI has been defined as the spatial planning of landscape systems at multiple scales and within varying contexts to provide open space, safeguard natural systems, protect agricultural lands, and ensure ecological integrity for cultural, social, and ecosystem benefits (Benedict and McMahon, 2002, 2006; Ahern, 2008). While the traditional definition of GI refers to areas of land that are least intervened by human action, in this expanded definition, we are deliberately including areas that are engineered to mimic natural processes and which provide cost-effective ecosystem services.

Although climate adaptation is a fairly new policy goal for GI (Gill et al., 2007; CCAP,

2011), three key characteristics qualify GI as a suitable tool for adaptation planning including multi-functionality (to match ecosystem beneﬁts with adaptation needs), multi-

13 scalar nature of the spatial elements, and a ‘no-regrets approach’. However, GI needs to

be matched to the character of the urban environment and coordinated across jurisdictions and planning scales to become an effective adaptation policy. In this chapter, we present a policy framework, the green infrastructure transect, that can help planners and policymakers identify appropriate GI policies for different urban environments and describe how these policies can create a regional adaptation planning framework.

1.3.4. Chapter Five: Estimating the potential vegetated pervious surface are for urban heat island adaptation strategies in the Boston Metropolitan Area

Green infrastructure (GI) is recommended as suite of adaptation of measures to reduce temperatures within urban contexts. The mitigation of the Urban Heat Island effect provides a test case for the effectiveness of GI as a future adaptation strategy.

Climate change is a global phenomenon that exacerbates local urban heat island (UHI).

UHI is directly correlated with urbanization as predevelopment land cover is transformed from hydrological active surfaces to impervious surfaces. For green infrastructure to have an impact on metropolitan climates, tracts of land (privately and publicly-held) should be preserved, acquired and envisioned as a metropolitan network of ecologically vegetated green infrastructure network providing maximum ecosystem benefits. To achieve that, maximum surface areas should be dedicated to vegetated surfaces. Yet the availability of space is not evident especially when considered in the context of urban expansion.

This chapter tests for the availability of surface area within land uses and across

the urban-rural gradient within the Metropolitan region of Boston, Massachussetts.

Several strategies are developed to intensify vegetated land cover (pervious surfaces) and

transform impervious surfaces (where possible) to vegetated surfaces within existing

14 patterns of development. A percent pervious (PP) metric is developed to characterize the region. Gradient zones of PP are defined and used to compare space opportunities across land use classes. Maximum opportunities are then tested to address the urban heat island phenomenon. A gradient of vegetated green infrastructure policies is then derived based on the availability of space across gradient zones. The results indicate that residential land uses carry the highest potential to intensify vegetated land surfaces while commercial land uses provide the highest potential for green roof implementation and impact of the UHI. The results for the gradient of policies indicate that transformation of impervious surfaces is possible within the urban core with more dependence on land based forested canopy towards the periphery of the study area.

1.3.5. Chapter Six: Conclusion: Integrating Green infrastructure into Adaptation

Planning

This chapter reviews the research in the dissertation as well as the limitations.

Green infrastructure policy implications are discussed for the purpose of increasing the potential adaptive capacity across urban regions. Original contributions are then identified with future research questions to be pursued.

15 CHAPTER 2

GREEN INFRASTRUCTURE, URBAN FORM AND ADAPTIVE CAPACITY: COMPLEMENTARITY, SYNERGY AND CONTRADICTION

2.1 Introduction

The purpose of this section is to identify the gaps in the knowledge that this study

will address, provide a literature background for the study and identify the variables of

concern. What we know is that CC is happening and the need to address the impacts on human and natural systems is becoming more urgent with increasing scientific evidence

(IPCC, 2007). While planning for CC has been predominantly focused on mitigation

(wheeler, 2008), adaptation is coming to the forefront of CC planning, especially within

the context of urban areas.

Urban regions are vulnerable and are already affected by CC due to historic

emissions (Wilby, 2006). Increased exposure and risk, and the deficiency of existing

urban systems to cope with known and uncertain magnitude of impacts, render urban

regions and their populations vulnerable to changing precipitation regimes, increased

temperatures (micro and regional climates), increased extreme daily temperatures

(intensification of the urban island heat effect), and rising sea level (Wilby, 2006, 2007).

As such, cities are just starting to address these impacts through initiatives that change

people’s preference for living (denser developments, reduction of power demand ),

reduce miles travelled (providing walking, cycling, and public transportation), provide cooling stations, mix uses in developments, and increase tree canopy (Lindley et al.,

2006). In addition to these measures (which coincide with sustainable planning and Smart

Growth principles making adaptation easier to mainstream), green infrastructure planning

16 is identified as an effective measure to prepare cities and their surrounding regions to CC impacts.

GI planning has been recognized at the regional scale to control sprawl (Benedict and McMahon, 2002, 2006), provide cultural and environmental benefits (Fabos and

Ryan, 2006) and at the local scale for environmental control (EPA, 2011). The focus of

GI on adaptation is new. To date, the literature has primarily focused on connecting the characteristics of GI planning with adaptation needs. We know that GI is a no-regrets approach to adaptation, it is complimentary with adaptation objectives, it is multi- functional thus providing social, economic, cultural, ecological and environmental benefits, and it is multi-scalar allowing applications from individual buildings (green roofs), streets (trees and swales), and parks to regional forests and reserves (Benedict and

McMahon 2003; Ahern, 2008; EPA, 2011). While these are very significant qualitative ingredients that have become the basis of many perspectives on the subject, few studies begin to provide practical guidance on how to adjust the urban (and regional) form to incorporate GI.

Many research connect aspects of the urban microclimate with GI (Ashley et al.,

2007; Gober et al., 20010), but only few connect adaptation, GI and spatial distribution into one study with practical useful results to planners2. Gill et al. (2007) use the city of

Manchester to characterize land cover within the urban city boundaries. They determine land-cover percentages within land-use categories as their primary input for climatic modeling. By using climatic and planning scenarios, they determine the existing adaptive capacity and the future potential based on IPCC climatic scenarios and planning scenarios

2 See GRaBS Project at the University of Manchester, UK. http://www.grabs-eu.org/

17 such as increasing green roofs, green walls, and tree planting. The results on adaptive capacity are percentage values for the whole urban area without any specificity to location, the impact of built urban form on this capacity, the threshold capacity of GI and the connection to preference of living.

The research by Gill et al. (2007) provides an inspiration for this study and highlights some of the gaps that are required. The comparative study by Forman (2008) on urban regions highlights the significance of GI at the regional scale and the delivery of ecosystem services essential to the wellbeing of people. Furthermore, by including the work of Hamin and Gurran (2008) on connecting urban form and climate change and the work by Brabec (2002, 2005) impervious surfaces within watersheds, the research will address the following gaps within the literature:

Provide assessment and measurement of spatial patterns of GI at the regional scale.

Provide practical data and information useful and applicable to spatial planning for climate change adaptation.

2.2 Resilience: An Interpretive Framework

The interpretive framework the study utilizes is the theory of resilience. The concept of resiliency originates from the field of ecology. Resiliency is defined as the capacity of a system to rebound after a natural or manmade disturbance (Holling, 1973;

Walker et al., 2004; Walker & Salt, 2006). This means an internal capacity to accommodate change and rebound to a state that is similar to the original state, but not necessarily identical. In many cases, where the magnitude of the disturbance is high, social-ecological systems may reach thresholds that alter their structure thus transforming their capacity to deliver ecosystem benefits.

18 This research considers urban regions as social-ecological systems. This means that human and natural systems within urban contexts are interconnected and interdependent and mutually affecting one another. Extending this analogy from the sciences to planning, there is a need to integrate and plan for both systems to consider mutual causes and effects. Therefore, the research brings together the work on urban regional patterns by Richard Foreman (2008) from landscape ecology and Brabec and

Lewis (2002, 2005) from resource efficiency along with recent literature on planning for climate change (Howard, 2009; Hamin et al., in review). In this manner, natural systems

(GI in the context of this research) and urban form are mutually considered as both causes and effects.

Walker and Salt (2006) propose three steps to manage for and enhance resilience of social-ecological systems. First, is to understand the drivers of the system under a certain condition. Slow drivers are coarse-scale variables often coupled with fine-scale fast variables. In the context of urban regions the coarse and fine scales drivers is the transformation of land from pervious and impervious surfaces, which is urbanization. The large scale driver is in this case is climate change which augments the conditions of the urban micro-climate.

Second, is to know the thresholds of drivers. When discussing GI, the limits of delivery of ecosystem services could characterize the threshold of adaptive capacity. The adaptive capacity is determined by the extent of area of pervious surfaces. The amount of pervious surface within an urban region determines how much water may percolate or be stored or the extent of canopy cover for temperature amelioration. Within urban regions, the amount of pervious surface is finite and constantly eroding, thus limiting the adaptive

19 capacity. Therefore, threshold is this context is directly related to the proportion of pervious to impervious surfaces within a given unit of study or area.

Third, is to enhance aspects of the system that enable it to maintain its resilience.

In this research, GI is considered as the element that enhances this resilience. GI is functions are determined by the extent of ecosystems provided such flood water storage, temperature reduction, and CO2 sink. By understanding the capacity of GI for adaptation, and planning at multiple scales to create a regionaal network of GI, the urban system is enhanced (in one way) to increase the resilient capacity of communities (Figure 2.1).

Step1: Identify Slow and fast drivers of change Climate Change and Urbanization Step 2: Understand thressholds of drivers impervious surfaces, extreme Temperatures

Step 3: Enhance aspects of the systeem Green Infrastructure (spatial patterns)

REESILIENCY

Figure 2.1: Three steps to achieve resiliency by managers and planners (Salt and Wallker, 2006).

In the context of this research, GI is used as the tool to enhance the urban system in face of the slow driver of change, climate change. In additiion, the capacity of the spatial elements of GI are limited by the fine scale driver of urbanization that creates cere tain limitation to the delivered services due to the conditions of surface transformation from pervious to impervious. The relationship beetween adaptive capacity, thresholds and prevalent conditions is the theoretical analogy that this research operates under.

20 2.3 Adaptation to Climate Change

Urban regions3 are identified as being simultaneously contributing to climate

change and being impacted by climate change (Stone, 2005; McEvoy et al., 2006; Wilby,

2007; Watkins et al., 2007; McPherson et al., 2008). Urban regions contribute to climate change in two ways. First, through the resulting greenhouse gas emissions which add to

the global stock of suspended particles in the atmosphere (IPCC, 2007); and second,

through non-emission contributions due to the morphological nature of urban areas of

altered land use and cover, as compared to predevelopment conditions (Gober et al.,

2010). Both factors force changes in local and regional climates (Stone, 2009) which

impact the wellbeing of communities and the ecological systems within.

Urban regions are impacted and are vulnerable to climate change impacts.

Citizens and ecological systems within are subject to the same vulnerabilities.

Vulnerability maybe defined as the combination of increased exposure and risk and the

deficiency of existing measures to cope with known and uncertain impacts. Impacts from

climate change as characterized by the IPCC (2007) report are increased or decreased

precipitation regimes, drought, increased temperatures, increased extreme daily

temperatures, and rising sea levels (IPCC, 2007) (See Table 2.1 for a full list of impacts

and indicators). The impact on urban areas simultaneously affects social and ecological

systems. For example, Guhthakurta and Gober (2007) show that an increase of daily low

temperature 1F (under UHI conditions) in the city of Phoenix increases water use by 290

gallons per typical single family unit.

3 Richard Forman’s (2008) definition of urban regions is used to mean a major city, the surrounding metropolitan area, suburbs, and any satellite cities in the vicinity, any heterogeneous area composed of buildings and open space, as well as natural amenities.

21 In addition to increased water use due to rising temperatures, the watersheds that

store and provide water to urban areas are and will be simultaneously impacted. For

example, it is projected that the watersheds that are responsible to store and provide water

in New York City will be affected by lower precipitation and drought conditions to

sustain the resource and provide sufficient quantities of water (Rosenzweig et al., 2007).

The same can be said to other ecosystem benefits that are provided by ecological systems within urban areas such as food, energy, clean air, and nutrient cycling (Forman, 2008).

The IPCC report (2007) clearly indicates the vulnerability of these systems and their potential high impact to social systems within urban regions. Therefore, impacts and causes are closely related at multiple scales effecting social and ecological systems.

To address the impacts of climate change through adaptation, structural and non- structural measures are proposed (Heinz, 2006; Wheeler et al., 2009). Structural measures are engineered systems that respond and protect from vulnerability due to exposure to hazardous locations such as flooding, erosion and sea rise levels in coastal urban regions.

For example, Barnett and Beckman (2007) discuss structural solutions, such as dams and levees, to protect against sea level rise in Boston, New York, Miami and San Francisco

Bay. Others, have proposed raising buildings above flood plains (Walker and Salt, 2006), redundancy of road networks (Kirshen et al., 2008), or levies to abate rising sea levels. In

many urban areas, structural measures may be the only viable solution to adapt, other than abandonment. Such options are expensive, but maybe justified on the basis of existing property values or other reasons. Structural solutions need to be evaluated based

on technical, financial, and political feasibility.

22 Non-structural measures are existing natural systems, constructed natural systems,

or simply un-built/undeveloped land; in other words the green infrastructure of an urban

region. While structural and non-structural measures may be needed to respond to climate

change impacts, specific contexts will determine what is needed, which combination and

what magnitude of each. Within this study, we focus on green infrastructure as a non-

structural measure and will be discussed in the next chapter.

In this research, I focus on non-structural measures to address climate change

impacts and enhance the urban system, namely GI. The two primary impacts of CC that I

seek to address and relevant to urban contexts are increased precipitation regimes (that

result in increased flooding) and temperature amelioration. The following section GI is discussed with relevant factors that need to be considered.

Impact Impact Description Reference Indicator Caused by Extreme Events Longer and Increasing summer temperatures will result in more Jendritzky more intensive frequent heat waves. In addition the duration of heat (2007), heat waves waves will extend and extreme temperatures will Laschewski increase. Heat waves and extreme temperatures are a (2008), and threat to human and animal health. They can also WHO (2005) cause damage to buildings and structures. Increase of Changes in seasonality and intensity of precipitation Zebisch et al. heavy rain and will result in more frequent heavy rainfall. More (2005) flash floods extreme weather events like thunderstorms will increase the likelihood of flash floods. Unlike river floods flash floods are a local phenomenon not limited to flood plains. Flash floods cause damage to buildings and private property. Increase of Changes in seasonality and intensity of precipitation Hennegriff et large river will result in more frequent river floods. River floods al. (2006) and floods cause damage to buildings and structures in flood Merkel et al. plains. (2008)

23 Increase of Melting ice and increasing water temperatures cause Rahmstorf storm surges sea level rise. More frequent storm events add to this (2007) and and result in more frequent and more extreme storm Woth (2006) surges. Storm surges can cause coastal erosion, damage to buildings and structures especially in ports. In cases of dam failure such damage will be severe and flooding will endanger human lives. Increase of More frequent heavy rainfall and changes in freeze- Glade et al. mass thaw cycles will destabilize hillsides as well as (2005), movements embankments and cause landslides and rock fall. Lang and Stronger variation in snowfall can change the Glade (2008), concurrency of avalanches. Mass movement events and McInnes cause damage to buildings and structures and endanger et al. human lives (2007) Increase of More frequent heat waves and changes in precipitation Badeck et al. forest fires cause drought and increase the danger of forest fires. (2004), Forest fires cause harm to the forestry industry. They Westerling can also damage buildings and structures as well as and Bryant wildlife habitats. (2008) and Girardin et al. (2009) More frequent Several extreme weather events can cause harm to Zebisch et al. destruction of road and rail networks. Even if public infrastructure is (2005) infrastructure not always directly damaged, traffic and transportation and Walsh et is delayed or disrupted al. (2007) Caused by slow changes Increased loss Changes in seasonality and intensity of precipitation Mathews et of soil by will result in more loss of soil by water erosion. al. (2006) water erosion Erosion reduces soil fertility but also affects water quality and causes increasing problems due to deposition. Changes in temperature as well as precipitation and Leuschner Loss of their seasonal variation are a kiumajor thread to and Schipka species and biodiversity. They will cause a rapid change in habitat (2004, Root biodiversity conditions which may lead to an increased loss of et al. species and biodiversity, especially in regions with a (2003), and high proportion of fragmented or isolated habitats. Walter et al. (2002) Increased Changes in seasonality and intensity of precipitation as Hergesell and fluctuation of well as changes in freeze-thaw cycles will cause more Berthold the pronounced fluctuations of ground water levels. This (2008),Kämpf groundwater could be exacerbated by erosion processes that reduce et al. (2008) level soil water storage capacity. The resulting drop in and Merkel et ground water levels in summer will increase drought al. (2008) situations, while rising ground water tables in winter endanger the structural integrity of buildings.

24 Fluctuation in Changes in the seasonal variation of precipitation will Zebisch et al. the result in more pronounced fluctuations of river (2005) availability of discharge. Shortage of water in drought situations can and Lehner et water for constrain production. Energy production can be al. (2005) industrial use especially affected

Desertification Table 2.1: Climate change impacts and their indicators

2.4 Green Infrastructure

Green infrastructure planning is predominantly a resource planning approach to conservation and protection of land with ecological and cultural benefits. Benedict and

McMahon (2002) define green infrastructure as the “natural life support system-an interconnected network of waterways, wetlands, woodlands, wildlife habitats and other natural areas; greenways, parks, and other conservation lands; working farms, ranches and forests; and wilderness and other open spaces that support native species, maintain natural ecological processes, sustain air and water resources and contribute to the health and quality of life for America’s communities and people (p.12)”. As a process, it is described as holistic, comprehensive, strategic, publicly implemented, participatory, and multi-disciplinary (Benedict and McMahon, 2002). Green infrastructure provides human and other species benefits such as enriched habitat and biodiversity; carbon and nitrogen cycles; carbon cycle; maintenance of natural landscape processes; cleaner air and water; increased recreational opportunities; improved health; better connection to nature and sense of place. In addition, green infrastructure is proposed as a growth management tool to control and direct patterns of development (Benedict and McMahon, 2002). When considering GI as an adaptation tool to climate change, every possible typology of un- built parts of the urban region should qualify in the definition of GI. Therefore the

25 definition of GI is expanded to include all possible potential to create and engineer GI

elements.

In addition to Benedict and McMahon’s (2002) definition, I include in the

expanded definition the following elements and typologies: hybrid systems such as green

roofs (Gill et al, 2007; Oberndorfer, et al., 2007) or vegetated facades of buildings;

interstitial spaces between buildings and streets (Corner, 2006); residual spaces from transportation and utility infrastructure (Garde, 1999); derelict land from industrial and military sites (Corner, 2004; Berger, 2006; Donadieu, 2006); urban agriculture; landfills

(Belanger, 2009); and private open spaces on residential property, schools, cemeteries, shopping malls, and parking spaces (Gill et al., 2007). The inclusion of these spaces to

form an urban green infrastructure network is intended to show the current capacity of

any urban region and its potential to increase the density of green infrastructure within

urban regions. This allows the provision of mitigation and adaptation measures at

multiple scales.

Varying planning strategies maybe employed at varying scales to develop such an

extensive network. For example, conservation of land through policy at the regional scale

for cultural, economic and ecological purposes (Benedict and McMahon, 2006); spatial

conservation approaches that define emerald networks (Forman, 2008) or biosphere

reserves (Solecki and Rosenzweig, 2004). Similarly, at the parcel scale, efforts such as

the Berlin biotope green area factor encourages owners to increase planted area and treat

run-off on site through financial incentives (BAF). The BAF incorporates hybrid systems

such green roofs and vertical vegetal surfaces. While the idea of incorporating green

infrastructure within urban regions may seem attractive, it is a space intensive measure.

26 Within space limitation in urban contexts, there is a need to intensify effort to find ways

and think of creative ways to ‘find’ space (Gill et al, 2007; Currie and Bass, 2008). This

study will address this aspect of the problem during the research process and will explore

the size and function of such a network along the urban-rural gradient.

There are four key ideas contained in the above definition that define green

infrastructure: network/connectivity, spatial arrangement and distribution, benefits, and

implementation tools. The ideas are briefly explained to set the context for the discussion

on scale and using green urbanism and landscape planning literature to point out to

varying meanings of these ideas. By discussing the five ideas, an overarching framework

for a hybrid definition of a GI across the urban-rural gradient is defined.

2.4.1 Network and Connectivity

The network aspect, and consequently connectivity of the system, underlies

ecological and social systems. One could argue that networks of open spaces have

evolved as systems to preserve and use open space in lieu of urban development. GI is

planned as a network of interconnected parts that forms a functioning whole (Randolph;

Benedcit & McMahon, 2002, 2006; McDonald et al.) with varying focus on ecological

and social concerns that depend on spatial contexts and planning objectives. Networks of open space simultaneously developed over time as urban park networks (The Emerald

Necklace in Boston and Chicago open space system by Fredrick Olmsted; Fabos and

Ryan, 2002) and regional systems (Elliot, McKaye and Lewis, 1996; Fabos, 1995).

Precedents to GI within the US and Europe have varyingly emphasized ecological and recreational values. As a multi-functional system, the emerald Necklace in Boston is the first connected and multi-functional system of open spaces that combines recreational,

27 preservation of natural landscape and management of the hydrological systems criteria

(Ndubisi 2002, Fabos 1985, Fabos 1996, 2004, Ahern 2004). Contemporary examples follow suite with varying emphasis on function and benefits. With primary ecological emphasis, an ecological network (predominantly in Europe) is a system of nature reserves and their interconnections that make a fragmented natural system coherent in order to support more biological diversity than in its non-connected form (Opdam et al., 2002;

Jongman 2004; Jongman et al., 2004; Jones-Walters, 2007). Greenways are also an example of networks of linear elements that predominantly focus on recreational and social benefits (Fabos and Ryan, 2006; Hellmund and Smith, 2006; Imam, 2006;

Ribbeirto & Barao, 2006) with lesser extent on ecological processes (Ahern, 1995, 2004).

Examples of greenways range from regional examples such as the New England

Greenway System (Fabos et. al, ) to multi-scalar systems that connect urban core areas to regional amenities such as the Confluence Greenway System in St. Louis to local rails for trails such as the Norwottuck trail in Amherst, MA. Another example of networks is

Greenbelts which originate in the UK and used at the regional level to control urban growth and provide recreational amenities for urban populations. It could be argued that it is the antecedent to the contemporary urban growth boundary used by Smart Growth.

While it is not explicitly intended to be a network, the circumscribing nature around urban cores prompts it to be so. The common aspect between all these precedents to GI is connectivity.

Inherent in networks is connectivity and linkage (Ahern, 1995; Benedcit &

McMahon, 2006; Fabos & Ryan, 2002; Nadubisi, 2004; Mell, 2009; Kambites & Owen,

2006). Physical and functional connectivity is key for networks with an ecological

28 objective to ensure species diversity and flow of natural processes (biotic and abiotic resources, Forman, 1995; Opdam, and Vos, 2002; Nadubisi, 2004; Jongman et al., 2004).

Connectivity within networks with a social and recreational focus ensures access to open spaces resources within urban cores and regional contexts. Linear elements of green transportation systems (biking and walking) define the linkages between parks and places of living (Fabos & Ahern, 1996; Chiesura, 2004; Conine, at al., 2004; Ribbeirto & Barao,

2006). Connectivity is also critical to ensure resource continuity and protection such as the plan for a GI network Barcelona, Spain (Forman, 2008). Connectivity functions at multiple scales for multiple reasons to ensure ecological health, accessibility, and resource protection. Connectivity depends on the individual spatial elements that form the network.

While connectivity is an integral characteristic of GI definition, when considered for adaptation purposes, it is not evidently applicable or necessary. In this research, the question of connectivity for adaptation purposes is left open and not addressed directly.

2.4.2 Multi-scalar spatial elements

The spatial elements that define a GI network vary with objective and context.

Within ecologically oriented plans, landscape ecological spatial principles of landscape configuration, namely patch-corridor-matrix system, are tools to designate core habitats, hubs, and movement corridors (Forman, 1995; Nadibisi, 2002; Zonneveld, 2002; Ahern,

2007). The elements that compose an ecological network are core areas, buffer zones, restoration areas, and ecological corridors (Jongman 2004, Jones-Walters 2007).

For example, that could include forests, lakes, river systems, wilderness reserves and grass fields. The basic variable that defines the elements of an ecological network are

29 species consideration and ecological flows. Elements that compose a GI network for human use and in proximity to urban areas, is a finer system of open spaces that includes a larger typology.

A typology of a GI within urban regions may be categorized into two categories: within urban cores and on the urban periphery (Schilling & Logan, 2008). A GI within cities may include neighborhood, city and regional parks; playgrounds and play fields; community gardens, recreational trails and urban greenways, underused private and public land, urban tree canopy including streets trees, public plazas, waterfronts, green roofs, abandoned land that can be converted to new green spaces, and commercial agricultural and forestry sites (Waldheim, 2006; Corner, 2002; Bunster-Ousa,

2001;Carroll, 2007). In addition residual and marginal spaces from transportation infrastructure contribute to urban GI capital (Garde, 1999;Mossop, 2006) . At the urban fringe, GI is comprised of larger green spaces that take advantage of lesser urbanization.

These may include waterways, wetlands, woodlands, wildlife habitats, greenways, large parks, conservation land, working farms, ranches, forests, and riparian flood plains

(Fabos and Ahern, 1996; Randolph, 2004; Hough, 1995, 2004). In addition, deregulated military sites, abandoned industrial sites, large logistical sites, spaces around and within transportation and utility infrastructural networks could be connecting linear elements at the urban fringe (Waldheim, 2006; Berger, 2006;Reed, 2006; Bellanger, 2006, 2007).

While GI within urban core and region areas are predominantly target human use and benefits, ecological health of these elements is key to maintain and ensure the social and cultural accrued benefits, thus the critical interdependency between ecological and social

30 systems. In this sense, a regional GI that connects the urban gradient has the potential to be truly multi-functional addressing social and ecological benefits.

Such a network is comprised of several typologies of spatial elements that are related to urban land use (Table 2.2) and to landscape ecological principles (Table 2.3).

The intent is to ensure that the multi-scalar network incorporates as many types possible and organized based on ecological principles set forth by green infrastructure planning.

The variability of size, and location within and outside urban cores and formal and informal un-built spaces provide the stock of land that potentially can be incorporated into such a network.

Region/Metropolitan City Neighborhood Parcel Area Wetlands Urban forests Parks Back yards Agricultural pasture Large Parks School yards Street Trees Natural reserves Urban Golf courses Unused lots agriculture Forests Network of Athletics fields Community gardens parks Rivers Cemeteries Spaces between building

Ponds Network of street trees Easements Lakes Gardens Parks Open space in residential developments Coastal parks Institutional building grounds Shoreline ecosystems Abandoned property Utility Lines Greenways Table 2.2: Typology of Green infrastructure spatial elements

31 Regional Urban Regional Urban Corridors Regional Urban Matrix Patches Parks Utility Lines Sub-urban residential Neighborhoods Golf Courses Drainage ways Industrial districts Wetlands Cultural greenways Waste Disposal Areas Natural reserves Ecological greenways Commercial Areas Forests River ways Mixed Use Districts Urban Forest patches Transportation Corridors Urban cores

Table 2.3: Typology of Green Infrastructure based on spatial configuration (After Forman’s (1987) Patch-corridor matrix concept and as suggested by Gill et al. (2007) and Ahern (2008)).

2.4.3 Multi-functionality as Ecosystem Benefits

Green infrastructure provides multiple services that include social, ecological,

and environmental4 benefits. These benefits accrue to people and other species with

varying magnitudes. Ecological benefits for species may include habitat conservation and connectivity, enhanced biodiversity, carbon and nitrogen cycling, maintenance of natural processes, and ensuring the continuity of water systems processes (Benedict &

McMahon, 2002, 2006; Millennium Ecosystem Assessment Report, 2006). The benefits to ecological systems mean a healthy functioning system, which in turn, ensures the provision of ecosystem services that people and the environment benefit from.

Environmental benefits pertaining to people and other species include storm

water pollutant reductions (Booth et al., 2002, 2004), enhanced groundwater recharge

(Brabec at al., 2002; Clausen, 2007), protection of soil, increased carbon sequestration

(McPaherson et al., 2008), adapting and mitigating to climate change (Gill et al., 2007;

4 I use the terms ecology and environment to differentiate what is termed biotic and abiotic, respectively, within natural resource fields. Ecological pertains to living species and environmental pertains to non-living such as soil, water, and air.

32 Currie & Bass, 2008; Mell, 2009; Hamin & Gurran, 2008), address flooding (Caroll,

2007) and improved air quality (EPA). The environmental and ecological benefits in turn translate into social and cultural benefits directly connecting people with open space.

Services derived by people from healthy GI translate to social, cultural, and economic benefits. Kambites and Owen (2006) and Schilling, J. & Logan (2008) provide an extensive listing by combining several sources. These include: increased recreational opportunities through access to parks, improved exercise and walking, increased water quality and quantity through best management practices, better connection to nature through engagement and access, improved sense of place through greening and social interaction, increased real estate values of property around open space, historical rootedness through heritage conservation, local food production, and ameliorating urban island heat impacts (Stone, 2010). While these benefits are not comprehensive, they provide an idea to the importance of GI within urban contexts.

The social, ecological, and environmental benefits are interconnected. Through ecological health, environmental and social benefits are derived. Through ecological and environmental services, social benefits are accrued by societies; and through best management practices and ecological and environmental consciousness and stewardship, ecological and environmental services are maintained. The best example of this feedback loop is climate change. Through anthropocentric emissions into the environment, ecological and environmental services are reduced due to changing climatic regimes. As such, the benefits that people accrue, such as water use and food production, are affected by reduced ecosystem services. Sustainable planning practices that consider ecology, environment and social wellbeing are part of GI’s principles.

33 2.4.4 Implementation

Green infrastructure is inherently a spatial approach to land-use planning and design. The objectives of planning a GI is to lay out (Benedict and McMahon, 2002,

2006; Kambites and Owen, 2006; Ahern, 2007) the system strategically as a whole ensuring integrity of the network. Furthermore, linkages between agencies, non- governmental organizations, and the private sector are key to ensure the spatial whole.

This means that GI should function across scales and jurisdictions.

There are two dominant approaches that follow the European spatial concept model (Ahern, 1999; 2007) and the US comprehensive planning5 model. The European approach defines a broad scale spatial concept, such as the ‘Green Heart’ in Holland, to capture the image- ability of people and ensure marketability of the proposal (Shrijnen,

2000; Tjallinji, 2002). Resources are assessed; goals defined at the nationals scale, then put out for stakeholder and community comments, then altered as necessary to reach a consensus. Alternatively, the US approach is based on deriving goals and visions through a public participation process that takes into account stakeholder needs and requirements.

Assessment of resources is carried out by specialists to identify systems condition and suitability. Alternative plans are derived and discussed to arrive to consensus. The main difference between both approaches is the scale of planning. In Europe, planning is national (top down), and as such, statuary bodies define a broad vision where local plans

(which may be participatory in nature) are plugged into the overall vision. In the US, planning is locally based (bottom up) with some level of state intervention. Land use

5 Comprehensive planning steps include: Identify issues, state goals, collect data, prepare plan, create implementation plans, evaluate alternatives, adopt plan, and implementation and monitoring.

34 decisions are directed at the town and city scale. At the state and national scales, rare coordination exists that defines the overall outcome and impact. In both approaches, assessment methods are similar based on geographic information systems and suitability analysis methods (McDonald et al.; Conservancy Fund) based on values derived from the project objectives (McHarg, 1969).

What is evident is the significance of combining both approaches. Within the US context and holding national planning aside (for now), a spatial concept at the local scale could be an approach to direct a participatory process, thus combining a top-down and bottom up approach. This idea is suggested because a participatory process is time consuming. A spatial concept that communities can react to and suggest comments may provide a better use of time and resources. Accordingly, conservation tools for GI implementation such as land acquisition, easement, flood plain management, smart growth tools, conservation land, land trusts, and partnerships with private entities may be developed.

What is apparent is the significance of scale and context in the reviewed models.

To account for varying objectives of land conservation, benefits, and spatial contexts to address land transformation and climate change impacts and causes, approaches at multiple scales that account for multiple variables are needed. To develop the rural gradient framework of GI, scale and context variability is central to account to the varying gradients of benefits, spatial configurations of patterns of use and GI, as well as to ensure the network aspect of GI at multiple scales and across gradient sectors. The discussion that follows attempts to discuss all the above aspects of GI in the context of

35 scale and context variability to redefine GI as a system of networks that functions across

scales and across contexts within the urban region.

2.4.5 Complementarity Between Green Infrastructure and Adaptation Measures

Green infrastructure is a suitable suite of measures that have the potential to

address climate change impacts. The environmental impacts of climate change such as

increases temperatures, run-off storage and infiltration due to extreme precipitation can

be potentially managed by multi-scale green infrastructure of patches and corridors. As an ecosystem based system, green infrastructure in the form of trees, for example, may reduce temperatures within urban contexts. Retarding basins and infiltration swales allow the delay of run-off and infiltration of excess rain water. This synergy renders green infrastructure as a complimentary (Table 2.4) response to climate impacts.

Green infrastructure measures already exist in many cities in the form of urban canopy, street trees, gardens, and parks. While these may not be conceptualized as green infrastructure providing benefits, the advantage is that cities are already doing what at least could be a first step. In other words, these ecosystem based measures may be considered as no-regrets measures that in addition to reducing climate change impacts, will also provide social, environmental, and cultural benefits. Therefore, the complementarity between green infrastructure and adaptation needs provide the impetus for cities and policy makers to enhance and intensify greening projects across urban areas.

36 Climate Change Impacts Ecosystem Benefits of Green Infrastructure Increased run-off and flooding due to Reduce and run-off through retarding, extreme precipitation events or rising storage and infiltration. Examples include sea levels swales and tree canopy. Local and global temperature increase Reducing surface and air local temperatures by shading and evapotranspiration. Anthropogenic emissions Improve air quality by reducing particles in

the air and act as CO2 sink thus mitigating climate change and reducing local urban heat island Rising sea level leading to flooding, Coastal/ river, natural and engineered erosion, and inundation. systems allow the absorption of any direct impact and protection to dwelling areas. Table 2.4: Complementarity of GI benefits and adaptation needs

2.5 Urban Form

The literature from climate change planning, land-use and urban planning suggest two ways to mitigate climate change within urban regions. First, through denser and more compact urban form and development (Lerch, 2007); and second, changing land uses towards uses that ameliorate local and regional temperatures, such open spaces with pervious surfaces (Stone, 2009). For this study, urban form is considered as a variable effecting the distribution of open space and thus will be included and controlled for through the selection process of the case studies.

Changing land-use and cover types aims to abate local and regional urban climate.

This is achieved by advocating planning and design that increases the quantity and quality of vegetated surfaces6 (Gurran and Hamin, 2008; Gill et al., 2007; Gober et al.,

6Vegetal surfaces (green infrastructure) are both a mitigation and adaptation measure, therefore a multi-functional and multi scalar system.

37 2010). The approach protects pockets of urban ecosystems and aims to promote vegetated

surfaces in creative ways within existing urban residential and commercial parcels,

infrastructural residual spaces, and tree canopy within transportation corridors. Vegetated

surfaces provide the potential to regain some of the lost predevelopment land cover to

mitigate and adapt to local climate changes (Stone, 2009). McPherson et al. (2008) show that local climates maybe ameliorated by increasing tree canopy cover. In their One

Million Trees Assessment Study of Los Angeles, they have identified and quantified storm water runoff reduction, energy-use reduction, atmospheric carbon dioxide reduction, and air quality improvement as direct benefits of urban tree cover. Therefore, by promoting land use and cover changes at multiple scales, GI may be identified as a land-use measure that primarily addresses adaptation and secondarily, mitigation.

To reduce GHG emissions, denser and more compact development may potentially conserve energy use and reduce miles travelled (Ewing et al., 2008). Alternative public commuter services and a strategic mix of uses may be needed to change people’s preferences to drive less and use public transport (Lerch, 2007; Wheeler, 2009).

Therefore, the literature is advocating the use of denser and more compact development.

Compact development may also reduce energy consumption generated by buildings and structures (Lerch, 2007). Multi-storey buildings and buildings that share walls use less energy for heating and cooling, thus producing less emission and energy waste (Lewis, 2006; Wilby, 2007). Compact and denser development may provide better mitigation results and coincides with Smart Growth and Sustainable urban planning.

When considering adaptation (especially for flooding and temperature reduction), the compact form is not an evident typology useful for adpatation.

38 The current discussion on the suitability of compact and dense urban form comes

from the literature on sustainable urban planning (Berke, 2007; Wheeler, 2008). Jabreen

(2006) suggests that nontraditional development, compact city, urban containment, and

eco-city are urban forms that contribute to sustainability. He finds that compact city ranks

highest when compared to the other forms based on density, diversity, mixed land use,

compactness, sustainable transportation, solar design, and ecological design. While

Jabreen (2008) focuses on the built form itself, Berke (2007) surveys historical green

planning urban forms at the regional scale. He identifies centrist, poly-centrist, and de- centrist models and compares them against five dimensions of green communities: human health, natural systems, spiritual renewal, livability, and fair share. He finds that the historical poly-centrist vision holds traction in contemporary planning thought. While

these two examples show that compact city form, at the urban or regional scale, maybe

the most suitable model for sustainable urban planning, they also suggest a conundrum of

space and density (Hamin and Gurran, 2008), when considered in conjunction with green

infrastructure.

Neuman (2005) refutes the argument of compact form based on evidence from the

literature. While he does not deny the benefits of compact and dense urban forms, and

rather than limiting the discussion on form, he argues that the process of living is more

indicative of the state of sustainability of urban areas than their form. He identifies five

intellectual traditions from which sustainability stems: capacity, fitness, resilience,

diversity, and balance. He offers four common points: health, contextually, relationships,

and systems thinking. What Neuman (2005) and others (e.g Forman, 2008) suggest are salient characteristics that would identify a certain urban form as sustainable or not.

39 Similarly, Alberti (1999) identifies centralization, density, grain, and connectivity as

spatial parameters of urban regions. The author assesses these urban patterns against

environmental performance such as sources, sinks, support systems, and human well-

being.

Within the context of the discussion of suitable urban form for climate change

planning, Hamin and Gurran (2008) offer an alternative urban form based on potential

performance. While the performance has not been empirically proven, they suggest that

“[T]the likely best urban form … must bring green space within settlements focused

along green transportation routes and flood plains (ribbons and corridors) rather than

large expanses (Hamin & Gurran 2008, p.242)”. This urban form may offer a possible

spatial response to climate change adaptation and mitigation.

2.6 Green Infrastructure, Space and Adaptation

The discussion on different urban forms and their suitability for climate change adaptation may be summarized as a problem of space. For green infrastructure to be effective as an adaptation measure, surface land area and/or surface built areas (engineered green infrastructure) should be allocated. This is not readily available in urban contexts specially when land value is high and competition for real estate is continuously on the rise.

Different urban forms provide varying possibilities to increase and expand green infrastructure surfaces. Brabec and Lewis (2002), identify four urban regional forms that provide an example for this discussion. They base their discussion on previous literature

(Lynch, 1954; Bacon, 1976) that resonate with more current research on the environmental impact of different urban regional forms (Berke, 2008; Forman, 2008;

40 Jabreen, 2008; Wheeler, 2009). The nuclear form, such as Calgary in Canada (Figure 2.2, bottom right) is characterized with a dense built center and a distinct definition with the surrounding land, be it agriculture or suburbs. Washington DC (Figure 2.2, bottom left) provides a clear example of a green city where pockets of densely built areas surround a larger nucleus connected by transportation and communication corridors with dispersed land in between. Clear definitions between built and un-built remain clear. A sprawled city such as Chicago (Figure 2.2, top left) extends over land in a homogeneous manner with patches of built and un-built areas somewhat equally distributed across the landscape. The stellar city as in Buenos Aires (Figure 2.2, top right) extends along transportation corridors with pockets of un-built areas in between.

When considering green infrastructure for adaptation, the primary premise is to maximize the protection and expansion of existing and potential green infrastructure spaces. The four urban types provide a glimpse of the varying possibilities for green infrastructure policies across urban regions. For example, the focus within a nucleated urban form may be on green roofs and street trees and small scale best management practices (in addition to existing parks and open space) that do not require extensive space because it is simply not available. On the other hand, a sprawled city may provide a larger array of green infrastructure policy possibilities (withstanding the demonstrated impacts of sprawl). Policies may vary from privately owned residential parcels, similar to the BOF to increasing patches of urban forest, protection of river actuaries and transformation utility lines. Tis demonstrates that re-equipping existing cities with green infrastructure measures for adaptation will vary across urban regions and very much dependent on the availability of space in the form of land or built surfaces.

Figure 2.2: Four types of regional urban form: Chicago, sprawl/constellation city (top, left); Buenos Aires, Stellar/Linear (top, right); Washington DC, Green City (bottom, left); and Calgary, Nucleus (bottom, right). (Urban remote Sensing: http://sedac.ciesin.org/urban_rs/ )

42 CHAPTER 3

WINDOWS OF OPPORTUNITY: ADDRESSING CLIMATE UNCERTAINTY THROUGH ADAPTATION PLAN IMPLEMENTATION

3.1 Introduction

Adapting cities to climate change is a pressing issue. However, creating a

feasible adaptation planning process is difficult given the uncertainties inherent in the

physical manifestations of climate change, as well as the modeling uncertainty in the timing and magnitude of the change. The result is that it is easier for policymakers to ignore climate change in their policymaking than risk being wrong, creating a significant

barrier to the implementation of climate adaptive actions (Dessai and Hulme 2007; Carter

2011). While reducing the underlying uncertainty will only occur through improvements

to climate science and modeling, reducing the impact of uncertainty can occur through improved policy and planning processes. While significant research attention has been paid to using scenario planning and vulnerability assessments to improve policy and reduce uncertainty, the implementation stage of adaptation planning also provides opportunities to reduce the impact of uncertainty. For planning purposes, what matters is the pace of change over the planning period, rather than some eventual end-point. But

the pace of local experience of climate change is difficult to project. As a result, municipal planning requires a highly flexible process that is designed to build

incrementally towards transformational policies, if conditions prove necessary.

In this paper we synthesize existing research and emerging practice to

conceptualize the ‘windows of opportunity’ planning model, bringing together adaptation

tipping points, incremental-to-transformational change, indicators, and phasing, focusing

on how to phase in the implementation of adaptation over time to allow flexible

43 responses to the pace of climate change and the effectiveness of interventions already undertaken, in a way that engages both scientific and local knowledge. We use case studies of two cities that are already approaching adaptation planning in this way to explore implementation of the model.

Overall this approach has several advantages. Perhaps most important, planned phasing matches investment to the stage of climate change that the community is experiencing. At the same time, it allows for incremental actions to build toward transformative change, while still benefiting from the advantages of a comprehensive process. This adaptive approach allows for testing of the efficacy of adaptive responses already undertaken. In-situ monitoring of early and no-regrets policies will help determine how effective they have been in reducing locally-experienced impacts of climate change, and thus inform what more needs to be done. Challenges remain—large infrastructural investments or abrupt climate changes may require large, one-time responses. But for the more gradual impacts of climate change, such as increased excessive heat days, more erratic and larger storm events, and extended drought and desertification, the model may be very helpful in providing a road map to move forward with adaptation regardless of uncertain conditions.

3.2 Climate Uncertainty and Urban Planning Policy

The level of climate change that is already underway is startling (Kintisch 2009;

Rahmstorf et al. 2007), and fairly consistently ‘ahead of schedule’ (McKibben 2008). In terms of mitigation, the hard fact is that global greenhouse emissions, far from declining, are still increasing. In 2010, the annual rate of emissions growth was 2.35%, higher than any of the previous five years (NOAA Earth System Research Laboratory 2011).

44 Emission levels are generally following the ‘high emissions scenario’ projected as a worst case by the IPCC in 2007 (European Environment Agency 2011), and there is no apparent movement towards global governance systems that would lead to significant reductions (Stafford Smith et al. 2011). Recent findings that incorporate the growth of emissions since 2007 suggest that the globe is currently headed toward 4°C (7.2 °F) of global average climate change, even if emissions reductions begin soon, and impacts will be worse in northern regions (Joshi et al. 2011).

3.2.1 Climate Change Timing Uncertainty

Research suggests that among the range of barriers to implementing adaptation policy, uncertainty over the level of change is a key reason for difficulties in getting policymakers to take action on climate change (Bedsworth and Hanak 2010; Moser and

Ekstrom 2010). From a policy perspective, climate uncertainty can be characterized as a function of magnitude, direction, and timing of change (Joshi et al. 2011). But given an average 20-year urban planning horizon, the pace and timing of change may be an even larger issue than where global climate eventually lands (Figure 3.1). As Figure

3.1demonstrates, a thirty-year plan (which is, admittedly, longer than most current policy horizons) could use a 1.2°, a 2.2°, or a 4.0° projection depending on the choice of global or regional forecast, and the longer the time horizon, the greater the uncertainty. And these are not worst-case projections--none assume any ‘abrupt’ climate shifts (Alley et al.

2003).

Figure 3.1: Timing of regional and global climate change impacts under high-emissions scenarios. The plots show smoothed and simplified approximations of the projections in three reports. The IPCC line is for A1FI scenario, which provides a best estimate of 4°C temperature change at 2090-2099 relative to 1980-1999 (IPCC 2007, p. 45). Betts, Collins et al. (Betts et al. 2011) use the IPCC A1FI scenario to estimate that a 4°C change by 2060-2070 is within the IPCC’s ‘likely’ range. In contrast, at the regional level for the Boston area, Frumhoff et al. (Frumhoff et al. 2007; Frumhoff et al. 2008) suggest that a 4°C change could be quite imminent.

3.2.2 Climate Variability and Climate Change

Within the one to two decade time-frame that most plans work, the locally- experienced impacts of climate change are likely to be relatively small compared to the impacts of natural climate variability (IPCC 2012). Particularly at the local level it may be easier for communities to unite in addressing existing climate variability than in addressing a threat such as climate change that is less directly experienced, and more politically charged. As a result, simultaneously addressing current climate variability and climate change may in many cases be a more policy-beneficial approach than focusing on

46 one or the other. In view of this, we sought to identify an approach that did not explicitly

require the separation of natural and anthropogenic-caused climate problems, and instead

focused on identifying a pathway that could assist communities in overcoming planning

barriers while still allowing for short and long term climate-change informed planning.

3.2.3 Uncertainty in Impacts at the Urban Micro-Climate Scale

At the urban level, the impact of climate events can be magnified (or reduced) by

the form and/or design of on-going urbanization processes (Hardoy and Pandiella 2009;

Schipper and Burton 2009; IPCC 2012), which create micro-climates that influence human climate-experience and ecological functions. One key variable is the amount of impervious surface. Higher imperviousness tends to lead to more flooding, more intense urban heat island effects, and increased desertification (Arnold and Gibbons 1996;

Brabec 2009; Rosenberg et al. 2010; Stedinger and Griffis 2011). These affect an environmental cycle that results in higher levels of particulates in the air; increased levels of pollutants, particularly ozone; decreases in floral and faunal diversity and numbers; and increasing destabilization of soils and floodplain systems (Stone, 2012). These in turn result in a higher incidence of human health problems (Few 2007; Shea et al. 2008), property damage and loss, and ecological degradation and species extinction (Nitschke and Innes 2008). The poor tend to be disproportionately effected by these changes as economic forces push them into areas that are highly impervious and flood prone with high heat indexes and unstable soils (United Nations Human Settlements Programme

(UN-Habitat) 2011).

Thus, if cities are built without attention to the climate impacts of development and the poor continue to be pushed into high risk areas, vulnerability to climate

47 variability increases regardless of climate change, and is magnified with it (United

Nations Human Settlements Programme (UN-Habitat) 2011). A city designed with

adequate green infrastructure to reduce urban heat island effects, with on-site stormwater

management accompanied by effective watershed management systems, and with climate-adapted buildings built on stable soils, is better positioned to manage climate

variability. These types of policies, which provide sustainable environmental and social

benefits, are widely held to be the place to start for reasons both obvious and subtle. The

obvious benefit is that they create better places to live without even having to argue the

climate question (Heltberg, Siegel, and Jorgensen 2009). The less obvious reason is that their micro-climate impacts may slow the need to undertake expensive larger-scale interventions (Stone 2012). Conceptually, good design can mean that a global climate temperature increase of 2°C may be locally experienced as the equivalent of 1°C, for example.

3.2.4 The Challenge of Uncertainty to Municipal Planning

Taken together, these factors make clear that uncertainty in the timing and local impacts of climate change are a real and significant challenge to the municipal climate adaptation process. The most common municipal and research response is to focus on climate scenarios, often using two – a high and low change – assuming a relatively straight line pace of change toward the projected degree (e.g., 2°C or 4°C) at the end of the planning window. This is clearly better than assuming climate stability, but does not fundamentally address the problem of timing, and the implications above, that climate change seems to be consistently ahead of schedule.

48 3.3 Municipal Adaptation Planning Processes

Cities across the globe are addressing these issues thrrough preparation of climate adaptation plans, as well as through ‘mainstreaming’ of policy wherein climate projections influence infrastructure calculations and the like directly, without any specific plan in place. Underlying the various approaches cities are taking are certain theoretical models for how planning should be done. The current best-practices approach to adaptation planning follows closely from traditionnal comprehensive plannning, but adds more focus on risk assessment, as shown in Figure 3.2 below.

Figure 3.2: The Adaptation Planning process as recommended by the U.S. National Research Council. Note the iterative process steps (light grey arrows) and focus on co-benefits in step 4. Adapted and redrawn from the US National Research Council, (2010).

To appropriately phase adaptation policies in ways that address current climate variability as well as on-going climate change requires impleementation processes that brring together the advantages of these three planning approaches: the big-picture view of traditional comprehensive planning, the specific goals and policy steps of incrementalism, and the continual testing and utility of adaptive management. One attempt to resolve these conflicts and achieve thee best of each is seen in the recent

49 guidance from the State of California to its cities and towns on adaptation, as shown in

Figure 3.3 below.

Figure 3.3: Local adaptation planning process advised in the State of California Adaptation Planning Guide (California Emergency Management Agency (Cal EMA) and California Natural Resources Agency 2012, p. 24). The focus in our research is particularly on Stage 9 (dark orange color) and the implementation of the iterative loop, but with implications to Stages 6 – 8 (light orange). The steps shaded in grey are part of the local vulnerability assessment (grey color) (Redrawn and adapted by authors. Original was adapted from Boswell et al, (2012).

The process of planning, even with the uncertainty of climate change, is fairly well understood because these models follow closely from the highly-developed practices of comprehensive planning. But in a typical stabble climate, thhe monitoring and phasing is not essential as in a climate-variable environment. As a resullt, our model particularly addresses this post-plan, implementation phase, where there is significant opportunity to improve practices.

50 3.3.1 Maladaptation in the Planning Process

Along with more opportunity for public engagement, one of the advantages of a

comprehensive process is that it allows for testing for maladaptation – defined by Barnett

and O’Neill (2010, p. 211) as “action taken ostensibly to avoid or reduce vulnerability to climate change that impacts adversely on, or increases the vulnerability of other systems, sectors or social groups.” They identify five distinct types of maladaptation: actions that

“increase emissions of greenhouse gases, disproportionately burden the most vulnerable, have high opportunity costs, reduce incentives to adapt, and set paths that limit the choices available to future generations.”

The first type of maladaptation listed above, where adaptive actions actually increase the use of fossil fuels, is a particular issue in developed countries and needs to be carefully managed (Howard 2009). A common example is residential air conditioning – it reduces the individual health impacts of heat waves, but at a long-term and shared cost of higher emissions. A particular challenge in phasing adaptation is the last issue in this set: remedies appropriate to 2°C degrees may interfere with 4°C degree adaptations—for example, investing in sea walls to address sea level rise appropriate to a 2°C global temperature rise may make it more difficult to persuade homeowners of the need for planned retreat of their built structures on the lot or indeed off the coast altogether as change moves toward the sea level rise associated with a 4°C rise in temperature. In all of these cases, a comprehensive, thoughtful approach will assist in avoiding maladaptation, but the time and resources necessary to create a plan mean that it will likely be rarely updated. This is a particular problem for climate change and its uncertainties.

51 3.3.2 Mainstreaming and Incremental Change

Mainstreaming, which tends to focus on incremental change, proposes that small steps be taken toward very specific policy goals, with little effort toward a fully

comprehensive approach (Lindblom 1959). Adaptive management builds on

incrementalism by focusing on the need for consistent testing, monitoring, and revision of policy as new information becomes available (Jacobson et al. 2009; Nelson, Adger, and

Brown 2007).

While perhaps more discussed than actually practiced, resilience theory demonstrates that rather than the unexpected, change is to be anticipated and tends to occur when thresholds are passed. As a result, planning needs to prepare social and ecological systems so that when stresses occur, systems can reorganize in a beneficial way to achieve a new and desirable system state (Folke 2006; Walker and Salt 2006).

This focus on small steps and the underlying processes brings significant advantages to the planning process, but may come at the cost of long-term vision and policy coherence.

3.4 Windows of Opportunity Model

The model (we propose is presented and its elements are discussed in the sections below (Figure 3.4).

Figure 3.4: Windows of Opportunity Climate Addaptation Model is based on phasing of policy to match new conditions. A critical path is determined by the community (black dotted line) where ATPs are identified along this path. The ATPs trigger moving from one phase of policies to the other, from no-regrets to transformational policies. Windowss of opportunity for planning at the initial stage and prior to phase transitions allows for plan revisions based on feedback loops where monitoring is carried out within short periods of time set by the community. The model application results in anticipating incremental additions to policy allowing matching of interventions to experienceed (or anticipated) climate conditions.

Commonly, comprehensive plans are deeply tied to time, with specific roll-out dates for actions. In a linear, predictable process this makes perfect sense. Under climate change’s conditions of timing uncertainty as described above, it is much less effective.

Because of this, a revised planning approach will focus on testing the actual environment and matching that to investments, rather than planning roll-out of policy based on specific years into the future. To achieve this, plans should identify incremental policies that can roll-out as needed, and that may cumulatively create a transformed state (Pittock, Jones, and Mitchell 2001; Wilson 2009). Following Park et al. (2012), transformation is defined

53 as policies “that fundamentally (but not necessarily irreversibly) results in change in the

biophysical, social, or economic components of a system from one form, function or location (state) to another” while incremental actions seek to maintain the essence and

integrity of an incumbent system (Park et al. 2012, , p. 119) (O’Brien 2012). In seeking to appropriately phase adaptation policies in ways that address current climate variability as well as on-going climate change, our model proposes a process that brings together the

advantages of these three planning approaches: the big-picture view of traditional

comprehensive planning, the specific goals and policy steps of incrementalism, and the

continual testing and utility of adaptive management.

3.4.1 Phasing Policies

The initial steps the community is likely to take are the no-regrets policies that

many authorities have identified as the appropriate place to start. The IPCC defines these

as “policies that would generate net social and/or economic benefits irrespective of

whether or not anthropogenic climate change occurs” (Glossary E-O 2007; see also

Callaway and Hellmuth 2007; Heltberg et al., 2009; Lempert and Collins 2007). One of

the key benefits to no-regrets policies in urban environments is that they may, as noted

above, reduce or delay the need to move to the next phase of the plan, with more

intensive response. At some point, however, movement to a more intense policy phase is

likely to be needed. In our model and increasingly in other studies, phasing is based on

thresholds (Walker and Salt 2006), or what (Kwadijk et al. 2010) call Adaptation Tipping

Points (ATPs) -- moments in time where the magnitude of climate change is such that the

current management strategies are no longer meeting local objectives, and new strategies

need to be put into place. ATPs are recognized through the use of indicators, defined as

54 statistical evaluative rubrics that reflect the status of a more complex system (Kates,

Travis, and Wilbanks 2012), or at the local level as proxies for the status of the system.

3.4.2 ATPs and Types of Indicators

The ATPs are representations of local conditions and values and climate as experienced through the lens of local human-ecological systems interactions. As a result, determination of what will constitute an ATP needs to engage a participatory, bottom-up perspective as well as utilizing existing sets of expert-derived indicators, and include natural and social/experiential indicators. Because of the difficulty of separating climate

‘noise’ from ‘signal’, scalar considerations (local versus regional), and the complex politics of decision making, using a suite of indicators in concert with local values will be more effective than any attempt to identify one threshold measure that indicates the need to move to the next phase. The indicators may be categorized into three types: climate related, social, and local urban environmental indicators. A portfolio of indicators can include those that are scientifically robust, and those that are more locally meaningful even if less scientifically robust (Boulanger 2008; Feiden and Hamin 2011).

Officially approved national/supra-national level climate-related indicators are beginning to be easily available. In the U.S., for example, NOAA has developed Global Climate

Change Indicators7 while the U.S. Global Change Research Program prepares a national

climate assessment and is developing a rigorous set of indicators including societal data.8

The Annual Greenhouse Gas Index prepared by NOAA provides a simple number of

cumulative global emissions (US EPA 2010). Europe has developed indicators for

7 See http://www.ncdc.noaa.gov/indicators/

8 See http://globalchange.gov/what-we-do/assessment

55 widespread use, especially the 2008 indicators report for climate impacts (European

Environment Agency, World Health Organization, and JRC European Commission

2008), the 2012 environmental indicators report for greenhouse emissions and

environmental conditions (European Environment Agency 2012), and the on-going data sets available on the EEA website. Data for developing countries is available through the

World Bank Climate Change Knowledge Portal,9 although with less detail than the US or

European initiatives.

Global level indicators, however, cannot take account of the local effects of

micro-climate, the positive outcomes of policy already implemented, or the changes in

context that occur outside the plan, such as new up-shore developments. As a result,

locally derived and relevant indicators are an important part of the process. Local

indicators provide the opportunity to engage community members, perhaps the climate

planning steering committee or other local board, as well as local staff in both defining

what is meaningful in the particular context of that plan, and in regularly measuring and

reviewing outcomes. While the more scientific indicators provide validity, locally

meaningful indicators assist in developing community support for and understanding of

the need for the policy change (Gasteyer and Flora 2000), and respond to the IPCC’s

2012 call for more direct inclusion of local knowledge in planning. Examples of local

environmental indicators might include five-year moving averages of the number of

extreme heat days in the region; number of bank-full and/or flood stage days; the increase

in level of mean high tide; levels of base flow in area rivers as an indicator of drought;

miles of beach impacted by storms; the number of individuals hospitalized for heat stress

9 For Brazil, for instance, see http://data.worldbank.org/country/brazil#cp_cc

56 or asthma; or similar indicators. The number of times a sea wall is overtopped per year, for instance, may be a locally-meaningful indicator that encourages action—but for that to happen, the record of occurrences must be made and annually reviewed. By focusing on local impacts (e.g. heat waves experienced) rather than only the causes (globally higher GHG) these kind of indicators help to overcome the uncertainty of microclimate effects, timing, and unanticipated conditions.

A key role of these indicators is to create prior agreement, or at least the opportunity for regular discussion, on what would constitute an ATP for the local community. Local values will determine whether they can tolerate three days when roads are flooded per year? Five? One event per year with more than 10 residents hospitalized due to heat stroke? These are not likely to be easy decisions, and require the community to engage very directly with deciding on acceptable levels of loss and risk management.

This also allows better integration of local knowledge systems into formal institutions.

3.4.3 Example: London and the Thames Barrier

London’s plan for the Thames barrier provides a recent example of combining phasing, indicators and ATPs. By maintaining a 1000-year flood standard, using the dual indicators of freeboard and storm surge, and incorporating various levels of projected sea level rise, the city of London identified the tipping points for increasing the height of the

Thames Barrier (see Figure 3.5). While this is an analogous application of the theoretical basis of our model, the model goes beyond this application to include multiple adaptation strategies, and includes value-based as well as quantitative measures.

Figure 3.5: An example of phased engineering adaptation for the Thames Barrier, responding to potential climate change trigger points (Reeder et al. 2009, 62)

3.5 Application in Case Study Communities

The application of the proposed model is illustrated annd grounded through two case study examples. The examples were selected from a larger pool of climate adaptation plans listed in available databases.10 Comprehensive adaptation plans were selected that explicitly recognized the uncertainty of magnitude and timing of climate change impacts and went beyond assessment to implementation, regardless of geographic or scalar considerations. We then chose case studies that incluuded as many components of

10 Georgetown Climate Center clearing house (http://www.georgetownclimate.org/adaptation/clearinghouse); ICLLEI adaptation resources(http://www.icleiusa.org/climate_and_energy/Climate_Adaptation_Guidance/free- climate-adaptation-resources?searchterm=climate+adaptation+plans); NOAA adaptation and action plans data base (http://collaborate.csc.noaa.gov/climateadaptation/Lists/Resources/AdaptationAction%20Plans.as px); and the United Nations Environment Programme(UNEP), climate change adaptation resources section (http://unfccc.int/adaptation/workstreams/nairobi_worrk_programme/items/6547.php)

58 the implementation model as possible including a form of phasing or gradual

implementation of adaptation policies and measures; triggering conditions or threshold

indicators; monitoring periods; and the provision of time periods and transitions for plan

update based on hard evidence that we term ‘windows of opportunity’ for planning

implementation. In both case studies, analysis was carried out to extract the components

based on the criteria set forth in Table 3.1.

Components Component Details Gradient of policy Adaptation measures included in plans are assessed and measures organized into a continuum of policies, from “no-regrets’ to (phasing) transformational. Windows of Initial period or period between phases allocated for extensive opportunity and revisions and update of implementation based on the shorter transition periods periods of monitoring feedback loops. Actions to be conducted by the planning team are identified and included. Triggering Are derived from the critical community paths and are threshold conditions indicators of impacts specific to the plan context. These triggers along with cost-benefit analysis instigate the move from one phase of policies to the other. Monitoring Time allocated to monitor implementation measures, monitor feedback loops triggering conditions, and incorporate technological advances and review of or parts of the plans. Monitoring periods are indicated in plans and accordingly included in the model application. Risk assessment: While not included in the model application diagrams, cost- Cost Benefit benefit analysis is discussed in each case to show the primary analysis focus of the magnitude of risk as well as an indicator (with triggering conditions) to move through policy phasing. Table 3.1: Components of the implementation model used to assess case study examples. Component details indicate the analysis conducted by the authors to identify and make explicit these components.

Two approaches were used to define the model components. A descriptive approach was utilized when the details existed in the plan allowing minimal processing of

information. A prescriptive approach was used when details were lacking and required additional processing and analysis. In each case study relevant components were

59 identified, simplified and then mapped against the community critical path. Adaptation measures were reorganized to fit the continuum of measures with specific attention given to ensure that later measures were built onto previous ones and supported future conditions, while maintaining the objectives of the plan. Where measures were missing, the authors propose complimentary measures (indicated accordingly).

The selected plans are unusual and particularly useful when presented according to the implementation model, because they specifically focus on implementation of adaptation by considering triggers for movement to the next stage of their adaptation plan. Thus they provide examples of policy phasing that may reduce (while not necessarily eliminating) the extent of uncertainty in the decision making process.

3.5.1 Managing Rising Sea Level Impacts: The City of Clarence, Tasmania, Australia

The Clarence City adaptation plan for coastal areas (Clarence City Council 2009) provides a road map for adaptation for risk prone areas within the city boundaries. The plan develops solutions that support the continued use of coastal areas while recognizing the need for long term protection, accommodation and retreat as sea levels rise (SLR).

“While use may be practical and desirable for many years, there will come a trigger when a response will be required to manage increasing risk (Clarence City Council 2009, iv)”.

In Clarence’s case, the triggering event, or ATP in our terminology, is specific levels of locally-experienced sea level rise. The planning is based on a community accepted worst-case scenario (critical path) of future conditions, but emphasizes “encouraging performance based responses that maintain acceptable levels of risk (Clarence City

Council 2009, v)”. The plan demonstrates that managing risk today through adaptation

60 measures can reduce impacts from sea level rise “from a factor of 10 up to a factor of

100, and economic costs of adaptation would be minor comppared to the damage avoided

(Clarence City Council 2009, iii).” Community participation and ratification is significant in the decision making process in spite of the challenges this poses to decision makers.

Interactive strategies that are participatory and combine bottoom-up and top-down approaches enhance a “community’s ability to cope and maxiimize community support for policy measures, especially in the case of drastic measures (Clarence City Council

2009, v).” The combination of these principles render the adaptation plan flexible and open for improvement and revisions as the effectiiveness of interventions becomes clear and impacts of climate change increase.

Figure 3.6: Phasing Model for Lauderdale Area, City of Clareence, Australia. The plan details include several components of the Windowss of Opportunity model. For this article, we focus on one high risk zone and apply the moddel (Figure 3.6)— Lauderdale and Roche Beach--out of the eighteen identified risk zones. Lauderdale is a 4,300m long low-lying sandy isthmus with dunes and housing developmeent. The plan identifies three major hazards for this area: storm surge and erosion, inundation, and rising water tables leading to failing septic tanks. Complicating the situation is current

61 coastal beach movement processes (not related to climate change), 11 which will occur at a faster pace under a changing climate condition.

The plan uses six primary and thirteen secondary variables to identify the extent

of exposure resulting in a risk priority ranking: areas currently at risk (next 25 years),

areas at medium risk (25-75 years), and areas with longer term risk (beyond 75 years).

Sea level rise (SLR)12 scenarios are developed for the whole city based on IPCC (2007,

17) emission projections with mid and high values for three milestone years: present

(zero SLR), 2050 (mid at 0.2m and high at 0.3m SLR), and 2100 (mid at 0.5m and 0.9m

SLR). Triggering conditions or ATPs are estimated for each zone based on the indicators of the current 100 year Average Return Interval (100yr ARI) for erosion/recession of the coastline, wave run-up, and inundation. ATPs will be triggered when the 100 yr ARI

would likely lead to significant damage to property or where more extreme events would

make emergency responses difficult. Identifying when an ATP is likely to occur or is

reached is dependent on continuous monitoring of selected indicators, climate science

developments, advancements of technology and community perception of risk. The

critical path and ATPs were deemed realistic and probable by experts working on the

plan and ratified by the community.

The first window of opportunity is used to refine the plan by conducting detailed

studies of all risk zones, secure immediate and long term funding, undertake cost-benefit

assessments of measures, and update climate information and indicators. Phases after

11 The plan identified any combination of the following coastal processes not related to climate change: adjust to past sea level rise (post ice age) or recent sea level rise, long shore drift, storm cut and rebuild, beach rotation, and changes in sea grass colonies that may trap or release sand ( Clarence city Council 2009, 52).

12 SLR based on Australian Height Datum (AHD) in Tasmania is based on mean sea level for 1972 at the tide gauges at Hobart and Burnie which was assigned the value of zero on the AHD.

62 implementation are the next windows of opportunity, in which the outcomes of policy

interventions on microclimates can be evaluated and the plan can be revised. Future plan revisions depend on ongoing monitoring of the selected indicators, implemented measures, lessons learned and evaluation of the ATPs. A five year period is suggested in the plan as a review period or the monitoring feedback loop in the model. The plan

explicitly recognizes the need for evidence based monitoring by observation and ground

measurement to understand the actual path of the indicators in relation to a changing

climate. The provision of initial, transition, and monitoring time periods allow for the

continuous updating of adaptation measures to respond to actual changes.

The adaptation measures address protection of dwellings and infrastructure, accommodate the changing coastline, and ultimately recommends retreat if ATPs indicate the necessity. The full set of measures included in the plan is clear enough to be reorganized into the continuum of measures in our model with no additional processing

required. These are included in the model application (Figure 4), providing a continuous

approach where implemented measures support future policies. An example for existing

properties protection is increasing the height and vegetation of existing sand dunes as no-

regrets measures that could transition into the construction of a sea wall, phased in based on ATPs. At present, using the indicator of wave run-up, the current level for a 100yrARI is 2.8m and dune average height (where present) is 3.5m. Minor sand nourishment to fill gaps and vegetation for stabilization will provide immediate protection. When the ATP for wave run-up of between 2.8m to 3.2m is reached, additional height will be added

(approximately 1-2m) to ensure protection. When the ATP of 3.7m is reached, topping existing dune height, then additional height of another one to two meters may be

63 required. If monitoring of the indicator shows that the future trajectory seems to be

exceeding set thresholds, then an ATP has been reached, and transitioning to sea walls is

triggered.

The plan provides a cost metric that is also used as an ATP, based on the cost of

adaptation measures per protected dwelling. As long as the cost of measure per dwelling

remains lower than the value of the property, the next phase of adaptation is deemed

feasible. While the plan explicitly states this principle and there is no provision of an

average property value to assess and include in the model diagram. The following

example may demonstrate the utility of the cost metric. The cost of sand nourishment at

present conditions for a 100yrARI is $136,000 per property for nineteen protected

properties. With an SLR of 0.3 and protecting 108 properties with sand nourishment, the

cost is $71,000 per property; for a 0.9 SLR and 195 protected properties, the cost

increases to $119,000 per property. When compared to the cost of a sea wall for the same number of protected dwellings, at present SLR 100year ARI, the cost per dwelling is

$974,000 and for an SLR of 0.9 (worst case scenario) is $174,000.

This cost metric is also used to reduce the extent of exposure to risk. Prior to the

25 year cut-off period, it is assumed that owners have located within these this risk area prone without knowledge of the associated future risks. After 25 years, and with awareness and communication campaigns set by the plan, residents will be assumed to have made a conscious choice to locate in a risk-prone area. At that time more costs will be allocated to individual properties. This should reduce the number of structures exposure to risk and thus future public costs for adaptation.

64 Using the proposed model, we organize the gradient of responses identified for

Lauderdale along no-regrets to transformative measure gradient in Figure 4. The no- regrets responses are critical for implementation as soon as possible, and as the plan notes will provide protection and maintain the coastline as an amenity. As monitoring of the indicators demonstrates changes in conditions, more intense, transformational measures such as sea walls and planned retreat kick in, assuring adequate responses. Mapping out the likely policies and the conditions indicating their need means that maladaptation is less likely, and costs can be better managed over time.

3.5.2 Managing flood impacts from Extreme Precipitation Events: The City of Copenhagen, Denmark

The Copenhagen Adaptation Plan is a state-of-the art document, developed in

2011 to address an array of climate change impacts with a particular focus on extreme precipitation events and rising sea level. Such extreme conditions are already occurring in

Copenhagen. The cloudburst event of July 2011 poured down 150mm of rain within two hours, a city record since measurements began in the mid-1800s. The result was an estimated insurance damage of 650-700 million Euros ($US 820-880 million) (EEA

2012). This focusing event helped the City of Copenhagen to expedite research and development of the comprehensive adaptation plan in a way that allows gradual and flexible adaptation over time (City of Copenhagen 2011).

The plan presents many exemplary practices, but for our purposes its main interest is its principle that adaptation should be flexible and staged. The plan is developed for incremental implementation with continuous monitoring and updating to include advancements in climate science, scenario projection methodologies, and

65 climatically responsive planning. The prioritization, implementation and extent of

effectiveness of adaptation measures are categorized based on three levels of intensity

(Table 3. 0.2): level 1, to reduce the likelihood of occurrence of an extreme event; level 2, to reduce the scale of impact; and level 3, to reduce the extent of vulnerability. The choice of the appropriate level is based on the feasibility of implementation within a specific zone of Copenhagen. For example, if the reduction of likelihood of an event

(level 1) is not feasible within a zone, then reducing the scale of impact (level 2) to manage damage is prioritized. If that is not deemed feasible, then reduction of vulnerability (level 3) becomes the dominant action. In addition to the intensity of measures, the choice of action will also depend on the geographic scale where the action is being implemented. Table 1 shows the relationship of the three levels of measures and the five geographic scales of planning relevant to Copenhagen: the region, the municipality, the district, the street, and the building. This approach ensures coordination and integration across planning scales and measure intensities, thereby better avoiding maladaptation and unnecessary investments.

66 Level1: Level 2: Level 3: Reduce Likelihood Reduce Scale Reduce Vulnerability  Delay quantities of rain in  Emergency preparedness  Protection of vulnerable catchment: Establishment of and infrastructure protection infrastructure Metro, S- detention basins within  Protection of vulnerable trains, tunnels, cultural catchment areas. infrastructure Metro, S- assets.  Pumping excess run-off to trains, tunnels, cultural sea assets.  Establish warning system for high waters Region Region  Disconnection of stormwater  Disconnection of stormwater  Planning**** using SUDS* using SUDS  Emergency Preparedness  Increased sewer capacity:  Planning**** New dimensional design  Warning Systems based on future new capacities.  Pumping of excess run-off to sea  Establishment of local dikes  Raise building elevations Municipality Municipality  Establishment of dikes  Decoupling of rainwater  Moving of vulnerable  Decoupling of rainwater using SUDS. functions and installations using SUDS  Emergency Management: Moving of vulnerable  “Plan B”** sandbags etc. functions to safe places  Establishment of dikes  Raised building elevation/threshold*** District  Local management of storm  Control of stormwater  Moving vulnerable water: Plan B solutions, runoff: disconnection of functions and installations separation of stormwater stormwater using SUDS. to safe places: moving from sewer Preparedness: sandbags etc., electrical cabinets for light  Raised building raised building regulation, pumping elevation/threshold elevation/threshold, sand stations etc. from low-lying bags points Street  Disconnection of stormwater  Backwater valves, sealed  Move vulnerable functions from sewer basements, preparedness, away from basement level  Backwater valve sandbags etc. (service rooms, electrical  raised building panels etc.) elevation/threshold Building Table 3. 0.2: Example of integrating adaptation measures across geographic scales and reduction of risk for flooding from extreme rain events and rising sea level (Adapted from City of Copenhagen 2011, 27 and 35). *SUDS: Sustainable Urban Drainage Systems. “SUDS consist of a number of different elements, all of which serve the purpose of managing stormwater locally. These may be elements that delay/store the water, that treat the water either before discharge to bodies of surface water or percolation of the stormwater. City of Copenhagen 2011, 26)”..** “Plan B” uses street surfaces as conveyance routes for excessive run-off.*** Raised threshold: Raising egress edges to prevent surface run-off water from entering building.**** Planning measures include : 1) “New sewer systems already have to be dimensionally designed today so that they cope with the projected volumes of rain and consequently meet the service objective. The dimensional design base has to

67 bee incorporated into all relevant municipal plans. 2) Separation of common sewer in SUDS solutions is to be promoted and implemented (City of Copenhagen 2011, 27)”.

To apply the Windows of Opportunity model as shown in Figure 3.7, several steps were required to process the data available in the plan. The extensive measures shown in Table 1 were re-categorized into three main headings with the corresponding continuum of measures: 1) reduction of quantities of flood water going into the sewer system (disconnection of storm water from combined sewer system, detention/retention basins, roofs of buildings for water collection, SUDS, etc.); conveyance of flood water

(redesign sewer system, pumping water to sea, etc.); protection of infrastructure and assets (raising levels, moving sensitive facilities, building dikes; and general emergency preparedness such as sand bags, “Plan B”, backwater valves, etc.). The set of measures run in parallel to ensure adequate climate proofing of Copenhagen.

Figure 3.7: Phasing Model applied to the Copenhagen Climate Adaptation Plan (2011)

68 Monitoring is planned for four-year periods. In addition to plan updates, technological advancements in climate science, and monitoring of indicators and their

ATPs, these windows of opportunity provide the chance to address context specific considerations related to Copenhagen. For example, the urban watershed that

Copenhagen rests within is under several administrative jurisdictions. Therefore, appropriate coordination and collaboration among these administrative entities is necessary to reduce the likelihood of extensive run-off from extreme climate events originating in these regional jurisdictions.

The plan’s ATPs are based on indicators of total flooded area (from extreme precipitation and wave surge) and sea wave surge. Similar to the Clarence case study, a financial metric is used to evaluate every step of the adaptation implementation. The risk index is included in the model as an additional criterion to move up the ladder of adaptation measures. The risk index is calculated as the difference of the public cost of adaptation measures and the cost of potential risks based on a specific ATP condition.

3.5.3 Case Study Discussion

The application of the model to these two exemplary cases demonstrates that

planning for the implementation of adaptation measures is possible regardless of the

uncertainty involved. While the plans vary in the areas addressed, context, and methodology, both plans recognize the need to move ahead with adaptation because the costs of ignorance are too high. In the context of uncertainty of information about the future, flexibility in adjusting plans, measures and methodologies is core to climate proofing communities. Organizing adaptation measures from no-regrets to transformational measures carries wide benefits in the current window of opportunity,

69 and incrementally adds measures as needed using information, advance technology, and monitoring of implemented projects to indicate when the next phase is required. To address the barrier of the high cost of adaptation, both cities anticipate moving gradually along a spectrum of integrated measures, allowing the opportunity to begin implementing while monitoring the need for next measures. Focusing on conditions rather than time frames reduces the barrier of uncertainty when it comes to adaptation policies.

3.6 Concluding Remarks

Urban areas need to build resilience to climate change and variability. Implicit in the approach presented in this article is the subtle but radical suggestion that phasing of policy be linked to locally experienced outcomes, rather than a strongly pre-defined plan that rolls out over time. This allows a focus on the local experience of environmental change and the outcomes of interventions put into place. Allowing this flexibility reduces one barrier in policy implementation, as policymakers’ fear of acting too precipitously is reduced. Action will only be taken when it is warranted – but plans are in place so that necessary action can be rapidly implemented.

Having a long-run view of an implementation path allows testing for maladaptation in proposed policies. And using a suite of indicators with pre-designated tipping points (ATPs) allows for the explicit inclusion of local knowledge, and reduces the need to differentiate between climate change and climate variability. Indicator sets need to be developed collaboratively amongst governmental levels, and in some instances be translated to common language such that communities can readily use them. National or state level agencies may wish to develop suggestions for local indicators, to help jump- start community considerations. While we have used an urban planning framework, the

70 basic approach of adaptive planning with pre-determined thresholds is also applicable to natural resource areas and conservation lands.

This analysis supports the literature’s emerging consensus on the importance of starting with no-regrets policies (Biesbroek et al. 2010; Juhola, Peltonen, and Niemi

2012), many of which are well-established best urban planning practices anyway. These are the policies of sustainable social and environmental development, including strategies for increasing green infrastructure in urban systems, increasing public and non-motorized transportation, and protecting ecosystems. In a given urban micro-environment, implementing these policies for cleaner, greener, healthier cities can slow the need for more radical transformations by directly addressing some of the impacts of climate change.

There is a great deal that is not addressed here. Perhaps the most pressing item is the difficulty of large dollar and long-time-frame investments, those that do not yield to gradual implementation. Permitting major water or shoreline interventions can take many years, and storm water piping lasts decades; for these major, long-term investments, future-climate-adapted policy based upon realistic climate change projections is needed now. Other challenges come from the need to balance scientific rigor and local meaningfulness in monitoring and choices of indicators; identifying appropriate portfolios is essential. Significant issues revolve around communicating with the stakeholders and elected officials, accustoming them to working around uncertainty and time concepts suggested by this approach to the planning process.

Continuing research on these issues is necessary.

71 Given the long time horizons of urban land use and infrastructure, it is essential that local officials begin including climate adaptation in their planning, but given the uncertainties inherent in climate projections, it is difficult for them to move forward. The strength of the approach presented in this paper is the ability to make incremental decisions about investments in climate change adaptations, but with a comprehensive view that minimizes maladaptation. And at this point in time, the imperative is to proceed, flexibly but thoughtfully.

CHAPTER 4

THE GREEN INFRASTRUCTURE TRANSECT: AN ORGANIZATIONAL FRAMEWORK FOR MAINSTREAMING ADAPTATION PLANNING POLICIES

4.1 Introduction

Although climate adaptation is a fairly new policy goal for GI (Gill et al., 2007;

CCAP, 2011), three key characteristics qualify GI as a suitable tool for adaptation planning including multi-functionality (to match ecosystem beneﬁts with adaptation

needs), multi-scalar nature of the spatial elements, and a ‘no-regrets approach’. However,

GI needs to be matched to the character of the urban environment and coordinated across

jurisdictions and planning scales to become an effective adaptation policy. In this chapter,

we present a policy framework, the green infrastructure transect, that can help planners

and policymakers identify appropriate GI policies for different urban environments and

describe how these policies can create a regional adaptation planning framework.

73 One of the primary principles of green infrastructure (GI) planning is to reconnect

communities in urban regions to natural environments (Lewis 1964; McHarg 1969; Noss

and Harris 1986; Benedict and McMahon 2002, 2006; Jongman 1995; Jongman et al.

2004; Fabos 2004). This is achieved through practices within and around cities that

identify, protect, and create spatial elements that provide ecosystem services that communities depend on (Benedict and McMahon 2006; Forman 2008). Development of

community parks and recreation trails, greenways, ecological networks, restored streams,

natural reserves, gardens, engineered natural systems, green roofs and facades, and

conserved agricultural land are all within the scope of GI. Furthermore, the same spatial

areas also provide urban cooling, storm water management, ﬂood water storage, ﬂora and

fauna habitat, and biking and walking routes. All of these urban functions must be

increased to build resilience to climate change. By connecting ecosystem beneﬁts to

community well-being (Nassauer 2006) and adaptation needs, GI planning may be

mainstreamed to become an integral part of adaptation planning policies.

A key advantage of the GI approach to adaptation is that it is already becoming an

accepted practice (Benedict and McMahon 2002; Ahern 2008). GI has become part of

current sustainable planning and design practices in many cities (EPA 2011; Newman

and Jennings 2008; Farr 2008). These initiatives function at multiple scales to improve

urban living conditions. These may include retention ponds and swales (at the parcel

scale), green streets and parks (at the neighborhood scale), increased tree canopies (at the

urban scale), and greenways (at the regional scale). As an accepted practice, GI is also a

‘no-regrets approach’ (Bedsworth and Hanak 2010) when considered as an adaptation

measure. As we move into the future, investment in GI policies will prove to be

74 beneﬁcial regardless of whether climate change scenarios materialize. For example, urban

greening results in cleaner air and cooler temperatures that would address current

problems (pollution and urban island heat effect) as well as ameliorate future increasing

temperatures. As a result, fairly minor changes to the technical speciﬁcations for GI

could, quite effectively, bring adaptation into mainstream practice. As GI is implemented to accommodate increased ﬂooding, ameliorate rising temperatures, or address the rise in sea levels, communities can take advantage of the cultural, social, and health beneﬁts of

cleaner and greener environments, regardless of the future magnitude of climate change

impacts.

Furthermore, the same characteristics that qualify GI as a spatial adaptation tool

within urban regions (notably GI’s multifunctional and multi-scalar properties) make it

difﬁcult to mainstream GI into adaptation planning. These characteristics create problems

in organizing intervention areas, jurisdictional coordination and implementation, and

trade-offs in economic beneﬁts and urban quality.

4.2 The Green Infrastructure Transect

To address these problems, we propose the green infrastructure transect (GI

transect) as a framework to utilize GI as an adaptation policy and to mainstream

adaptation into current planning practices. The GI transect is a conceptual tool that

integrates GI measures across varying urban contexts and across planning scales. It builds

on transect concepts from ecology, landscape planning, and urban planning13. We

13 In the early twentieth century, the natural transect became one of the foundational tools of ecological research. The evidence that certain flora and fauna flourished symbiotically together within specific mineral and climactic environments became the ethical basis for the protection of species. Patrick Geddes (1854–1932) adopted the ecological transect as

75 speciﬁcally use the urban transect as a stepping stone to develop this framework.

The urban transect (Duany and Talen 2002) was devised as an urban planning tool

to plan and design physical environments according to peoples’ preferences of where to

live and work. The urban transect identiﬁes six zones (urban core, urban center, general

urban, suburban, rural, and natural) with distinct physical boundaries as units of study.

These zones form a planning model applicable within many urban contexts. The zones provide the basis for a neighborhood structure based on walkable streets, mix of uses, transportation options, and traditional architectural styles and housing diversity. The strength of the urban transect is in describing the appropriate built forms and identifying interventions within each urban zone in a simple and comprehensible manner. The concept falls short of specifying the respective open spaces and natural functions that respond to the speciﬁc urban contexts and needs within each transect zone.

In contrast, the natural transect used by ecologists and biologists is a scientiﬁc method of assessment of habitat. It is based on the fundamental principles of relationships and interdependencies between eco-zones and used to assess the physical, biological, and natural processes within and across eco-zones. Contrary to the urban transect, it does not specify distinct spatial zones. Rather, the characteristics of different local ecosystems deﬁne different habitat zones and the relationship between them. This same principle is later adopted within landscape and regional planning to assess and understand

a model to devise the ‘valley section’ (Geddes 1915). Taken from ridgeline to shoreline, the ‘valley section’ shows natural conditions with their associated human presence and occupation to show a gradation of human preference for location and work. Based on Geddes, Lewis Mumford’s (1895–1990) concept of human ecology was used to develop a decentralized regional vision of metropolitan areas (Mumford 1937). Ian McHarg (1920–2001) applied the natural transect for land conservation in landscape planning showing transitions and relationships within and across natural eco-zones (McHarg 1969).

76 relationships between land, and natural and human systems in order to plan and manage

natural resources within urban regions (McHarg 1969; Picket 2004;Berger 2006).

Overall, the GI transect combines the general principles of urban and natural transects into a single assessment model. The primary characteristics are three: first, the simultaneous consideration of human and natural systems as a mutual cause-and-effect relationship effecting the functional capability of GI (pervious and impervious surfaces as indicators); second, the designation of urban zones as unique spatial contexts that may impact the adaptive capacity of communities within; and third, the explicit consideration that GI is an interconnected system that transcends administrative and political boundaries.

This interconnectedness of GI serves as an impetus and analogy to integrate adaptation policies across the GI zones increasing the local capacity for adaption. We qualify this level of policy integration as ‘horizontal integration’. The term is meant to generate targeted GI recommendations speciﬁc to each GI zone and coordinate them across

boundaries14 (within scales). This is achieved by mapping and assessing each GI zone against a set of criteria to be able to recommend targeted GI measures. Six GI zones are

identiﬁed and are intended to represent an alternative model (to the urban transect) of

contemporary urban regions. These include coastal (if present), urban core, urban,

transition (the middle ground), suburban, and peri-urban zones. In addition, we use the

following criteria to assess each GI zone: vulnerability assessment using spatial data

14 Cross-jurisdictional coordination was identiﬁed as a primary concern when assessing the 4,000 GI networks across the conterminous USA for their ecological connectivity where 10% of the hubs and links cross administrative and political boundaries. When down-scaling the same observation to regional and local scales, forest stands, water bodies, and wetlands are not restricted to regional, city, town, or property boundaries.

77 (physical and social), identifying the primary climate change impact based on spatial configuration and character, identifying the spatial character of each GI zone, determining the spatial configuration of pervious and impervious surfaces (existing and potential GI), determining GI typology relevant to each zone, and recommending appropriate GI measures within each GI zone (Figgure 4.1). The sequential process of assessment begins with vulnerability assessment and concludes with recommendations providing a specific policy focus to local communities for adaaptation and the possible responses through GI.

Figure 4.1: The Green Infrastructure Transect: Concept and Organization

Furthermore, several existing GI recommendations relevant to adaptation policies call for the protection of forest stands and wetlannds or increasing tree canopy or engineering swales and rain garden systems. These GI spatial elements are not restricted to regional, city, town, or property boundaries as they are subject to conditions (i.e. topography, geology, and hydrology) beyond the control of governing bodies. Therefore, analysis and assessment should consider recommendations within each zone and the outward extensions of GI. By mapping adjacent spatial conﬁgurations, horizontal integration is attained. This enhances the adaptation capacity of local communities through coordination of policies. Yet, it does not account for coordination across jurisdictional boundaries and planning scales neccessary for regional resiliency.

78 Developing a network of GI increases the resiliency of a region. It provides alternative evacuation routes, species migration routes, CO2 sinks, ﬂood water storage, buffer zones against rising sea water and reduction in regional temperatures. To achieve a coordinated regional network requires the integration of planning scales (neighborhood, urban, regional) into a single regional planning framework providing a platform for communication and coordination across jurisdictions. We term this integration across scales as ‘vertical integration’.

Vertical integration provides the mean to respond to the multi-scalar and hierarchal nature of GI by considering current planning processes. GI networks are hierarchal especially when planned within urban contexts. When considering GI for storm water management, connectivity of GI elements should be considered across the hierarchy of urban planning scales (street or parcel neighborhood, city, and urban region)

(Kato 2010). For example, several streets with bio-swales and retention ponds in residential yards at the neighborhood scale can constitute a green corridor at the city scale which, in turn, with city parks can be part of a regional park system (Jim and Chen 2003;

Girling and Kellett 2005). But each individual GI element (parcel to regional scales) is planned and implemented differently, depending on the context, size, and planning process. Vertical integration provides a way to unify these processes under a hierarchal single framework that leads to a regional vision.

Integration across scales is necessary to increase the adaptive capacity at both the regional and local levels. The adaptive planning meta-model developed by Kato (2010), for a planning framework to manage GI, is an example of such a process. It is an iterative process designed for the US context. Similar to the GI hierarchy, neighborhood plans that

79 are participatory in nature form the basis of an urban plan. The sum of the several urban plans could deﬁne a vision for a region. In the US context, a bottom-up approach

(participatory) could lead to a regional vision. The reverse (top-down planning) may also hold true when considered within other planning and administrative contexts. Regardless of whether the vision (top-down) or local planning (bottom-up) comes ﬁrst, the intention here is to advocate for a two-way and iterative approach that includes both and provides the ﬂexibility and adaptability to respond as circumstances arise and change.

Figure 4.2: Multi-tiered GI adaptation planning framework: horizontal and vertical integration. The underlying concepts behind the GI transect point to the spatial, contextual, and administrative interdependencies governing mainstreaming adaptation planning. Vertical and horizontal integration are the primary elements of the GII transect that integrate local and regional plans into a single and ﬂexible adaptation planniing framework. To make these ideas concrete, we apply the three-step approach of vulnnerability assessment, characterization of existing GI, and GI policies recommendations to the Boston metropolitan region.

80 4.3 Boston Metropolitan Region

The metropolitan region of Boston occupies the eastern shoreline of the state of

2 Massachusetts in the USA. It covers a land area of approximately 12,000 km , housing

4.5 million people with an average density of 366 persons per square kilometer (Census

Bureau 2010). The metropolitan region incorporates 120 towns and 8 regional

jurisdictions within its boundary (Census Bureau 2010). It is characterized by an urban

core (Boston) as the center of governance, business, and transportation. From the urban

core to the periphery, residential sprawl of varying densities along transportation

corridors and around commercial centers is interspersed by forest, wetlands, river basins, and, to lesser extent, agricultural land (Figure 4.3 and Figure 4.4). At the planning level,

the state of Massachusetts (MA) has adopted and is implementing smart growth

principles to control development and preserve natural and cultural assets.15 Part of the

smart growth initiative is the Climate Action Plan (CAP 2007, 2010). The plan is focused

on mitigation measures to reduce emissions from buildings, transportation, waste

management, and land use. In the 2010 update of the plan, recommendations for

adaptation were included as part of addressing causes and effects of climate change.

The NECIA (2007) report on climate change impacts within the New England

region shows that Massachusetts climate will resemble the southern states of the Eastern

Coast of the USA.16 Taking the year 2000 as the baseline, the report demonstrates that the

15 Since planning is locally based and participatory, the state of Massachusetts may only advance these planning principles through ﬁnancial incentive means. Towns and cities may develop their comprehensive zoning, recreation and open space, and economic development plans based on smart growth principles in return for ﬁnancial incentives. 16 Under the high emissions scenario, the Massachusetts climate will likely resemble that of the current Florida climate and under a lower emissions scenario will resemble the current weather of Northern Carolina.

81 ⁰ metropolitan region of Boston will experience increase in temperatures by 4–7 C in the

⁰ winter and 3–8 C in the summer, rising sea level of 25–60 cm, and increased precipitation byy 20–30%. To address these impacts, the City of Boston identiﬁed guidelines for adaptation planning (CAP 2010) that include, in addition to economic and social measures, spatial measures that focus on GI.

Figure 4.3: Metropolitan region of Boston: spatial distribution of pervious and impervious surfaces

Figure 4.4: Metropolitan region of Boston: green infrastructure across town boundaries

The adaptation recommendations for Boston (CAP 2010) set priorities and deﬁne the required information and planning priorities and approaches. Out of the thirteen recommendations, many focus on GI principles such as greening the city, green roofs, sustainable water management, and protection and increase of large tracts of vegetated surfaces. In addition, planning cross-jurisdictions and scales iis identiﬁed as a priority to increase the adaptive capacity of the urban region.

In the process of transforming these adaptation recommendations into actions, we apply the GI transect to assess the applicability of the multi-ttiered organizational framework to Boston. In the assessment stage, we map vulnerability, climate change impacts, and the physical environment across thee GI zones (Figure 4.5). Vulnerability is mapped using the following spatial data layers from Mass GIS17: topography, open space,

17 Mass GIS is a spatial data portal managed by the state of Massachusetts that provides a

83 roads, location within the watershed, and socio-economic data for each location. Climate impact is mapped according to the NECIA (20077) report showing the magnitude and focus within each zone. Aerial images are used to map the urban character identifying the physical environment of work and housing.

Figure 4.5: Green infrastructure transect application to Boston region, step one: vulnerability assessment

We found that the coastal zone is predomiinantly impacted by rising sea level, the urban to transition zones are affected by a high magnitude of increased temperatures and

ﬂooding, and the peripheral zones are impacted, at a lower magnitude, by temperature rise and ﬂooding. The exposure to physical risks is further exasperated by the effect of the urban heat island effect (UHI) and the gradation of impervious and pervious surfaces across the GI transect. The compounded impacts of climate change and the physical free download service of available data layers across the state. See http://www.mass.gov/mgis/.

84 characteristics of the urban region of Boston are grounds to consider different adaptation planning focuses for communities across the GI transect. To bbe able to devise and recommend GI policies within existing pervious ssurfaces, which address the variation of vulnerability, we map the existing distribution of GI across the zones.

To map the spatial distribution of GI across the zones, we also use Mass GIS data. We overlay the following layers: impervious surfaces, digital terrain, open space layers

(public domain), waterways, forests, roads, and administrative boundaries.

Figure 4.6: Green infrastructure transect application to Boston region, step two: existing green infrastructure patterns

We find that open space and un-built18 land increases in area as we move towwards the periphery (Figure 4.6). What significantly increases, and not usually included in the inventory of GI, are un-buuilt spaces within the prriivate domain (yards, gardens, and school grounds). Since ecosystem benefits are not bounded by administrative limits (Fabos,

18 Un-built land is considered as potential to increase green infrastructuree area within an urban region

2004) and increase proportionally with GI area,19 it is critical to ensure that GI policies

simultaneously address land within the private and public domains.

The final step is to identify and recommend appropriate GI policies across the GI transect zones. We distinguish clear complementarities between GI benefits, community needs, and vulnerability requirements (Figure 4.7). We list the typologies of GI elements that already exist within each zone or those that could potentially be introduced or enhanced. Ecosystem benefits that are complementary to community needs and climate impacts are also listed in accordance with the spatial typology. By overlaying information from steps one and two, we begin to identify the potential GI policies. For example, the coastal areas will benefit from planned retreat where vulnerable built areas across the coast may gradually be transformed into landscapes for recreation. The resulting coastal

landscapes become non-structural20 defenses incorporated as recreational and ecological

landscape features. Therefore, the policy here would focus on preserving and intensifying

all existing GI elements and to deﬁne a long-term plan to allow time for legal procedures

and ﬁnancial compensation to take place for the coastal zone transformation. Within the

urban zone of the GI transect, policies should address increased temperatures

(compounded by UHI) and retention of water run-off. Existing parks and open space,

green roofs, green facades, and street planting are spatial elements that should be

increased through revisions to building regulations, open space plans, and environmental

19 Ecosystem beneﬁts are directly proportional to the amount of land available for GI: the more forested land, the more the potential for temperature control, and the more the golf courses and open land, the more water storage may be achieved.

20 Non-structural defenses are based on naturally occurring or engineered defenses such as wetlands, marshes, sand coasts, and eastern dams.

86 policies. Through the Biotope Area factor21, the city of Berlin is an example where zoning and ﬁnancial incentives result in an increase in tree canopy and ‘at the sourcee’ water management. Towards the periphery, policcies that enhance connectivity and preserve, conserve, and increase forests, large parks, natural reserves, and biospheres are integral for run-off storage, species migration, temperature control, and water inﬁltration to ensure ecosystem services at the regional scale.

Figure 4.7: Green infrastructure transect application to Boston region, step three: identiﬁcation of GI policies

21 See the City of Berlin, Senate Department for Urban Devellopment: http://www.stadtentwicklung. berlin.de/umwelt/landscchaftsplanung/bff/index en.shtml.

87 To ensure consistency across local GI policies with the Boston region, vertical

and horizontal integration of policies is utilized to coordinate and implement planning

projects across town jurisdictions. Planning in Massachusetts is predominantly

participatory and happens at the local (town) scale. This means that parcel and

neighborhood scale plans should build up to form an overall town plan that explicitly

considers GI measures for adaptation. The open space plans that are mandatory to US

towns could be extended beyond recreation to incorporate ecological and adaptation

plans. Town plans then need to build the overall regional vision. This may be achieved by

expanding the mandate of regional planning bodies beyond transportation and economic

development towards a more active role to coordinate and integrate local plans. Even

more, regional bodies should be responsible to monitor and develop regional climate

projections that help in providing the vision for regional and local adaptation plans. A

hierarchal organizational structure that works in both directions (from local to regional or from regional to local) ensures that all constituents and measures serve an intended local role within a larger regional approach. The proposed structure that we have presented may be a ﬁrst step in integrating local adaptation planning across scales and jurisdictions using current and accepted knowledge.

4.4 Conclusion

Adaptation policies run the risk of a piecemeal, systematized approach. It is easy to prescribe a green roof here and a rain garden there and hope that they will add up to a proper systematic approach. However, the challenges of adaptation are too signiﬁcant for this to be effective. Framing GI planning through the transect approach provides a way to

88 conceptualize a whole system of GI spatial elements, identify coming climate challenges, and plan to integrate local policies at site scale with adaptation needs at the neighborhood, city, and regional scales. In this chapter, we brieﬂy used Boston as a case study to demonstrate how the GI transect may be applied and how it can assist in interpreting and framing overall GI for adaptation. We conclude that GI will be an effective adaptation policy when it is matched to the physical character of urban environments (urban, suburban, and rural) and the needs of communities they are intended to serve. This approach is a ﬁrst step in mainstreaming adaptation planning using current GI practices.

89 CHAPTER 5

PERCENT PERVIOUS: AN ASSESSMENT METHOD OF GREEN INFRASTRCTURE SPACE OPPORTUNITIES - AN APPLICATION TO THE BOSTON METROPOLITAN GRADIENT FOR URBAN HEAT ISLAND ADPATATION

5.1 Introduction

When planning to retrofit cities with measures that address impacts of local and global climate change, green infrastructure is included in adaptation plans as an effective strategy to reduce impacts of climate change. It is the complementarity between climate impacts (increasing temperature, flooding due to extreme precipitation events, and rising sea level) and ecosystem services that render green infrastructure as a no-regrets climate measures. As an ecosystem based approach to climate proofing, green infrastructure depends on biological and ecological processes between soil, water, vegetation, and climate to deliver ecosystem services. Green infrastructure is especially effective for temperature amelioration in urban climates (REF). Green infrastructure measures that reduce air and surface temperatures are highly dependent on the extent of coverage of vegetative surfaces be it ground cover, shrubs or tree canopy. Assuming that all biological components of an ecosystem are functioning well, the primary variable in the calculation of ecosystem services is the amount of area that vegetative cover occupies

(Figure 5.1). This means that more vegetated surfaces will result in more ecosystem services. But within urban contexts, high land values and competition for real estate render urban land a highly contested commodity. Dedicating maximum possible area for green infrastructure to effectively ameliorate the urban climate is not readily available,

but should be planned and pursued. As a direct result of this competition, different land

use classes impact the availability of possible current and future vegetative surfaces in

90 three ways: determine the morphology of the surfaces that potentially could be vegetated due to foot print size and outdoor uses; define type of vegetation suitable for that land use; and determine the future permanency of dediicated surfaces for vegetation, especially when accounting for more drastic climate change impacts and further urbanization. The problem of sufficient space within urban contexts is more acute when considering the application of green infrastructure to reduce the immpacts of the urban heat island.

Figure 5.1: Primary variable for the provision of ecosystem services is the area of hydrologically active surfaces (Stone, 2012) or vegetated surrfaces.

The urban heat island (UHI) is considered as the earliest documented form of human induced local climate change (Oke, 1976). The UHI is a phenomenon specific to urban areas and is defined as the measure of excess heating expressed in terms of the horizontal temperature difference between the city and its surrounding context (Oke, 1979; Kuttler,

2008; EPA, 2008). The UHI is a direct result of the urbanization process. Urbanization causes surface change from its original landscape form (preddominantly vegetated) to hard and reflective surfaces that absorb energy and emit heat to the surrounding air, increasing surface and air temperatures. Within land use based measures, green infraastructure measures that are based on vegetated surfaces are eligible to mmitigate the UHI and adapt to its impacts because they specifically reverse this trend in land cover change. For that to

91 be effective there should be a substantial increase in hydrologically active surfaces

(Stone, 2012) that absorb the heat and reduce the air temperature. The biological processes in vegetation qualify them as naturally occurring solution to the heat problem.

For vegetated surfaces to become operational as a planning tool they need to be envisioned as a green infrastructure network across the whole urban area or region

(Rosenzeig et al., 2006). Such an ecological network may include urban forests, green roofs, and transformed impervious surfaces to accommodate vegetation. In other words it is a network that maximizes the allocation of land and built surfaces that are explicitly accounted as an infrastructure that provides heat reduction services.

To achieve these two planning objectives, land and built surfaces surfaces provide varying opportunities for the development of a green infrastructure network that is effective in reducing temperatures. Patches of forested or dense tree areas less than three

(Bowler et al, 2010) or four hectares (Rosenzweig et al., 2006) may not be effective in extending the benefits of heat reduction beyond patch limits. Similarly, flat building roofs less than 200 meters may also not contribute to the heat reduction measures. When considering types and sizes of effective surfaces for UHI, impacts of urbanization in general and the impact of land use types on the morphology of vegetated surfaces (Figure

5.2), the allocation of sufficient surface area for an effective green infrastructure network within an urban region becomes a challenge for planners. The problem persists as to how move forward with green infrastructure policy within the constraints of available space within urban contexts.

Figure 5.2: Impact of land use classes on space availability and morphology of pervious surfaces.

To address this conundrum of available surfacce area, this chapter develops an assessment method to estimate potential space for green infrastructure vegetative surfaces across land use types. The Boston Metropolitan Area is used as an application case study and urban heat island as a treatment to limit the study and to identify specific green infrastructure measures suitable for the treatment. Pervious surface is considered as surrogate to existing and potential vegetated surffaaces. A percent pervious land metric is developed and used to assess the extent of available space acrross the urban-rural gradient and available pervious surfaces across different land uses. The objective is further narrowed down to select suitable surface areas that could be effective for UHI reduction.

Based on the results, a gradient of green infrastructure policies is proposed to assist in the development of green infrastructure policies centered on the gradient of available pervious surfaces. The focus is on assessing opportunities of surface area (space) to increase tree canopy cover and transform impervious surfaces to vegetated surfaces. We utilize readily available spatial data from Mass GIS (see footnote 22). This study considers UHI as a surrogate to global climate change based on the fact that measures that aim to reduce temperature reduction in urban areas are synergetic with adaptation

93 measures (Alcoforado at al., 2008). Impervious surface is considered as the direct result

of the different land use classes and is surrogate to the urbanization process.

The proposed method aims to answer the following research questions: 1) How does

pervious land surface vary across the metropolitan gradient? (2) How does pervious land

surface opportunity vary across land uses along the gradient? (3)When considering UHI,

what is the maximum opportunity of pervious surfaces for green infrastructure measures

across land-use categories? And (4) How does the percent pervious area metric inform

green infrastructure policy decisions when considering UHI treatment? This study considers UHI as a surrogate to global climate change based on the fact that measures that aim to reduce temperature reduction in urban areas are synergetic with adaptation measures (Alcoforado at al., 2008). Impervious surface is considered as the direct result of the urbanization process that limits availability of pervious surfaces. Land use classes

are surrogate to the urbanization process.

There are two objectives for the steps taken in the method: First, to answer the

research questions and second to conduct the analysis in a relatively expert-free approach

using readily available spatial data and basic geographic information systems (GIS)

skills. The first objective will be thoroughly discussed in the following sections. The

second objective proposes an easy to apply assessment and estimation method of

available space for green infrastructure addressed to planners in public and/or private

planning agencies. This is in contrast to more specialized and advanced techniques such

as remote sensing that are time consuming and costly. Mapping vegetative covers and

specifically tree canopy cover (TCC), requires specialized high resolution imagery with

expert knowledge of multiple techniques of classification, (Irani and Galvin, 2003; Goetz

94 et al., 2003; McPherson et al., 2011). When applying multiple reclassification tools in

conjunction with the normalized difference vegetation index (NDVI) and

evapotranspiration mapping to large geographic areas, extracting this high resolution

information is taxing on time, equipment, and resources. While remote sensing is

becoming standard practice for scientists, specialized firms, and educational institutions,

it is still not standard practice for planning agencies. Such work is usually contracted to

third parties. On the other hand, basic GIS know-how has become standard skill for

planners to manage spatial data sets for states, towns and cities are common practice. The

basic data sets used in this study are readily available in many states, and specifically in

the North East of the US (Mass GIS, VCGI, NH GRANIT, MEGIS, and RIGIS 22), the geographic context of this study. Therefore, a method for the use of planners that utilizes readily available base data, knowledge and skills may be helpful in developing reliable, fast, and inexpensive estimates of green infrastructure opportunities within urban regional contexts.

The remainder of this chapter introduces the relationship between urban heat and urban form and extracts suitable green infrastructure measures for UHI. The method is then discussed followed with a detailed account of the results. The significance of the findings and caveats of this research are discussed in the concluding section.

22 Mass GIS is the Massachusetts portal for geographic information systems; VCGI is Vermont Center of Geographic Information; NH GRANIT is the New Hampshire statewide GIS data clearinghouse; MEGIS is Maine office of GIS; and RIGIS is the Rohde Island Geographic Information System.

95 5.2 The Urban Heat Island and Urban Form

The urban heat island (UHI) is a phenomenon specific to urban areas. The UHI is

defined as the measure of excess heating expressed in terms of the horizontal air temperature difference between the city and its surrounding context (Oke, 1979; Kuttler,

2008; EPA, 2008). Urban areas manifest hotter surface and air temperatures than their

surroundings due to the imbalance in the energy budget. The air or surface temperature

difference is a result of the modification of the pre-development landscape into surfaces

that are impermeable and prone to energy absorption during the day (long wave energy)

and heat release during the night (short wave energy) (Oke, 1971). The urban heat island

is identified by measuring surface or air temperatures. Surface temperatures have an

indirect but significant influence on air temperatures (EPA). For example, parks and

vegetated areas have cooler surface area and contribute to cooler air temperatures. Within

highly built areas, surface temperatures are much higher and contribute to hotter air

temperature. Because air mixes within the atmosphere the relationship between surface

and air temperatures is not constant (Figure 5.3).

Figure 5.3: Typical Surface and air temperature differences during the day and at night (Source; EPA http://www.epa.gov/hiri/images/UHI_profile-rev- big.gifhttp://www.epa.gov/hiri/images/UHI_profiile-rev-big.giif )

The character and intensity of UHI formaattion has been documented through the field of urban climatology (Chandler, 1962; Oke,, 1982; Taha, 1997). UHI is typically present day and night, but tend to be strongest during the day when the sun is shining.

On average, the difference in daytime surface temperatures between developed and rural areas is 10 to 15°C (18 to 27°F) and the difference in nighttime surface temperatures is typically smaller, at 5 to 10°C (9 to 18°F) (Roth et al., 1989; Voogt and Oke, 2003). In addition to city location (i.e. coastal or inland), topography, and wind direction, the phenomenon differs in magnitude and intensity by city size (OOke, 1973) aand urban

97 morphology (Kobayashi & Takamura, 1994; Nunez & Oke, 1977). Both characteristics

are a result of the urbanization process.

Urbanization has consequences on urban form that highly contributes to UHI

(Gartland, 2008;Kuttler, 2008; Alcoforado et al., 2008; Stone 2012a & b). Mills (2007)

distinguishes between two definitions of urbanization: the stock effect, which is created

by the physical presence of the city; and the flux effect, which is the outcome of activities

associated with cities, in other words, the urban system. Both effects have consequences

on the UHI.

The stock effect is the physical outcome of urbanization which is seen in the

impermeable land cover and closely spaced buildings. In many cities, the urban land

cover decreases from the center to the periphery with less defined edge boundaries due to

intertwining non-urban land cover (parks, forest, wetland, etc.). In its three dimensional

form, cities in the western hemisphere tend to have taller buildings in the urban center

with gradual decrease in height towards the periphery. While the concept of a dense and

concentrated city may reduce energy consumption and emissions (total miles travelled

and better insulation) (i.e. contributing less to air pollution), in return, a compact urban

form reduces total vegetated cover (i.e. reducing the evaporative cooling and shade

capacities).The combination of impervious surfaces and the three dimensional urban form tend to exasperate hotter temperatures through four attributes of the city: reduction in evaporative cooling, low surface reflectivity, re-absorption of reflected radiation by vertical surfaces and the contribution of hot waste heat to air temperatures from mechanical and electrical systems (Stone, 2012).

98 The flux effect or the urban system links the central employment built-up area

with its economic and population hinterland. This link is governed by networks of

transportation and communication connecting different settlements and facilitating flows

of resources, information and people. The physical manifestation of the urban system is

an urban form with settlements spread over large tracts of land connected by

infrastructural systems interspersed by non-built land covers. Tis definition corresponds

to an urban region (Lewis and Brabec, 2005; Forman, 2008) or a metropolitan area (US

Census, 2010). The impact of dispersed and distributed urban form results in increased emissions, energy use and resource consumption that negatively contribute to the urban climate through air pollution and particulates in the air. The combination of radiating heat from impermeable surfaces and the contribution of noxious gases interact and result in a feedback loop exasperating the UHI effect.

The feedback loop may be described as the continuous interaction between land surface characteristics, urban activity and the sun’s energy (Figure 5.4). The stock effect from urban and suburban areas within a metropolitan region increase impervious surfaces and reduce vegetated surfaces by increasing number and spread of buildings, roads and parking surfaces. This process reduces albedo 23 of the urban region and air temperatures

within street canyons increase. These land cover changes impact the urban climate by

decreasing evaporative cooling, transpiration as well as increasing heat absorption,

radiant heat and heat production (Oke, 1979, 1987; Gartland, 2008; Kuttler, 2004,2008).

As a result of the flux effect, the impact of air pollution and the resulting heavy air from

23 Albedo: When sunlight hits an opaque surface, some of the sunlight is reflected. This fraction is called the albedo (a). The rest of incoming energy is absorbed and designated as (1-a). Low-a surfaces become much hotter than high-a surfaces (Akbari, 2005).

99 transportation and energy use tend to confine hottter air closer to the surface and within the urban canopy increasing the intensity of the UHI across the whole region. This negative feedback loop results in three different types of UHI: surface heat island (SHI), heat islands of the urban canopy layer (HIUC), and the urban boundary layer heat island

(UBH). The SHI is the heating of urban surfaces and the HIUC is the resultant heating of the air between the ground surface and top of buiildings. In both cases the impact is local.

UBH is the further mixing of air beyond the urban canopy with impacts in the downwind direction beyond the urban core (Kutller, 2008). The different types become more extreme and acute depending on the time of day, seasons, and climate conditions (Further discussion on UHI types is provided in Appendix A). The variability and different types of UHI occurring at different times of the day render UHI a dynamic phenomenon.

Figure 5.4: Process of UHI formation in relation tto urbanizattion process.

100 The UHI varies in intensity depending on regional climatic conditions and

contexts. Oke (1982) developed generalizations of the intensity of UHI that show the dynamic and uncertain nature of the phenomenon. The UHI intensity decreases with

increasing wind speeds as these provide direct breeze to residents and shifts the urban

boundary layer and plume downstream. This releases entrapped heat and replaced with

cooler air resulting in reduced temperatures (Figuerola and Mazzeo, 1998; Magee et al.,

1999; Morris et al., 2001; Unger et al., 2001). Cloud cover tends to reduce the intensity of

UHI because cloud cover screens the incoming sun energy which is the primary source of

absorbed and released heat (Ackerman, 1985; Ripley et al., 1996; Morris and Simmonds,

2000). UHI intensity is best developed in the summer due to the higher intensity of the

sun and reduced amounts of rain (reducing hydrant cooling), at least in the northern

latitudes (Philandras et al., 1999; Morris et al., 2001). UHI intensity is greatest at night

due to the nature of the phenomenon where impervious material absorbs the sun energy

during the day, stores it and radiates it back in short waves during the night (Ripley et al.,

1996; Jauregui, 1997; Magee et al., 1999; Mont´avez et al., 2000; Tereshchenko and

Filonov, 2001; Kuttler, 2008). It follows that UHI may disappear by day or the city may

be cooler than the rural environs (Tapper, 1990; Steinecke, 1999). In conjunction with the

above, cities with a larger footprint amplify the intensity of UHI due to the larger surface

area of impervious surfaces that increase absorbed sun energy and released heat.

Increased population size also increases radiant heat through increased car use and energy

consumption (Park, 1986; Yamashita et al., 1986; Hogan and Ferrick,1998). Furthermore,

rates of heating and cooling are greater in the surrounding area than built-up area because

of the difference of surface characteristics between both contexts (Johnson, 1985).

101 UHI may be designated as an example of local climate change, as it is the best

documented instance of human induced climate modification (Oke, 1987; APA, 2007;

EPA, 2008; Stone, 2012). Climate change, broadly speaking, refers to any significant change in measures lasting for an extended period resulting from natural processes or anthropogenic reasons (EPA, 2008). Local climate changes resulting from the UHI fundamentally differ from global climate changes in that their causes are different and

impacts are limited to the local scale and decrease with distance from their source. Global climate changes, such as those caused by increases in the sun’s intensity or greenhouse gas concentrations, are not locally or regionally confined. The impacts from urban heat islands and global climate change are often similar. For example, some communities may experience longer growing seasons due to either or both phenomena (Alcoforado and

Andrade, 2008). UHI and global climate change can both also increase energy demand, particularly summertime air conditioning demand, and associated air pollution and greenhouse gas emissions, depending on the electric system power fuel mix (EPA, 2008).

The influence of UHI on global warming is not the same as the converse.

Alcoforado and Andrade (2008) conducted a literature review to understand this relationship. Their review indicates that the influence of the UHI on global warming is minimal because urban areas cover less than one percent of the Earth’s land area (Oke

1997), and the amount of energy released by man is much less significant than the energy received by the earth from the sun. But, cities are a very important source of anthropogenic greenhouse gases (Crutzen 2004; Lamptey et al. 2005; Makar et al. 2006;

Kahn 2006) and thereby contribute indirectly to global warming (Crutzen 2004;

Sherwood 2002) by exerting a slight influence upon the computation of global warming,

102 especially in studies utilizing fine grained spatial data (Brázdil and Budíková, 1999;

Beranová and Huth, 2005; Quereda-Sala et al., 2000).

On the other hand, the impacts of global warming (including its impacts upon human well-being and health, various ecosystems, and on levels of energy and water

consumption) may be exacerbated in metropolitan areas. Depending both on their latitude

and regional climate, global warming will either improve or worsen livability conditions

within metropolitan areas (Oke 1997; Stone 2005). From the point of view of the human bio-climate, the high-latitude cities will probably improve, and low- and mid-latitude cities, especially in the summer, will probably be worse. In general, global warming will increase temperatures in metropolitan areas regardless of latitude or climatic context.

Therefore, warmer cities are likely to experience an increase in the levels of air pollution and water consumption. From that point of view all cities will probably be in a worse condition. Regarding energy consumption, colder climatic zones will have improved conditions and, in the winter, those at intermediate latitudes. However, additional climate-related problems may arise in high-latitude cities as a consequence of global warming. The consequences of global warming will exhibit considerable regional variability and will depend on the future frequencies of weather types. For example, an increase in vertical instability (urban plume, explained in types of UHI in Appendix A) associated with higher temperatures can partially offset urban warming (Alcoforado and

Andrade, 2008). While some cities may benefit from increased temperatures, the overwhelming impacts are negative in nature. Fluctuations in intensity of the UHI on daily or seasonal basis and coupled with extreme heat events with decreased moisture content impact the well-being and health of communities.

103 Elevated temperatures, particularly during the summer, can affect a community’s

environment and quality of life. Increased summer time temperatures increase energy demand for cooling adding pressure to the electricity grid during electricity peak demand

periods as well as increasing heat loss from the use of cooling equipment. This demand

increases 1.5 to 2 percent for every 0.6°C (1°F) increase in summertime temperature

(EPA,2008). The implication is disruption of people’s activities and waste of money for maintenance of buildings and infrastructure (Gartland, 2008). In urban centers, five to ten percent of electricity consumption is used to compensate for the UHI effect (Akbari,

2005). This increased demand of electricity causes higher levels of air pollution and greenhouse gas emissions. Electricity generation in the United States is dependent on fossil fuel combustion which is increased during peak demand periods emitting pollutants such as carbon monoxide and sulfur dioxide. Furthermore, increased emissions and higher temperatures tend to increase the level of ground level ozone formation (EPA,

2008; Stone 2008).

Increased daytime surface temperatures, reduced nighttime cooling, and higher air pollution levels associated with urban heat islands can affect human health by contributing to general discomfort, respiratory difficulties, heat cramps and exhaustion, non-fatal heat strokes, and heat-related mortality (EPA, 2008; Stone 2012). UHI can also exacerbate the impact of heat waves, which are periods of abnormally hot, and often humid, weather. Sensitive populations, such as children, older adults, and those with existing health conditions, are at particular risk from these events. For example, in 1995, a mid-July heat wave in the Midwest caused more than 1,000 deaths (Taha et al, 2004).

More recently, the heat wave of 2003 in Europe reached unprecedented high values

104 unseen in almost 350 years of keeping climatic records (Stone, 2012). In the UK, the registered nighttime air temperatures in London reached 6-9⁰C higher than those recorded for rural locations south of London. This event claimed 600 more deaths than usually accounted for during August (COL, 2006). On August 11, 2013 temperatures in

Switzerland reached an unimaginable 42⁰C (107⁰F). In Paris, night time temperatures did not go below 27⁰C (80⁰F). This resulted in an increase in hospital admissions reaching 20 to 30 percent in the first two days of the event. It was also observed that the majority of fatalities were older citizens 65 years and above and a disproportionate number living alone. Satellite24 images during the height of the heat wave in Europe registered many areas being almost 11⁰C (20⁰F) above normal temperatures for the same period.

Furthermore, the Centers for Disease Control (CDC, 2004) estimates that from 1979 to

1999, excessive heat exposure contributed to more than 8,000 premature deaths in the

United States exceeding mortalities resulting from hurricanes, lightning, tornadoes, floods, and earthquakes combined.

Furthermore, the UHI impacts urban and aquatic ecosystems. The barren construction techniques that foster heat islands tend to be unattractive, unappealing and unhealthy for urban flora and fauna (Gartland 2008). Increased temperatures tend to foster early plant bloom and lead in many instances to extinction of local species or the prevalence of invasive species that are hardier to warmer temperatures. Aquatic ecosystems are degraded by surface UHI by thermal pollution. Surface run-off from pavement and roofs tends to be higher in temperature by 27⁰C-50⁰C than air temperatures

(EPA, 2008). Field measurements from one study showed that runoff from urban areas

24 Temperature anomalies (degrees below or above normal) in Europe on August 31, 2003, NASA.

105 was about 11°C -17°C hotter than runoff from a nearby rural area on summer days when

pavement temperatures at midday were 11°C -19°C above air temperature. When the rain

came before the pavement had a chance to heat up, runoff temperatures from the rural

and urban areas differed by less than 2°C (Roa-Espinosa et al., 2003). This excess heat is

transferred quickly downstream by conventional conveyance systems effecting the

metabolism and reproduction of many aquatic species (EPA, 2008).

The UHI is an urban specific phenomenon. Its impacts affect people, other species

and the environment. Strategies to reduce urban heat islands produce multiple benefits

that lower surface and air temperatures thus reducing health risks, energy demand, air

pollution and greenhouse gas emissions. The UHI compares in its impacts to global

climate change at the local scale. Therefore advancing measures to mitigate the UHI also

address adaptation to global climate change impacts (EPA, 2008; Stone, 2005, 2012,

2012a).

5.3 Effective Vegetated Green Infrastructure Measures for the Urban Heat Island

In environmental and urban planning, measures to mitigate the UHI are centered

on land-based measures that change the urban land cover (EPA; Akbari, 2003; Stone,

2012). Properties of urban materials, in particular low solar reflectance, high thermal

emissivity, and high heat capacity, influence urban heat island development, as they determine how the sun’s energy is reflected, emitted, and absorbed, and consequently radiated back into the air (Oke, 1988, 1997; Akbari, et al., 2001; Akbari, et al., 2003;

Rosenzweig et al., 2006; EPA, 2008; Stone, 2005,2012). Measures that disrupt this process of energy absorption, storage and release of heat into the urban atmosphere are predominantly addressed in three ways: 1) changing characteristics of materials to

106 increase albedo by increasing reflectivity or absorption of water to reduce latent heat; 2)

obstructing the UHI process by blocking the sun’s energy through shading or increasing

cooling by ensuring that water in its multiple forms is present in ample quantities in the

soil and air; and 3) technological advances that reduce heat waste from electrical and

mechanical equipment (Akbari, 2001). These measures include cool roofs; cool

pavements; green roofs and facades; and increasing vegetative cover in all its forms25.

Experimental and modeling studies of land-based mitigation strategies have found that the combination of several measures can slow warming trends when implemented extensively throughout urbanized regions (Stone, 2012). Variable combinations of tree planting and vegetative cover (including green roofs), albedo enhancement of surface materials, and reductions in waste heat emissions were found to reduce city-wide air temperatures from 1⁰C to 7⁰C (2⁰F and 13⁰F) (Kikegawa, Genchi,Kondo, & Hanaki,

2006; Lynn et al., 2009; Rosenzweig, Solecki, & Slosberg, 2006; Taha, 1997; Zhou &

Shepherd, 2010). Of the three classes of land-based UHI mitigation, tree planting and other vegetative strategies are generally found to be the most effective, with surface reflectivity and waste heat strategies typically accounting for lower reductions in near surface air temperatures, depending upon the spatial extent of coverage and the regional landscape type (Rosenzweig, Solecki, & Slosberg; Gill et al., 2007; Hart & Sailor, 2009;

Lynn et al., 2009; Zhou & Shepherd, 2010). Furthermore, vegetated green infrastructure is a relatively inexpensive measure to install and maintain in the long run (Akbari, 2005).

This is true when considering other benefits that accrue when using vegetative green

25 Akbari (2005) provides a detailed discussion on each of three strategies.

107 infrastructure and the presence of administrative bodies in cities that already manage

urban greening projects.

Vegetated green infrastructure reduces surface and air temperatures due to a

combination of physical and biological properties (Akbari, 2001; EPA). The physical

properties of tree and shrub canopy provide shade to hard surfaces reducing surface

temperatures. By reducing temperatures, evaporation is also reduced maintaining

moisture in the air and soils. The presence of higher moisture content in the air and

surfaces is a critical in moderating temperatures. As a growing medium for vegetation,

soils have lower albedo coefficients reducing reflected energy resulting in reduced air

temperatures in the immediate surroundings. Evapotranspiration is the process of uptake

of water from soils and rain interception into the atmosphere (Jasechko et al., 2013). This

process results in evaporative cooling enhancing the urban breeze in general and cooling

surrounding air (Akbari, 2002; EPA). The process of carbon intake and release of oxygen

vegetation reduces air pollutants and cleans the air (Akbari et al., 2001). The removal of

pollutants reduces entrapment of hot air reducing the impact of the UHI. This process

also allows vegetated cover to become a CO2 sink contributing to the process of climate change mitigation (EPA). These physical and biological processes are further explained in Figure 5.5. To effectively impact the urban climate, vegetated green infrastructure cover should be increased across any area to derive the maximum ecosystem benefit of

heat reduction.

108

Figure 5.5: Diagram showing how increasing physical and biological properties of vegetated green infrastructure decreases the urban island heat.

Individual or groupings of trees, shrubs, and green roofs provide effective measures to reduce surface and air temperatures. The compounded effect of multiplee measures across large areas will effectively increase the chances of temperature reduction. In their simulation study to mitigate UHI in the Ciity of New York,

Rosenzweig et al. (2006) conclude that an “ecological infrastructure, a combined strategy of urban forestry and living roofs, has the greatest city-wide temperature impact (p. S-8)”.

For example, the range of city-wide temperature rreduction of different mitigation scenarios26 ranges is from 0.1⁰C (0.1⁰F) for open space planting (typical park type of

26 Strategies are in bold and UHI mitigation scenarios are enumerated as per Rosenzweig et al. (2006, S-5). Urban Forestry:1) Urban Forestry/Grass-to-Trees (open space planting); 2) Urban Forestry/Street-to-Trees (curbside planting); 3) Urban Forestry/Grass + Street-to-Trees (open space + curbside planting). Light Surfaces: 4) Light Surfaces/Roof-to-High Albedo (light roofs); 5) Light Surfaces/Impervious-to-High Albedo (roofs + sidewalks/streets). Living Roofs: 6) Living RRoofs/Roof-to-Grass. Ecological Infrastructure: 7) Urban Forestry/Grass + Street-to-Trees and Living Roofs. Urban Forestry + Light Roofs: 8) Urban Forestry/Grass + Street-to-Trees and Light

109 planting) to 0.4⁰C (0.7⁰F) for ecological infrastructure (vegetation planted for its

ecosystem services). At three in the afternoon (peak heat time), the impact ranges from

0.1⁰C (0.2⁰F) for open space planting to 0.7⁰C (1.2⁰F) for ecological infrastructure.

While these values may not seem significant when considering that the UHI for New

York is approximately 4⁰C (air temperature), they are averaged over all heat-wave days

and times and therefore generalized. To consider also that these values were determined

from four sample zones that were analyzed in detail and regressed to the whole city. The

actual localized temperature reduction is higher when considering specific local zones

rather than the overall urban area (Rosenzweig et al., 2006).

In the context of using green infrastructure to adapt cities to climate change, Gill et al. (2007) conducted simulation studies of different green infrastructure measures for

the city of Manchester, U.K. using the energy exchange model. Future climate and

several vegetated surface scenarios were developed. The study developed a baseline

condition for the period from 1961 to 1990 with temperature projections for 2020, 2050,

and 2080 with low, medium and high emissions scenarios for each year (IPCC, 2007).

These projections were conducted for different land uses with varying vegetated cover:

forest and agriculture (high evaporative cover), to residential with 66 percent evaporative

cover, and urban core areas at hundred percent built up with no evaporative cover. The

results indicate that increasing vegetative cover (trees and shrubs) and green roofs reduce

surface temperatures, consequently reducing air temperature. For example, in high-

density residential areas, maximum surface temperatures in 1961–1990 with current form

Roofs. Combination of All: 9) 50% Open Space + 50% Curbside + 25% Living Roofs + 25% Light Roofs.

110 are 27.9°C. Adding 10 per cent green cover decreases maximum surface temperatures by

2.2°C in 1961–1990, and 2.4°C to 2.5°C by the 2080s Low and High emissions

scenarios, respectively. Thus, maximum surface temperatures decrease by 0.7°C by the

2080s Low and increase by 1.2°C by the 2080s High, in comparison to the 1961–1990

current form case. Adding green roofs to all buildings in land uses where impervious

surfaces dominate can also have a dramatic effect on maximum surface temperatures,

keeping them below the 1961–1990 current form case for all time periods and emissions

scenarios. The difference made by the green roofs becomes greater with the time period and emissions scenario. For example, in 1961–1990, greening roofs results in maximum surface temperatures of 24.6°C in town centers, a decrease of 6.6°C compared to the current form case of 31.2°C. By the 2080s High, greening roofs in town centers results in

temperatures of 28°C, 7.6°C less than if roofs are not greened and 3.3°C less than the

1961–1990 current form case.

In all the scenarios presented for the city of Manchester, the determinant factor for

temperature reduction is the total amount of vegetated surfaces. Gill et al (2007) also

argue that a network based on landscape ecological principles (patch-corridor-matrix

model developed by Forman (1985) improves the urban watershed hydrology (by storage

and retention of precipitation from intense events) which in turn supports healthy

vegetation and increases evaporative cooling.

Both simulation studies suggest that increasing vegetated cover in its many forms

in the form of a vegetated green infrastructure will have a high impact on the urban

climate within highly urbanized urban contexts. If the same conclusions are extended to

the metropolitan scale by extending the network of vegetated green infrastructure to

111 include metropolitan or regional system, the same would arguably apply. A metropolitan scale network would add to the landscape typologies that would not be otherwise included within an urban core focus. These may include river systems, forests and forest stands, agricultural land, grass land, vegetation within suburban home yards, and natural reserves. While these typologies maybe distant from the highly urbanized centers, the contribution would be in the form of ameliorating the regional climate by providing cleaner air, removing pollutants, cooling air downstream, dissipation of latent heat and the contribution to regulating the urban canopy effect of UHI. Furthermore, allocation of larger tracts of land for green infrastructure beyond the urbanized limit would also provide escape for residents during extreme heat events as well as refuge for other animal species.

When considering future conditions of expanding urban regions and projected climate change impacts, a metropolitan scale network ensures the preservation of tracts of vegetated surfaces for future UHI reduction and climate change adaptation. By planning today for a future metropolitan network, policies that protect, conserve, and preserve forests or agricultural land into the future will ensure that cities continue to be climate proofed into the future. A metropolitan green infrastructure network that extends from the roof of a building to the large tracts of forest and river systems across different land uses would contribute to local UHI mitigation and global climate change adaptation.

5.4 Framework of Strategies to Increase Vegetated Green Infrastructure Surfaces

As an urban phenomenon, the UHI has direct impacts on the livelihood of comminutes at large. The ecosystem benefits derived from vegetated surfaces provide a cost effective and no-regrets approach to climate proof cities. Trees, the urban canopy,

112 vegetated surfaces and green roofs are found to be most effective in reducing surface and air temperatures. Extensive implementation of these measures in the form of a vegetated green infrastructure network provide the basis for effective temperature reduction resulting in reduced mortality rates, energy use, and cleaner air.

These principles summarized above define a theoretical framework (Figure 5.6) that addresses the assessment of availability of space as the primary determinant to achieve a multi-scale vegetated green infrastructure network within urbanized regions.

The primary impetus is to increase the amount of vegetated surfaces to address current

UHI and future potential climatic impacts. The framework defines four strategies that respond to the literature findings and include: conservation and protection; intensification and expansion; transformation; and impact reduction. The framework uses the term pervious surfaces (McPherson, 2012) to suggest un-built areas with soil that potentially can support vegetation (naturally or through amending soils); and impervious surfaces to suggest all built form (buildings, roads, and parking) and considered as the primary delimiter of space. The strategies are devised as planning tools to assess the potential increase of vegetated surfaces within the constraint of space scarcity within urban contexts.

Conservation and protection is the Protection of existing tree canopy such as forest and the conversion of pervious surface such agriculture into forest, when conditions of ownership and use allow. Large stands of trees and forest cover in general, provide multiple benefits. These benefits include reduction of evaporative cooling from pervious surfaces, shade and urban breeze enhancements ameliorate. Large stands of trees enhance local wind patterns in cities where cooler air over vegetated surfaces

113 replaces warmer air in adjacent city neighborhoods (Akbari, 2002). The effective size of

a stand of trees to begin to impact surrounding air is ranges between three to four hectares. In a metadata analysis of the literature on vegetative cooling, Bowler et al.

(2010) found that three hectares is the effective size of a tree stand or urban forest where

for temperature reduction extends beyond the patch size. This impact is reduced as the

distance increases from tree cover location. Rosenzweig et al. (2006, 2011) have also

found that a four hectare tree stand begins to impact surrounding areas. This means that

the more trees are present the more benefits are accrued and the morphology of patches

within a region define the extent of the benefits accrued beyond the stand itself.

Intensification and Expansion is the addition of trees, shrubs, and ground cover to existing pervious surfaces to ensure maximum delivery of ecosystem services. Expansion

is the increase of total pervious surfaces dedicated for vegetated green infrastructure.

Transformation is the replacement of impervious surfaces with pervious surfaces

capable of supporting vegetation and/or the superimposition on impervious surfaces of

vegetal surfaces such as green roofs. Green roofs provide benefits by cooling the upper

limits of the urban canopy and reduce energy consumption in buildings. Green roofs in some cases reduce surface temperature by 30-60°C and ambient (air) temperature by 5°C

when compared to conventional black roofs (EPA). Three examples demonstrate the

effectiveness of green roofs. These are organized by ascending scale: neighbored, city,

and urban region. In Portland, Oregon, a study estimated that a neighborhood with 100%

green roofs could reduce the UHI effects by 50-90 percent (Dunnett and Kingsbury,

2004). Similarly, an Environment Canada study determined that greening 6 percent of

available roof space in the city of Toronto would reduce summer temperatures by 1°C to

114 2°C overall (Ligeti, 2007). Additionally, a study in New York City estimates that a 0.4°C reduction in the regional UHI effect can be achieved with the installation of green roofs on 50 percent eligible roofs across the entire city (Rosenzweig, 2006). While these studies differ in scale, regional landscape type and local climate conditions, the lesson is that green roofs are effective in reducing the surface and air impacts of the UHI when substantial coverage of green roofs is simulated for the scale of each individual study area. It follows that transformation of impervious surfaces is a valid strategy to reduce the

UHI.

Impact reduction is protecting surfaces using vegetated material through by shading impervious surfaces or ensuring all un-built areas are planted. Trees and shrubs provide shading of impervious and pervious surfaces. Shading by vegetation blocks the sun’s energy reducing surface temperature of impervious surfaces, evaporation from pervious surfaces, and energy consumption that reduces heat waste generation. Shading impervious surfaces such as streets, sidewalks and walkways reduce surface temperature transmission of latent heat into the urban canopy layer. The smooth nature of urban materials and usually darker color of concrete and asphalt increase absorption of energy and reduces reflection of the sun’s energy. Shading reduces this absorption and consequently the release of short wave energy during the night. When considering parking surfaces, these are surfaces that radiate large amount of heat to their sheer size and concentration within commercial uses. Davis et al. (2010) found in an estimate of parking footprint in the Illinois region that parking surfaces are overdesigned and could amount to double the actual use of parking spaces during peak hours. McPherson (2001) in a study of a parking shading ordinance in Sacramento, California, that parking lots are

115 6% over designed and 25% of parking spaces are unused. Based on these studies, the potential to increase tree shading and vegetative cover seemss to be possible within parking surfaces.

Figure 5.6: Strategies to increase pervious surfaces.

This framework suggested above defines the overall methodological approach to assess the potential of space availability for vegetated green infrastructure. The followings section develops the method of inquiry in detail within the connstraints of urbanization and the UHI.

116 5.5 Method

The purpose of the method is to identify opportunities of surfaces within the

Metropolitan Region of Boston (MB) that potentially can be designated as a vegetated

green infrastructure. Percent pervious is introduced as a land matric to characterize the

MB and test for availability of space. Maximum opportunities are then derived to address

the UHI as a treatment of the MB extending the notion of pervious surfaces beyond land

areas. Using the framework of strategies discussed in the previous section, the method

tests for the availability of space by identifying opportunities based reducing land cover

variables within the MB to two: Pervious surfaces, representing all un-built surfaces; and

impervious surfaces; representing buildings, roads, and parking surfaces and considered as the primary delimiter of available space. These two variables vary according to land use types and accordingly provide varied opportunities and maximum opportunities across the MB.

The method is grounded in the theoretical Green Infrastructure Transect

(Abunnasr and Hamin, 2012) and the urban-rural characterization and assessment method that connects ecosystem benefits and land-use classes (Gill et al., 2008; Alberti, 2009;

Radford and James, 2013). Using readily available spatial data sets from Mass GIS

(geographic information systems spatial data portal for the state of Massachusetts), the

following steps were carried out: (1) Define study area, (2) Calculate Percent Pervious,

(3) develop pervious surface data set, (4) characterize the metropolitan area into zones of

percent pervious, (5) assess pervious surface opportunities for vegetated green infrastructure, (6) assess maximum pervious surface opportunities effective for the urban

117 heat island reduction, and (7) define a gradient of green infrastructure policies (based on surface opportunities) suitable to address the urban heat island.

Figure 5.7: Steps included in the method.

5.5.1 Study Area

The Consolidated Metropolitan Statistical Area (CMSA) of Boston,

Massachusetts (MA) is 7,230km2 (1,786,571 Acres) with an estimated population 4.27 million in 2005. The CMSA of Boston is comprised of four Metropolitan Statistical areas

(MSA) that include Boston, Lowell, Lawrence, annd Brocton ((US Census bureau, 2012).

The total number of towns is 161 with an average town/city surface area of 16.6km2 ranging from 3.3km2 for Nahant to 266.2km2 for Plymouth (Figure 5.8) For the remainder of this article, the study area will be referred to as Metropolitan Boston (MB).

MB is an example of a metropolitan area The land use percentage of the total study are in

2005 (reclassified to match 1951-1999 classification) is forest (41%), Urban (33%), water

(18%), agriculture (3%), recreation and open land (2%). When compared to 1971

118 statistics, there is a loss in agriculture of (-54%), forest (-14%) and open land (-77%) and an increase in urban use (18%) and recreation (39%) (Mass GIS). When considering surface cover (reduced to two variables) across all land uses within MB, pervious surface comprises 5,255km2 (72%) and impervious surfaaces 1,131.5kkm2 (16%). When averaged across the whole MB, 16% impervious surfaces may not be considered high27. When considering these percentages across a rural-urban gradient, several zones within the MB will include percentage of imperviousness of greater than 50%. While climatic and precipitation conditions (range of average yearly rainfall between 850 to 1100 mm) support forest cover and population did not exceed three percent increase during the

2000-2010 period, slow but significant loss of forest and agriculture land is evident (Mass

GIS). This reduction in pervious surface supportiive of vegetation has and is resulting in the loss of substantial potential in reducing air and surfaces temperatures.

Figure 5.8: Study Area boundaries

27 A 10% imperviousness within a watershed is considered as the threshold where water quality and consequently, ecological quality is degraded (Brabec et al., 2002).

119 5.5.2 Percent Pervious

Percent pervious is introduced as a metric to assess the availability of space for vegetative green infrastructure. Pervious surface is defined as land surfaces that are covered by soil and include all grass and bare soil (McPherson et al, 2011). Such surffaces can or potentially can be amended to support vegetation. In this study, pervious surface is used as a surrogate to all types of vegetative cover: trees, shrubs, and ground cover.

Impervious surfaces is utilized to represent all built surfaces aand as the primary delimiter of space (Figure 5.9).

Figure 5.9: Reduction of variables of interest to two primary surrogate indicators, pervious and impervious surfaces.

Percent pervious (PP) is defined as the peercentage of pervious surfaces from total land area, excluding water. Water surfaces are excluded as the focus is on vegetative cover that requires soil. Equation (1) defines the calculation method for PP where ‘p’ is total area of pervious surfaces within the study area and, ‘imp’ is the total area of

120 impervious surfaces within the same study area. A secondary named pervious-to-

impervious ratio was also developed to assess the magnitude of the relationship between

pervious and impervious surfaces. This metric was only used in the characterization of

MB as it assisted in defining and refining the PP gradient zones.

∑ PP= x 100% (1) ∑

In the context of metropolitan scale planning, PP becomes a useful metric to

estimate the potential surfaces that can be planted for a specific ecosystem benefit. The

PP is derived from readily available spatial data sets allowing planners with minimal

environmental or ecological knowledge to conduct estimates of the potential vegetated surfaces using current GIS skills and know-how. without utilizing specialized skills and

knowledge required for remote sensing The reduction to a single metric avoids

specialized and sometimes complex processes to calculate, for example, tree canopy

cover or evapotranspiration as a surrogate for vegetated surfaces (Gill et al., 2008;

McPherson, 2012). This is not to say that these methodologies are not relevant. On the

contrary, remote sensing becomes a second step required at a finer scale after the initial

assessment using the PP.

5.5.3 Pervious Data Set

Four data layers corresponding to the most effective land based adaptation

measures to temperature reduction were developed. These land cover based data layers

are: pervious surfaces (Per-poly), roads, buildings and parking. The raw data was

downloaded (from December 2012 to March 2013) free of charge from MassGIS.

121 MassGIS provides extensive metadata explaining methodology, third party data

providers, and support. The reference date of the study is 2005, the date of the aerial

images that were used to develop the impervious surface data layer and land-use classes.

All effort was done to use data layers closest to this date. When this was not available,

removal of additional polygons were masked out using the pervious and impervious

polygon layers discussed later. Impervious surfaces, administrative boundaries,

conservation and recreation open space, forest stewardship, land-use, building footprint,

assessor’s maps level 3 and roads were cropped or compiled to correspond to the MB

boundary. Two main steps were carried: 1) derive polygon data layers for pervious and

impervious surfaces, and 2) transfer attributes from land use and town data to each of the

required data layers.

The pervious and impervious polygon layers were developed from converting the

raster impervious surfaces layer to polygons. Sixteen raster tiles, with one meter pixel

resolution (derived from 50cm resolution aerial imagery) were used to cover the study

area. Each tile was converted to polygons separately to reduce processing time and

aggregated into a single data layer corresponding to the MB boundary. Square edges

resulting from the raster origin were simplified. Polygons less than 1m2 (corresponding to pixel size) were considered as error and were removed to match the resolution of the original data. The conversion process resulted in two sets of polygons within the data layer corresponding to the binary classification28 by MassGIS. Grid code one corresponds

28 Mass GIS classification of impervious and impervious surfaces during data set extraction: Impervious surfaces include: (1) All constructed surfaces such as buildings, roads, parking lots, brick, asphalt, concrete ;( 2) Also included are areas of man-made compacted soil or material such as mining or unpaved parking lots (no vegetation present). Non-impervious surfaces include: (1) All vegetated areas, natural

122 to the polygons representing impervious surface and zero polygons representing pervious surfaces. Each set of polygons was extracted into a separate layer. Due to classification limitations (shadow, dark spots in the aerial imagery) of the original raster layer, some polygons corresponding to orthogonal structures do not correspond to the building footprint configuration. An assessment of this error was carried out by superimposing the footprint polygon over the impervious polygons. Residual polygons falling outside or inside the building footprint polygons were extracted and compared. The area of polygons falling outside the building foot prints were 3% more than the area of the polygons falling within. Accordingly the error margin is minimal and both sets of residual polygons approximately cancel each other. Rectified such inconsistencies would have been time consuming. The inconsistencies were maintained in the data sets.

The attributes of the land-use classes were transferred to the buildings, Per-poly and IMP-poly using GIS analysis tools. The land use layer includes 33 classes which were maintained with no further classification. The same process was not repeated for the roads data layer because when two different land use polygons meet at a street, the adjacency is defined by the road center line. This results in multiple polygons for each road within several land use classes, rendering the data unusable for this study.

Accordingly the road data layer did not include land use designation.

A data layer comprising all 161 towns and cities was compiled and used to transfer town location and data to all four data layers. The result of this data processing is

five data layers, each corresponding to pervious buildings, roads, and parking and the

and man-made; (2) Water bodies and wetland area; (3) Ski runs; (4) Natural occurring barren areas (i.e. rocky shores, sand, bare soil)

123 fifth is the study area including towns. Table 5.1 lists the five data sets and their

acronyms. The overall data set is referred to as pervious-adapt (PER-ADPT).

Data Layer Data Layer Attribute information Scale Name acronym Pervious PER Town name, land use attributes Study area Surfaces Roads RD Town name, land use attributes Study area Buildings BLDG Town name, land use attributes Study area Parking PRKG Town name, land use attributes Study area Study Area SA Town name, land use attributes Table 5.1: Derived data sets used in the study from raw spatial data downloaded from Mass GIS.

5.5.4 Characterizing the Study Area

An urban-rural gradient (hereof referred to as the metropolitan gradient) was

constructed to characterize the study area based on the PP land metric. The aim of the

characterization exercise is to develop gradient zones across the MB to assess the area

variation across the region. The percent impervious has been previously used to define

urban-rural gradients (Gill et al., 2008, Radford and James, 2013) as a measure of extent

of impact of urbanization on ecosystems. The percent pervious (PP) and pervious-to-

impervious (PER: IMP) ratio is explicitly used in the characterization process to highlight

the potential of greening of urban areas and to measure the magnitude of available space

for vegetation. The resulting gradient zones also serve as the unit of study of the

subsequent analysis steps.

The PP and PER: IMP are calculated by deriving the pervious and impervious surface areas for each of the 161 towns and cities. PP is calculated as the percentage pervious surface of total land area, excluding water. PER: IMP was calculated by dividing both entities and used as a measure of space availability for vegetation when

124 compared to impervious surfaces. The town or city boundary rather than the watershed

was considered in these calculations since the focus of the analysis is on vegetation and

not hydrological systems.

The two metrics were assigned to each town then mapped in GIS to determine the

gradient zones. The pervious surface gradient zones were determined by comparing three

interpretations of the data to account for the incremental change of pervious surfaces: (1)

natural breaks within the data for PP and PER: IMP, (2) aggregated frequency count

based on single values of PER:IMP, and (3) the table matching PER:IMP and PP values.

The process of using both metrics allowed for a fine determination of the gradient zones.

To understand the reasons behind the regional distribution, PER: IMP and PP were

compared to population density data and road network configuration. PER: IMP and PP ratios were plotted for each town against distance gradient graphs using the City of

Boston as the origin. The city of Boston is the center of the MB (U.S. Census Bureau,

1999) and where higher temperatures are most felt. The plots were analyzed and compared to similar distance plots of total population density of each town. The gradient map was compared to the road network. Results are discussed in terms of pervious space availability across the gradient and its relationship to impervious surfaces, road network and population density.

5.5.5 Assess Pervious Surface Opportunities

There are four dimensions for the definition of opportunity in this study. First, the un-accounted vegetated surfaces across land uses that are usually not considered part of an ecological system. Second, surfaces within privately-owned land uses, such as residential and commercial, where there is no administrative control (when compared to

125 public parks, for example) on the type and extent of vegetative land cover. Third, large tracts of land, such as forests or wetlands, that are ecologically active but with no means of protection for future permanency. Fourth, surfaces that could become ecologically active vegetative surfaces in the future but are threatened by current and projected urban expansion.

The above definition is applicable to any green infrastructure planning condition.

PP is explored within land use classes and across gradient zones derived in the characterization step.

5.5.5.1 Percent Pervious Across Land-use Classes

Land use classes are first categorized into three general levels of perviousness: highly pervious, pervious and impervious, and highly impervious (predominantly water related uses) (Table 5.2). These categories were defined after carefully studying the metadata of land use classes identified by MassGIS and the methodology that specifies the components included in each land use. All water related land uses are considered impervious and excluded from the analysis ( Brabec, et al., 2002) except for forested wetland which was included in the forest category because of the existence of forest canopy. Water is excluded since the focus of the study is on vegetation that requires soil for growth. Pervious land use classes were further divided into two categories: forest and agriculture. The forest coverage is of high resolution and covers all tree stands of not less

2,000m2. Agricultural patches are considered future opportunity and included in the final analysis. The land use classes that include pervious and impervious surfaces were all included in the analyses of the next step.

126 Land-use Classes Status Land-use Classes Status Impervious only Pervious and Impervious Include Water Exclude Water-Based Recreation Include Non-Forested Wetland Exclude Golf Course Include Saltwater Wetland Exclude Marina Include Cranberry Bog Exclude Multi-Family Residential Include Saltwater Sandy Beach Exclude High Density Residential Include Pervious only Medium Density Include Residential Forest Forest/incl. Low Density Residential Include Forested Wetland Forest/incl. Very Low Density Include Residential Cropland Agriculture/inlc. Urban Public/Institutional Include Pasture Agriculture/inlc. Commercial Include Orchard Agriculture/inlc. Industrial Include Nursery Agriculture/inlc. Transportation Include Brush-land/Successional Agriculture/inlc. Cemetery Include Pervious and Impervious Junkyard Include Power line/Utility Include Mining Include Participation Recreation Include Transitional Include Spectator Recreation Include Waste Disposal Include

Table 5.2: Categorization of perviousness of land-use classes

Land use classes are categorized as highly pervious or pervious-impervious are

analyzed for their potential based on the percent pervious metric. Percent pervious

opportunity is defined as the percentage of unused or unprotected land cover category

from the total of the same category. The analysis is conducted for the four green

infrastructure strategies within each land use class across the six gradient zones. For the

phase of the analysis percent pervious of total land use class per zone is calculated to assess the extent of opportunity contribution of each land use class to the total zone

opportunity.

127 5.5.5.2 Opportunity in Pervious Land-use Classes: Forest and Agriculture

The opportunity is derived by identifying protected and conserved forest and agricultural land. Forest in the land use data layer includes all contiguous stands of trees present across all land uses. Forest may include stands of trees within residential or commercial property.

The assumption is that pervious surfaces that already support vegetation should be protected. Forest and agricultural land that are not under any level of protection are considered potential to conserve. The conservation and recreation open space data layer is used to transfer the ‘level of protection’ attribute to PER. Three protection levels out of

five are considered: ‘perpetuity’, ‘limited’, and ‘term limited’. Although ‘limited’ and

‘term limited’ have time constraints in protection terms, it is assumed that protection will

continue. Forest and agricultural land-use classes that are protected are considered as part

of a functioning green infrastructure network. Unprotected land surfaces are considered future opportunity.

5.5.5.3 Opportunity in Pervious and Impervious Land-use Classes

The opportunity within pervious surfaces in the remaining land-use classes (with

pervious and impervious surfaces) is defined as the percentage of pervious surface that

could possibly be dedicated for an ecological vegetative cover after deducting land area for outdoor people or activity use as defined by the land-use class. To achieve this, further classification of land use classes was carried out based on PP and average pervious patch size within each land use class across each gradient zone. PP of each land use class measures the extent of contribution of each land use class to the total pervious surface within the gradient zone. In addition, pervious patch size is an indicator of

128 availability of space based on tree planting standards resulting from the land use class.

Both metrics were calculated and compared for each land use class and across gradient

zones. Six categories were identified based on these metrics: intense urban use,

residential, recreation, urban public/institutional, service, and transitional (Table 5.3).

Land-use classes were then assigned a coefficient of use (CU) and a coefficient of tree

planting (CT). CU is a measure of how much of the pervious surface is used by people or

the activity as determined by the land-use class. It ranges from zero to one, where one

indicates that all pervious surfaces can be allocated for vegetative cover and zero

indicates no potential for increasing vegetative cover where all pervious surfaces are

dedicated to the land-use activity.

To determine the CU, pervious data layers is spatially related to the assessor’s

maps level 3 (Mass GIS) to simultaneously include the land use attribute with the

property boundaries. This procedure allows the identification of land use classes within

property boundaries. The resultant layer was superimposed over the same aerial images

used to classify the land use data to visually inspect and describe space available within

each land use. A first quick inspection identified pervious surfaces within five land use

classes with evident CU values of either zero or one. ‘Transitional’ land use class as it

describes properties in transition from one land use class to another.

The remainder sixteen land use classes were then closely inspected and measured

using GIS. This was carried out in a systematic manner by laying a 0.25 km2 (Radford

and James. 2013) grid across each gradient zone with each quadrant . The CT was then

calculated and assigned for each land use class.

129 The mature tree size allocation was assigned for each land use class based on three sizes used by McPherson et al. (2011): small (4.6 m crown diameter), medium

(9.1m crown diameter) and large (5.2m crown diameter) requiring a minimum pervious surface footprint for soil of 1.5, 3.3, and 9.3 m2, respectively. Each size was allocated for each land use class based on the average patch size derived for each land use class for each gradient zone. The total qualifying potential PP for each land use across each gradient zone was calculated based on these criteria and compared.

Land use Categories CU CT Intense urban use: low use – low % pervious Industrial 1.0 1.0 Junkyard 0.3 0.3 Mining 0 0.0 Transportation 0.2 0.2 Commercial 1.0 1.0 Marina 0.1 0.1 Residential : Medium to high use -high % pervious High Density Residential 0.1 0.1 Multi-Family Residential 0.2 0.2 Medium Density Residential 0.2 0.2 Low Density Residential 0.3 0.3 Very Low Density Residential 0.4 0.4 Urban Public/Institutional: High use intensity - medium %pervious Urban Public/Institutional 0.5 0.5 Recreation: High use Intensity - high %pervious Participation Recreation 0.1 0.1 Spectator Recreation 1 1.0 Water-Based Recreation 0.15 0.15 Golf Course 0.15 0.15 Service: Low use intensity - high % pervious Cemetery 0.2 0.2 Nursery 0.15 0.15 Open Land 0.8 0.8 Powerline/Utility 0 0 Waste Disposal 0.15 0.15 Excluded Transitional n/a n/a Table 5.3: Classification scheme of coefficient of use (CU) and coefficient of tree planting (CT) for each land use class.

130 5.5.6 Assess Maximum Opportunities of Landover for the Urban Heat Island

Following the definition of opportunity in section 5.5.5, maximum opportunity is defined as the maximum possible pervious area (land surfaces and transformed impervious surfaces) that can be dedicated for vegetative cover based on the effective measures of a specific condition. In this context, this means green infrastructure measures

that are effective for UHI as identified in the literature review. In addition, pervious

surfaces are not only limited for pervious land surfaces, but also include green roofs and

surface transformations of impervious surfaces to pervious. Then maximum opportunity

is identified as pervious surfaces that can realistically be transformed into a vegetative

cover network that are effective to reduce the UHI.

5.5.6.1 Patches of Vegetated Forest Surfaces

Maximum opportunity in these two land-use classes is identified based on the

effective patch size of pervious surfaces. Further selection of forest patch size based on a

3Hectare (30,000m2) forest patch size is considered a priority for protection since

contiguous patches of trees begin to impact temperatures of surrounding areas

(Rosenzweig et al., 2006; Bowler et al., 2010; EPA). Total qualifying areas for forest and

agricultural land were calculated and percent pervious was derived. The percent

opportunity was graphed for each zone and compared. Pervious surfaces within the forest

land use class that comply with the identified size and are protected are considered as functioning infrastructure. Patches that comply with the size requirement but not are protected are considered as maximum potential opportunity, similarly for agriculture.

131 5.5.6.2 Green Roofs across land-use classes

Green roof opportunities were identified based on roof shape and footprint size.

Assessors’ maps provide the ‘building style’ attribute. These attributes were transferred

to the BLDG data layer by spatial relationship. All building styles that have pitched or curved roofs were discarded as being unsuitable for green roofs (cape cod, bungalow, colonial, conventional, raised cape, raised ranch, ranch/split, ranch gabled, and three family). As a result, building footprints within all residential classes were excluded expect for high density. From the remaining land use classes, buildings with footprints less than 50m2 were discarded as these become ineffective and uneconomical when

accounting for 30% of the roof area for building equipment and other service

requirements (LEED). Buildings in the following land use classes were considered as

opportunities for green roofs: commercial, golf courses, high density residential,

industrial, marina, passive recreation, transport, urban public/institutional, and water

based recreation. Total area of qualifying buildings was calculated and multiplied by a

coefficient of 0.7 to account for roof equipment and services. Percent of qualifying

building area of total buildings was calculated and average foot print size. Results were

plotted and compared.

5.5.6.3 Pervious Surfaces Across Road Categories

Road opportunities are defined as the potential to transform impervious surfaces

to green streets and the extent of increasing tree cover within easements of roads and

streets. The analysis of potential opportunities within roads is conducted at the gradient

zone level and not land use classes due to limitations in the land use data layer, as

previously explained.

132 The selection of streets or roads appropriate for each green infrastructure measure are based on the type of use identified by the Massachusetts Department of

Transportation (MassDOT) and dimensional requirements for healthy tree planting

discussed in section 2.4.4. There are six primary types: 1) limited access highway, 2)

Multi-lane Highway (no limited access) 3) Other numbered highway, 4) Major Road

Arterial, 5) Minor Road Arterial, 6) Ramp. The street types are also an indicator of speed of travel in decreasing order form type one to six.

Road types from four to six were selected for tree planting intensification within easements. The lower speeds (<45miles/hr.) with adjacent dense tree cover may not hinder driving safety. The selected road types were further selected by the available width of pervious surface within the easement and the suitability for tree planting. The available pervious surface was calculated by subtracting the width of the impervious surface area

(car travel surface, shoulders, and sidewalks) from the easement width. Roads were further selected by excluding pervious surfaces within easement that are less than two meters, based on large tree pit requirements. Area of pervious surface opportunity was calculated by multiplying the width of the easement by the total length of each street.

Total areas for each gradient zone were calculated to derive percent pervious and percent of qualifying roads of each road type.

5.5.6.4 Parking Surface Opportunities Across Land-use Classes

Parking opportunities are defined as the potential to transform paved surfaces into planting areas for surface transformation or shade trees. Pervious surfaces within parking spaces are accounted for within the land use types of pervious surfaces discussed previously. The percentage value that is used to estimate potential transformation is 30%

133 of total impervious parking area. This is a measure is conservative when compared to the

35% vacant parking lots during peak hours and 6% over ordinance requirements

(McPherson et al., 2001) in the area of Sacramento. Similarly, Davis et al. (2010) estimated across four northern mid states that available parking spaces are double the number of actual cars in use.

Further processing of the PKG was required. The easement width buffered RD, buffered train; BLDG layers were used to spatially erase the corresponding impervious elements. The airport point layer from MassDOT was used to identify airport locations.

These were removed by mask. The remainder impervious polygons include small property driveways and parking lots. The 10 Ha (10,000m 2) minimum threshold area

sizes that reflects large parking area was arrived at after 26 test iterations to remove

smaller polygon sizes. The 30% percent metric was then applied to all parking polygons.

Percent pervious of land use class and average parking lot size were calculated, graphed

and compared.

5.5.7 Gradient of Green Infrastructure Strategies

Area calculations of opportunity pervious surface inform the gradient of

opportunities by matching green infrastructure measures with PP of within land-use

classes and across gradient zones. Areas of pervious surface opportunities in each land

use class are assigned a corresponding planning code based on Table 5.4. It is necessary

to conduct this step to be able to sum up total opportunities for vegetated green

infrastructure policies. All PP values for each land use class within each gradient zone are

summed up based on the same code designation. Total PP for each zone is calculated for

each green infrastructure strategy. The values are then assigned to gradient zone

134 boundaries and mapped. A series of maps and graphs for each green infrastructure measure within each gradient zone are analyzed and compared. A final map output is generated that sums total strategy opportunities based on space availability is discussed.

GI Strategies GI Measures Contribution to Code temperature reduction Conserve Increase Protection: Maintain Cooling of urban breeze C existing forest tree canopy and reduce surface protect existing privately owned temperature; maintain forested land, especially patches of moisture in soil and air. 4Hecatre area. Expand Four Hectare patches‐Contiguous Cooling of urban breeze E Contiguous intensification: Establish new reduce surface patches forest cover on existing pervious temperature; maintain surfaces such as open land or moisture in soil and air. agricultural land (when opportunity arises). Pursue aggressively to increase tree cover expanse Intensify PER: Use Factor‐Intensify: Cooling of urban breeze I Increase tree canopy within reduce surface pervious surfaces in private temperature; maintain property and street easements moisture in soil and air. Transform Green roofs, green streets Decrease albedo, reduce T impervious energy use and cool to pervious upper urban canopy Reduce Street and Parking shading: Provide shade and R impact Provide pervious surfaces within decrease albedo. parking, streets and residential property to reduce impact of UHI

Table 5.4: Relationship between planning strategies, green infrastructure strategies, and contribution to temperature reduction

5.6 Results

5.6.1 Gradient Zones of Percent Pervious Across Study Area

The characterization of the MB resulted in six zones (Figure 5.10) based on the gradation of percent pervious metric. Zone one is characterized by pervious surfaces that are less than fifty percent of total land area. Zones two to five increase in increments of

135 ten percent. PP in zone six is greater than ninety percent. Table 5.5 provides a listing of the six zones and a comparison to three other studies. The perrcent impervious was used to characterize the urban area of Manchester in the UK (Gill et aal., 2008; Radford and

James, 2013) and an urban area in Illinois (Davies et al., 2010). The values from the studies were reversed to show percent pervious for comparison. The six zones include: intense urban, urban core, urban, sub-urban, peri-urban, and semi-rural. Since the objective is to test for availably of space and provide maximum benefits of green infrastructure, the increments of ten percent between zones provide a reliable measure of the possible opportunities for vegetation increasee.

Figure 5.10: Six Gradient Zones by percent pervious and pervious-to-impervious ratio

136 Characterization Results Comparable Studies Zone Name Pervious-to- Percent Gill et al. (2008) - Davis et al. (2010) impervious Pervious Radford and Zone (PER:MP) (PP) James (2013) 1 Intense PER: IMP≤1PP≤50% PP ≤50% Urban PP≤25% High Urban urban 2 Urban core 1.090% PP>95% Rural PP>97% rural IMP>9.0 Table 5.5: Results of the Boston Metropolitan Area characterization and comparison to similar studies.

When comparing the distribution of the six PP gradient zones across the metropolitan area, there is a clear and direct relationship with population density and major road network layout distributions (Figure 5.11). The lowest percent pervious values (Intense urban, urban core, and urban) correspond to the higher population densities within and immediate surroundings of major urban concentrations such as the

Cities of Boston and Lowell. The mid-range values of percent pervious (urban, suburban, and peri-urban) correspond to medium density population concentrations along major interstate highways. For example, towns along the I-90 corridor and the I-93 towards Lawrence and MA-Route 3 towards Lowell are clear areas in transition. While the Boston Metropolitan Area may be characterized as an urban core with low pervious surface concentration, there is a clear indication that the whole region is becoming more urbanized with loss of clear boundaries between urban and non-urban land, a characteristic of the New England Towns. Some anomalies exist such as the towns of

137 Holliston and Dover south of I-90 and Carlisle north of I-90, where high percent pervious values exist within medium or low percent pervious values. This may be attributed to the fact that these towns are bedroom communities, with high level of conservation and protection of land. In general, the semi-rural regions are located on the peripheral edge of the study area either towards the north-western perimeters, north-eastern coast and the southern area bordering Rohde Island.

Figure 5.11: Metropolitan gradient zones comparred to population density an major road network

5.6.2 Pervious and Impervious Surfaces within All Land-use classes and Across Gradient Zones

The overall land cover for MB is characterized by 72 per cent pervious surfaces,

16 percent impervious and 12 percent water surfaces of the total study area. The overall

PP for the study area is 80 percent. While these percentage values suggest a predominance of pervious surfaces indicating a high potential for vegetation cover

138 increase, plotting the PP for each town across the zones offers a finer interpretation of the

PP distribution along the urban-rural gradient (Fiigure 5. 12). Pervious surface cover is generally increasing as distance increases from Boston city, the center of MB (Census

Bureau, 2012). The pervious and impervious surfaces are negatively correlated across the gradient zones. Pervious surfaces range from 38 percent in zone one to 79 percent in zone six and impervious surface is at 57% for zone one and 7% in Zone 6.

Figure 5. 12: Distribution of pervious, impervious and water surfaces acrross the urban- rural gradient. Of the land uses that are totally pervious in surface, 49.5 percent is forest and brush-land and 3.2 percent are agriculture. The remaining 20.3 per cent of pervious surfaces are distributed across the remaining land use classes. The total impervious surfaces (16 percent of total MOB area) are composed of 48 percent roads (including sidewalks) and

24 per cent building footprint and the remainder 28 percent is parking spaaces and driveways.

Pervious surfaces across the 33 land uses (LU; Figure 5.13) show a varied distribution of PP within land use classes and across gradient zones. The results indicate the three highest ranking LUs with highest PP values are low, medium and high density

139 residential. The sum of PP values these three LUs is increasing from zone 1 to zone 6.

The lowest ranking LUs with lowest PP values less than 0.75% across all zones are marina, mining, nursery, orchard, transitional, waste disposal, and water based recreation.

The highest three ranking PP values for LUs across the six zones are reported here. The sum of the first three ranking PP values within LU classes is 49 % for zone 1, 62% for zone 2, 46% for zone 3, 56% for zone 4, 64% for zone 5, and 49% for zone 6.

The first three ranking PP values for LUs include multi-family residential (27%), high density residential (13%) and urban public/institutional (9%) for zone 1; high density residential (26%), multi-family residential (20%)), and forest (16%) for zone 2; forest

(20%), medium density residential (20%), high density residential (16%) for zone 3; forest (31%), medium density residential (18%), and high density residential (7%) for zone 4 forest (44%), low density residential (11%), and forested wetland (9%) for zone

5; and forest (53%), forested wetland (10%), andd low density residential (7%) for zone 6.

Figure 5.13: Distribution of percent pervious across land uses within the MB.

140 When analyzing the data on impervious surfaces, the distribution of road area is constantly increasing across the gradient zones with zones five and six demonstrating a

sharp decrease of six and three percent from zone four, respectively. Zone one is 28

percent covered with road and side walk surfaces and zones two, three, and four are at 22

percent, 17percent and 11percent respectively. Across all gradient zones, the percentage

of road surface area is half the total impervious within each zone.

Results for buildings show that the trend of building footprint cover is decreasing across

the gradient with a range of 6 to 1 percent from zone one to zone six, respectively. When

compared to the total impervious surface area within each zone, surface area of buildings

shows an increasing trend from zone one to zone four ranging between 10 percent to 22

percent, respectively; while zones five and six are close in value, 15 and 14 percent,

respectively; still, both show a higher value by 5 and 3 percent than zones one and two,

respectively.

5.6.3 Pervious Surface Opportunities Across Land-uses

The results for pervious surfaces across land uses and across zones indicate that

residential land use types (all densities) offer the highest presence of pervious surfaces

followed by recreation and urban public/institutional land-uses, open land and then

functions related to commercial and industry. Land-uses with total pervious surfaces are

not discussed in this section. Opportunities are defined within land-uses that include

pervious and impervious land covers.

141 The first three ranking classes are reported here and any additional land use types that are of significance are also mentioned (

Figure 5.14). The sum of percentage coverage of qualifying pervious surface area for zone 1 is 4.2% distributed as follows: minor rroad arterial (1.5%), limited access highway (1.3%), and other numbered highway (11.3%). The suum of percentage coverage of qualifying road area for zone 2 is 4.8% distributed as follows: minor road arterial

(1.9%), limited access highway (6.5%), and major road arterial (1.2%). The sum of percentage coverage of qualifying road area for zone 3 is 8.4% distributed as follows: minor road arterial (6.5%), major road arterial (1..5%), and limited access highway

(0.44%). The sum of percentage coverage of qualifying road area for zone 4 is 3.35% distributed as follows: minor road arterial (1.5%), limited acccess highway (1.4%), and major road arterial (0.44%). The sum of percentage coverage of qualifying road area for zone 5 is 2.7% distributed as follows: minor road arterial (2.1%), other numbered highway (0.3%), and major road arterial (0.0.3%). The sum of percentage coverage of qualifying road area for zone 6 is 1.5% distributeed as follows: minor road arterial (1.9%), limited access highway (6.5%), and major road arrterial (1.2%).

142

Figure 5.14: Percent pervious surfaces across land-use types within each gradient zone and ranking of the highest three values

5.6.4 Maximum Opportunities of Pervious Suurfaces for the Urban Heat Island

5.6.4.1 Preserve and Intensify: Large Patches of Contiguoous Pervious land Surfaces

Large patches of four hectares or more proovide the opportunity to impact the surrounding ambient temperatures. The results suggest that low, medium, and to some extent, high density residential land-use classes provide the highest opportunity across all gradient zones. These are followed by golf courses and participation recreation. The values for PP that correspond to the four hectare patches range from zero percent for industrial, junkyard, mining, marinas, water-based recreation across all zones to a maximum value 11% for medium density residential in zone two.

The first three ranking land use classes foor the PP values that correspond to the four hectare patches are reported here (Figure 5.15). For gradient zone one participation recreation (1.1%) and transportation (0.7%) rank first and second respectively. High

143 density residential, golf courses, and open land rank third at an equal PP value of 0.63%.

Zone two land-use classes are high density residential (4.4%), followed by golf courses

(1.1%) and participation recreation (0.8%). Zone three is dominated by residential land

use classes and includes medium density (6.2%), high density (2.95), and low density

(0.8%). Similarly, residential land uses dominate zone four and include medium density

(11.1%), low density (3.3%), and high density (1.9%). Zone five land uses classes are

low density residential (9.5%), medium density residential (4.9%), and golf courses at

1.7%. Similarly, zone six is dominated by residential and golf course land uses classes.

The values are low density residential (5.0%), very low density residential (1.9%), and

golf courses at 1.4%.

What is surprising about these results is that in zones one and two, where PP is

lowest, four hectare patch opportunities reside in public or recreation land use classes.

This is an unexpected result since the characterization of zones by PP indicates that the

zones one and two have the least available space. It is worthwhile noting that the reported

values for zones one and two still represent a small fraction of the total land area. In zones three and four where middle PP values are established, opportunities reside in residential land-use classes and for zone six low residential and golf courses provide the highest opportunity percentages. In addition, zones three to six show a significantly higher PP value of four hectare patches when compared to zones two and three.

144

Figure 5.15: Percent pervious values for four hectare patches across land uses and their ranking cross the six zones.

5.6.4.2 Intensify: Patches of Pervious Land Surfaces Across Land-use classes when Applying Coefficient of Use

Opportunities to intensify vegetative cover exist across all land uses. By applying the use coefficient to each qualifying land use class, net PP provides the opportunity to intensify vegetative cover without jeopardizing the outdoor space use of the land-use class. The overall results indicate that residential classes rank first across all gradient zones followed by commercial and open land. The results for the first three ranking classes are reported here (Figure 5.16).

PP values for multi-family residential class (2.1%) dominates in zone one followed by open land (0.9%) and industrial (0.8%). Zone two is somewhat similar with

145 multi-family residential class (2.0%) followed by commercial (1.5%). Rank three is shared by industrial (1.0%) and public/institutional (1.0%) land use classes. Zone three is ranked by medium density residential (1.9%) and commercial (1.6%) second. Rank three is shared by industrial (0.9%) and public/institutional (0.9%) land use classes. Zone four includes medium density residential (1.9%) in first rank, commercial (1.6%) in second rank and the third rank shared by low density residential (1.4%) and commercial (1.5%).

Zone five includes low density residential (4.7%) in first rank, open land (1.8%) in second rank, and very low density residential (1.2%) in third rank. Zone six is similar to zone five in land uses classes but differs in rank order: low density residential (3.2%), very low density residential (2.6%), and open land (2%).

Figure 5.16: Comparative table and ranking of PP as defined by use-coefficient across land-use classes.

5.6.4.3 Transforming Building Roofs to Greenn Roofs

146 Out of all the land use categories in the study area, building footprints in nine land

uses classes qualify for surface transformation to green roofs (Figure 5.17). The land-use

classes are: commercial, high density residential, urban public/institutional, industrial,

transportation, participation recreation, marina, water based recreation, and golf courses.

These land uses are dominated with large footprints and mostly flat roofs. Some land use

types such as high density residential, golf courses, marina, and water based recreation

may include buildings with pitched roofs and especially in the zones more distant from

the center. With the available data, it was not possible to determine roof type of every

single building.

The results indicate that building foot print areas across land uses are dominated

by commercial in zone 1, high density residential in zones two, three, and four; and with

industrial buildings in zones five and six significance reduction across zones. The first

three ranking area percentages are reported as percentage of total zone area. 21.4 per cent

is the sum of the highest three LUs for zone 1 and multi-family residential (10.4%), high

density residential (7.6%) and urban public/institutional (3.5%). 31 percent is the sum of

the highest three LUs for zone 2 and include high density residential (18.6%), multi- family residential (10%), and participation recreation (1.1%). 33 percent is the total of the highest three LUs for zone 3 and include medium density residential (15.5%), high density residential (12.5%), multi-family residential (5.2%). 35 percent is the total of the highest three LUs for zone 4 and include medium density residential (20%), low density residential (8%), and high density residential (7%). 32 percent is the total of the highest three LUs for zone 5 and include low density residential (19%), medium density residential (9.6%), and low density residential (3.2%). 27 percent is the total of the

147 highest three LUs for zone 6 and include low density residenttial (15.6%), urban very low residential (8.3%), and open land (3%).

Figure 5.17: Qualifying potential building roofs for green roof implementation

5.6.4.4 Intensify and Shade for Trees: Pervious Surface Opportunities in Road Easements

The data shows that the overall opportuniities for pervious surfaces within road easements is predominantly in class five roads (minor road arterials) followed by class one roads (limited access highways). The first three ranking classes are reported here

(Figure 5.18). The sum of percentage coverage of qualifying road area for zone 1 is 4.2% distributed as follows: minor road arterial (1.5%), limited acccess highway (1.3%), and other numbered highway (1.3%). The sum of perccentage coverage of qualifying road area for zone 2 is 4.8% distributed as follows: minor rroad arterial (1.9%), limited access highway (6.5%), and major road arterial (1.2%). TThe sum of percentage coverage of qualifying road area for zone 3 is 8.4% distributeed as follows: minor road arterial (6.5%),

148 major road arterial (1.5%), and limited access highway (0.44%%). The sum of percentage coverage of qualifying road area for zone 4 is 3.35% distributed as follows: minor road arterial (1.5%), limited access highway (1.4%), and major road arterial (0.44%). The sum of percentage coverage of qualifying road area for zone 5 is 2.7% distributed as follows: minor road arterial (2.1%), other numbered highway (0.3%), and major road arterial

(0.0.3%). The sum of percentage coverage of qualifying road area for zone 6 is 1.5% distributed as follows: minor road arterial (1.9%), limited acccess highway (6.5%), and major road arterial (1.2%).

Figure 5.18: Copmartive duagram and table of PP potential within road easements categorized by road class

5.6.4.5 Transform Roads to Vegetated Surfaces

The roads that qualify for transformation to pervious surfaces all land uses and gradient zones are within the lowest speed category, namely, minor arterial. The results indicate that class five has the highest potential accross all six zones (Figure 5.19). The

149 qualifying road impervious surface from zone onne to six are 4.3%, 8.7%,, 1.8%, 3.9%, 1, and1% , respectively. When accounting for half the road surfaces that possibly may become pervious, these percentage values are reduced by hallf as follows, from zone one to six: 2.2%, 4.3%, 0.9%, 2%, 0.5%, and0.5%, respectively. For this to acctually occur, a traffic study at multiple scales needs to conduct.

Figure 5.19: Class five roads (local arterial) that could potentially be transformed to green streets

5.6.4.6 Transform for Tree Shading: Pervious Surface Opportunities in Parking Surfaces

Similar to building footprints, the qualifying parking spaces are within five land use types. The results indicate that parking surfaces within commercial and industrial land uses across all gradient zones are the highest contributors, followed by urban/public institutional and transportation. Spectator recreation registered the lowest possible contribution across all zones.

The first three ranking classes are reported here (Figure 5.20). The sum of percentage coverage of qualifying road area for zone 1 is 5.8% distributed as follows: transportation (3.3%), commercial (2.8%), and industrial (2.7%), and The sum of percentage coverage of qualifying road area for zone 2 is 5.5% distributed as follows:

150 industrial (2.5%), commercial (2%), and urban public/institutional (0.89%). The sum of percentage coverage of qualifying road area for zone 3 is 6.4% distributed as follows: commercial (3.2%), industrial (2.2%), and public institutional (1.09%). The sum of percentage coverage of qualifying road area for zone 4 is 5.5% distributed as follows: commercial (2.6%), industrial (2%), and spectator recreation (0.94%). The sum of percentage coverage of qualifying road area for zone 5 is 3% distributed as follows: industrial (1.3%), commercial (1.2%), and urban public/institutional (0.6%). The sum of percentage coverage of qualifying road area for zone 6 is 1.3% distributed as follows: urban public/institutional (0.5%), industrial (0.4%), and commercial (0.35%).

Figure 5.20: Opportunities to transform impervious surfaces tto vegetated surfaces within qualifying landn -use classes

151 5.6.5 Gradient of Green Infrastructure Policy Opportunities

When summing all PP values for the different strategies within each land use

class for each gradient zone, the extent of opportunity for each green infrastructure

matches surface temperature distribution for UHI for MB. The results indicate that

opportunity for the different measures varies differently across gradient zones. The total

PP value for four hectare patches increases from zone 1 (5.1%) to zone 3 (15.2%), levels

out for zones 4 and 5 at 21%, and dips for zone 6 (13.3%). For opportunities across

pervious surfaces when applying the use-coefficient, PP values are somewhat level for all

zones ranging from 8.7% for zone 1 to 10% for zones 5 and six. Intensifying tree canopy

within road easements shows a sharp increase in zone 3 with lower values for zones 1 and

6. The PP value slightly increases from zone 1(4.1%) to zone 2 (4.8%), then sharply

reaches 8.4% in zone 3 with sharp drop in zone 4 (3.3%) value reaching 1.5% for zone 6.

When considering transforming local arterial roads to green streets, PP values doubles from zone 1 (4.1%) to zone 2 (8.7%). A gradual decrease is then observed from zone 3

(2%) to 1% in zone 6. Green roof show a gradual decrease from zone 1 (9.9%) to zone 6

(0.3%) with a single anomaly for zone 3 at 23.4%. PP values for transforming impervious surfaces to pervious surfaces impact reduction gradually and consistently decreases from zone 1(10.6%) to zone 6 (1.4%) (Figure 5.22)

When compared to the UHI, green roofs, parking transformation and intensifying vegetation within pervious surfaces based on use coefficient provide the highest opportunity to mitigate the UHI within zone 1. Four hectare patches, transforming roads to pervious surfaces, and intensifying vegetation within pervious surfaces based on use- coefficient show highest potential for zone 2. Four hectare patches, intensifying

152 vegetation within pervious surfaces based on use--coefficient, and green roofs provide the highest opportunities across zone 3. Four hectare patches, intensifying vegetation within pervious surfaces based on use-coefficient and transforming parking surfaces show the highest potential for increasing vegetation cover. Opportunities for zones 5 and 6 reside in four hectare patches, intensifying vegetation within pervious surfaces based on use- coefficient, and vegetation within road easements (Figure 5.21).

Figure 5.21: Spatial distribution of each green infrastructure measure type across the gradient zones compared to the urban heat islandd distribution for MB. From top left, clockwise: The urban heat island of Boston City ( source: Urban Heat Islands, http://www.urbanheatislands.com/); three hectare patches; opportunity by coefficient of use; opportunity for parking shading and transformation opportunities; opportunities for road easement transformation; and opportunities for green roofs.

153 Strategies in general may be categorized as follows: inntensifying vegetation within pervious surfaces based on use-coefficient shows opportunity across all zones and has been ranked as first or second. Four hectare patches provvide the second highest frequency across five zones (zone 2 to zone 6), and the remainder is distributed across all zones. What is of significance is that where the intensity of the UHI (zones 1 to 3) transformation of impervious surfaces (roads, buildings and parking) provide substantial opportunities.

Figure 5.22: Opportunities for green infrastructure measures across gradient zones.

When considering adaptation to climate change, accounting for forest and agricultural land uses becomes a future opportunity and challenge to maintain in the near and distant future. When including forest and agriculture in the results, there is a clear indication (Figure 5.23) that forest cover will play a significant role in climate proofing

Boston., as the PP values consistently are increassing from zone 1(6.2%) to zone 2 (67%).

154 the priority here would be to devise policies that protect and preserve these land uses into the future.

Figure 5.23: Opportunities for green infrastructure measures across gradient zones including forest and agricultural land use classes.

When considering which land uses provide the highest opportunities for the different green infrastructure strategies, residential uses are highest in rank (Table 5.6).

The results also allow for the prioritization of policy across gradient zones based on space availability within each zone (Table 5.7)

Green Infrastructure Policies Opportunity Land-uuses Contiguous tree patches All residential categoories Tree intensification All residential, commercial, open land Green Roofs High density residential, commercial, industrial Street shading and green streets Class 5, local roads Parking tree shading & Transform Industrial and commmercial Table 5.6: Highest provision of opportunity across land-use classes by green infrastructure policy.

155

Priority 1 Priority 2 Priority 3 Zone 1 Parking Green Roofs Tree Intensification shading/transform Zone 2 Tree Intensification Street shade & green Contiguous tree patches streets Zone 3 Green Roofs Contiguous tree patches Tree Intensification Zone 4 Contiguous tree patches Tree Intensification Parking shading/transform Zone 5 Contiguous tree patches Tree Intensification Parking shading/transform Zone 6 Contiguous tree patches Tree Intensification Street shading Table 5.7: Green infrastructure prioritization scheme across gradient zones.

5.7 Discussion and Conclusion

The research findings presented here are significant because they begin to give an

order of magnitude of the available opportunities to incorporate varied green

infrastructure measures. The literature has made the connection between green

infrastructure and ecosystem benefits that accrue to communities and their environment

(Pauleit and Duhme, 2000). Small scale best management practices have been developed

and implemented predominantly at the local or parcel scale (EPA). There is growing

recognition in the fields of planning of the importance of green infrastructure in

improving livelihoods (APA, 2012). Planning and design documents specify green infrastructure as tools to improve physical urban contexts. While benedict and McMahon

(2002) provide steps to develop urban and metropolitan scale networks, the challenge

remains to plan and implant these networks within already established urban. Prior to

devising policy, considering available space that could be transformed, intensified, or

preserved should be the first step in developing responsive policy. The results from this

156 research may be considered as an initial step in identifying where possibilities reside,

what measures are applicable and what are the policy priorities that are needed to

advance the development of a vegetated green infrastructure network.

The assessment method presented here aimed to provide a methodology to

identify space opportunities of vegetated green infrastructure network. It is addressed to

planners and policy makers. Using a method based on basic GIS analysis tools using

readily available data sets provides an easy and relatively quick way to define policy

priorities. The characterization of a metropolitan region based on the percent pervious

land metric allows planners to identify where opportunities and constraints are for

implementation. This becomes critical when considering the UHI, for example, where

specific environmental performance is required. Reducing temperatures will require a

certain extent and coverage of vegetated surfaces. The proposed method permits the

identification of the possible surface areas that can be dedicated for the provision of

ecosystems services. It follows, that the area metrics that result from this method allow for the calculation of the environmental performance, as the surface area of land cover is

a primary input. These calculations will ultimate determine the extent that vegetated

surfaces may be able to reduce the UHI. Such calculations allow ecologists and

environmental planners to prioritize policies and assess the extent of reliance on

vegetated surfaces for environmental performance.

The priority for planners is to ensure the functionality of green infrastructure

and/or green space and to conserve what exists. In the context of Boston, protecting the

36% of the total forested areas within the study area, that is situated within private and

public ownerships should be a priority. Then it becomes possible to intensify and expand

157 green cover by defining opportunities across land uses. By using the percent pervious metric as the basis of the assessment method combined with the patch-corridor-matrix model (Forman, 1988), an integrated approach may be developed to advance a regionals network. This will require further exploration into the configuration aspect of green infrastructure spatial elements.

There are several improvements that can be done to refine the proposed method.

Additional data layers should be added to ensure to refine the assessment and account for variable that impact the provision of vegetated surfaces. Soil data and existing tree canopy cover derived from remotely sensed images, will improve and refine the results to address factors that directly impact vegetative surface implementation.

This research focused on the assessment of space opportunity for vegetated green infrastructure for the UHI. The UHI is considered as a form of local climate change\ and represents a sample of what global climatic change may bring. The impacts of global climate change are comparable to the UHI as both have direct local impact on communities. Measures that are effective for the UHI are also applicable to adaptation to climate change impacts. Green infrastructure is one of the most promising adaptive strategies to climate proof cities. But this needs to be recognized in the planning process at multiple scale and explicitly considered within the constraints of available space.

158 CHAPTER 6

INTEGRATING GREEN INFRASTRUCTURE AND ADAPTATION PLANNING

6.1 Prologue

This dissertation emanates from keenness to find practical solutions to problems

that we face today and will face in the future. In more specific, it is addressed to planners

who are constantly seeking out new ways of dealing with problems. Today’s

complexities in planning need to simultaneously consider population increase, fast and

vast urbanization, depletion of resources, food security and climate change prompt us as planners, to pursue practical, tangible, cost-effective, and imageable solutions that our audience expects us to deliver. Yes, this is what we all wish: perfect solutions. But if not

seeking the best, most appropriate and “out-of-the-box” ideas is not our aim, then what

is?

The discussion brought forward in this dissertation was approached in this spirit:

to find ways to move the discussion in planning on green infrastructure implementation

and adaptation planning into a realm that begins to propose a continuum of theoretical

ideas and practical solutions that inform planners and improve the wellbeing of society.

Robert Young (2011) expressed this concern when he suggested that the underinvestment

in green infrastructure at the metropolitan scale, similar to other infrastructural systems,

has left planners with little experience to manage green infrastructure initiatives. I heed

his call and conclude this work with a brief review of the chapters, a discussion on green

infrastructure and adaptive capacity and propose a framework that integrates the ideas in

the chapters.

159 6.2 Green Infrastructure and Adaptive Capacity

I have argued that green infrastructure should become a significant component of climate change adaptation planning and policy. The three primary chapters identify explicit contributions that address different components of this argument. The model proposed for incremental adaptation implementation identifies planning issues that are core to adaptation and relevant to green infrastructure. For green infrastructure to be

effective it should be implemented at multiple scales, gradually and in a flexible manner

allowing adjustments as climate change materializes. An incremental approach with a

comprehensive vision is the subject of the proposed green infrastructure transect to

mainstream adaptation policies. The proposed green infrastructure transect integrates

adaptation policies at multiple scales and suggests a method to prioritize green

infrastructure strategies according to context and scale. This prompted the development

of an assessment method that aims to evaluate the maximum potential of space that could be allocated for green infrastructure implementation at the metropolitan scale. The three contributions form a framework that explicitly advocates for an ecosystem approach to adaptation through green infrastructure measures and policies.

Chapter three argued for an implementation method for adaptation plans based on future uncertainty of timing and magnitude of climate change impacts. It is based on phasing adaptation policies according to triggering indicators. Moving from no-regrets

policies to transformational ones depends on the extent that future climate projections

materialize. The aim was to provide a way where policy makers may be readily to accept of adaptation agendas through long term phasing of policies and resources. The risk for policy makers is that they will be accountable if climate change intensifies and no action

160 is carried out. A long term vision with clear but flexible objectives will allow for incremental implementation that is flexible to changing climatic, economic, and political environments. The results of testing the model on current case studies suggest that phasing of adaptation measures from no-regrets to transformational may reduce uncertainty of magnitude and timing of climate impacts by having a flexible and adjustable plan that evolves as climate impacts materialize; investment is spread across longer time frames allowing for monitoring to indicate the need to increase or reduce investment in adaptation; and reducing uncertainty of future information by beginning to plan today and update frequently as information is available and technological advancements are achieved. This long term and phased approach is critically relevant to green infrastructure as it requires long periods of times to establish as well as the challenge to account accrued services within current economic valuation systems.

In chapter four a theoretical framework modeled on green infrastructure principles to integrate adaptation policies is developed. The green infrastructure transect is a framework for vulnerability assessment based on a variation of physical and social contexts. The primary proposition is that green infrastructure implementation should be matched with social and physical conditions of communities. According to location along the transect and the prioritized climate impact, green infrastructure policies are planned according to suitable measures appropriate to contexts. Integration across scales

(vertical) and contexts (horizontal) becomes necessary to coordinate green infrastructure policies for adaptation across administrative jurisdictions and scales. When applied to the metropolitan Boston, the results suggest that the limitations of space and character along the transect will prompt varying green infrastructure measures. Prioritizing climate

161 change impacts across the transect will prompt communities to develop policies and plans

that suite their needs (horizontal integration) while contributing to an overall vision of

future conditions developed by regional bodies (vertical integration). By encouraging

integrated local and regional planning, allocation of green infrastructure resources could

be matched to needs, provide redundancy of systems when required, and define enhance

the quality of life in general.

One of the primary observations from chapter four is that space for green

infrastructure implementation varies across urban-rural gradient. Chapter five addresses

this issue and a method is proposed to estimate surface areas across land uses usable for

green infrastructure to form a network at the metropolitan scale. Within urban contexts,

un-built surfaces are highly determined by land uses. The morphology of available space,

soils, and vegetation types vary according to land use types. Percent pervious (PP) is

developed as a land metric to identify the variations across land use types. The area

estimation method is applied to the metropolitan area of Boston (MOB) taking the urban

heat island as an application example to identify suitable green infrastructure measures.

The MOB was characterized using the PP into gradient zones of perviousness. PP within each land use type in each gradient zone is identified allowing the comparison of PPs across land uses and zones. A gradient of potential green infrastructure policies is then developed to assist policy makers in defining the potential of green infrastructure policies. Four primary green infrastructure measures were found to be effective in reducing temperatures: contiguous tree canopy, green roofs, street and parking tree shading. Four strategies were also set: conservation of what is present, intensify pervious surfaces, transform impervious surfaces and reduce impact. When considering the urban

162 heat island, the results suggest the potential of developing a network across the urban

region varies based on space availability and type of green infrastructure measure. For

example, where highest intensities of UHI dominate (in and close to urban core), green

roofs, parking transformation, and street shading are highly possible to implement

compared to contiguous patches of forest or pervious land surfaces. Farther from the core where the UHI is less prominent, the results indicate a reversal of space potential availability. The significance of these findings suggest that there is a clear need to expedite implementation of specific green infrastructure measures within the urban core to address the UHI. On the other hand, longer term planning to preserve and expand patches of pervious land surfaces for future urban expansion should be accounted for starting today.

The phased adaptation planning model based on triggering conditions coupled with a green infrastructure transect that matches measures and contexts across scales and the exploration of potential surface area allocation for green infrastructure aim to enhance the ability of communities, cities and urban regions to better cope with climate change impacts. In more specific, the propositions advocate an explicit integration of green infrastructure in adaptation planning as no-regrets policies where adaptation could begin now with multiple benefits accrued to the future. While green infrastructure may not be sufficient to climate proof cities, it highly contributes to increasing the adaptive capacity from individuals to communities to regions.

As cities begin to pursue adaptation plans that increase resilience to climate change impacts, they aim to enhance their adaptive capacity across multiple sectors to address social, environmental, and economic vulnerabilities. Environmental

163 vulnerabilities are a challenge to address due to the local-to-regional extent of impacts

and the multi-faceted infrastructural responses required to address them. As an ecosystem based approach, green infrastructure is a suitable set of measures and policies that enhance the adaptive capacity of specific communities and the urban area at large. The

complimentary between ecosystem services and climate impacts, no-regrets and multi- level benefits render it as a necessary component to enhance the adaptive capacity.

In general, enhancing the adaptive capacity is achieved by adjusting existing

systems and resources to cope with short and long term impacts. In human dominated systems, as is the case in an urban metropolitan system29 , the ability to adjust is

dependent on recognizing a specific threat (or threats) followed by policies and planning initiatives that respond to this threat over time (Walker and Salt, 2008). In the context of climate change impacts, adaptive capacity is the “the ability of a system to adjust to climate change (including climate variability and extremes) to moderate potential damages, to take advantage of opportunities, or to cope with the consequences (IPCC

2007, p. 35)”. Two key ideas are relevant to green infrastructure. First, green infrastructure can simultaneously act in multiple adaptation capacities; and second, it is necessary to adjust and expand open space planning to include un-built land and natural assets to become a green infrastructure network to maximize the environmental benefit.

29 Urban System is a network of towns and cities and their hinterlands which can be seen as a system since it depends on the movements of labor, goods and services, ideas, and capital through the network. Crucial to the interactions within the system are efficient systems of transport and communication (Forman, 2008)

164 The first idea is that green infrastructure can function in three adaptation capacities: anticipatory30, autonomous31, or planned32 (IPCC, 2007). It is anticipatory by the mere fact that open spaces exist within metropolitan regions. Remnants of the natural landscape, in the form of parks and private yards, are already providing ecosystem services at the local scale regardless of the explicit recognition of climate change. But this is dependent on the autonomous adaptation of natural elements within these open spaces. These services depend on the health of vegetation and soils and the availability of water. Increasing temperatures may lead to tree canopy death due to lower hardiness or change in soil composition or reduction in water availability. Such impacts prompt ecological succession resulting in species migration towards the north (Lerner and Allen,

2012). But climate change seems to be increasing at a faster pace than anticipated with projections reaching 2⁰C by mid-century and 4⁰C by 2100 (Frumhoff et al., 2007). Such a time frame is arguably too short for natural adaptation to occur. As a planned system for climate adaptation, green infrastructure could account for the inherent vulnerability of the network itself as well as ensuring the continued delivery of ecosystem benefits. For that to occur, adjustments to open space planning must be made, and very soon.

30 Anticipatory adaptation is “adaptation that takes place before impacts of climate change are observed; also referred to as proactive adaptation (IPCC 2007, p.35)”.

31 Autonomous adaptation is “adaptation that does not constitute a conscious response to climatic stimuli but is triggered by ecological changes in natural systems and by market or welfare changes in human systems; also referred to as spontaneous adaptation (IPCC 2007, p.35)”.

32 Planned adaptation is “adaptation that is the result of a deliberate policy decision, based on an awareness that conditions have changed or are about to change and that action is required to return to, maintain, or achieve a desired state (IPCC 2007, p.35)”.

165 The second idea is the adjustment or transformation of open space into a green infrastructure network providing maximum ecosystem benefits that specifically address impacts of climate change. In the USA, open space planning is centered around conservation and recreation of predominantly public land. Land and natural assets in metropolitan areas include both publicly and privately-held open spaces. Arguably, these assets are not counted as part of an infrastructural metropolitan network but as disparate and distinct elements that provide localized ecosystem services. When accounting for ecosystems services necessary for adaptation, land and natural assets within private and public land should qualify to be part of a green infrastructure system. To transform these assets into an urban metropolitan green infrastructure capable of effectively contributing to adaptation, spatial adjustments to GI planning need to be considered.

To effectively use green infrastructure as an adaptation measure, the extent of current and future ecosystem benefits must be understood and measured. Under future scenarios, recent studies have measured the extent of reduction of temperature and flooding (Gill et al., 2007) and removal of pollutants and CO2 (Currie & Basse, 2008) of green roofs, trees, and shrubs in Manchester (UK) and Vancouver (Canada) respectively.

These studies have focused on small-scale BMPs within urban cores. Small-scale BMPs contribute to enhancing site and neighborhood level climatic and environmental conditions and demonstrate the effectiveness of green infrastructure, therefore, enhancing local adaptive capacities. But if the understanding of green infrastructure is expanded to include local, neighborhood, urban and regional natural and engineered systems, then the extent of expanding the contribution of green infrastructure to the adaptive capacity is augmented. For example, the metropolitan Area of Boston boasts the provision of high

166 level of open space and forested cover (51%). Yet, only 15% of these forests are

protected ensuring continuity of ecosystem provision into the future. Residential

property (from single homes to housing projects) include large tracts of un-built surfaces.

Residential pervious surfaces account 25% of total land area (excluding water). These are predominantly lawn areas that could be transformed into functioning green infrastructure, while maintaining recreational open space. The two examples in Boston demonstrate that the potential to account land as part of green infrastructure to increase the adaptive capacity is available. What is required is a reconceptualization of what open space constitutes to function as an infrastructure.

Few studies explicitly connect ecosystem services33 and adaptation beyond the urban or sub-urban cores. It is true that the focus of climate change adaptation is at the local scale, as this where the impacts are directly felt, at the individual level (ECLEI,

YR). Yet, there are multiple benefits to considering the metropolitan or even the regional

scale when considering GI for adaptation. Recognizing large and small patches and

corridors across a metropolitan region for the capacity as a CO2 sink (Stone, 2012) or

reduce regional temperatures (Gartland, 2008) or act as water storage from excess

precipitation (Gill et al., 2007), are complimentary to small scale BMPs within the urban

core. Studies that focus on the benefits of forest canopies (Nowak and Greenfield, 2012,

2012a), natural resources (Forman,T.T., 2008), and the impact of detention and retention

33 Ecosystem services (values and functions) are biological and ecological processes that include, but not limited to, the water cycle, carbon cycle, nutrient cycle, air and water pollutant removal and photosynthesis. The benefits accrued to communities from these services maintain the quality and quantity of water (Brabec, 2009; 2002); clean the air and act as a CO2 sink (McPherson, 2002); provide recreation spaces; improve health conditions (REF); and ameliorate temperatures of urban climates (Akbari, 2002; Gartland, 2007).

167 basins at the watershed scale primarily focus on total ecosystem benefits without connecting to adaptation. If sufficient mass of forests, natural reserves, river watersheds,

in addition to large scale parks, golf courses, etc. are maintained, temperatures within the

metropolitan climates may be ameliorated. The point here is that adaptation to climate

change impacts has been focused on solutions from within the urban core. It is time that we begin looking beyond to compliment the local approach. Green infrastructure can play the role of ameliorating metropolitan or urban-regional climate at least in the early stages towards adaptation planning.

Approaching open space (or the un-built realm) planning in a metropolitan region as an interconnected system of patches and corridors within a matrix (Forman, T.T.,

1997, 2008; Ahern, 2007) provides a strategic approach to utilize GI as an adaptation measure. A GI system that extends landscape ecological values and envisioned as a whole to respond to metropolitan adaptation needs is a framework that incorporates all land resources as potential contributors (Gill et al., 2007). The patch-corridor-matrix concept parallels metropolitan infrastructural terminology, similar to transportation (i.e. hubs, connectors, and destinations).Designating specific land typologies based on their ecosystem benefit and suitability to respond to climate change impacts allows the development of a network that is incorporated into the overall metropolitan infrastructural system. This incorporates existing land resources as well as planning for the future development of the GI system based on projected climate threats. These land resources are available today and should be protected, enhanced, or expanded. When considered as a metropolitan system, towns and cities may be able to concentrate their land resource planning efforts depending on the contribution to the overall system and

168 adaptation targets. A metropolitan system should incorporate the plethora of current policies, programs and regulations (DCR, DEP, Smart growth, conservation) that ensure permanency of vegetated patches and corridors. These need to be further developed to explicitly connect them to climate change adaptation. Still, it can be argued that these also require an extensive period of time to implement. This is true. But starting inventory, assessment, and adjustment of an existing stock of land with the purpose of developing the GI system may be a more viable approach than beginning with what is not there or requiring establishment through new policies and regulations. Furthermore, lower land values beyond the urban core provide opportunities to aggressively pursue land purchases. When considering urban expansion in the future and the likely increase in climate impacts, land areas beyond the core will probably be transformed in character to emulate urban core areas. This may lead to loss of more vegetated, ecological and natural surfaces that inherently provide adaptation ecosystem services. So thinking about future conditions and locking land today for future green infrastructure is another way that open space planning should be adjusted and thought of more along the line of a necessary and fundamental infrastructure. Furthermore, investing today in land at current prices may reduce the likelihood of large scale investments in the future. Thus an incremental and phased approach is arguably the way forward to invest in a future green infrastructure system across metropolitan areas to climate proofing the future today.

6.3 Green Infrastructure Framework for adaptation at the metropolitan scale

In this context, planning green infrastructure at the metropolitan or urban-region scale begins to immediately accrue adaptation benefits early in the adaptation process. At the same time, a vision of a green infrastructure system provides flexibility to pursue

169 expansion and development of the green infrastructure system. In this manner, adaptation

using green infrastructure may begin today, monitoring its effectiveness while expanding

the existing stock of land into a multi-scalar infrastructural green infrastructure system

with multiple ecosystem benefits pertinent to adaptation. To achieve this, the following

principles are proposed to formulate a framework that adjusts open spaces to become a

green infrastructure network capable of enhancing the adaptive capacity of urban regions.

These principles are envisioned to contexts similar of the North East of the United States.

Principle 1 – Develop a green infrastructure regional vision: An urban region

should articulate a metropolitan green infrastructure spatial vision to ensure that local

initiatives of green infrastructure contribute to the overall adaptation targets. The regional

dimension is critical because climate variability and impacts occur at this scale. At the

same time, when considering urban spread, reserving land assets is critical for future

climate proofing. Targets for adaptation set at the regional scale will ensure that local and

participatory implementation contribute to transform green infrastructure to a utility-like

infrastructure in public policy terms, that is included within a municipal budgeting

(Benedict and McMahon, 2006) and actively pursued. This may be achieved by

establishing a planning or coordinating body (similar to transportation or water resource

management authorities) empowered with defining opportunities and cross jurisdictional

coordination of green infrastructure resources. For example, in Metropolitan Boston an

integrative body may be established to coordinate works across the Department of

Conservation and Recreation, department of Environmental protection, the regional planning bodies, and Department of Transportation.

170 Principle 2 - Integrate top-down and bottom-up planning: Building on a vision, participation of communities is necessary for smooth and meaningful implementation.

Communities empowered with developing and contributing to vision building will be more readily accepting of initiatives in their neighborhood. Suitable measures that respond to peoples’ place of living to ensure longevity, adoption, and implementation of policies.

Principle 3 – Integrate ecosystem services accounting: Explicit acknowledgement and accounting of how singular and aggregated ecosystem benefits accrue from vegetated land surfaces at the site, neighborhood, urban and metropolitan scales should be continuously done to determine extent of contribution to the adaptive capacity.

Identifying the targets of performance and thresholds of provision of ecosystem services are fundamental to establishing a system and an infrastructure that can respond to climate impacts. Being able to determine how much tree canopy is available and size of tree crowns will assist in determining how much rain water could be intercepted from an extreme precipitation reducing run-off. Conducting such inventories periodically will support monitoring efforts of climate change and continuously enhance and improve the adaptive capacity.

Principle 4 – Expand green infrastructure surfaces across cities: Adopt a proactive approach to intensifying urban tree canopy, natural vegetated surfaces, and engineered vegetated surfaces on existing and newly acquired plots of land, regardless how small or where located. Strategies such as protecting existing vegetated surfaces, transforming impervious surfaces, creating green roofs and transforming derelict land become core strategies in green infrastructure planning. Therefore accounting for typologies of land

171 use and land cover on privately or publicly owned land become crucial to maximize

ecosystem benefits as the relationship of ecosystem services and vegetated surface area is

directly proportional.

Principle 5 – Account for system vulnerability: The acknowledgement that green

infrastructure is inherently vulnerable when considering climate change. Changing

temperatures and drought regimes will impact the flora of a region, likely leading to the

death of forest cover and resulting in higher impacts from desertification and heat

indexes. Anticipating this possible change would require replacing vegetation species by

more drought tolerant types to ensure to ensure longevity and biodiversity, ensuring

maximum possible ecosystem services in the short and long terms. Knowing how much ecosystem services (adaptive capacity) may be derived within a specific context is crucial to long term adaption planning.

Principle 6 – Develop a continuum of green infrastructure policies: Identification

and categorization of GI measures and policies into a continuum that builds from local

small scale measures into a more transformative regional network. In the context of GI, a

transformative policy could be the acquisition of forested land by eminent domain to

ensure continuity of forest cover. What is fundamentally different for green infrastructure, is that even transformational measures remain no-regret win-win solutions as cities become greener. The purpose is to actively pursue strategies that respond to climate change intensification in the future

Principle 7 – Incremental and phased implementation: the continuum of measures will also need to be implemented in an incremental and phased manner. For green infrastructure expansion, this may be the only way forward as challenges to valuation

172 remain as well as the long term nature of transformation of land uses and covers over time. Gradual implementation begins with accounting what is available, build a vision based on climate impacts, define extent of green infrastructure contribution, and then begin phased implementation.

Principle 8 – Multi-disciplinary planning and managing bodies: Adaptation and green infrastructure planning span environmental and biological sciences, social sciences, planning and design. Management and economic planning are also core fields. To achieve a planning result that addresses social needs and climatic impacts and that is well managed and economically feasible will require a team of professionals, administrators, policy makers and community representatives with commitment to achieve these objectives. This is not an easy task that questions how conventional planning and implementation have been carried out. The task to bring together a climate scientist, an anthropologist,, an ecologist, an architect, a landscape architect, planner, an analyst, a policy maker, a financial manager, and a community representative to achieve tha goal of adaptation planning will require new ways to organize and manage that accounts for the multi-layered process.

There are several reasons why the above principles are useful to planners and policy makers concerned with using green infrastructure for adaptation. Understating the current adaptive capacity provides the framework to inventory existing GI spatial typologies and their respective land covers, thus providing a single data layer for assessment. Developing a GI system at the regional scale provides a systems’ understanding of the opportunities and constraints. As a system, GI should become a critical infrastructural component of the overall regional metropolitan infrastructure

173 system and considered as equal to other utility or infrastructural systems. By deriving the current adaptive capacity from GI, thresholds of how much GI may contribute to adaptation planning may be established. This allows a determination of whether GI is sufficiently capable to address the context specific climate change impacts or supplementary policies and measures are needed to be bundled with GI. A regional

approach provides the possibility of understanding cross jurisdictional opportunities and

constraints providing the possibility for effective policy coordination across administrative boundaries. Lastly, all the above compelling reasons will contribute to developing clearer guidelines to implement policy, planning, and design of GI enhancing

the efficiency of the GI system itself and increasing the resilience of any metropolitan

region.

6.4 Future Research Work

The research in this dissertation has provided insights and understandings of

green infrastructure and climate change adaptation. As much as this has been a learning

and educational experience, the process has also brought about many questions that

complement this research and/or develop it further. The ideas and questions suggested

below define a long term research agenda that connects practice and research; planning

and implementation; and green infrastructure with adaptation planning.

6.4.1 Adaptation Planning

Adaptation Planning: The recent surge of adaptation plans may define the way

forward for adaptation. Advances in theoretical frameworks combine comprehensive and

risk management planning defining approaches to adaptation. This research will aim to

174 analyze the extent of application of theoretical work into adaptation plans and how the theoretical frameworks could be developed from lessons learnt in practice. A sample of adaptation plans developed by cities across the world (developed and developing countries) will be analyzed to draw generalizations that define previous experience and show the way forward for future adaptation plans and these should move forward. The initial objectives include:1) Conduct literature review of current adaptation theoretical frameworks 2) define common principles and themes across case studies, 3) develop a common framework between a majority of adaptation plans, 4) define the basis that plans are carried upon, 5)Compare case studies to theoretical frames, 6) identify and categorize types of adaptation measures based on type of climate impact and social/economic/political contexts, 7) understand and compare decision rules used within adaptation plans, and 8) develop cross cutting generalizations and themes that include timing, triggering thresholds, and cost implications.

On a more theoretical level, investigate the meaning and relationship between sustainability and adaptation to climate change. Are these mutually complimentary or not? Are the objectives the same or differ? What are the overlaps and the contradictions?

Such a study may provide a way to better define sustainability in the context of climate change and to mainstream adaptation under the sustainability umbrella. This may change the perception of policy makers to adaptation and if seen as multi-objective benefiting communities rather than a political agenda, action on the adaptation front may be pursued more aggressively.

175 6.4.2 Green Infrastructure Planning

Green infrastructure is being advanced and developed as an adaptation set of measures. Research and experiments are being conducted to operationalize the use of green infrastructure in adaptation plans. The extent of reliance on green infrastructure and the expected outcome is not clear. Similar to the first research objective set above, this project will set to survey adaptation for their extent of use of green infrastructure. The research will address and investigate the following: 1) how is GI defined in the context of adaptation? 2) How much do the adaptation plans rely on green infrastructure for effective response? 3) What are the implementation steps that have been taken to implement green infrastructure strategies? 4) What are the gaps apparent in the adaptation plans to further the inclusion and implementation of GI strategies for adaptation?

Chapter five of this dissertation focused on the urban climate as a surrogate for larger scale climate adaptation. Green infrastructure spatial typologies were defined across different land-uses at the regional scale that are suitable to mitigate the urban heat island in Boston Metropolitan area. This same research is to be expanded to include other climate change impacts such as flooding from increased precipitation, the threat of rising sea level on coastal areas and the potential role of green infrastructure. Such an approach will “complete” the study on Boston area to develop a green infrastructure regional scale adaptation framework that accounts for the dominant and relevant climate impacts in metropolitan Boston. Furthermore, the research will investigate the relationship of spatial distribution of green infrastructure elements and their potential to address vulnerable communities and improve their adaptive capacity. A combination of landscape ecological

176 metrics, spatial statistics and socio-economic data will be sued to develop an effective regional scale adaptation plan.

The proposed green infrastructure transect in chapter four is an initial proposal with the potential to develop it into an operational green infrastructure planning tool with multiple objectives. The procedural framework and practical aspects will require further development and verification. Literature from landscape ecology, urban planning, growth management, and landscape planning will be analyzed to further develop the conceptual idea.

Develop the work completed in chapter five of this dissertation into a comparative study between several US cities. Such a study will aim to understand the relations ship between regional urban form and the distribution of pervious surfaces (surrogate for green infrastructure). Patterns at the regional, urban and neighborhood scales may be identified as being potential best practices for retrofitting cities for climate adaptation. In addition to the general objectives, several issues will also be pursued: 1) compare the distribution of pervious surfaces across the urban-rural gradient of selected case studies,

2) develop patterns and common insights on pervious land distribution across land-uses and compare, and 3) develop generalized land metrics that account for the urban form and pervious land distribution.

6.4.2 Metropolitan Planning and Urban Form

The research was conducted within the context New England, US. While this was done for several practical reasons, several lessons and methods may be could be transferred to the Middle East. This projected project will aim to analyze the applicability of environmental planning in general, and the use of green infrastructure within the

177 context of planning in the Middle East. Case studies from cities will be analyzed for their

planning systems and the potential to introduce environmental planning as a core part of the planning system. This is particularly important because planning remains very deterministic, top-down and environmental planning is still in its infancy. In specific, the research will: 1) analyze metropolitan areas in the Middle East for the planning system,

political and economic context and willingness to incorporate environmental planning, 2) identify and study any cities with an environmental dimension in its planning structure, and 3) How do the climatic, geographic, and social context impact the application of green infrastructure within the ME context. It is important to note while the Middle East is imagined as a single entity in the Western mind, the contexts vary putting forward several challenges to circumvent.

6.5 Epilogue

Lastly, while this is not a future research objective, the research conducted in this dissertation has prompted me to consider the significance and importance of green infrastructure and climate change as symbols of what humanity’s value have come to. I would like to reflect on one aspect not directly related to green infrastructure but stems from some conclusions that I have developed during reading and research.

There is still much more to be done on green infrastructure and adaptation. But the mere fact of combining these two elements under a planning paradigm signifies two important issues: green infrastructure is based on biological and ecological processes and adaptation is about the climate. These are two components necessary for life, to our existence.

178 What is surprising is that there are few adherents to the natural, ecological and

biological potential of green infrastructure. The benefits seem obvious, but may not be

appreciated and understood in our current society. In an age where every activity, object

and action is immediately transformed into monetary value, the intangible or

unquantifiable values should be accounted for and considered in our inventory of

consumption. I am not claiming that monetary value is not important in our day and age,

but there are values and benefits that are important to us as humanity but may not be

tagged. So what do we do? The simple answer is account for it, consider it and protect until the day we are able and ready to value nature and the environment in a more concrete way that could be understood by humanity. What is most surprising is that we, as humans, have emanated from this same nature that we seek to destroy.

I hope that the ideas and thoughts presented here help to elevate the understanding and significance of the place that we live in.

179 APPENDICES

180 APPENDIX A

TYPES OF URBAN HEAT ISLAND

Three types of UHI may occur at different times of the day and cover different areas of the metropolitan area (Kutller, 2008). Surface heat island (SHI) and heat islands of the urban canopy layer (HIUC) are affected by the ground and caused by high surface temperatures. They occur due to the following mechanisms: anthropogenic heat from building sides; greater shortwave absorption due to canyon geometry, decreased net long- wave loss due to reduction of sky view factor by canyon geometry (sky view factor is the ratio of the amount of the sky “seen” from a given point on a surface to that potentially available); greater daytime heat storage (and nocturnal release) due to thermal properties of building materials greater sensible heat flux due to decreased evaporation resulting from removal of vegetation and surface waterproofing; and convergence of sensible heat due to reduction of wind speed in canopy. SHI and HIUC mainly occur in built-up areas and therefore have clearly defined boundaries and HIUC affect the atmosphere between the surface and mean roof height (Oke, 1979).

Urban boundary layer heat islands (UBH) form above the canopy layer as a result of heat transfer (entrainment of heat from warmer canopy layer), artificial heat input

(anthropogenic heat from roofs and stacks), and increased absorption of radiation by atmospheric pollutants (shortwave radiative flux convergence within polluted air) with resulting thermal re-emission (entrainment of heat from overlying stable air by the process of penetrative convection). This type of heat island already extends so far upwards into the atmosphere above a city that it is propagated downwind by the overall wind patterns and gives the well-known “urban plume”. The UBH occurs over built-up

181 areas and extends over surrounding areas downward of wind direction (Oke, 1979;

Kuttler 1997, 2008). The variability and different types of UHI occurring at different times of the day render UHI a dynamic phenomenon.

Figure 5.24: Urban Boundary Layer (Source: Kuttler, 2008)

The UHI varies in intensity depending onn regional climatic condittions and contexts. Oke (1982) developed generalizations of the intensity of UHI that show the dynamic and uncertain nature of the phenomenon. The UHI intensity decreases with increasing wind speeds as these provide direct brreeze to residents and shifts the urban boundary layer and plume downstream. This releeases entrapped heat and replaced with cooler air resulting in reduced temperatures (Figuerola and Mazzeo, 1998; Magee et al.,1999; Morris et al. ,2001; Unger et al.,2001). Cloud cover tends to reduce the intensity of UHI because cloud cover screens the incoming sun energy which is the primary source of absorbed and released heat (Ackerman,1985; Ripley et al.,1996; Morris

182 and Simmonds,2000). UHI intensity is best developed in the summer due to the higher

intensity of the sun and reduced amounts of rain (reducing hydrant cooling), at least in

the northern latitudes (Philandras et al., 1999); Morris et al., 2001). UHI intensity is

greatest at night due to the nature of the phenomenon where impervious material absorbs the sun energy during the day, stores it and radiates it back in short waves during the night (Ripley et al., 1996; Jauregui,1997; Magee et al., 1999; Mont´avez et al., 2000;

Tereshchenko and Filonov, 2001; Kuttler, 2008). It follows that UHI may disappear by day or the city may be cooler than the rural environs (Tapper, 1990; Steinecke,1999). In conjunction with the above, cities with a larger footprint amplify the intensity of UHI due to the larger surface area of impervious surfaces that increase absorbed sun energy and released heat. Increased population size also increases radiant heat through increased car use and energy consumption (Park, 1986; Yamashita et al., 1986; Hogan and Ferrick,

1998). Furthermore, rates of heating and cooling are greater in the surrounding area than built-up area because of the difference of surface characteristics between both contexts

(Johnson, 1985).

183 APPENDIX B

THE TABLES

Population data Basic calculated Areas (m2) (m2) (m2) (m2) (m2) (m2) (m2) Town Name Water n 2005 y Area Area of Area of Area of Surfaces Surfaces Pervious n change n change Town/Cit Populatio Populatio POP 2000 POP 2010 s Surfaces s Surfaces 2000-2010 Estimated Imperviou Town # 1 ABINGTO 14605 15985 1380 N 15.295, 26.390.84 4.596.484 2.019.58 19.774.78 00 7,07 ,75 1,92 0,40 2 ACTON 20331 21924 1593 21.128, 52.564.53 7.253.183 3.302.20 42.009.14 00 2,79 ,02 0,14 9,63 3 AMESBUR 16450 16283 -167 Y 16.366, 35.547.16 5.025.558 5.341.93 25.179.67 00 8,40 ,83 3,82 5,75 4 ANDOVER 31247 33201 1954 32.224, 83.274.58 13.843.72 6.163.97 63.266.88 00 3,62 7,97 3,34 2,31 5 ARLINGTO 42389 42844 455 N 42.616, 14.203.93 5.859.636 873.792, 7.470.508 00 7,83 ,90 35 ,58 6 ASHLAND 14674 16593 1919 15.634, 33.342.67 4.935.325 2.175.89 26.231.45 00 7,45 ,51 4,17 7,77 7 AVON 4443 4356 -87 4.400,0 11.754.17 2.792.296 1.342.41 7.619.465 0 5,97 ,52 3,79 ,66 8 AYER 7287 7427 140 7.357,0 24.613.33 4.506.749 2.488.56 17.618.02 0 4,65 ,30 1,10 4,25 9 BEDFORD 12595 13320 725 12.958, 35.840.94 6.399.591 3.706.43 25.734.92 00 9,73 ,56 4,20 3,97 1 BELLINGH 15314 16332 1018 0 AM 15.823, 48.919.13 6.829.099 3.138.82 38.951.20 00 5,20 ,58 8,77 6,85 1 BELMONT 24194 24729 535 1 24.462, 12.210.76 4.215.878 190.123, 7.804.761 00 3,28 ,28 71 ,29 1 BERKLEY 5749 6411 662 2 6.080,0 42.944.09 3.360.598 3.200.59 36.382.90 0 9,93 ,96 2,11 8,86 1 BERLIN 2380 2866 486 3 2.623,0 34.117.22 2.566.841 1.896.94 29.653.43 0 0,00 ,13 3,92 4,95 1 BEVERLY 39862 39502 -360 4 39.682, 39.915.20 9.376.823 1.978.10 28.560.27 00 8,47 ,21 9,68 5,58 1 BILLERIC 38981 40243 1262

184 5 A 39.612, 68.158.85 13.317.15 103.641, 54.738.06 00 5,15 2,82 83 0,50 1 BLACKST 8804 9026 222 6 ONE 8.915,0 29.474.86 3.382.997 1.205.07 24.886.79 0 6,80 ,66 1,28 7,86 1 BOLTON 4159 4897 738 7 4.528,0 52.003.96 3.176.749 3.537.70 45.289.50 0 0,66 ,58 7,03 4,05 1 BOSTON 8 588.95 617.59 28.63 603.27 129.913.8 72.411.82 6.055.07 51.446.95 7,00 4,00 7,00 6,00 49,95 3,52 3,71 2,72 1 BOXBORO 4868 4996 128 9 UGH 4.932,0 26.951.58 2.377.793 2.290.11 22.283.67 0 4,41 ,16 2,09 9,16 2 BOXFORD 7921 7965 44 0 7.943,0 63.204.67 4.154.998 5.933.65 53.116.02 0 5,54 ,49 2,07 4,98 2 BRAINTRE 33828 35744 1916 1 E 34.786, 37.369.75 10.663.59 3.926.95 22.779.20 00 5,56 9,31 3,12 3,13 2 BRIDGEW 25185 26563 1378 2 ATER 25.874, 73.451.63 8.604.039 12.205.7 52.641.79 00 1,19 ,58 99,71 1,90 2 BROCKTO 94304 93810 -494 3 N 94.057, 55.724.09 19.891.66 2.176.53 33.655.88 00 1,46 4,44 8,25 8,77 2 BROOKLI 57291 58732 1441 4 NE 58.012, 17.676.10 6.887.936 217.607, 10.570.55 00 3,60 ,79 13 9,68 2 BURLINGT 22876 24498 1622 5 ON 23.687, 30.723.42 8.189.952 1.211.32 21.322.14 00 2,28 ,52 3,85 5,91 2 CAMBRID 101355 105162 3807 6 GE 103.25 18.400.88 10.753.45 1.879.64 5.767.786 8,00 7,15 6,39 4,08 ,68 2 CANTON 20775 21561 786 7 21.168, 50.541.64 8.492.730 6.909.40 35.139.50 00 0,10 ,98 3,66 5,46 2 CARLISLE 4717 4852 135 8 4.784,0 40.207.03 2.370.547 4.985.93 32.850.55 0 4,49 ,33 0,38 6,78 2 CARVER 11177 11509 332 9 11.343, 102.880.6 11.445.96 28.305.8 63.128.83 00 48,17 9,05 42,62 6,50 3 CHELMSF 33858 33802 -56 0 ORD 33.830, 59.787.67 12.102.83 3.420.77 44.264.05 00 3,83 8,84 9,06 5,93 3 CHELSEA 35116 35177 61 1 35.146, 5.730.029 4.323.118 47.986,0 1.358.925 00 ,66 ,62 1 ,03 3 COHASSE 7261 7542 281 2 T 7.402,0 26.048.18 2.866.451 2.543.03 20.638.70 0 4,25 ,49 1,69 1,07 3 CONCORD 16993 17668 675 3 17.330, 66.904.55 6.144.843 6.598.87 54.160.83

185 00 7,65 ,11 6,36 8,18 3 DANVERS 25212 26493 1281 4 25.852, 35.705.00 9.823.624 2.721.01 23.160.36 00 9,18 ,08 6,24 8,86 3 DEDHAM 23464 24729 1265 5 24.096, 27.669.92 6.169.084 3.438.25 18.062.58 00 5,45 ,77 6,44 4,24 3 DIGHTON 6175 7086 911 6 6.630,0 57.418.09 4.278.902 6.957.35 46.181.83 0 1,61 ,85 2,32 6,44 3 DOVER 5558 5589 31 7 5.574,0 39.933.24 2.789.778 2.518.11 34.625.35 0 4,03 ,53 1,46 4,04 3 DRACUT 28562 29457 895 8 29.010, 55.522.00 8.635.846 3.800.31 43.085.83 00 5,05 ,77 8,72 9,56 3 DUNSTAB 2829 3179 350 9 LE 3.004,0 43.387.64 1.659.098 2.380.15 39.348.39 0 7,35 ,25 8,53 0,57 4 DUXBURY 14248 15059 811 0 14.654, 62.340.79 5.872.869 11.394.3 45.073.57 00 7,73 ,59 51,70 6,44 4 EAST 12974 13794 820 1 BRIDGEW 13.384, 45.428.30 5.268.365 6.130.12 34.029.81 ATER 00 9,99 ,70 7,81 6,48 4 EASTON 22299 23112 813 2 22.706, 75.713.80 8.005.411 11.062.1 56.646.27 00 3,60 ,07 16,23 6,30 4 ESSEX 3267 3504 237 3 3.386,0 36.924.44 1.704.579 10.233.5 24.986.31 0 5,06 ,35 54,78 0,93 4 EVERETT 38037 41667 3630 4 39.852, 8.904.846 6.127.534 65.146,2 2.712.166 00 ,86 ,33 8 ,25 4 FOXBORO 16246 16865 619 5 UGH 16.556, 53.987.52 7.713.648 4.675.56 41.598.31 00 9,59 ,95 6,06 4,58 4 FRAMING 66910 68318 1408 6 HAM 67.614, 68.642.06 16.482.78 4.671.13 47.488.14 00 0,19 0,74 2,40 7,05 4 FRANKLIN 29560 31635 2075 7 30.598, 69.998.12 10.206.00 3.482.27 56.309.84 00 6,01 9,62 1,12 5,27 4 GEORGET 7377 8183 806 8 OWN 7.780,0 34.054.31 3.511.600 3.793.27 26.749.43 0 2,33 ,30 5,35 6,68 4 GLOUCES 30273 28789 -1484 9 TER 29.531, 69.154.95 8.168.946 7.935.62 53.050.38 00 9,49 ,57 9,58 3,34 5 GROTON 9547 10646 1099 0 10.096, 87.479.47 4.927.314 7.937.22 74.614.93 00 0,74 ,18 0,07 6,49 5 GROVELA 6038 6459 421 1 ND 6.248,0 24.338.16 2.484.271 3.402.66 18.451.22 0 9,30 ,29 8,14 9,87

186 5 HALIFAX 7480 7518 38 2 7.499,0 45.028.53 3.313.086 13.738.2 27.977.17 0 4,81 ,48 70,85 7,48 5 HAMILTO 8315 7764 -551 3 N 8.040,0 38.592.24 2.644.280 7.305.53 28.642.43 0 3,68 ,08 1,37 2,23 5 HANOVER 13164 13879 715 4 13.522, 40.478.06 6.252.058 4.329.04 29.896.96 00 8,49 ,67 5,94 3,88 5 HANSON 9515 10209 694 5 9.862,0 40.748.13 4.071.997 10.236.7 26.439.40 0 6,49 ,66 35,49 3,34 5 HARVARD 5981 6520 539 6 6.250,0 70.327.48 5.478.348 4.831.33 60.017.80 0 4,64 ,60 0,94 5,10 5 HAVERHIL 58969 60879 1910 7 L 59.924, 92.457.99 15.251.63 9.074.98 68.131.38 00 8,43 1,83 5,45 1,15 5 HINGHAM 19882 22157 2275 8 21.020, 58.912.61 8.614.047 3.990.24 46.308.31 00 4,75 ,32 7,46 9,97 5 HOLBROO 10785 10791 6 9 K 10.788, 19.070.45 3.156.003 1.290.83 14.623.61 00 4,52 ,48 5,50 5,54 6 HOLLISTO 13801 13547 -254 0 N 13.674, 49.344.83 4.759.852 4.514.26 40.070.72 00 7,13 ,51 2,10 2,52 6 HOPEDAL 5907 5911 4 1 E 5.909,0 13.836.17 2.167.438 923.340, 10.745.39 0 5,47 ,56 53 6,38 6 HOPKINTO 13346 14925 1579 2 N 14.136, 72.137.07 7.303.724 7.303.72 57.529.62 00 4,96 ,82 4,82 5,32 6 HUDSON 18113 19063 950 3 18.588, 30.726.32 5.921.626 2.714.36 22.090.33 00 5,75 ,35 2,09 7,31 6 HULL 11050 10293 -757 4 10.672, 7.726.340 2.368.711 1.240.49 4.117.139 00 ,72 ,03 0,30 ,39 6 IPSWICH 12987 13175 188 5 13.081, 85.641.53 5.381.461 21.437.4 58.822.66 00 7,40 ,58 09,95 5,87 6 KINGSTON 11780 12629 849 6 12.204, 49.210.26 6.912.890 6.001.50 36.295.86 00 1,84 ,14 7,38 4,32 6 LAKEVILL 9821 10602 781 7 E 10.212, 93.548.11 6.266.380 28.520.6 58.761.07 00 6,10 ,73 65,20 0,17 6 LANCAST 7380 8055 675 8 ER 7.718,0 72.477.53 5.747.417 5.313.14 61.416.96 0 3,44 ,89 6,86 8,69 6 LAWRENC 72043 76377 4334 9 E 74.210, 19.238.17 10.157.22 1.294.58 7.786.363 00 3,74 2,00 8,18 ,56 7 LEXINGTO 30355 31394 1039

187 0 N 30.874, 43.093.41 9.075.687 1.958.80 32.058.92 00 3,64 ,31 5,01 1,32 7 LINCOLN 8056 6362 -1694 1 7.209,0 38.800.94 3.519.122 4.425.80 30.856.01 0 3,94 ,04 9,94 1,96 7 LITTLETO 8184 8924 740 2 N 8.554,0 45.351.16 5.644.772 4.570.11 35.136.28 0 8,62 ,88 4,54 1,20 7 LOWELL 105167 106519 1352 3 105.84 37.630.65 15.760.32 2.616.96 19.253.37 3,00 7,87 2,36 1,28 4,23 7 LYNN 89050 90329 1279 4 89.690, 29.910.09 12.486.40 2.469.36 14.954.32 00 4,17 3,84 7,81 2,52 7 LYNNFIEL 11542 11596 54 5 D 11.569, 27.046.60 4.052.119 3.959.55 19.034.93 00 1,42 ,25 0,66 1,51 7 MALDEN 56340 59450 3110 6 57.895, 13.127.93 7.205.063 106.982, 5.815.885 00 1,50 ,81 57 ,12 7 MANCHES 5228 5136 -92 7 TER 5.182,0 20.164.50 2.237.103 2.237.10 15.690.29 0 0,47 ,52 3,52 3,43 7 MANSFIEL 22414 23184 770 8 D 22.799, 52.902.10 8.789.856 4.236.25 39.875.99 00 4,54 ,47 0,81 7,26 7 MARBLEH 20377 19808 -569 9 EAD 20.092, 11.326.81 4.197.372 244.656, 6.884.783 00 2,54 ,84 64 ,06 8 MARLBOR 36255 38499 2244 0 OUGH 37.377, 57.241.19 12.011.94 4.717.39 40.511.85 00 1,63 1,08 2,16 8,39 8 MARSHFIE 24324 25132 808 1 LD 24.728, 73.955.19 8.649.453 12.806.9 52.498.77 00 7,07 ,39 63,74 9,94 8 MAYNAR 10433 10106 -327 2 D 10.270, 13.903.41 2.483.314 1.171.32 10.248.76 00 0,00 ,32 9,12 6,56 8 MEDFIELD 12273 12024 -249 3 12.148, 37.955.73 4.044.229 6.294.66 27.616.83 00 0,41 ,03 1,40 9,98 8 MEDFORD 55765 56173 408 4 55.969, 21.880.87 8.569.100 1.105.60 12.206.16 00 8,05 ,21 7,96 9,88 8 MEDWAY 12448 12752 304 5 12.600, 30.216.56 3.942.380 2.224.06 24.050.11 00 6,71 ,22 7,03 9,46 8 MELROSE 27134 26983 -151 6 27.058, 12.348.56 4.189.917 320.695, 7.837.952 00 5,78 ,68 63 ,47 8 MENDON 5286 5839 553 7 5.562,0 46.586.50 3.055.503 3.099.94 40.431.05 0 2,60 ,09 1,87 7,64 8 MERRIMA 6138 6338 200 8 C 6.238,0 23.057.17 2.264.470 1.652.92 19.139.78

188 0 8,08 ,85 0,34 6,89 8 METHUEN 43789 47255 3466 9 45.522, 59.557.81 12.552.75 4.886.47 42.118.59 00 3,69 2,48 0,46 0,75 9 MIDDLEB 19927 23116 3189 0 OROUGH 21.522, 187.001.1 13.705.70 45.231.3 128.064.0 00 09,86 9,30 87,68 12,88 9 MIDDLET 7744 8987 1243 1 ON 8.366,0 37.520.73 3.707.048 5.514.13 28.299.55 0 0,69 ,93 0,44 1,32 9 MILFORD 26799 27999 1200 2 27.399, 38.909.66 7.923.312 2.664.76 28.321.58 00 5,27 ,71 6,68 5,88 9 MILLIS 7902 7891 -11 3 7.896,0 31.760.62 2.833.861 6.267.99 22.658.76 0 4,08 ,30 4,81 7,97 9 MILLVILL 2724 3190 466 4 E 2.957,0 12.863.78 1.132.085 575.124, 11.156.57 0 2,68 ,51 76 2,41 9 MILTON 26062 27003 941 5 26.532, 34.464.52 6.177.461 2.025.31 26.261.74 00 4,19 ,78 6,69 5,72 9 NAHANT 3632 3410 -222 6 3.521,0 3.249.309 880.816,5 564.033, 1.804.459 0 ,47 4 12 ,81 9 NATICK 32170 33006 836 7 32.588, 41.337.98 8.733.016 4.335.59 28.269.37 00 1,60 ,28 3,29 2,03 9 NEEDHAM 28924 28886 -38 8 28.905, 32.947.76 7.576.580 3.708.45 21.662.73 00 6,82 ,40 4,64 1,78 9 NEWBURY 6717 6666 -51 9 6.692,0 62.628.19 3.481.428 20.836.6 38.310.11 0 8,76 ,72 57,91 2,13 1 NEWBURY 17189 17416 227 0 PORT 17.302, 22.644.70 4.721.518 1.838.19 16.084.99 0 00 9,41 ,19 6,50 4,72 1 NEWTON 83829 85146 1317 0 84.488, 47.019.52 16.662.43 1.553.45 28.803.62 1 00 0,88 1,41 9,88 9,59 1 NORFOLK 10460 11227 767 0 10.844, 39.875.15 3.768.750 4.516.76 31.589.63 2 00 1,89 ,41 3,77 7,71 1 NORTH 27202 28352 1150 0 ANDOVER 27.777, 71.785.07 9.781.970 8.332.17 53.670.92 3 00 6,42 ,20 7,25 8,97 1 NORTH 13837 14892 1055 0 READING 14.364, 34.936.71 5.638.685 4.140.52 25.157.50 4 00 0,79 ,46 0,10 5,23 1 NORTON 18036 19031 995 0 18.534, 75.918.56 7.176.126 13.870.9 54.871.49 5 00 1,77 ,09 43,51 2,17 1 NORWELL 9765 10506 741 0 10.136, 54.975.19 4.762.735 8.026.38 42.186.07 6 00 8,71 ,60 4,61 8,50

189 1 NORWOO 28587 28602 15 0 D 28.594, 27.193.61 8.645.782 3.037.61 15.510.21 7 00 6,79 ,70 9,46 4,63 1 PEABODY 48134 51251 3117 0 49.692, 43.520.40 13.387.71 2.660.99 27.471.69 8 00 6,41 2,42 6,92 7,07 1 PEMBROK 16927 17837 910 0 E 17.382, 60.993.78 6.861.154 14.825.9 39.306.63 9 00 2,37 ,12 96,33 1,92 1 PEPPEREL 11142 11497 355 1 L 11.320, 60.068.09 4.469.198 3.324.88 52.274.01 0 00 9,32 ,29 7,39 3,64 1 PLAINVIL 7683 8264 581 1 LE 7.974,0 29.776.04 4.362.360 2.434.61 22.979.06 1 0 2,24 ,17 5,47 6,60 1 PLYMOUT 51701 56468 4767 1 H 54.084, 266.178.9 27.253.34 25.408.2 213.517.2 2 00 22,91 1,16 89,59 92,16 1 PLYMPTO 2637 2820 183 1 N 2.728,0 39.121.04 2.869.928 9.925.07 26.326.03 3 0 1,82 ,03 8,52 5,27 1 QUINCY 88025 92271 4246 1 90.148, 44.840.77 14.967.95 4.284.32 25.588.49 4 00 7,05 7,03 5,98 4,04 1 RANDOLP 30963 32112 1149 1 H 31.538, 26.971.97 6.235.209 2.994.65 17.742.10 5 00 5,03 ,32 9,37 6,34 1 RAYNHA 11739 13383 1644 1 M 12.561, 53.774.79 7.207.969 7.343.19 39.223.63 6 00 7,24 ,02 4,59 3,63 1 READING 23708 24747 1039 1 24.228, 25.874.49 5.185.104 5.109.91 15.579.47 7 00 7,04 ,69 9,85 2,50 1 REVERE 47247 51755 4508 1 49.501, 15.961.56 7.739.138 2.956.47 5.265.954 8 00 4,47 ,45 1,71 ,31 1 ROCKLAN 17670 17489 -181 1 D 17.580, 26.209.25 4.944.818 3.919.18 17.345.25 9 00 0,36 ,19 1,51 0,66 1 ROCKPOR 7767 6952 -815 2 T 7.360,0 18.173.63 4.944.818 1.008.03 12.220.78 0 0 4,24 ,19 0,75 5,30 1 ROWLEY 5500 5856 356 2 5.678,0 48.015.90 3.119.261 11.304.8 33.591.76 1 0 2,20 ,83 73,89 6,48 1 SALEM 40402 41340 938 2 40.871, 21.576.22 8.090.837 842.777, 12.642.60 2 00 2,19 ,92 32 6,95 1 SALISBUR 7827 8283 456 2 Y 8.055,0 40.926.01 4.288.956 12.364.2 24.272.78 3 0 2,88 ,94 66,04 9,90 1 SAUGUS 26078 26628 550 2 26.353, 29.371.04 7.798.816 3.896.46 17.675.76 4 00 1,98 ,94 2,81 2,23 1 SCITUATE 17863 18133 270

190 2 17.998, 44.862.40 5.859.523 7.253.73 31.749.14 5 00 9,24 ,74 7,49 8,01 1 SHARON 17408 17612 204 2 17.510, 63.334.44 6.022.301 7.243.89 50.068.25 6 00 8,65 ,95 1,29 5,41 1 SHERBOR 4200 4119 -81 2 N 4.160,0 41.841.24 2.149.712 4.452.59 35.238.93 7 0 0,89 ,59 7,76 0,54 1 SHIRLEY 6373 7211 838 2 6.792,0 41.177.88 3.362.522 2.464.59 35.350.76 8 0 0,36 ,25 1,83 6,28 1 SOMERVIL 77478 75754 -1724 2 LE 76.616, 10.698.16 8.241.635 32.124,2 2.424.400 9 00 0,36 ,79 1 ,36 1 SOUTHBO 8781 9767 986 3 ROUGH 9.274,0 40.205.35 4.838.714 4.900.08 30.466.55 0 0 8,50 ,78 5,38 8,34 1 STONEHA 22219 21437 -782 3 M 21.828, 17.213.71 4.953.268 1.823.69 10.436.74 1 00 4,04 ,57 5,54 9,93 1 STOUGHT 27149 26962 -187 3 ON 27.056, 42.644.40 7.768.140 2.987.62 31.888.64 2 00 6,87 ,01 5,07 1,79 1 STOW 5902 6590 688 3 6.246,0 46.608.28 2.895.983 5.617.03 38.095.26 3 0 6,51 ,26 3,28 9,97 1 SUDBURY 16841 17659 818 3 17.250, 64.092.08 6.371.480 7.416.37 50.304.22 4 00 1,29 ,88 5,30 5,11 1 SWAMPSC 14412 13787 -625 3 OTT 14.100, 8.030.877 3.047.403 359.355, 4.624.118 5 00 ,99 ,32 95 ,72 1 TAUNTON 55976 55874 -102 3 55.925, 124.959.8 18.444.40 16.946.4 89.568.99 6 00 92,70 3,66 94,27 4,77 1 TEWKSBU 28847 28961 114 3 RY 28.904, 54.771.58 9.493.179 4.785.50 40.492.89 7 00 0,86 ,14 7,57 4,15 1 TOPSFIEL 6141 6085 -56 3 D 6.113,0 33.164.42 2.647.882 4.920.64 25.595.90 8 0 7,89 ,05 4,83 1,01 1 TOWNSEN 9198 8926 -272 3 D 9.062,0 85.405.10 4.077.107 3.560.09 77.767.90 9 0 5,51 ,39 0,26 7,86 1 TYNGSBO 11081 11292 211 4 ROUGH 11.186, 46.800.58 5.013.704 4.030.45 37.756.41 0 00 2,87 ,61 9,61 8,65 1 UPTON 5642 7542 1900 4 6.592,0 56.547.64 3.670.105 4.224.41 48.653.11 1 0 1,91 ,19 7,35 9,37 1 WAKEFIEL 24804 24932 128 4 D 24.868, 20.704.63 6.275.487 2.877.79 11.551.35 2 00 6,13 ,26 5,53 3,34 1 WALPOLE 22824 24070 1246 4 23.447, 54.554.21 8.494.618 5.951.08 40.108.51

191 3 00 9,63 ,02 5,76 5,85 1 WALTHA 59226 60632 1406 4 M 59.929, 35.649.43 12.848.81 3.059.77 19.740.84 4 00 0,24 9,44 0,61 0,19 1 WAREHA 20335 21822 1487 4 M 21.078, 96.101.33 12.269.63 17.201.0 66.630.60 5 00 6,28 2,14 94,83 9,31 1 WATERTO 32986 31915 -1071 4 WN 32.450, 10.681.44 5.258.087 351.803, 5.071.556 6 00 6,61 ,05 29 ,27 1 WAYLAN 13100 12994 -106 4 D 13.047, 41.063.12 4.122.368 6.780.27 30.160.48 7 00 3,72 ,36 4,88 0,48 1 WELLESL 26604 27982 1378 4 EY 27.293, 27.269.57 6.671.186 1.634.15 18.964.23 8 00 6,39 ,58 6,29 3,52 1 WENHAM 4440 4875 435 4 4.658,0 21.085.80 1.656.085 5.339.10 14.090.60 9 0 0,09 ,66 9,37 5,06 1 WEST 6634 6916 282 5 BRIDGEW 6.775,0 40.594.00 4.178.756 7.956.17 28.459.07 0 ATER 0 7,16 ,81 2,50 7,85 1 WEST 4149 4235 86 5 NEWBURY 4.192,0 38.137.76 2.044.766 5.106.82 30.986.17 1 0 8,26 ,87 2,28 9,11 1 WESTFOR 20754 21951 1197 5 D 21.352, 81.226.63 9.143.410 9.236.07 62.847.14 2 00 4,07 ,85 7,16 6,06 1 WESTON 11465 11261 -204 5 11.363, 44.885.98 5.657.816 2.871.25 36.356.91 3 00 1,63 ,36 1,42 3,85 1 WESTWOO 14117 14618 501 5 D 14.368, 28.965.19 5.177.484 1.755.70 22.032.00 4 00 2,71 ,27 0,69 7,75 1 WEYMOU 53988 53743 -245 5 TH 53.866, 46.183.62 11.623.50 5.271.99 29.288.11 5 00 0,40 5,21 9,39 5,80 1 WHITMAN 13882 14489 607 5 14.186, 18.022.24 3.457.571 2.270.97 12.293.69 6 00 5,92 ,35 8,10 6,47 1 WILMING 21367 22325 958 5 TON 21.846, 44.399.12 9.700.889 4.835.30 29.862.93 7 00 3,03 ,87 3,05 0,11 1 WINCHES 20810 21374 564 5 TER 21.092, 16.439.01 4.991.372 868.279, 10.579.36 8 00 5,81 ,61 24 3,96 1 WINTHRO 18303 17497 -806 5 P 17.900, 5.897.573 2.363.820 874.648, 2.659.105 9 00 ,54 ,07 10 ,37 1 WOBURN 37258 38120 862 6 37.689, 33.522.27 12.057.32 1.594.61 19.870.32 0 00 4,04 8,84 6,73 8,47 1 WRENTHA 10554 10955 401 6 M 10.754, 58.538.61 6.190.167 4.783.81 47.564.63 1 00 7,35 ,87 2,27 7,21

192 Totals 1453 4.209.1 4.354.4 05 4.281.8 7.237.619 1.131.525 852.421. 5.253.672 66,00 71,00 25,00 .944,17 .341,45 794,70 .808,02 Average 26.143, 27.046, 902,5 26.595, 44.954.16 7.028.107 5.294.54 32.631.50 89 40 2 19 1,14 ,71 5,31 8,12 Table 5.8; List of towns included in the study area with basic population, area, and land cover data.

193

LULC Code-2005 LULC Area (m2) Perce Description of LULC categories (MassGIS) Code nt of Study Area Cropland 1 1,73% Generally tilled land used to grow row crops. 125.110.023, Boundaries follow the shape of the fields and 06 include associated buildings (e.g., barns). This category also includes turf farms that grow sod Pasture 2 1,25% Fields and associated facilities (barns and 90.185.069,5 other outbuildings) used for animal grazing 3 and for the growing of grasses for hay. Orchard 35 0,17% Fruit farms and associated facilities 12.257.193,3 7 Nursery 36 0,13% Greenhouses and associated buildings as well 9.695.242,02 as any surrounding maintained lawn. Christmas tree (small conifer) farms are also classified as Nurseries. Agr. 237.247.527, 98 Forest 3 41,08 Areas where tree canopy covers at least 50% 2.972.983.78 % of the land. Both coniferous and deciduous 8,46 forests belong to this class. Brushland/Success 40 0,24% Predominantly (> 25%) shrub cover, and ional 17.373.151,8 some immature trees not large or dense 2 enough to be classified as forest. It also includes areas that are more permanently shrubby, such as heath areas, wild blueberries or mountain laurel Forest 2.990.356.94 0,28 Open Land 6 1,21% Vacant land, idle agriculture, rock outcrops, 87.512.503,5 and barren areas. Vacant land is not 4 maintained for any evident purpose and it does not support large plant growth. Powerline/Utility 24 0,59% Powerline and other maintained public utility 42.848.973,0 corridors and associated facilities, including 3 power plants and their parking areas. Open land 130.361.476, 57 Saltwater Sandy 25 0,51% DEP Wetlands (1:12,000) WETCODEs 1, 2,

Beach 36.851.736,2 3, 6, 10, 13, 17 and 19 7 Participation 7 0,96% Facilities used by the public for active Recreation 69.278.739,8 recreation. Includes ball fields, tennis courts, 1 basketball courts, athletic tracks, ski areas, playgrounds, and bike paths plus associated parking lots. Primary and secondary school recreational facilities are in this category, but university stadiums and arenas are considered

194 Spectator Recreation. Recreation facilities not open to the public such as those belonging to private residences are mostly labeled with the associated residential land use class not participation recreation. However, some private facilities may also be mapped. Spectator 8 0,07% University and professional stadiums Recreation 5.005.543,52 designed for spectators as well as zoos, amusement parks, drive-in theaters, fairgrounds, race tracks and associated facilities and parking lots. Water-Based 9 0,03% Swimming pools, water parks, developed Recreation 2.163.038,86 freshwater and saltwater sandy beach areas and associated parking lots. Also included are scenic areas overlooking lakes or other water bodies, which may or may not include access to the water (such as a boat launch). Water- based recreation facilities related to universities are in this class. Private pools owned by individual residences are usually included in the Residential category. Marinas are separated into code 29. Golf Course 26 0,87% Includes the greenways, sand traps, water 62.782.557,5 bodies within the course, associated buildings 5 and parking lots. Large forest patches within the course greater than 1 acre are classified as Forest (class 3). Does not include driving ranges or miniature golf courses Marina 29 0,02% Include parking lots and associated facilities 1.755.207,53 but not docks (in class 18 Recreati on 177.836.823, 54 Multi-Family 10 3,12% Duplexes (usually with two front doors, two Residential 225.855.059, entrance pathways, and sometimes two 32 driveways), apartment buildings, condominium complexes, including buildings and maintained lawns. Note: This category was difficult to assess via photo interpretation, particularly in highly urban areas. High Density 11 3,68% Housing on smaller than 1/4 acre lots. See Residential 266.335.544, notes below for details on Residential 98 interpretation Medium Density 12 6,91% Housing on 1/4 - 1/2 acre lots. See notes Residential 500.108.414, below for details on Residential interpretation 68 Low Density 13 8,57% Housing on 1/2 - 1 acre lots. See notes below Residential 620.566.580, for details on Residential interpretation 64 Very Low Density 38 1,94% Housing on > 1 acre lots and very remote, Residential 140.652.680, rural housing. See notes below for details on 91 Residential interpretation.

195 Urban 31 1,73% Lands comprising schools, churches, colleges, Public/Institutional 125.468.684, hospitals, museums, prisons, town halls or 96 court houses, police and fire stations, including parking lots, dormitories, and university housing. Also may include public open green spaces like town commons. Commercial 15 2,32% Malls, shopping centers and larger strip 167.942.285, commercial areas, plus neighborhood stores 29 and medical offices (not hospitals). Lawn and garden centers that do not produce or grow the product are also considered commercial. Industrial 16 1,84% Light and heavy industry, including buildings, 133.049.169, equipment and parking areas 41 Transportation 18 1,57% Airports (including landing strips, hangars, 113.709.881, parking areas and related facilities), railroads 57 and rail stations, and divided highways (related facilities would include rest areas, highway maintenance areas, storage areas, and on/off ramps). Also includes docks, warehouses, and related land-based storage facilities, and terminal freight and storage facilities. Roads and bridges less than 200 feet in width that are the center of two differing land use classes will have the land use classes meet at the center line of the road (i.e., these roads/bridges themselves will not be separated into this class) Cemetery 34 0,43% Includes the gravestones, monuments, 31.357.599,7 parking lots, road networks and associated 3 buildings Junkyard 39 0,06% Includes the storage of car, metal, machinery 4.463.256,68 and other debris as well as associated buildings as a business Mining 5 0,29% Includes sand and gravel pits, mines and 20.745.274,0 quarries. The boundaries extend to the edges 9 of the site’s activities, including on-site machinery, parking lots, roads and buildings. Transitional 17 0,39% Open areas in the process of being developed 28.270.644,4 from one land use to another (if the future 7 land use is at all uncertain). Formerly identified as "Urban Open" Waste Disposal 19 0,14% Landfills, dumps, and water and sewage 10.472.701,7 treatment facilities such as pump houses, and 5 associated parking lots. Capped landfills that have been converted to other uses are coded with their present land use. Urban 2.388.997.77 8,48 Forested Wetland 37 8,24% DEP Wetlands (1:12,000) WETCODEs 14,

596.196.348, 15, 16, 24, 25 and 2 34 Water 20 3,34% DEP Wetlands (1:12,000) WETCODEs 9 and

241.865.643, 22

196 95 Non-Forested 4 4,08% DEP Wetlands (1:12,000) WETCODEs 4, 7,

Wetland 295.238.654, 8, 12, 23, 18, 20, and 21. 78 Saltwater Wetland 14 1,45% DEP Wetlands (1:12,000) WETCODEs 11

104.763.752, and 27 16 Cranberry Bog 23 1,03% Both active and recently inactive cranberry 74.454.759,0 bogs and the sandy areas adjacent to the bogs 8 that are used in the growing process. Impervious features associated with cranberry bogs such as parking lots and machinery are included. Modified from DEP Wetlands

(1:12,000) WETCODE 5. Water 1.312.519.15 8,31

7.237.319.70 5,16 Table 5.9: Table of land-use classes used in the study (MassGIS).

197

Data Set Data Resolution and source of Information Date Type Community Vector Created by Mass GIS from latitude and 2011 Boundaries polygons longitude coordinates found in the 68- (Cities and volume Harbor and Lands Commission Town Towns) from Boundary Atlas and cross checked with USGS Survey Points quad sheets; error ±12feet. MA State Vector, MassGIS; Digitized from 1:25,000 linework 1991 Outlines Polygons from U.S. Geological Srvey Mylar map sheets Boston Vector, US Census Bureau 1999 Metropolitan Polygons http://www.whitehouse.gov/omb/inforeg_statpo Statistical Area licy#ms boundaries New England Vector, Mass GIS; No information available 2007 Boundaries Polygons Regional Vector, Mass GIS; Derived from town layer and list of 2007 Planning Polygons towns included within regional planning agency Agencies (RPA) Level 3 Vector, Mass GIS; Assessors surveys and Upda Assessors' Polygons documentation, high resolution ted Parcels 2013 Impervious Raster Mass GIS; 1 meter pixel size; reclassified from 2005 Surface 50 cm resolution near-infrared imagery derived from an aerial Vexcel Ultra Cam sensor. Land Use (2005) Vector, Delineating and coding the data was carried out Refer Polygons using semi-automated methods for the ence classification of the 4-band 2005 ortho imagery date with resolution of 50cm using attributes from 2005 the manually-compiled 1999 data, and assessor Avail parcels and other ancillary data. Smallest able mapping unit is 1 acre with exceptions of ¼ 2007- acres where assessor parcel maps were used to 2009 refine residential areas. Building Vector, Two-dimensional roof outlines for all buildings 2012 Structures (2-D, Polygons larger than 15m2 (150 square feet), interpreted from 2011-2012 by third party contractor using LiDAR sensor Ortho Imagery) color imagery with 30 cm. Mass DOT Vector, Center lines, from transportation department Upda Roads Arcs surveys including detailed information such as te width of paving, easement width, type, class, 2012 usage, destination cities and states etc. Protected and Vector, From the Department of conservation and Upda Recreational Polygons recreation. Digitized from survey maps. ted Open Space 2013

198 Forest Vector, From the Department of conservation and 2008 Stewardship Polygons recreation. Digitized from survey maps. Program Properties BioMap2 Vector, Mass GIS; The Massachusetts Natural Heritage Upda Polygons & Endangered Species Program and The Nature te Conservancy’s Massachusetts Program 2010 developed BioMap2 in 2010 as a conservation plan to protect the state’s biodiversity. http://www.nhesp.org/ Color Ortho Raster; 50cm, pan sharpened. 2005 Imagery (2005) JPEG or MrSID; Table 5.10: Original data set used to develop study data layers (adapted from MassGIS web site, http://www.mass.gov/anf/research-and-tech/it-serv-and-support/application- serv/office-of-geographic-information-massgis/datalayers/layerlist.html)

199

Land use CU CT Tree size Description of available pervious surface and classes and related remarks Categories Intense urban use: low use – low % pervious Industrial 0.0 1.0 Small- Spaces between buildings, roads and parking with no medium obvious use. Soil would require amendment Junkyard 0.7 0.3 Small- Areas not covered by cars and metal junk. Soil would medium require amendment Mining 1.0 0.0 n/a Not preferred since ground activities will impact tree health. Once capped, surface area 100% planted Transportatio 0.8 0.2 Small Spaces between buildings and roads with no obvious n use. Soil would require amendment Commercial 0 1.0 Small Spaces between buildings and roads with no obvious use. Soil would require amendment Marina 0.9 0.1 Small- medium Residential : Medium to high use -high % pervious High Density 0.9 0.1 Small Narrow spaces between building structure and Residential boundary limits and potential single trees in backyards, depending on configuration of property. All remaining surfaces as lawn or shrub are contributing to temperature amelioration. Multi-Family 0.8 0.2 Small- Variable spaces available especially around common Residential medium parking. All remaining surfaces as lawn or shrub are contributing to temperature amelioration. Medium 0.8 0.2 Small Spaces available along side and front setbacks as well Density as backyard All remaining surfaces as lawn or shrub Residential are contributing to temperature amelioration. Low Density 0.7 0.3 Large Spaces available alongside and front setbacks as well Residential as backyard. All remaining surfaces as lawn or shrub are contributing to temperature amelioration. Very Low 0.6 0.4 Large Spaces available alongside and front setbacks as well Density as backyard All remaining surfaces as lawn or shrub Residential are contributing to temperature amelioration. Urban Public/Institutional: High use intensity - medium %pervious Urban 0.5 0.5 Medium- This land use category includes formal open space Public/Institu large and spaces around institutional buildings. To tional accommodate for the low use of spaces around buildings and intense use of public parks, an equal value is included for both coefficients. Recreation: High use Intensity - high %pervious Participation 0.9 0.1 Medium These categories are delineated in the land use layer Recreation around the actual playing fields. Any forest or scrub

200 land is included in other land use categories such as forest and brush land. Spectator 0 1 Small- Spaces between buildings, roads and parking with no Recreation medium obvious use. Soil would require amendment Water-Based 0.85 0.15 Small- At front and near water edges and around facilities Recreation medium when present. Golf Course 0.85 0.15 Medium- Spaces available between par centerlines. If design large standards are changed, then more available land for tree planting could be accommodated. Service: Low use intensity - high % pervious Cemetery 0.8 0.2 Medium- These categories are delineated in the land use layer large around the actual burial areas. Any forest or scrub land is included in other land use categories such as forest and brush land. Nursery 0.85 0.15 Medium- These categories include the surfaces where large vegetation is planted or stored and Christmas tree plantations. Space is available at front and around buildings for planting. Open Land 0.2 0.8 large No obvious use and potentially 100% for tree planting. A factor of 0.8 is included to account for the presence of rock outcrops. Powerline/Ut 1.0 0 none The presence of electrical cables does not allow the ility growth of trees. Yet, as pervious surfaces covered with ground cover they contribute to temperature reduction. Waste 0.85 0.15 Medium- Space available at entrances and around facility Disposal large buildings. This is available land for 100% tree planting when capped. Excluded Transitional n/a n/a n/a Final land use is not identified Table 5.11: Detailed information on calculation of coefficient of use by land use class

201 BIBLIOGRAPHY

(APA) America Planning Association (2007). How cities use parks for climate change management, City Parks Forum Briefing Papers. Chicago, IL: American Planning Association. 4pp. Online available at http://www.planning.org/cityparks/briefingpapers/climatechange.htm. Last accessed March 01,2013 .

(MEA)Millennium Ecosystem Assessment Report (2005).World Bank. www.milleniumassessment.org last accessed February 2013.

Abunnasr,Y. and Hamin,E.(2012). The Green Infrastructure Transect: An Organizational Framework for Mainstreaming Adaptation Planning Policies. In Resilient Cities 2, Local Sustainability 2, K. Otto-Zimmermann (ed.), Springer.

Adger, W.N., Arnell, N.W., and Tompkins, E.L. (2005). Successful adaptation to climate change across scales. Global Environmental Change, 15, 77-86.

Ahern, J. (2008). Green Infrastructure for Cities: The Spatial Dimension. In Cities of the Future: Towards Integrated Sustainable Water and Landscape Management, Novotny, V. & Brown, P. (Ed.), IWA publishing, London.

Akbari, H. (2005). Energy Saving Potentials and Air Quality Benefits of Urban Heat Island Mitigation. 2008 from http://www.osti.gov/bridge/servlets/purl/860475- UlHWIq/860475.PDF Last Accessed March 2013.

Akbari, H., 2002. “Shade trees reduce building energy use and CO2 emissions from power plants,” Environmental Pollution, 116:S119-S126.

Akbari, H., M. Pomerantz, and Taha, H. (2001). Cool Surfaces and Shade Trees to Reduce Energy Use and Improve Air Quality in Urban Areas. Solar Energy 70(3): 295-310.

Akbari,H., Rose,L.S., and Taha, H. (2003).Analyzing the land cover of an urban environment using high-resolution orthophotos. Landscape and Urban Planning, 63, 1-14.

Alberti, M. (1999). Urban Patterns and Environmental Performance: What Do We Know? Journal of Planning Education and Research, 19, 151-163.

Alberti, M. (2005). The Effects of Urban Patterns on Ecosystem Function. International Regional Science Review, 28(2),168–192

Alberti, M. (2009). Advances in urban ecology: integrating Humans and Ecological Processes in urban ecosystems. Springer, New York, NY.

202 Alcoforado, M.J. & Andrade, H. (2008). Global Warming and the Urban Heat Island. p.249-262. In: Marzluff, J.M., G.Bradley, E. Shulenberger, C. Ryan, W. Endlicher, U. Simon, M. Alberti, & C. ZumBrunnen (eds.), Urban ecology: An international perspective on the interaction between humans and nature, New York, Springer.

Alley, R.B., Marotzke, J., Nordhaus, W.D., Overpeck, J.T., Peteet, D.M., Pielke, R.A., Pierrehumbert, R.T., Rhines, P.B., Stocker, T.F., Talley, L.D., Wallace, J.M. (2003) Abrupt Climate Change. Science 299, 2005-2010.

Amati,M. and Taylor,L. (2010).From Green Belts to Green Infrastructure. Planning, Practice & Research, 25(2),143–155.

Arnfield, A.J. (2002). Two decades of urban climate research: A review of turbulence, exchanges of energy and water, and the urban heat island. International Journal of Climatology, 23: 1-26.

Arnold, C.L.J., Gibbons, C.J. (1996) Impervious Surface Coverage: The Emergence of a Key Environmental Indicator. Journal of the American Planning Association 62, 243-258.

Ashley, R. et al. (2007). Adaptable Urban Drainage: Addressing Change in Intensity, Occurrence and uncertainty. Built Environment.1(33): 71-84.

BAF: Biotope area factor. Berlin City, Senate Department for Urban Planning. (March 27, 2010) http://www.stadtentwicklung.berlin.de/umwelt/landschaftsplanung/bff/index_en .shtml

Bailey, T.C & Gatrell, A.C. (1995). Interactive Spatial Data Analysis. Prentice Hall, London.

Barnett, J. & Beckman, J. (2007). Coastal cities face rising seas. Planning, 73 (8), 34-37.

Barnett, J., O Neill, S. (2010) Maladaptation. Global Environmental Change 20, 211- 213.

Bedsworth, L.W., Hanak, E. (2010) Adaptation to Climate Change --A Review of Challenges and Tradeoffs in Six Areas. Journal of the American Planning Association 76, 477 -495.

Belanger, Pierre (2009). Landscape as infrastructure. Landscape Journal 28(1),79-95.

Benedict, M. A., & McMahon, M. T. (2002). Green infrastructure: Smart conservation for the 21st century. Renewable Resources Journal, 20, 12–17.

203 Benedict, M. A., & McMahon, M. T. (2006). Green infrastructure: Linking landscape and communities. Washington, DC: Island Press.

Beranová, R.; Huth, R. (2005) – Long-term changes in the heat island of Prague under different synoptic conditions. Theoretical and Applied Climatology, 82 (1–2): 113–118.

Berger, A. (2006). Drosscape: Wasting land in urban America. New York: Princeton Architectural Press.

Berke, P.R. (2007). The Evolution of Green Community Planning, Scholarship, and Practice. Journal of the American Planning Association, 74(4), 393-407.

Betts, R.A., Collins, M., Hemming, D.L., Jones, C.D., Lowe, J.A., Sanderson, M.G. (2011) When could global warming reach 4 degrees C? Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 369, 67-84.

Bickman, L., Rog, D., Hedrick, T. (1998). Applied Research design: A practical Approach. In Handbook of Applied Social Research Handbook, Bickman Ed. Sage Publications. 5-38

Biesbroek, G.R., Swart, R.J., Carter, T.R., Cowan, C., Henrichs, T., Mela, H., Morecroft, M.D., Rey, D. (2010) Europe adapts to climate change: Comparing National Adaptation Strategies. Global Environmental Change 20, 440-450.

Blakely, E. (2007). Urban Planning for Climate Change (Working Paper). Lincoln Institute of Land Policy.

Blaschke, T.(2010). Object based image analysis for remote sensing, ISPRS Journal of Photogrammetry and Remote Sensing, 2-16.

Bolstad, P. (2006). GIS Fundamentals: A First text on geographic Information Systems, Second Edition. Eider Press, White Bear Lake, Minnesota.

Bolund,P. and Hunhammar,S. (1999). Ecosystem Services In Urban Areas. Ecological Economics 29,293–301.

Boswell, M.R., Greve, A.I., Seale, T.L. (2012) Local Climate Action Planning. Island Press, Washington DC.

Boulanger, P.M. (2008) Sustainable Development Indicators: A Scientific Challenge, a Democratic Issue. SAPIENS 1, 245-259.

Bowler, D.E., Buyung-Ali, L., Knight,T.M., Andrew S. Pullin,A.S., (2010).Urban greening to cool towns and cities: A systematic review of empirical evidence. Landscape and Urban Planning, 97,145-155.

204 Brabec, E. (2009) Imperviousness and Land Use Policy: Toward an Effective Approach to Watershed Planning. Journal of Hydraulic Engineering 14, 425-433.

Brabec, E., Schulte, S., and Richards, P.L.(2002). Impervious Surfaces and Water Quality: A Review of Current Literature and Its Implications for Watershed Planning. Journal of Planning Literature,16(4), 499-514.

Brabec, Elizabeth & Geoffrey McD. Lewis. (2002). “Defining the Pattern of the Sustainable Urban Region,” C. A. Brebbia, J. F. Martin-Duque and L. C. Wadhwa, eds. The Sustainable City. Southampton, England: WIT Press.

Brabec, Elizabeth. (2008). Imperviousness and Land Use Policy: Toward an effective approach to watershed planning. Journal of Hydrologic Engineering 14 (4): 425-433.

Brázdil, R.; Budíková, M. (1999) – An urban bias in air temperature fluctuations at the Klementinum, Prague, the Czech Republic. Atmospheric Environment, 33 (24– 25): 4211–4217.

Brody,S.D, Zahran,S., Grover,H., and Vedlitz,A.(2008). A spatial analysis of local climate change policy in the United States: Risk, stress, and opportunity. Landscape and Urban Planning, 87 ,33-41.

Brooks, M.P. (2002) Planning Theory for Practitioners. Planners Press: American Planning Association, Chicago.

Brooks, N., Adger, W.N., and Kelly, P.M. (2005). The determinants of vulnerability and adaptive capacity at the national level and the implications for adaptations. Global Environmental Change, 15, 151-163.

Brown, R.D. (2011). Ameliorating the effects of climate change: Modifying microclimates through design. Landscape and Urban Planning, 100, 372-374.

Busha, C. H., & Harter, S. P. (1980). Research methods in librarianship, techniques and interpretation. New York: Academic Press.

California Emergency Management Agency (Cal EMA), California Natural Resources Agency, (2012) California Climate Change Adaptation Policy Guide. California Emergency Management Agency and California Natural Resources Agency, Mather, CA.

Calthorpe, Peter (1996). The Next American Metropolis: Ecology, Community and the American Dream. Princeton, Princeton Architectural Press.

CAP. (2007/2010). The city of Boston’s climate action plan. City of Boston. December 2007. http:// www.abettercity.org/environment/bostoncap.html. Accessed 18 June 2011.

205 Carter, J.G. (2011) Climate change adaptation in European cities. Current Opinion in Environmental Sustainability 3, 193-198.

Carter, T. and Jackson, C.R. (2007). Vegetated roofs for stormwater management at multiple spatial scales. Landscape and Urban Planning, 80 (1),84-94.

Cathy Aniello, Ken Morgan, Arthur Busbey and Leo Newland, Mapping micro-urban heat islands using LANDSAT TM and a GIS, Computers & Geosciences, 21(8), 965-967

CCAP. (2011). The value of green infrastructure for urban climate adaptation. Center for Clean Air Policy, February 2011. http://www.ccap.org/docs/resources/989/Green Infrastructure FINAL. pdf. Accessed 18 June 2011.

CCP (2008).Cities for Climate Protection Australia Adaptation Initiative: Local Government Climate Change Adaptation Toolkit. Prepared by ICLEI Oceania www.iclei.org/oceania Last Accessed February 2013.

CDC. 2004. Extreme Heat: A Prevention Guide to Promote Your Personal Health and Safety. Centers for Disease Control and Prevention. http://www.bt.cdc.gov/disasters/extremeheat/heat_guide.asp. Last Accessed March 2013.

Chang, H. C. (1974). Library goals as responses to structural milieu requirements: A comparative case study. Unpublished doctoral dissertation, University of Massachusetts, Amherst.

Chiesura,A.(2004). The Role of Urban Parks for the Sustainable City. Landscape and Urban Planning 68, 129–138

City of Copenhagen (2011) Copenhagen Climate Adaptation Plan. City of Copenhagen, Copenhagen.

Clarence City Council (2009) Climate change impacts on Clarence coastal areas, Final Report. Clarence, TA City Council, Clarence, Tasmania.

Clark,C., Busiek,B. and Adriaens,P. (2010). Quantifying Thermal Impacts of Green Infrastructure: Review and Gaps. Cities of the Future/Urban River Restoration, Water Environmental Federation. LINK

CNT(2010).The Value of Green Infrastructure: A Guide to Recognizing Its Economic, Environmental and Social Benefits. Center for Neighborhood Technology (CNT). http://www.allianceforwaterefficiency.org/uploadedFiles/Resource_Center/Libr ary/Green_building/CNT-AmR-The-Value-of-Green-Infrastructure-2010.pdf Last Accessed January 2013.

206 COL (2010). Rising to the Challenge - The City of London Climate Change Adaptation Strategy. First Published May 2007 Revised and Updated January 2010. http://www.cityoflondon.gov.uk/services/environment-and- planning/sustainability/Documents/pdfs/SUS_AdaptationStrategyfinal_2010up date.pdf Last Accessed February 2013.

COL (2012). Tree Strategy Supplementary Planning Document, City of London, Local Development Framework (Adopted 15th May 2012). http://www.cityoflondon.gov.uk/services/environment-and- planning/planning/heritage-and-design/trees-and-open-spaces/Pages/Trees.aspx Last Accessed February 2013.

COL(2006). London’s Urban Heat Island: A Summary for Decision Makers. Mayor of London. www.london.gov.uk/mayor/environment/climate_change/ .

Condon, P. M., Cavens, D., & Miller, M. (2009). Urban Planning Tools for Climate Change Mitigation (Policy Focus Report. Lincoln Institute of Land Policy

Connors, J.P., Galletti, C.S., and Chow, W.T.L. (2013). Landscape configuration and urban heat island effects: assessing the relationship between landscape characteristics and land surface temperature in Phoenix, Arizona. Landscape Ecology 28, 271-283.

Corburn,J. (2009). Cities, Climate Change and Urban Heat Island Mitigation: Localising Environmental Science. Urban Studies, 46(2), 413-427.

Corner, J. (2004). Not unlike life itself: Landscape strategy now. Harvard Design Magazine (21) (Fall-Winter): 32-4.

Crutzen, P.J. (2004).New Directions: The growing urban heat and pollution “island” effect - impact on chemistry and climate. Atmospheric Environment, 38 (21): 3539–3540.

Currie, B. A., & Bass, B. (2008). Estimates of air pollution mitigation with green plants and green roofs using the UFORE model. Urban Ecosystems, 11(4), 409-422.

Currie, B.A and Bass, B. (2008).Estimates of air pollution mitigation with green plants and green roofs using the UFORE model. Urban Ecosystems, DOI 10.1007/s11252-008-0054-y

Davis, A.Y., Pijanowski, B.C., Robinson, K.D., and Kidwell, P.B. (2010). Estimating parking lot footprint in Upper Great Lakes Region of the USA. Landscape and Urban Planning, 96 (2), 68-77.

Davoudi,S., Crawford,J. and Mehmood,A. (2009). Planning for Climate Change Strategies for Mitigation and Adaptation for Spatial Planners. Earthscan, London.

207 De Graaf.R.E.(2008). Reducing flood vulnerability of urban lowland areas. Proceedings of the 11th International Conference on Urban Drainage, Edinburgh, Scotland, UK, 2008

De vries, J. and Wolsink, M. (2009). Making Space for Water: Spatial Planning and Water Management in the Netherlands. In Planning for Climate Change: Strategies for Mitigation and Adaptation for Spatial Planners, Davoudi, S., Crawford, J., and Mehmood, A. (Ed.). Earthscan, London, Sterling VA.

Dessai, S., Hulme, M. (2007) Assessing the robustness of adaptation decisions to climate change uncertainties: A case study on water resources management in the East of England. Global Environmental Change 17, 59-72.

Donadieu, P. (2006). Landscape Urbanism in Europe: From brownfields to sustainable urban development. Journal of Landscape Architecture (Autumn): 36-45.

Duany, A., & Talen, E. (2002). Transect planning. Journal of American Planning Association, 68(3), 245–266.

Duany, Andres, Elizabeth Plater-Zyberk and Jeff Speck (2005). Smart Growth: New Urbanism in American Communities.

DuMont, R. R. (1975). The large urban public library as an agency of social reform, 1890-1915. Unpublished doctoral dissertation, University of Pittsburgh, Pennsylvania.

EEA(2012).Urban adaptation to climate change in Europe: Challenges and opportunities for cities together with supportive national and European policies. European Environment Agency. Office for Official Publications of the European Union, Copenhagen.

Eisenhardt, K. M. (1989). Building theories from case study research. Academy of Management Review, 14(4), 352-550.

Ellis, J.B. (2013). Sustainable surface water management and green infrastructure in UK urban catchment planning. Journal of environmental Planning and Management, 56 (1), 24-41.

Emory, C. W., & Cooper, D. R. (1991). Business research methods. (4th ed.). Boston, MA: Irvin.

EPA (2008). Reducing Urban Heat Islands: Compendium of Strategies. Washington, D.C.: United States Environmental Protection Agency, Office of Atmospheric Programs. U.S. Environmental Protection Agency. Online available at http://www.epa.gov/heatisland/resources/compendium.htm. Last accessed 15 Feb 2013.

208 EPA (2010).Green Infrastructure Case Studies: Municipal Policies for Managing Strom water with Green Infrastructure. U.S. Environmental Protection Agency Online available at http://www.epa.gov/greeninfarstrcture.htm. Last accessed 15 Jan. 2013.

EPA . U.S. Environmental Protection Agency . "Heat Island Effect." Online available at http://www.epa.gov/heatisland/. Last Accessed February 2013.

European Environment Agency (2011) Observed global fossil fuel CO2 emissions compared with six scenarios from the IPCC. European Environment Agency.

European Environment Agency, (2012) Environmental Indicator Report 2012: Ecosystem resilience and resource efficiency in a green economy in Europe. European Environment Agency, Luxembourg.

European Environment Agency, World Health Organization, JRC European Commission, (2008) Impacts of Europe's changing climate : 2008 indicator- based assessment. Office for Official Publications of the European Communities, Luxembourg.

Ewing, R., Bartholomew, K., Winkelman, S., Walters, J., & Chen, D. (2007).Growing Cooler: The Evidence on Urban Development and Climate Change. ULI–the Urban Land Institute.

F´abos, J. G. (2004). Greenway planning in the United States: Its origins and recent case studies. Landscape and Urban Planning, 68, 321–342.

Fabos, J.Gy. & Ryan, R.L. (2006). An Introduction to greenway Planning Around the World. Landscape and urban Planning. 76(1-4), 1-6.

Farr, D. (2008). Sustainable urbanism: Urban design with nature. Hoboken: Wiley.

Feiden, W., Hamin, E.M. (2011) Assessing Sustainability: A Guide for Local Governments. American Planning Association (Planners Press), Chicago.

Few, R. (2007) Health and climatic hazards: Framing social research on vulnerability, response and adaptation. Global Environmental Change 17, 281-295.

Folke, C. (2006) Resilience: The emergence of a perspective for social-ecological systems analyses. Global Environmental Change-Human and Policy Dimensions 16, 253-267.

Forman (2010). Urban Ecology and the arrangement of Nature in Urban regions. In Ecological urbanism. Ed. Mostafavi, M and Doherty, G. Harverd University, Graduate School of Design. Lars Muller Publishers, Germany.

209 Forman, R. T. T. (2008). Urban regions: Ecology and planning beyond the city. Cambridge: Cambridge University Press.

Foster, J., Lowe, A. and Winkelman, S. (2011). The Value of Green Infrastructure for Urban Climate Adaptation. Center for Clean Air Policy. Available Online at http://ccap.org/assets/THE-VALUE-OF-GREEN-INFRASTRUCTURE- FOR-URBAN-CLIMATE-ADAPTATION_CCAP-February-2011.pdf Last Accessed December 2012.

Frumhoff, P.C., McCarthy, J.J., Melillo, J.M., Moser, S.C., Wuebbles, D.J., (2007) Confronting Climate Change in the U.S. Northeast: Science, Impacts, and Solutions. Synthesis report of the Northeast Climate Impacts Assessment (NECIA), Cambridge, MA.

Frumhoff, P.C., McCarthy, J.J., Melillo, J.M., Moser, S.C., Wuebbles, D.J., Wake, C., Spanger-Siegfried, E. (2008) An integrated climate change assessment for the Northeast United States. Mitigation and Adaptation Strategies for Global Change 13, 419-423.

Fussel,H.M. (2007). Adaptation planning for climate change: concepts, assessment approaches, and key lessons. Sustainable Science 2:265–275

Gaber, John, and Sharon Gaber. 2007. Qualitative Analysis for Planning & Policy: Beyond the Numbers. Chicago: American Planning Association.

Gallopin, G.C. (2006). Linkages between vulnerability, resilience, and adaptive capacity. Global Environmental Change, 16, 293-303.

Garde, A.M. (1999). Marginal Spaces in the Urban Landscape: Regulated or Incidental Open Spaces? Journal of Planning Education and Research, 18, 200-210.

Gartland, L. (2008). Heat Islands: Understanding and Mitigating Heat in Urban Areas. Earhscan, London.

Gasteyer, S., Flora, C.B. (2000) Measuring ppm with tennis shoes: Science and locally meaningful indicators of environmental quality. Society & Natural Resources 13, 589-597.

Geddes, P. (1915). Cities in evolution: An introduction to the town planning movement and to the study of civics. London: Williams & Norgate.

Gedzelman, S.D., Austin, S., Cermak, R., Stefano, N., Partridge, S., Quesenberry, S., and Robinson, D.A. (2003). Mesoscale aspects of the urban heat island around New York City. Theoretical Applied Climatology. 75(1-2), 29-42

Georgi, N. J. and K. Zafiriadis (2006). "The Impact of Park Trees on Microclimate in Urban Areas." Urban Ecosystems 9(3): 195-209.

210 Gerring, J. (2007). Case study research : Principles and practices. New York: Cambridge University Press.

Gill, S.E., Handley, J.F. Ennos, A.R., and Pauleit, S. (2007). Adapting Cities for Climate Change: The Role of the Green Infrastructure. Built Environment, 33(1),115-133.

Gill, S.E., Handley, J.F., Ennos, A.R., Pauliet, S., Theuray, N., and Lindley, S.J. (2008). Characterizing the urban environment of UK cities and towns: A template for landscape planning. Landscape and Urban Planning, 87 (2),210-222.en, N.

Girling, C., & Kellett, R. (2005). Skinny streets and green neighborhoods: Design for environment and community. Washington, DC: Island Press.

Gober et al. (2010). Using Watered Landscape to manipulate Urban Island Heat Effect. Journal of the Amercian Planning Association. 76(1), 109-121.

Gober, P., Brazel, A., Quay, R., Myint, S., Grossman-Clarke, S., Miller, A., & Rossi, S. (2010). Using Watered Landscapes to Manipulate Urban Heat Island effects. Journal of the American Planning Association, 76(1), 109-121.

Golden,J.S., Brazel,A., Salmond,J. and Laws,D. (2006). Energy and Water Sustainability: The Role of Urban Climate Change from Metropolitan Infrastructure. Journal of Green Building 1(3),124-138.

Goldhor, H. (1972). An introduction to scientific research in librarianship. Urbana, IL: University of Illinois.

Gonzales, G. A. (2005). Urban sprawl, global warming and the limits of ecological modernization. Environmental Politics, 14(3), 344-362.

Goode,D. (2006). Green Infrastructure: Report to the Royal Commission on Environmental Pollution. http://www.greeninfrastructurenw.co.uk/climatechange/doc.php?docID=168 Last Accessed March 2013.

Green, N. and Handley, J. (2009). Patterns of settlement compared. In Planning for Climate Change: Strategies for Mitigation and Adaptation for Spatial Planners, Davoudi, S., Crawford, J., and Mehmood, A. (Ed.). Earthscan, London, Sterling VA.

Guhathakurta, S. and Gober, P. (2007). The Impact of the Phoenix Urban Heat Island on Residential WaterUse. Journal of the Planning Association, 73(3), 317-329.

Hacker, J. N., & Holmes, M. J. (2007). Thermal comfort: Climate change and the environmental design of buildings in the United Kingdom. Built Environment, 33(1), 97-114.

211 Hall, J.M., Handley, J.F., and Ennos, A.R. (2012). The potential of tree planting to climate-poof high density residential areas in Manchester, UK. Landscape and Urban Planning, 104 (2),27-33.

Hamel, J. (with Dufour, S., & Fortin, D.). (1993). Case study methods. Newbury Park, CA: Sage.

Hamin, E., and Gurran, N. (2008). Urban form and climate change: Balancing adaptation and mitigation in the U.S. and Australia. Habitat International (33), 238-245.

Hardoy, J., Pandiella, G. (2009) Urban poverty and vulnerability to climate change in Latin America. Environment and Urbanization 21, 203-224.

Harris, S., & Sutton, R. (1986). Functions of parting ceremonies in dying organizations. Academy of Management Journal, 19, 5-30.

Hart, M. A., & Sailor, D. J. (2009). Quantifying the influence of land-use and surface characteristics on spatial variability in the urban heat island. Theoretical and Applied Climatology, 95(3), 397–406.

Heinz. (2006). A survey of climate change adaptation and planning. Washington DC: The H. John Heinz III Center for Science, Economics and the Environment.

Heltberg, R., Siegel, P.B., Jorgensen, S.L. (2009) Addressing human vulnerability to climate change: Toward a `no-regrets' approach. Global Environmental Change 19, 89-99.

Henry, James A. and Steven E. Dicks (1987). "Association of Urban Temperatures with Land Use and Surface Materials." Landscape and Urban Planning 14: 21-29.

Holling, C. S. (1973). Resilience and stability of ecological systems. Annual Review of Ecological Systems, 4, 1-23.

Holman-Dodds,J.K., A. Allen Bradley,A.A., and Potter,K.W (2003). Evaluation of Hydrologic Benefits of Infiltration Based Urban Storm Water Management. Journal of The American Water Resources Association, February Issue.

House-Peters,L.A. and Chang, H. (2011). Modeling the impact of land use and climate change on neighborhood-scale evaporation and nighttime cooling: A surface energy balance approach. Landscape and Urban Planning, 103, 139-155.

Howard, J. (2009). Chapter 3: Climate change mitigation and adaptation in developed nations: A critical perspective on the adaptation turn in urban climate planning. In Planning for Climate Change . Ed. Davoudi, S, Crawford, J., & Mehmood A. . Earthscan, UK.

212 Huang, G., Zhou, W., and Cadenasso, M.L. (2011). Is everyone hot in the city? Spatial pattern of land surface temperatures, land cover and neighborhood socioeconomic characteristics in Baltimore, MD. Journal of Environmental Management 82, 184-197.

Huang, J., Lu,X.X., Sellers,J.M.(2007) A global comparative analysis of urban form: Applying spatial metrics and remote sensing. Landscape and Urban Planning, 103, 139-155.

Huq S, Rahman A, Konate M, Sokona Y, Reid H (2003) Mainstreaming adaptation to climate change in least developed countries (LDCs). Technical report, International Institute for Environment and Development, London (LDCs). Technical report, International Institute for Environment and Development, London

Huq, S., Kovats, S., Reid, H., & Satterthwaite, D. (2007). Editorial: Reducing risks to cities from disasters and climate change. Environment and Urbanization, 19(3), 3-15.

ICLEI (2007).Preparing for Climate Change: A Guidebook for Local, Regional, and State Governments. Center for Science in the Earth System (The Climate Impacts Group) Joint Institute for the Study of the Atmosphere and Ocean University of Washington, King County, Washington and ICLEI – Local Governments for Sustainability. Available at http://www.cses.washington.edu/db/pdf/snoveretalgb574.pdf Last Accessed January 2013.

IPCC(2007). An Assessment of the Intergovernmental Panel on Climate Change: Synthesis Report.

IPCC, (2007) Climate Change: Synthesis Report. Intergovernmental Panel on Climate Change.

IPCC, (2012) Summary for Policymakers. , in: Field, C.B., V. Barros, T.F. Stocker, D. Qin, D.J. Dokken, K.L. Ebi, M.D. Mastrandrea, K.J. Mach, G.-K. Plattner, S.K. Allen, M. Tignor, and P.M. Midgley (Ed.), Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation: A Special Report of Working Groups I and II of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK and New York, NY USA, pp. 1- 19.

IPCC. (2007). Climate change: Synthesis report No. 4th Assessment Report) Intergovernmental Panel on Climate Change.

Jabreen, Y.R. (2006). Sustainable Urban Forms: Their Typologies, Models, and Concepts. Journal of Planning Education and Research. 26, 38-52.

213 Jacobson, C., Hughey, K.F.D., Allen, W.J., Rixecker, S., Carter, R.W. (2009) Toward More Reflexive Use of Adaptive Management. Society & Natural Resources 22, 484-495.

Jasechko, S., Zachary D. S, Gibson, J.J., S. Jean Birks, S.J., Yi Yi, and Peter J. Fawcett, P.J. (2013). Terrestrial water fluxes dominated by transpiration. Nature. doi:10.1038/nature11983. Retrieved 4 April 2013

Jim, C. Y., & Chen, S. S. (2003). Comprehensive greenspace planning based on landscape ecology principles in compact Nanjing city, China. Landscape and Urban Planning, 65(3), 95–116.

Jongman, R. H. G. (1995). Nature conservation planning in Europe: Developing ecological networks. Landscape and Urban Planning, 32, 169–183.

Jongman, R. H. G., K¨ulvik, M., & Kristiansen, I. (2004). European ecological networks and greenways. Landscape and Urban Planning, 68, 305–319.

Jongman,R.H.G. and Pungetti,G.(2004).Ecological Networks and Greenways Concept, Design, Implementation. Cambridge University Press, Cambridge.

Joshi, M., Hawkins, E., Sutton, R., Lowe, J., Frame, D. (2011) Projections of when temperature change will exceed 2 [deg]C above pre-industrial levels. Nature Clim. Change 1, 407-412.

Juhola, S., Peltonen, L., Niemi, P. (2012) The ability of Nordic countries to adapt to climate change: assessing adaptive capacity at the regional level. Local Environment: The International Journal of Justice and Sustainability, 1–18.

Kahn, M.E. (2006) – Green Cities. Urban Growth and the Environment. Brookings Institution Press, Washington, D.C.

Kates, R.W., Travis, W.R., Wilbanks, T.J. (2012) Transformational adaptation when incremental adaptations to climate change are insufficient. Proceedings of the National Academy of Sciences of the United States of America 109, 7156- 7161.

Kato, S. (2010). Green space conservation planning framework for urban regions based on a forest bird-habitat relationship study and the resilience thinking. Ph.D. dissertation, Umass, Amherst; AAT 3409604.

Kato, S., Ahern, J. (2008) 'Learning by doing': adaptive planning as a strategy to address uncertainty in planning. Journal of Environmental Planning and Management 51, 543-559.

214 Kazmierczak, A. and Carter, J. (2010). Adaptation to climate change using green and blue infrastructure: A data base of case studies.GRABS project, University of Manchester, U.K.

Kazmierczak, A., and Cavan, G. (2011). Surface water flooding risk to urban communities: Analysis of vulnerability, hazard and exposure. Landscape and Urban Planning, 103, 185-197.

Kikegawa, Y., Genchi, Y., Kondo, H., & Hanaki, K. (2006). Impacts of city-blockscale countermeasures against urban heat-island phenomena upon a building’s energy-consumption for air-conditioning. Applied Energy, 83(6), 649–668.

Kintisch, E. (2009) Projections of Climate Change Go From Bad to Worse, Scientists Report. Science 323, 1546-1547.

Kirkpatrick, J.S. and M.D. Shulman (1987) A statistical evaluation of the New York City – Northern New Jersey urban heat island effect on summer daily minimum temperature. National Weather Digest, 12(1), 12.

Kirshen, P., Ruth, M., & Anderson, W. (2008). Interdependencies of urban climate change impacts and adaptation strategies: A case study of Metropolitan Boston USA. Climate Change, 86, 105-122.

Klein, R. J. T., Lisa, E., Schipper, F., & Dessai, S. (2005). Integrating mitigation and adaptation into climate and development policy: Three research questions. Environmental Science & Policy II, 8(6), 579-588.

Klein, R., Nichols, R.,Thomalla,F. (2003). Resilience to natural hazards: How useful is this concept?. Environmental Hazards 5, 35-45

Klok, L., Zwart, S., Verhagen, H., and Mauri, E. (2012). The surface heat island of Rotterdam and its relationship with urban surface characteristics.

Kurn, D.M., S.E. Bretz, B. Huang, and H. Akbari (1994). The potential for reducing urban air temperatures and energy consumption through vegetative cooling. Pacific Grove, CA: American Council for an Energy Efficient Economy Summer Study on Energy Efficiency in Buildings. 31pp. Online available at http://www.epa.gov/heatisland/mitigation/trees.htm. Last Accessed March 2013.

Kuttler, W. (2008). The urban Climate-Basic Applied Aspects. p.233-248. In: Marzluff, J.M., G.Bradley, E. Shulenberger, C. Ryan, W. Endlicher, U. Simon, M. Alberti, & C. ZumBrunnen (eds.), Urban ecology: An international perspective on the interaction b etween humans and nature, New York, Springer.

215 Kwadijk, J.C.J., Haasnoot, M., Mulder, J.P.M., Hoogvliet, M.M.C., Jeuken, A.B.M., van der Krogt, R.A.A., van Oostrom, N.G.C., Schelfhout, H.A., van Velzen, E.H., van Waveren, H., de Wit, M.J.M. (2010) Using adaptation tipping points to prepare for climate change and sea level rise: a case study in the Netherlands. Wiley Interdisciplinary Reviews: Climate Change 1, 729-740.

La Rosa, D. and Privitera, R. (2013). Characterization of non-urbanized area of land-use planning of agricultural and green infrastructure in urban contexts. Landscape and Urban Planning, 109 (1), 94-106.

Lamptey, B.L.; Barron, E.J.; Pollard, D. (2005). Impacts of agriculture and urbanization on the climate of the Northeastern United States. Global and Planetary Change, 49 (3–4): 203–221.

Landsberg,H. (1981).The Urban Climate. Volume 28 in the International Geophysics Series. Academic Press, a subsidiary of Harcourt Brace Jovanovich, Publishers, New York

Lawson, V. (1971). Reference service in university libraries, two case studies. Unpublished doctoral dissertation, Columbia University, New York.

Lerch, D. (2007). Post Carbon Cities: Planning for Energy and Climate Uncertainty. Post Carbon Institute.

Lerner, J. and Allen, W.L. III (2012). Landscape-Scale Green Infrastructure Investments as a Climate Adaptation Strategy: A Case Example for Midwest United Sates. Environmental Practice, 14 (1),45-56.

Lewis, G.McD., and Brabec, E. (2005). Regional land pattern assessment: development of a resource efficiency measurement method. Landscape and Urban Planning, 72, 281-296.

Lewis, Geoffrey McD. (2006). Evaluating effects of land pattern with regional characteristic curves: Land cover indicators and estimating building energy use from thermal image data. PhD Dissertation. University of Michigan School of Natural Resources and Environment.

Lewis, Geoffrey McD., and Elizabeth Brabec. (2005). Development of regional measurement methods: the context for quantifying influence and impact. Landscape Review 10 (1 & 2): 44-48.

Lewis, P. H., Jr. (1964). Quality corridors for Wisconsin. Landscape Architecture, 54, 100–107. McHarg, I. (1969). Design with nature. Garden City: The Natural History Press. Mumford, L. (1937). What is a city? New York: McGraw-Hill. Nassauer, J. (2006). Landscape planning and conservation biology: Systems thinking revisited.

216 Lindblom, C.E. (1959) The Science of "Muddling Through". Public Administration Review 19, 79-88.

Lindley, S. J., Handley, J. F., Theuray, N., Peet, E., & McEvoy, D. (2006). Adaptation strategies for climate change in the urban environment: Assessing climate change related risk in UK urban areas. Journal of Risk Research, 9(5), 543-568.

Lynn, B. H., Carlson, T. N., Rosenzweig, C., Goldberg, R., Druyan, L., Cox, J., et al. (2009). A modification to the NOAH LSM to simulate heat mitigation strategies in the New York City metropolitan area. Journal of Applied Meteorology and Climatology, 48(2), 199–216.

Mastrandrea, K.J. Mach, G.-K. Plattner, S.K. Allen, M. Tignor, and P.M. Midgley (Ed.), Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation. Cambridge University Press, Cambridge, UK and New York, NY USA, pp. 437-486.

Makar, P.A.; Gravel, S.; Chirkov, V.; Strawbridge, K.B.; Froude, F.; Arnold, J.; Brook, J. (2006) – Heat flux, urban properties, and regional weather. Atmospheric Environment, 40 (15): 2750–2766.

Mass GIS. Office of Geographic Information, Commonwealth of Massachusetts, Information Technology Division. http://www.mass.gov/anf/research-and- tech/it-serv-and-support/application-serv/office-of-geographic-information- massgis/datalayers/overview.html . Accessed during the period from October 2012 to April 2013.

Maxwell, J. (1998). Designing a Qualitative Study. In Handbook of Applied Social Research Handbook, Bickman Ed. Sage Publications. 69-100.

McAdams, D. C. (1979). Powerful actors in public land use decision making processes: A case study in Austin, Texas. Unpublished doctoral dissertation, University of Texas, Austin.

McClure, C. R., & Hernon, P. (Eds.). (1991). Library and information science research: perspectives and strategies for improvement. Norwood, NJ: Ablex.

McEvoy,D., Lindley,S., and Handley,J. (2006). Adaptation and mitigation in urban areas: synergies and conflicts. Municipal Engineer 159 (ME4),185–191

McPherson G, Simpson J, Xiao Q, Wu C (2008). Los Angeles 1-Million tree Canopy Cover Assessment. USDA, Forest Service, general Technical Report, ref: PSW- GTR-207.

McPherson, E.G. (2011). Sacramento’s Parking Lot Shading Ordinance: Environmental And Economic Costs Of Compliance. Landscape and Urban Planning, 57 ,105- 123.

217 McPherson, E.G., Simpson, J.S., Xiao, Q., and Wu, C. (2011). Million trees Los Angeles canopy cover and benefit assessment. Landscape and Urban Planning, 99 ,40-50.

McPherson, E.G., Simpson, Peper,P.J., Maco,S.E., and Xiao, Q., (2005). Municipal Forest Benefits and Costs in Five US Cities. Journal of Forestry, December.

Mcpherson,G.,Nowak,D.,Heisler,G.,Grimmond,S.,Souch,C.,Grant,R.,And Rowntree,R.(1997). Quantifying Urban Forest Structure, Function, and Value: The Chicago Urban Forest Climate Project. Urban Ecosystems, 1997, 1, 49–61

Mell, I.C. (2008). Can green infrastructure promote urban sustainability? Engineering Sustainability, 162, 23-34.

Mell, I.C. (2008a). Green Infrastructure: concepts and planning. FORUM Ejournal 8 (June): 69-80 Newcastle University.

Meyer,B.C. and Loibl,W. (2010). Climate change and spatial planning. Landscape and Urban Planning, 98 ,139-140.

Miles, M. B., & Huberman, A. M. (1984). Qualitative data analysis: A sourcebook of new methods. Beverly Hills, CA: Sage.

Millennium Ecosystem Assessment Report (2005). Ecosystems and Human Well-being: Synthesis. Island Press, Washington, DC.

Miller, F. (1986). Use, appraisal, and research: A case study of social history. The American Archivist: 49(4), 371-392.

Mills, G. (2007). Cities as agents of global change. International Journal of Climatology, 27, 1849-1857.

Milly, P.C.D., Betancourt, J., Falkenmark, M., Hirsch, R.M., Kundzewicz, Z.W., Lettenmaier, D.P., Stouffer, R.J. (2008) Stationarity Is Dead: Whither Water Management? Science 319, 573-574.

Mitchel, A. (2009). The ESRI Guide to GIS: Analysis, Volume2: Spatial Measurement & Statistics. ESRI Press, Redlands, CA.

Moser, S.C., Ekstrom, J.A. (2010) A framework to diagnose barriers to climate change adaptation. Proceedings of the National Academy of Sciences 107, 22026- 22031.

National Research Council (2010) Adapting to the Impacts of Climate Change. The National Academies Press, Washington, DC.

218 Naumann, Sandra, McKenna Davis, Timo Kaphengst, Mav Pieterse and Matt Rayment (2011): Design, implementation and cost elements of Green Infrastructure projects. Final report to the European Commission, DG Environment, Contract no. 070307/2010/577182/ETU/F.1, Ecologic institute and GHK Consulting.

NECIA. (2007). Confronting climate change in the U.S. Northeast: Science, impacts, and solutions: A report of the Northeast Climate Impacts Assessment. Union of Concerned Scientists. http://www.climatechoices.org/ne/resources ne/nereport.html. Accessed 19 June 2011.

Nelson, D.R., Adger, W.N., Brown, K. (2007) Adaptation to Environmental Change: Contributions of a Resilience Framework. Annual Review of Environment and Resources 32, 395-419.

Neuman, M. (2005). The Compact City Fallacy. Journal of Planning Education and Research, 25, 11-26.

Newman, P., & Jennings, I. (2008). Cities as sustainable ecosystems: Principles and practices. Washington, DC: Island Press.

Nitschke, C.R., Innes, J.L. (2008) A Tree and Climate Assessment Tool for Modelling Ecosystem Response to Climate Change. Ecological Modelling 210, 263-277.

NOAA Earth System Research Laboratory, (2011) The NOAA Annual Greenhouse Gas Index (AGGI). U.S. Department of Commerce, National Oceanic & Atmospheric Administration, NOAA Research, Earth System Research Laboratory Global Monitoring Division, Boulder, CO.

Noss, R. F., & Harris, L. D. (1986). Nodes, networks, and mums: Preserving diversity at all scales. Environmental Management, 10(3), 299–309.

Nowak, D.J. and Greenfield, E.J. (2012a). Tree and impervious cover change in U.S. cities. Urban Forestry & Urban Greening, 11, 21-30.

Nowak, D.J., and Greenfield, E.J. (2012b). Tree and impervious cover in the United States. Landscape and Urban Planning, 107 (1),21-30.

Nowak, D.J., Rowntree, R.A., McPherson, E.G., Sasinni, S.M., Kerkmann, E.R., and Stevens, J.C. (1996). Measuring and analyzing urban tree cover. Landscape and Urban Planning, 36, 49-57.

Nussbaum, S., Menz, G. (2008). Chapter 3:eCognition Image Analysis. In Object-Based Image Analysis and Treaty Verification. Springer Science+Business Media B.V.

219 NYC (2011).NYC Green Infrastructure Plan: A Sustainable Strategy for Clean Waterways. City of New York, Offcie of the Mayor. http://www.nyc.gov/html/dep/html/stormwater/nyc_green_infrastructure_plan.s html Last Accessed March 2013.

O’Brien, K., M. Pelling, A. Patwardhan, S. Hallegatte, A. Maskrey, T. Oki, U. Oswald- Spring, T. Wilbanks, P.Z. Yanda, (2012) Toward a sustainable and resilient future, in: Field, C.B., V. Barros, T.F. Stocker, D. Qin, D.J. Dokken, K.L. Ebi,

Oberndorfer, E., et al. (2007) "Green Roofs as Urban Ecosystems: Ecological Structures, Functions, and Services." Bioscience ,57(10), 823.

Oberndorfer, E., Lundholm,J., Bass, B.,Coffman,R.R.,Doshi,H.,Dunnett,N.,Gaffin,S.,Kohler,M., Liu,K., And Rowe,B.(2007). Green Roofs as Urban Ecosystems: Ecological Structures, Functions, and Services. BioScience,57, 823-833.

Oke, T.R. (1988). Boundary Layer Climates.2nd Ed. Routledge, London.

Oke, T.R., 1997. Urban climates and global change. In: Perry, A.; Thompson, R. (eds) Applied Climatology: Principles and Practice. London: Routledge: 273–287.

Onishi, A., Cao, X., Ito, T., Shi, F., and Imura, H. (2010). Evaluating the potential for urban heat-island mitigation by greening parking lots. Urban Forestry & Urban Greening, 9, 323-332.

Pacala, S. and Socolow, R. (2004). Stabilization Wedges: Solving the Climate Problem for the Next 50 Years with Current Technologies. Science 13 August 2004:Vol. 305. no. 5686, pp. 968 – 972.

Paris, M. (1988). Library school closings: Four case studies. Metuchen, NJ: Scarecrow Press.

Park, S.E., Marshall, N.A., Jakku, E., Dowd, A.M., Howden, S.M., Mendham, E., Fleming, A. (2012) Informing adaptation responses to climate change through theories of transformation. Global Environmental Change 22, 115¬

Patton, M. Q. (1980). Qualitative evaluation methods. Beverly Hills, CA: Sage.

Pauleit,S. and Duhme,F. (2000). Assessing the environmental performance of land cover types for urban planning. Landscape and Urban Planning 52, 1-20

Pauleit,S., Ennos,R.,and Golding,Y. (2005).Modeling the environmental impacts of urban land use and land cover change—a study in Merseyside, UK. Landscape and Urban Planning 71, 295-310

220 Pendall, R., Foster, K. A., and Cowell, M. (2007). Resilience and regions: Building understanding of the metaphor. Berkeley, CA: IURD University of California.

Pham, T., Apparicio, P., Seguin, A., Landry, S., and Gagnon, M. (2012). Spatial distribution of vegetation in Montreal: an uneven distribution or environmental inequity. Landscape and Urban Planning, 107 (3), 214-224.

Pickett, S. T. A., Cadenasso, M. L., & Grove, J. M. (2004). Resilient cities: Meaning, models, and metaphor for integrating the ecological, socio-economic, and planning realms. Landscape and Urban Planning, 69(4), 369–384.

Pittock, A.B., Jones, R.N., Mitchell, C.D. (2001) Probabilities will help us plan for climate change. Nature 413, 249-249.

Pizarro, R. (2009). Urban Form and climate Change: Towards Appropriate Development Patterns to Mitigate and Adapt to Global Warming. In Planning for Climate Change: Strategies for Mitigation and Adaptation for Spatial Planners, Davoudi, S., Crawford, J., and Mehmood, A. (Ed.). Earthscan, London, Sterling VA.

Powell, R. R. (1985). Basic research methods for librarians. Norwood, NJ: Ablex.

Pozzi, P. & Small, C. (2001). Exploratory Analysis of Suburban Land Cover and Population Density in the U.S.A. Paper 35, IEEE/ISPRS joint Workshop on Remote Sensing and Data Fusion over Urban Areas,Rome, Italy, 8-9 November, 2001

Pozzi, P. & Small, C. (2002). Vegetation and population density in urban and Suburban Areas in the U.S.A. Conference proceedings of Third International Symposium of Remote Sensing of Urban Areas Istanbul, Turkey, June 2002.

Preston, B.L. and Stafford-Smith, M. (2009). Framing vulnerability and adaptive capacity assessment: Discussion paper. CSIRO Climate Adaptation Flagship Working paper No. 2. http://www.csiro.au/org/ClimateAdaptationFlagship.html Last Accessed February 2013.

Quereda-Sala, J.; Olcina, A.G.; Cuevas, A.P.; Cantos, J.O.; Amoros ,A.R.; Chiva, E.M. (2000). Climatic warming in the Spanish Mediterranean: Natural Trend or Urban Effect. Climatic Change, 46 (4): 473–483.

Radford, K.G. and James, P. (2013). Changes in the value of ecosystem services along the rural-urban gradient: A case study of Greater Manchester, UK. Landscape and Urban Planning, 109 ,117-127.

Rahmstorf, S., Cazenave, A., Church, J.A., Hansen, J.E., Keeling, R.F., Parker, D.E., Somerville, R.C.J. (2007) Recent Climate Observations Compared to Projections. Science 316, 709.

221 Rannow,S., Loibl,W., GreivingS., Gruehn,D., and Meyer,B.C.(2010). Potential impacts of climate change in Germany—Identifying regional priorities for adaptation activities in spatial planning. Landscape and Urban Planning, 98,160-171.

Reyer,C., Guericke,M., and Ibisch,P.L. (2009). Climate change mitigation via afforestation, reforestation and deforestation avoidance: and what about adaptation to environmental change? New Forests, 38,15–34.

Richard, W., Palmer, J., and Kolokotroni, M. (2006). Increased Temperature and Intensification of the Urban Heat Island: Implications for Human Comfort and Urban Design, Built Environment 33(1), 85-96.

Roa-Espinosa, A., T.B. Wilson, J.M. Norman, and Kenneth Johnson. 2003. Predicting the Impact of Urban Development on Stream Temperature Using a Thermal Urban Runoff Model (TURM). National Conference on Urban Stormwater: Enhancing Programs at the Local Level. February 17-20. Chicago, IL. http://www.epa.gov/nps/natlstormwater03/31Roa.pdf Last Accessed February 2013.

Robert L. Wilby, R.L. (2003). Past and projected trends in London’s urban heat island.Weather,58,251-260.

Romero Lankao, P. (2008). Urban Areas and Climate Change: Review of Current Issues and Trends. Issues Paper for the 2011 Global Report on Human Settlements. http://www.isse.ucar.edu/romero-lankao/

Romero Lankao, P. and Tribbia, J.L. (2009). Assessing patterns of vulnerability, adaptive capacity and resilience across urban centers. Proceeding of the Fifth Urban Research Symposium: Cities and Climate Change, responding to an urgent agenda , Marseille, France. World Bank. Web site accessed December 2012: http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTURBANDEVE LOPMENT/0

Rosenberg, E.A., Keys, P.W., Booth, D.B., Hartley, D., Burkey, J., Steinemann, A.C., Lettenmaier, D.P. (2010) Precipitation Extremes and the Impacts of Climate Change on Stormwater Infrastructure in Washington State. Climatic Change 102, 319-349.

Rosenzweig, C., Major, D. C., Demong, K., Stanton, C., Horton, R., & Stults, M. (2007). Managing climate change risks in New York city's water system: Assessment and adaptation planning. Mitigation and Adaptation Strategies for Global Change, 12(8), 1391.

Rosenzweig, C., Solecki, W.D., Hammer, S.A., and Mehrotra, S. (2011). Climate Change and Cities: First Assessment Report of the Urban Climate Change Research Network. Urban Climate Change Research Network (UCCRN), Center for Climate Systems Research, Earth Institute, Columbia University.

222 Cambridge University Press, Cambridge.

Rosenzweig,C., Solecki, W.D. , Slosberg, R.B. (2006). Mitigating New York City’s Heat Island with Urban Forestry, Living Roofs, and Light Surfaces, New York City Regional Heat Island Initiative Final Report. New York State Energy Research And Development Authority.

Roth, M., T. R. Oke, and W. J. Emery. (1989). Satellite-derived Urban Heat Islands from Three Coastal Cities and the Utilization of Such Data in Urban Climatology. International Journal of Remote Sensing, 10, 1699-1720.

Ruth,M.(2003).Modeling Infrastructure Vulnerabilities and Adaptation to Climate Change in Urban Systems: Methodology and Application to Metropolitan Boston. Paper presented at the European Regional Science Association Meetings, Jyväskylä, Finland, August 2003

Schindler, D. (1996). Urban youth and the frail elderly: Reciprocal giving and receiving. New York: Garland.

Schipper, E.L.F., Burton, I., (2009) Understanding adaptation: Origins, concepts, practice and policy, in: Schipper, E.L.F., Burton, I. (Eds.), The Earthscan reader on adaptation to climate change. Earthscan, London ; Sterling, VA, pp. 1-14.

Seabrook, L., Mcalpine, C.A., and Bowen, M.E. (2011). Restore, repair or reinvent: Options for sustainable landscapes in a changing climate. Landscape and Urban Planning, 100, 407-410.

Shaw, R., Colley, M., and Connell, R. (2007). Climate Change Adaptation by Design: a guide for sustainable communities. TCPA, London.

Shea, K.M., Truckner, R.T., Weber, R.W., Peden, D.B. (2008) Climate Change and Allergic Disease. Journal of Allergy and Clinical Immunology 122, 443-453.

Simons, H. (1980). Towards a science of the singular: Essays about case study in educational research and evaluation. Norwich, UK: University of East Anglia, Centre for Applied Research in Education.

Small, C. (2002). Satellite Remote Sensing of Urban Structure and Evolution. Lamont Doherty Earth Observatory, Columbia University www.LDEO.Columbia.edu/~small/Urban.html

Smit, B. and Wandel, J. (2006). Adaptation, adaptive capacity and vulnerability. Global Environmental Change, 16, 282-292.

223 Smith, M.S., Horrocks, L., Harvey, A., Hamilton, C. (2011) Rethinking adaptation for a 4 degree C world. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 369, 196-216.

Solecki, William and Rosenzweig, Cynthia (2004). Biodiversity, Biosphere Reserves, and the Big Apple: A Study of the New York Metropolitan Region. Annals New York Academy of Sciences, 1023, 105-124.

Spatari,S., Yu,Z., and Montalto,F.A. (2011).Life cycle implications of urban green infrastructure. Landscape and Urban Planning, 159,2174-2179.

Stake, R. E. (1995). The art of case study research. Thousand Oaks, CA: Sage.

Stedinger, J.R., Griffis, V.W. (2011) Getting from Here to Where? Flood Frequency Analysis and Climate. Journal of the American Water Resources Association 47, 506-513.

Stone, B. (2005). An emerging Role for Planners in the Climate Change Debate. Journal of the American Planning Association 71(2), 13-25.

Stone, B. (2005b). Urban heat and air pollution: An emerging role for planners in the climate change debate. Journal of the American Planning Association, 71(1), 13-25.

Stone, B. (2008). Urban sprawl and air quality in large US cities. Journal of Environmental Management 86, 688–698

Stone, B. (2012) The City and the Coming Climate: Climate Change in the places we live. University of Cambridge Press, Cambridge.

Stone, B. Jr. (2009). Land Use as Climate Change mitigation. Environmental Science 7 Technology, 43(24), 9052–9056

Stone, B., and Norman, J.M. (2006). Land use planning and surface heat island formation: A parcel-based radiation flux approach. Atmospheric Environment 40,3561–3573.

Stone, B., Mednick, A. C., Holloway, T., & Spak, S. N. (2007). Is compact growth good for air quality? Journal of the American Planning Association, 73(4), 404-418.

Stone, B., Vargo, J., and Habeeb, D. (2012). Managing climate change in cities: will climate action plans work? Landscape and Urban Planning, 107(3),263-271.

Stubbs,S. (2008).Natural Green Space and Planning Policy: Devising a Model for its Delivery in regional Spatial strategies. Landscape Research, 33(1), 119-139.

224 Sun, R. and Chen, L. (2012). How can urban water bodies be designed for climate adaptation? Landscape and Urban Planning, 105 (1),27-33

Swisher, R., & McClure, C. R. (1984). Research for decision making, methods for librarians. Chicago: American Library Association.

Taha, H. (1997). Urban climates and heat islands: Albedo, evapotranspiration, and anthropogenic heat. Energy and Buildings, 25(2), 99–103.

Taha, H. and L.S. Kalkstein, S.C. Sheridan, and E. Wong. 2004. The Potential of Urban Environmental Controls in Alleviating Heat-wave Health Effects in Five US Regions. Presented at the American Meteorological Society Fifth Conference on Urban Environment. 25 August. See also NOAA. 1995. Natural Disaster Survey Report: July 1995 Heat Wave. http://www.nws.noaa.gov/om/assessments/pdfs/heat95.pdf Last Accessed March 2013

Taylor, R. S. (1967). Question-negotiation and information-seeking in libraries. Bethlehem, PA: Center for the Information Sciences.

Thi Van & Xuan Bao (2009). Study of the Impact of Urban Development on Surface Temperature Using Remote Sensing in Ho Chi Minh City, Northern Vietnam. Geographical research, Institute of Australian Geographers. 48(1),86– 96.geor_607 86..

Tratalos,J., Fuller,R.A., Warren,P.H., Davies,R.G.,and J. Gaston,K.J. (2007). Urban form, biodiversity potential and ecosystem services. Landscape and Urban Planning, 83,308-317.

Tzoulas,K., Korpela,K., Venn,S., Yli-Pelkonen,V., Ka´zmierczak,A., Niemela,J., and James,P. (2007). Promoting ecosystem and human health in urban areas using Green Infrastructure: A literature review. Landscape and Urban Planning 81, 167–178

United Nations Human Settlements Programme (UN-Habitat) (2011) Cities and Climate Change: Global Report on Human Settlements 2011. Earthscan, London.

United States Census Bureau, Data section, http://2010.census.gov/2010census/. Accessed 18 June 2011.

US EPA, (2010) Climate Change Indicators in the United States. United States Environmental Protection Agency, Washington, DC.

Van de Voorde, T., Jacquet, W., and Canters F. (2011). Mapping form and function in urban areas: an approach based on urban metrics and continuous impervious surface data. Landscape and Urban Planning, 102 (1), 143-155.

225 Verbeeck, K, Orshoven, J.V., Martin Hermy, M. (2011). Measuring extent, location and change of imperviousness in urban domestic gardens in collective housing projects. Landscape and Urban Planning 100, 57–66

Voogt, J.A. and T.R. Oke. (2003). Thermal Remote Sensing of Urban Areas. Remote Sensing of Environment, 86, 370-384.

Waldheim, C. and Berger, A. (2008). Logistics landscape. Landscape Journal 27(2), 55- 95.

Walker, B., Holling, C. S., Carpenter, S. R., & Kinzig, A. (2004). Resilience, adaptability and transformability in Social–ecological systems. Ecology and Society, 9(2), 5-13.

Walker, B.H., Salt, D. (2006) Resilience thinking : sustaining ecosystems and people in a changing world. Island Press, Washington, DC.

Walmsley,A. (1995). Greenways and the making of urban form. Landscape and Urban Planning 33, 81-127.

Watkins, R., Palmer, J., and Kolokotroni, M. (2007). Increased temperature and intensification of the urban heat island: Implications for human comfort and urban design. Built Environment, 33(1), 85-96.

Watkins,R., Palmer,J. and Kolokotroni,M. (2007). Increased temperature and Intensification of the Urban Heat island: Implications for Human Comfort and Urban Design. Built Environment 33(1), 85-96.

Weber,T., Sloan,A.,Wolf,J. (2006). Maryland’s Green Infrastructure Assessment: Development of a comprehensive approach to land conservation. Landscape and Urban Planning 77, 94-110.

Weiss, C.H., & Bucuvala, M. J. (1980). Social science research and decision- making. New York: Columbia University Press.

Weng, Q. & Quattrochi, D.A. (2007). Urban Remote Sensing. CRC Press, Boca Raton, London, New York.

Wheeler, S. (2008). State and Municipal Climate Change Plans. JAPA. 74 (4): 481-496. Environment, 33(1), 31-45.

Whitford,V.,Ennos,A.R., and Handley,J.F. (2001).”City Form and Natural Process”- indicators for the ecological performance of urban areas and their application to Merseyside,UK. Landscape and Urban Planning 57, 91-103.

Wholey, J. S., Hatry, H. P., & Newcomer, K. E. (Eds.). (1994). Handbook of practical program evaluation. San Francisco: Jossey-Bass.

226 Wickham, J.D., Ritters, K.H., Wade, T.G., and Vogt, P. (2010). A national assessment of green infrastructure and change for the conterminous United States using morphological image processing. Landscape and Urban Planning, 94,186-195.

Wilby,R.L. (2007).A Review of Climate Change Impacts on the Built Environment. Built Environment 33(1), 31-45.

Wilson, E., (2009) Use of Scenarios for Climate Change Adaptation in Spatial Planning, in: Davoudi, S., Crawford, J., Mehmood, A. (Eds.), Planning for Climate Change: Strategies for Mitigation and Adaptation. Earthscan, London, pp. 223- 235.

Wolf, K. L. (2003). Ergonomics of the City: Green Infrastructure and Social Benefits. In C. Kollin (ed.), Engineering Green: Proceedings of the 11th National Urban Forest Conference. Washington D.C.: American Forests.

Yin, R. K. (1984). Case study research: Design and methods. Newbury Park, CA: Sage.

Young, R.F. (2011). Planting the Living City: Best Practices in Planning Green Infrastrctures-Results from Majro U.S. Cities. Journal of the American Planning Association, 77(4), 368-381.

Young, R.F. and McPherson, E.G. (2013). Governing metropolitan green infrastructure in the United States. Landscape and Urban Planning, 109 (1),67-75.

Zahran, S., Grover,H.,Brody,S.D. and Vedlitz,A. (2008). Risk, Stress, and Capacity: Explaining Metropolitan Commitment to Climate Protection. Urban Affairs Review, 43,447-473

Zhou, & Shepherd, J. M. (2010). Atlanta’s urban heat island under extreme heat conditions and potential mitigation strategies. Natural Hazards, 52(3), 639–668.

Zhou, W., Huang, G., and Cadenasso, M.L. (2011). Does spatial configuration matter? Understanding the effects of land cover pattern on land surface temperature in urban landscapes. Landscape and Urban Planning, 102 (1),54-63.

227