The Use of Integrated Urban Models in the Nordic Countries Summary and Documentation of an Online Survey

Ryan Weber & Veronique Larsson

NORDREGIO WORKING PAPER 2014:1

The Use of Integrated Urban Models in the Nordic Countries

Summary and Documentation of an Online Survey

Ryan Weber & Veronique Larsson The Use of Integrated Urban Models in the Nordic Countries Summary and Documentation of an Online Survey

Nordregio Working Paper 2014:1

ISBN 978-91-87295-18-8 ISSN 1403-2511

Nordregio P.O. Box 1658 SE-111 86 Stockholm, Sweden [email protected] www.nordregio.se www.norden.org

Editors: Ryan Weber & Veronique Larsson Cover photo: Johannes Jansson / norden.org. A portside urban quarter in Reykjavik, Iceland.

Nordic co-operation Nordic co-operation is one of the world’s most extensive forms of regional collaboration, involving Denmark, Finland, Iceland, Norway, Sweden, and the Faroe Islands, Greenland, and Åland. Nordic co-operation has ﬁ rm traditions in politics, the economy, and culture. It plays an important role in European and inter- national collaboration, and aims at creating a strong Nordic community in a strong Europe. Nordic co-operation seeks to safeguard Nordic and regional interests and principles in the global community. Common Nordic values help the region solidify its position as one of the world’s most innovative and competitive.

The Nordic Council is a forum for co-operation between the Nordic parliaments and governments. The Council consists of 87 parliamentarians from the Nordic countries. The Nordic Council takes policy initiatives and monitors Nordic co-operation. Founded in 1952.

The Nordic Council of Ministers is a forum of co-operation between the Nordic governments. The Nordic Council of Ministers implements Nordic co-operation. The prime ministers have the overall responsibility. Its activities are co-ordinated by the Nordic ministers for co-operation, the Nordic Committee for co-operation and portfolio ministers. Founded in 1971.

Nordregio – Nordic Centre for Spatial Development conducts strategic research in the ﬁ elds of planning and regional policy. Nordregio is active in research and dissemina- tion and provides policy relevant knowledge, particularly with a Nordic and European comparative perspective. Nordregio was established in 1997 by the Nordic Council of Ministers, and is built on over 40 years of collaboration.

Stockholm, Sweden, 2014 Contents

Foreword ...... 9

Summary ...... 11

1 Introduction ...... 13 1.1 Cities, environment and urban planning ...... 13 1.2 What are integrated urban models and why should we use them?...... 13

2 Nordic Survey on Integrated Urban Models ...... 15 2.1 Method ...... 15 2.2 General proﬁ le of the authorities that responded ...... 18 2.3 The use of integrated models – diffusion in the Nordic countries ...... 18 2.3.1 Types of IUMs in use in the Nordic countries ...... 20 2.3.2 Basic technical details of the models used in the Nordic region ...... 20 2.3.3 What land use attributes and policy priorities are considered? ...... 21 2.3.4 How are urban models developed, implemented and used? ...... 21 2.4 What about those authorities not using an integrated model? ...... 23

3 Conclusions ...... 27

4 Directions for Future Work ...... 29

5 References ...... 31

Foreword

How can urban planning contribute to create sustaina- companies’ growth and jobs. ble and attractive urban areas and at the same time sup- Th e Nordic working group is therefore interested in port green growth in the Nordic city-regions? how spatial planning knowledge and instruments can Th is question is of course not answered by a single contribute to the policy goals of green growth of Nordic approach focusing on the spatial planning dimension city-regions. Th is includes knowledge on and develop- alone but has to be approached in a broader and multi- ment of modelling tools for urban areas to undertake dimensional manner. Focusing on the spatial planning spatial development in ways that facilitate policy goals dimension is however fundamental in planning and de- and decisions associated with green growth. veloping sustainable urban regions. In this report Review of Land-Use Models Nordregio Th e Nordic working group for green growth – sustain- emphasises the benefi ts of integrated approaches to able urban regions aims to develop policies, strategies planning. Th rough new sophisticated technologies and and planning tools in close collaboration with the Nor- models, more and more factors can be considered si- dic city-regions to support practitioners working with multaneously, which can provide more accurate future urban development on a regional and local scale. Th is predictions of diff erent planning strategies and policies. involves bridging economic development, social well- I hope that the report’s focus on integrated land use being, physical planning, demographic changes, cli- models can provide useful new information for policy mate and environmental perspectives – which, when decision-making on future urban planning questions. considered together, represent the complexity of gov- erning urban development. Rasmus Kristian Pejter Rasmussen, Danish Ministry of Th e purpose of the Nordic working group is to pro- the Environment mote the development of sustainable and attractive Nordic cities and partly to promote these solutions Chairman of Nordic working group for green growth in and outside the region for the benefi t of Nordic – sustainable urban regions.

NORDREGIO WORKING PAPER 2014:1 9

Summary

Integrated urban models (IUMs) can help us gain a and with the realisation that the results they provide better understanding of urban dynamics by creating are still meaningful. Th erefore, the term “IUM” refers simplifi ed future scenarios that can be manipulated to the six cases where we think truly integrated urban through simulation. Th ey enable virtual experimenta- models are used, while the term “urban models” refers tion, allowing the potential impact of new infrastruc- the 16 cases identifi ed initially. ture, technologies or policies to be determined, which Th e skewed results are most likely due to the defi - can support strategic planning. Models can also act as nition not suffi ciently emphasising that models are powerful tools for facilitating participatory, collabora- tools combining two key components: the integration tive decision-making. Th e visual outputs provide a se- of multiple land use, transport and socio-economic at- ries of “what if” scenarios, dependant on the strategic tributes and the ability to produce future land use de- policy decisions that are made. velopment scenarios. Nevertheless, it is surprising how Th is working paper presents the results of a survey few authorities use IUMs given the strong history of carried out by Nordregio’s researchers (with prepara- comprehensive planning in Nordic countries. Th is ap- tory support from WSP Analysis and Strategy). It was pears to be due to a lack of awareness of the uses and distributed to municipal and regional planning au- benefi ts of these models, which makes it important that thorities in the 29 largest Nordic city-regions. A total new information be communicated to planners and of 50 responses were received, with at least one from policymakers effi ciently. either a municipal or regional authority in each city- Urban models generally operate by restricting land region. development in certain areas (and promoting it in oth- Th e aim of the survey was to learn more about the ers), depending on the existing land use characteristics, use of integrated urban modelling tools by Nordic au- planning goals and socio-economic demand for devel- thorities through the following questions: opment. Th e survey showed that transport/mobility  To what extent are integrated urban models being infrastructure, heritage buildings and public spaces used by Nordic planning authorities? such as green space were the most commonly cited  What are the basic technical details of the models development restrictions (areas that cannot be devel- being used? oped). Conversely, transport accessibility, in terms of  How are these models developed, implemented and roads and public transport, as well as residential and used as planning support systems? non-residential building density were the land uses most commonly cited by urban models as attracting Main ﬁ ndings development. When studying the responses provided regarding Of the 50 responses, 39 came from municipalities and the development outcomes prioritised by urban mod- 11 from regional authorities. Th ere were 22 responses els, the survey showed that the planning policies inte- from Sweden, 12 from Denmark, eight from Norway, grated into the models most frequently prioritise the seven from Finland and one from Iceland. strategic densifi cation of existing built-up areas and Initially, 16 of the 50 authorities stated that they use areas close to public transport. Interestingly, there IUMs; however, when the details of these models were seemed to be less of a focus on developing “underused clarifi ed using subsequent survey questions, it was de- areas”. termined that only six of the 50 authorities use truly Looking at the developmental and operational as- integrated urban models. Th is clarifi cation presented pects of urban models, the majority were developed us with a problem regarding how to interpret the re- using outside knowledge and resources – via a com- maining results of the survey. Th erefore, we decided mercial product or through cooperation with con- to move forward with the fi nding that, while only six sultancies or research institutions. In the majority of of 50 respondents use truly integrated urban models, cases, respondents stated that staff at the authority in 16 respondents use some degree of urban modelling

NORDREGIO WORKING PAPER 2014:1 11 question cannot augment or update the model. Th is in- menting IUMs ought to be – the provision of a tool to dicates that urban planners who work “on the ground” support decision-making that stimulates dialogue and for municipalities or regions oft en do not possess the debate among a broad group of stakeholders. expertise to operate such models.. While 16 respondents stated that urban models have While some enlightening answers were provided to been applied in their area, 34 of the 50 respondents said questions regarding the attributes, basic features and they have not implemented an IUM. Some of the re- development of urban models, more detailed questions spondents in the latter group felt that their municipal- on methodology and operational details appeared to ity or region is too small to justify the use of an IUM, be too detailed for most of the respondents. Also, the the appropriate land use and transport data are too overall lack of response to the question about the costs hard to access and/or the costs are unjustifi ably high. associated with urban models shows that defi ning their However, the most common rationale that respondents actual costs is a diffi cult task. Rather than being one- provided for not using IUMs was that they were not dimensional, costs are oft en distributed across various familiar enough with IUMs and their benefi ts and do budget lines, such as development, implementations, not have the technical skills to use them. updating, consultancy costs, internal labour costs, etc. Th e survey’s fi ndings indicate that more and bet- Th e survey also indicated that internal knowledge ter information on the use and benefi ts of IUMs could and experience, high cost and data issues also act as be provided to planners and other related practition- barriers to the use of urban models by authorities. ers. Th is information should focus on how to integrate Th e notion of collaboration and cooperation across these tools into current practices as planning support municipal boundaries was also addressed in the sur- systems and on the practical benefi ts of applying the vey. Th e results showed that, among the ten munici- models. Consequently, a more in-depth analysis in- palities which stated they used models (the other six volving those respondents who are in the process of cases were regional authorities), only three models applying or have successfully applied such models were developed completely independently. Th e other would make an important contribution to the IUM seven municipalities stated that the urban models were knowledge base. Not only would it be important to developed in some form of cooperation with other learn more about issues such as costs, data availability, municipalities or their respective regional authorities. learning and knowledge development, but it would also Th erefore, the degree of cooperation appears not only be valuable to learn more about exactly how the mod- to be quite positive, it also introduces the question of els are being used by planners; for what purposes, by whether IUM projects could be seen as an opportunity whom and with what end goals or objectives in mind. to support new and strengthened forms of cooperation Th is type of information provides insight into the ex- between municipalities and regional authorities. Fur- pectations placed on the use of models that can provide thermore, cooperation among and between authorities newcomers with a better view of how they might ben- touches on what the most important objective of imple- efi t from these tools.

12 NORDREGIO WORKING PAPER 2014:1 1 Introduction

1.1 Cities, environment and urban understanding of the future impacts of diff erent planning policies, supported by knowledge about economic planning theories and social behaviours (Berglund, 2014). Physically, cities are expanding faster than ever and, Th is report presents the fi ndings of a survey that was compared to rural areas, are increasingly responsible carried out by Nordregio between 14 November 2013 for the vast majority of national growth in developed and 10 January 2014. Th e aim of the survey was to learn countries. However, parallel to this growth, the spatial more about the use of integrated urban modelling tools planning of city development is becoming increasingly by Nordic authorities through the following questions: complex – having to support growth opportunities To what extent are integrated urban models being while balancing the need to focus on important chal- used by Nordic planning authorities? lenges such as demographic transformations, social What are the basic technical details of the models problems and unrest, resource depletion, climate being used? change management, transport congestion and under- How are these models developed, implemented and developed mobility, and diffi cult trade-off s in land use used as planning support systems? decisions. Furthermore, these natural and human factors are highly interconnected, with dynamic feed- 1.2 What are integrated urban backs that require comprehensive approaches to understanding the challenges and opportunities for models and why should we use development. them? Worldwide, cities are responsible for almost 75% of Integrated urban models (IUMs) combine multiple ur- global resource consumption and contribute to a ma- ban attributes to produce future land use development jor proportion of the environmental impacts generated scenarios. Th ese attributes revolve around the interplay by the unsustainable production and consumption between land use and transport development, and are of resources (UNEP, 2011). Because of these impacts, processed in the model to propose optimal locations cities are obvious targets for policies and actions that for future development. Attributes that may be includ- make genuine progress on environmental change (Bai, ed in an IUM: McAllister, Beaty, & Taylor, 2010). While part of this relates to the fact that decision-makers at municipal  Residential buildings – location and density and regional levels have an increasingly large role in  Population growth shaping economic growth strategies (Katz & Bradley,  Non-residential buildings – location and density 2013), the reality is that cities and their administrations  Job growth can act as leaders on multiple sustainability-related is-  Roads sues (Jensen, Christensen, & Hansson, 2011). Among  Green space protection the most important of these issues, municipalities of-  Public transport network ten have the main capacity in two fi elds with impor-  Environmental risk management tant sustainability implications: land use and transport  Cycling network planning. Th erefore, the tools used to support planners  Cultural heritage sites in their work with comprehensive policy and planning campaigns can play a crucial role in determining both Th ese types of models have been developed over the the resource effi ciency and attractiveness of urban ar- past 65 years and have had many diff erent names, e.g. eas (Jensen, Christensen, & Hansson, 2011). Th ese tools land use models, integrated land use models, land use include integrated land use-transport models – acting transport interaction (LUTi) models, integrated urban as planning support systems – which provide a better models, etc. We will proceed using the term “integrat-

NORDREGIO WORKING PAPER 2014:1 13 ed urban models” (IUMs) in order to be as broad as nity characteristics and other quality-of-life attributes. possible and to acknowledge the highly-integrated na- Planners rely on a number of diff erent tools to help ture of land use planning. A common feature of each of them accommodate the range of factors and considera- these models is that they, to diff ering degrees and tions they face. Apart from IUMs, others include com- through diff ering processes, integrate the land use/ prehensive and detailed plans, regional development built environment and transport dimensions to pro- plans, environmental impact assessments (EIA), stra- vide an urban development scenario. Th is integration tegic environmental assessments (SEA), life cycle anal- is based on the well-understood relationship between yses (LCA), ecological footprints, green accounting, the development of the transport network and the de- multi-criteria assessment methods, indicator systems, velopment of the rest of the urban system, where “land frameworks, certifi cation schemes and guidelines, all use patterns infl uence travel needs, mobility patterns, of which are developed for a range of single or multi- and the evolution of transportation infrastructure; and sectoral aspects of planning (Jensen & Elle, 2007). the transportation system, in turn, infl uences where IUMs help planners by: people engage in activities and how urban form changes.” (EUNOIA, 2012, p. 9). In other words, land use and  Providing a better understanding of the urban dy- transport – as the location of things and the movement namic system. between them – are the essence of spatial patterns in  Highlighting the trade-off s implied in diff erent ur- cities. ban development options in an integrated way by creat- Over the past 25 years, our understanding of how ing a simplifi ed reality that can be manipulated through to plan cities has shift ed, from top-down ordering and simulation. equilibrium to concern for local detail, complexity  Enabling virtual experimentation that allows plan- and change that is driven from the bottom-up (Batty, ners to test, demonstrate and evaluate the potential 2013a). Th is paradigm shift has been driven by two key consequences of policy decisions and other actions on factors: the incredible technological improvements in land use patterns (i.e. multiple simulations based on computing and geographic information technology diff erent development thresholds or planning strate- and the impact of the “dissenting voices” of urban- gies). ists who have led us to an understanding that cities are  Producing visual scenarios that can facilitate par- complex, heterogeneous hubs of diversity, novelty and ticipatory, collaborative decision-making (in policy surprise (Batty, 2008). and design roles). Th e visual outputs provide a series of Alongside this shift , urban growth has continued “what if” scenarios, dependant on the strategic policy faster than ever, resulting in intense competition for decisions that are made. space and ever increasing congestion. Planners therefore face the diffi cult task of dealing with the complex, Rather than being defi nitive statements about the fu- interconnected and ever-changing context of urban ture, these representations tend to narrow down the development. Th ey oft en struggle with the challenge number of possible policy interventions without mak- of balancing the demands of growth with the desire ing a predictive statement about which is the only, or to preserve the natural environment, unique commu- optimal, solution.

14 NORDREGIO WORKING PAPER 2014:1 2 Nordic Survey on Integrated Urban Models

Few studies have focussed on the extent to which IUMs to ensure that all new corresponding municipalities are used in practice and there is not one available that and regions were covered in the survey. All core and is set in a Nordic context. We therefore wanted to know hinterland municipalities in each city-region received where, how and why these models are being used in the the survey (see Map 1), as did each of the regional au- Nordic region, or why they are not being used. While thorities. acquiring this information we also tried to get a sense Th e respondents at the municipalities and regional of the opportunities for future use. authorities who received the survey were identifi ed in several ways. Some of the contacts were provided by their respective country’s association of local authori- 2.1 Method ties, while others were retrieved through individual Preparation of the survey involved a three-step process internet searches. Th is resulted in a varied contact list in order to ensure that it targeted the most relevant in- that included head planners, Geographic Information formation about the development, implementation and System GIS experts and even the general inboxes of use of IUMs in the Nordic countries. First, Nordregio some planning departments. prepared a draft version of the survey based on their Ideally, the hope was that those receiving the sur- existing expertise and a detailed literature and internet vey would forward it to their colleague who was best review of land use modelling. Th is was done using the equipped to provide a response. However, it became online platform Google Drive, which was chosen be- clear that the questions were oft en too diverse (cover- cause of its user fl exibility and graphic qualities. Next, ing expertise held by multiple people rather than just the draft was reviewed by the consultancy WSP Analy- one individual) to be answered by a single respondent sis and Strategy, who have a well-established history of and required additional information from multiple developing and implementing integrated land use people. Th is is most likely due to the fact that questions models. Th is review led to an iteration of the fi rst draft , addressed a conceptual, analytical and organisational which was then sent to three municipalities during a space between GIS and urban planning and because in pilot phase. Th is phase provided valuable feedback re- large organisations it is oft en multiple individuals who sulting in the survey being shortened and some of the work with specifi c aspects of these fi elds. It thus turned questions being revised to produce the fi nal version. out to be diffi cult for one respondent to provide a com- Th e survey was distributed to 140 municipalities prehensive overview of how tools were used in other and regional authorities in the Nordic countries on departments. 10 November 2013. By 10 January 2014, a total of 50 Considerable eff ort was put into obtaining responses responses had been received and representative cover- and following up on certain responses deemed par- age of the core city-regions was achieved by retrieving ticularly interesting or that required clarifi cation. For answers from at least one of the planning authorities instance, some respondents had indicated that they (either the municipality or the regional authority) in used IUMs, whereas it was subsequently made clear each of the targeted regions. that the models used were of a simpler kind, usually Th e targeted planning authorities were selected by based on more straightforward GIS spatial analysis. identifying the largest city-regions in the Nordic coun- Th is divergence had an impact on the statistical results tries and their functional urban areas (FUAs) accord- obtained from the survey and required us to make a ing to the OECD data and methodology (see Map 1). distinction between those respondents who appear to While the OECD data dates from 2006 (prior to the be using “truly integrated urban models” and those municipal reforms in Denmark and Finland), the new who use some degree of modelling, but in what is prob- municipal structure was overlaid on the older version ably a more simplistic and less integrated context. We

NORDREGIO WORKING PAPER 2014:1 15 therefore decided to move forward on the understand- integrated way. ing that, while those respondents reporting that they Lastly, while the responses do not provide a suffi - use IUMs in the fi rst instance have adopted some sort cient sample size to make defi nitive statements about of integrated approach to spatial development/physi- the characteristics of IUM use in the Nordic countries, cal planning, only a subset of these have gone as far to the 50 responses do provide insights from at least one implement what we refer to as “true IUMs”. While the of the planning authorities (either the municipality or survey identifi es those cases where true IUMs appear to the regional authority) in each of the targeted FUAs. be used, the remainder of the analysis is based on the Consequently, we can use the results to make indicative wider group of respondents who stated they use an IUM statements about potential IUM usage patterns in the in the fi rst instance. Nordic countries. Th e results thus provide an interest- We therefore make a distinction; the term “IUM” ing overview of the current situation and insight into only refers to those cases where we think true IUMs the work that can be done to support more effi cient use are being used. We refer to all the other cases simply as of models in the future. “urban models” that consider land use in some kind of

M ap 1: Nordic city-regions targeted by the survey.

16 NORDREGIO WORKING PAPER 2014:1 Nordic City-Regions

Tromsø

ordreg

io & N

LS forFinland admi

nistrat

e Oulu boundari

Umeå

Kuopio Trondheim

Jyväskylä

Tampere

Lahti

Bergen Turku Helsinki

Oslo Uppsala Västerås Larger urban areas Stockholm Stavanger Örebro Core Municipalities

Norköping Hinterland Municipalities Kristiansand Linköping Smaller urban areas Core Municipalities Göteborg Jönköping Borås Hinterland Municipalities Identification of Nordic city-regions is based on the Aalborg OECD classification of functional urban areas (FUAs). "Larger urban areas" are city-regions with populations over 200,000, while "Smaller urban areas" are city- Århus regions with between 50,000 and 200,000 inhabitants. Helsingborg While the OECD classification identifies Linköping København Malmö and Norrköping as two seperate "Small urban areas", Odense they are identified here as one "Larger urban area". Oulu is also considered a "Smaller urban area" by the OECD, but because it's population is now over the 0 100 200 threshold it is as a "larger urban area" on this map. National boundary km § Regional boundary Data Source: OECD, 2008

NORDREGIO WORKING PAPER 2014:1 17 2.2 G eneral proﬁ le of the authori- While the main focus of the survey was on inte- ties that responded grated models, one question asked about the use of single models. Th is was based on the report by Berglund We received 50 responses to the survey, providing the (2014), which identifi ed three types of model systems desired representative coverage of the largest city-re- that can be combined into modular model systems: gions in the Nordic countries we had targeted. Of these 1. Single urban land use models, mainly used for plan- responses, 39 were from municipalities and 11 were ning purposes, as well as for policy analysis and assess- from regional authorities. Twenty-two responses came ment. Oft en detailed in terms of data and spatial reso- from Sweden, twelve from Denmark, eight from Nor- lution. way and seven from Finland. One response came from 2. Single land use change models, mainly used in envi- an Icelandic authority (within the Reykjavik city-re- ronmental sciences. Oft en detailed in terms of data and gion) and, even though this was not one of the targeted spatial resolution. city-regions, it was decided to include this response to 3. Land use and transportation models, which repre- spread our coverage to an additional Nordic country. sent a holistic, systemic approach to urban and trans- One of the fi rst questions concerned the size of the portation systems. Oft en more generalised in terms of responding planning authority. Even though the sur- data and spatial resolution. vey focussed on larger FUAs, in practice these are com- Th e last of the general profi le questions thus asked prised of upwards of 30 individual municipalities. As about which urban issues are monitored and/or strate- shown in Figure 1, over half of these municipalities gized using the fi rst two types of “singular” mod- have a population under 100,000. While these munici- els. Figure 2 shows that “Transport and Mobility”, palities are not small in a Nordic context (for instance, “Risks (e.g. fl ooding, etc.)” and noise are the dominant only fi ve Nordic city-regions have more than 500,000 themes, while “Housing” was only identifi ed by ten re- inhabitants and only 12 have a population greater than spondents. 200,000), they could be considered as small in the context of the scale in which IUMs should be implemented. Th is was indicated in the report by Berglund (2014), 2.3 The use of integrated models which described how the use of IUMs tends to be most – diffusion in the Nordic countries benefi cial in larger city-regions, where there is most In the survey, we identifi ed IUMs by saying that, “We likely to be pressure on land resources and complex, view ‘land use’ as refl ecting the development of the integrated transport infrastructure. built environment, which is the urban space where Another question sought to establish the extent to people live, work, commute and recreate on an every- which geographical information systems (GIS) are used day basis. Th erefore, land use models are technical by Nordic planning authorities, as this is required if tools for systematically analysing the impacts of poli- municipalities or regions are to adopt the use of IUMs. cies, plans, projections, decisions, etc. on the built envi- In line with expectations, almost every region and mu- ronment of urban areas. Th e notion of ‘integrated’ re- nicipality maintains a GIS database of land use and/ fl ects models that combine multiple attributes of the or demographic data to aid them in strategic planning.

Population distribution of the respondents

16 under 50,000 50,000 - 100,000 100,000 - 200,000 12 200,000 - 300,000 300,000 - 400,000 8 over 400,000 Number of respondents 4

Figure 1: Population distribu- 0 Municipality Region tion of the municipal and regional authorities that responded.

18 NORDREGIO WORKING PAPER 2014:1 Use of individual modelling tools for planning support

Urban issues 0 5 10 15 20 25 30

Transport and mobility

Risks (e.g. flooding, etc.)

Noise

Public space, parks and/or green space

Cultural heritage

Access to public services

Air quality

Housing Human qualities (e.g. attractiveness, safety, etc.) Other Figure 2: Urban issues covered by modelling tools. built environment and socio-economic development as from ESRI... “and “We use GIS programs to integrate parameters, restrictions or potentials of the model (e.g. diff erent types of information on maps, for example, combining multi-mode transport accessibility with public transport stops or fl ooding risks crossed with proximity to work space and preservation of green ar- property databases. We use MapInfo for this”. While eas).” many IUMs are largely constructed using GIS soft - It was surprising to note that 16 authorities initially ware, such as ArcGIS and its Spatial Analysis exten- stated that they use IUMs. From the results of the gen- sion, the fact that respondents couldn’t identify the eral profi le questions, it is notable that: names of the commercial or open-source models being used led to hesitation as to whether they are actually  Eleven of the 15 municipalities with at least 100,000 using IUMs. residents stated that they use at least one IUM, while In retrospect, it seems that our defi nition didn’t do a only two of the 23 municipalities with fewer than good enough job of emphasising that we were referring 100,000 residents said they use one. to models that produce future land use development  Eight of 27 municipalities (30%) with two or fewer scenarios; and that there was confusion about the dif- transport modes in their public transport network stat- ferences between single and integrated models. How- ed that they use an IUM, whereas four of the nine cities ever, because IUMs are probably not widely used, it is with at least three public transport modes said they use also likely that there isn’t a well-developed knowledge an IUM. Th is roughly corresponds to Berglund’s (2014) base on the characteristics, use and benefi ts of them in fi nding that integrating land use development with a the Nordic countries (cf. Berglund, 2014). complex and integrated transport infrastructure is an As a result, follow-up clarifi cation was carried out important rationale behind the use of integrated mod- with most of the 16 respondents who originally stated elling. that use/have used an IUM. We determined that six  Two of ten authorities in Denmark, four of seven in of the 16 respondents appear to be using an IUM that Finland, fi ve of eight in Norway and fi ve of 21 in Swe- provides future development scenarios: two in each of den stated that they used IUMs. While there are a Sweden, Finland and Norway. Th ree of these were mu- number of potential contributing factors that could ex- nicipalities and three were regional authorities. plain these diff erences (not least the small sample size), As mentioned in the methodology, we therefore de- the use of IUMs may be less prevalent in Denmark cided to proceed on the understanding that, while 16 than the other Nordic countries. regions/municipalities use some sort of integrated approach to spatial development/physical planning, only However, closer examination of the results showed that six of these have gone as far as to implement what we respondents did not have a consistent understanding of refer to as true IUMs. However, the remainder of the IUMs. For example, while some of the responses said survey evaluates the use of “urban models” based on that they do use integrated modelling, when asked the the 16 respondents who stated that they use these types names of these models answers included: “ArcGIS of modelling tools. analysis on planning and land use”, “We use ArcGIS

NORDREGIO WORKING PAPER 2014:1 19 Figure 3: Use of IUMs by the responding authorities. from a national dataset, combined with local data about how people live and travel. It is used to meas- Use of integrated models by the responding authorities ure distances and serves as an aid to locating new developments, schools or commercial areas, with 35 the ability to use diff erent scenarios to suggest the

30 most eff ective location. While only two instances were found during the survey, further investigation 25 revealed that it has been implemented in at least 13 20 diff erent regions and municipalities in Norway (Skogstad Norddal, 2014). 15

Number of respondents 10 YKR-aineistot: YKR is the Finnish Monitoring System of Spatial 5 Structure, developed by the Finnish Environment 0 Institute, SYKE. It is a raster model (250x250 m no yes yes, don't yes, after multiple know follow-up resolution) that can analyse data about societal structure or local information about the number of inhabitants and workplaces, building density and 2.3.1 Types of IUMs in use in the car ownership. Responses to the survey showed it Nordic countries was used by the municipality of Oulu and the Vars- inais-Suomi region. Of the six respondents using IUMs, three separate models were identifi ed. Th e following section off ers some of their basic characteristics: 2.3.2 Basic technical details of the models used in the Nordic region Integrated Planning Model (IPM): Th e basic technical details of the urban models are Th e IPM was developed by the consultancy WSP their overall methodological frameworks, soft ware re- Analysis & Strategy in 2006 and has been imple- quirements, spatial resolution and data input require- mented by two regional authorities: Stockholm and ments. Th e fi ndings showed that: Skåne. It is available free of charge, albeit with per- mission from the Stockholm County Council, and Questions were too detailed for most respondents: user support can be provided by WSP Sverige. It fol- Th e technical questions were too detailed for most lows the cellular automata principles and is an Arc- of the respondents. In the rare cases an answer was GIS-based land use model that provides detailed provided, the respondents tended to be GIS experts graphical output of future land use, with one of its rather than planners. Th is in itself is an important strengths being the ability to handle densifi cation fi nding as it shows that applying an urban model of already built-up areas. Its main applications are typically requires contributions from a number of in regional planning and future land use and it al- planning-related experts – experts who may rarely lows users to create and compare scenarios refl ect- work in close collaboration with each other. In turn, ing policy choices. It is modular soft ware that is this suggests the importance of good inter-depart- part of an integrated model system with the region- mental collaboration to facilitate urban model de- al economic model rAps and the transport model velopment. LuTRANS/SAMPERS (Berglund, 2014).

Overall methodological framework: Areal- og transportplanleggingsmodellen (ATP model): Cellular automata (CA) models are clearly the most Th e ATP model is a transportation and localisation common, followed by a few respondents who noted model used in Norway. It was developed as a re- that their urban models are based on equilibrium search project and is freely available. However, only principles. As noted in the parallel report by WSP the consultancy Asplan Viak has the ability to up- Analysis & Strategy (Berglund, 2014), this is inter- date and develop it further. Asplan Viak also pro- esting because, while CA is an important type of vides analyses for municipalities or counties that framework, it is by no means the only approach. In prefer to contract out these services. contrast, many simple yet eff ective IUMs use a rule- Th e model is based on the transport network taken

20 NORDREGIO WORKING PAPER 2014:1 based method that determines the future state of a threshold). land parcel based on strategic planning goals. Figure 4 and Figure 5 show the attributes that are considered as restrictions or opportunities by diff er- Soft ware requirements: ent urban models. Restrictions based on mobility in- Commercial GIS programmes, including ArcGIS frastructure, cultural heritage, public and green space, and MapInfo, are clear standards, while open- existing building densities and risks appear to be the source soft ware and custom-built programmes most common. within the models themselves are also used by a As in the case of development restrictions, a major- number of authorities. ity of the identifi ed urban models prioritise urban development according to transport accessibility (roads Spatial resolution: and public transport), existing and future demand for Equally-sized cells representing land use are the residential space and population development, public main spatial representation. Th ese are likely to be at and green space provision and cycling infrastructure. a fairly high scale – up to 100 metres resolution. Th ematically, the most common policy priorities emphasised by the identifi ed urban models were: “stra- Data requirements and availability: tegic densifi cation of existing built up areas” and “areas While seven of the respondents noted that the data close to public transport”. Interestingly, there appears was available free of charge, four noted that most of to be less of a focus on developing underused areas. it is purchased from external providers and six However, it is important to point out that, by running mentioned that the data is available internally. In- a model multiple times, planners can get visual repre- terestingly, of the four who stated that they pur- sentations of how diff erent policy priorities could play chase external data, three were users of the IPM and out in space, thus providing planners and policymak- ATP models, which indicates that true IUMs likely ers with more information with which to evaluate their require data to be purchased. options, which is an overarching and fundamental benefi t of these types of models in general. 2.3.3 What land use attributes and policy priorities are considered? 2.3.4 How are urban models devel- Land use and socio-economic attributes are typically oped, implemented and used? modelled through transition rules that determine Most of the models identifi ed in the survey were devel- where and to what extent urban development takes oped with the help of external knowledge and resourc- place. Th ese rules restrict urban development in spe- es, i.e. using a commercial product or through cooper- cifi c locations while prioritising it in others. A basic ation with consultancies or research institutions example would be selecting “protected green space” as (Figure 7). Similarly, Figure 8 shows that, in the major- a restriction (an area that cannot be developed) and va- ity of cases, staff at the authorities in question rarely cant land next to a new train station as an opportunity have the specifi c expertise required to augment or up- (an area that can be developed up to a certain density date the preferred urban model.

Land use attributes considered as development restrictions by the models

Land use attributes 0 2 4 6 8 10 12

Mobility infrastructure

Cultural heritage

Public space, parks and/or green space

Residential building density

Non-residential/office building density

Risks (e.g. flooding, etc.)

Public services

Property/cadastral structure

Human qualities (e.g. attractiveness, safety, etc.)

Rental vs. Ownership in housing

Air quality Noise Figure 4: Land use attributes Land or building market value considered as development restrictions by urban models.

NORDREGIO WORKING PAPER 2014:1 21 Figure 5: Land use attributes considered as development opportunities by urban models.

How models are used in a planning perspective is de- nicipal or regional boundaries do not refl ect the termined by the context and will vary considerably de- way cities function or it could be that models have pending on how governance processes unfold in prac- been developed by national governments for use at tice. Th us, it is a diffi cult question to answer using a the regional or local level. Municipal respondents survey. Nevertheless, Figure 9 shows that two thirds of were therefore asked how their preferred model has the responses stated that their urban model is used to been integrated with their neighbouring munici- support strategic planning on an ongoing basis. Th is palities. Only four of 11 municipalities stated that refl ects the typical regenerative nature of strategic they apply their model only to their municipal area plans, which oft en go through circular processes of (or parts of it). In contrast, three other municipali- continuous development, implementation, re-formula- ties noted that they apply their model in collaboration and updating. tion with neighbouring municipalities and three Issues relating to cooperation, cost and constrains others share it with their respective regional au- were also addressed in this section of the survey. thority. Not only does the degree of cooperation appear to be quite positive, it also introduces the ques- Cooperation among municipal authorities: tion of whether IUM projects could be seen as an Sharing the development or use of an urban model opportunity to support new forms of cooperation could be due to a number of factors – it could help on urban spatial development issues. to reduce costs, it can help address the fact that mu- Cost: Th e diffi culty of accounting for the costs associated Figure 6: Land use policy priorities emphasised by urban with developing and implementing urban models models. was emphasised by the overwhelming lack of

Land use development outcomes responses to our questions on the issue. As sug- emphasised by the preferred model gested by Berglund (2014), costs can vary consider-

17% ably. For instance, an open-source model is free, but there are costs associated with data collection, 31% operating, training and maintenance, whereas models requiring the use of consultancies will have a completely diff erence cost structure. Only two 17% respondents were able to provide an estimate of the cost, both users of the IPM model. One region said

4% the cost of implementing the model was SEK 150,000, while the other said that it cost SEK 1.45 million to develop and implement; an illustration 31% of the large variations in estimated costs. Strategic densification of existing built-up areas Priority for development of underused areas Strategic densification in areas close to public transit Achieving mixed land use Other

22 NORDREGIO WORKING PAPER 2014:1 Figure 7: How was the preferred model developed? Figure 8: Are staff in the authority’s planning department able to augment or update the model?

How was the preferred model developed? Are staff in the authority's planning department able to augment or update the 6% model?

25%

25% 37%

63%

13% 31% Internally In co-operation with a consultant In co-operation with a research institute or university Commercial product (3rd party) Yes No Non-commercial product (e.g. open source)

Constraints: 2.4 What about those authorities Seven respondents provided answers about the key constraints associated with the use of their pre- not using an integrated model? ferred model. Two other respondents noted that Th e 34 respondents who stated that their authority their urban model is diffi cult to use, and the high does not use IUMs were asked an alternate set of ques- cost of developing and/or running the model was tions. First, they were asked if they would consider us- also mentioned twice. Data constraints were also ing one in the future; 16 provided feedback, with seven identifi ed in four cases, including issues relating to saying yes, fi ve maybe and four saying no. Some of the accessibility, heavy data slowing down the model most interesting comments were: and diffi culties updating data.  “Hard to say. We do not know the market for this Figure 9: How is the prefered model used by your authority? kind of tool.” “If the tools were more intuitive and easier to use, and if the data were more accessible and well-

How is the preferred model used by your described.” authority  “Not at this time. It is oft en time-consuming to learn a new tool. Th e tools we use today are adequate for the small scale we work at. It’s rare that issues are so complex we need modelling, and in those cases we do need

35% it (for example, noise issues), we use consultants in- stead.”  “Our municipality has 2,400 inhabitants, and only one person working in the planning fi eld, therefore, I am not sure we will consider using a land use model.”

Each of the above comments refl ects how the respective authority does not currently have the technical knowledge to develop or apply such a tool and that obtaining that knowledge might be too costly, relative to the per- ceived or expected benefi ts. In addition, two of the

Used regularly as a strategic decision-making tool comments specifi cally note how their municipalities are too small and land issues are rarely complex enough Used as an input for a single project or policy document to warrant the use of such models.

NORDREGIO WORKING PAPER 2014:1 23 Another comment specifi cally referred to the cost of plan. When this happens, the full merits of the tools and bureaucratic process involved in investing in such (for supporting discussion and deliberation) may not tools: be properly acknowledged. Lastly, in an attempt to gain a better understanding  “Yes, but it’s been hard to convince the leadership of of the reasons why they do not use IUMs, respond- the administration.” ents were asked to rate, between 1 (strongly disagree) and 5 (strongly agree), how strongly they agreed with But most interesting was the following comment: fi ve statements about why their authority does not use IUMs. Th e results in Figure 10 allow for the following  “Some years ago (2003), we developed a plan using refl ections: operationalised goals for future expansion. Th e operationalised objectives were used to prioritise diff erent  On average, respondents disagree with the assertions land uses. GIS was an important tool. A model was “municipal size (too small)”, “cost (too high)” and “data built. Th e material was very interesting as a basis for availability (too diffi cult)” more than they agree with discussion, but could not be used directly to produce a them. However, the polarised responses to each state- plan. We are not currently considering the future use of ment (i.e. strongly agree or strongly disagree) show that integrated modelling tools.” these could be important reasons for some authorities. Th e last quote perhaps implies that the value of these For example, we compared the responses with the size types of IUMs can be misinterpreted, focusing solely of the authority and found that those strongly agreeing on the production of part of an urban development with size being an important rationale were small mu-

Figure 10: A series of questions asking respondents the extent to which they agree with statements regarding why IUMs are not used by their authority. Responses in blue represent an average response below three, i.e. an average disagreement with the statement. Responses in orange represent a response above three, i.e. an average agreement.

Our municipality/region is too small to justify the use of Data is too difficult to access integrated urban models 0 2 4 6 8 10 12 0 5 10 15 Scale Scale 1 - strongly disagree 1 - strongly disagree

2 2

3 3

4 4

5 - strongly agree 5 - strongly agree

The use of a modelling system is too expensive for our Our staff does not have the technical skills to use the modelling municipality/region tools 0 2 4 6 8 10 12 0 2 4 6 8 10 12 Scale Scale 1 - strongly disagree 1 - strongly disagree

2 2

3 3

4 4

5 - strongly agree 5 - strongly agree

We are not familiar with integrated land use models or their benefits 0 2 4 6 8 10 12 Scale 1 - strongly disagree

5 - strongly agree

24 NORDREGIO WORKING PAPER 2014:1 nicipalities (most of them under 100,000 inhabitants), the use of such models. while those strongly disagreeing had a much larger population. Population size thus appears to be an im- In the last question in the survey, respondents were portant factor in determining whether models are asked if they would like to be provided with a report used. synthesising the results of the survey. Th e high number  We quite clearly see that unfamiliarity with IUMs of positive responses suggests that they would like and their benefi ts and a lack of the technical skills to more information about using models and that they manage IUMs scored above the 3.0 threshold. Th is in- fi nd the information interesting and relevant to their dicates that these are likely to be key factors limiting work.

NORDREGIO WORKING PAPER 2014:1 25 26 NORDREGIO WORKING PAPER 2014:1 3 Conclusions

In relation to the general objectives of the survey, a nities in the models tell us a lot about what policies are number of conclusions can be drawn: being prioritised by planners. Development restrictions seem to be most commonly based on existing To what extent are IUMs being used by Nordic planning building densities, the location of mobility infrastruc- authorities? ture, existing buildings/areas of cultural heritage and existing public and green space. Similarly, urban devel-  Th e survey results indicate that the use of IUMs is opment appears to be promoted based primarily on very limited. Although 16 respondents initially stated transport accessibility (through roads, public transport that they use IUMs, further analysis of their responses and cycling infrastructure), population growth, future indicates that only six of the 50 respondents use/have demand for residential space and building density used one of the three integrated models identifi ed: thresholds. Th e core focus on building characteristics IPM, YKR and the ATP model. and mobility infrastructure seems to refl ect the essence  Two of ten authorities in Denmark, four of seven in of the built environment, as the locations in the city Finland, fi ve of eight in Norway and fi ve of 21 in Swe- and the way in which we move between these locations. den stated that they used IUMs. While there are a  Respondents identifi ed most with the idea that they number of potential contributing factors that could ex- use models to promote strategic densifi cation of exist- plain these diff erences (not least the small sample size), ing built-up areas and areas close to public transport. the use of IUMs may be less prevalent in Denmark However, it was interesting to note the comment by one than the other Nordic countries. of the respondents who stated that their model empha-  In terms of those who responded that their authority sises whichever development outcome is identifi ed by a does not use an IUM, Figure 10 showed that a lack of particular policy strategy. For example, by running a familiarity with IUMs and of the technical expertise to model multiple times, visual representations of how implement them seem to be the dominant factors lim- diff erent development scenarios will play out in space iting their wider use. can be obtained.

What are the basic technical details of the models being How are these models developed, implemented and used used? as planning support systems?

 Th e general lack of consistency and the tendency for  Th e survey results supported the fi nding by Berglund respondents to skip these questions is what initially (2014) that the costs of implementing an IUM can vary made us aware that many planners in our target group considerably depending on purchase, consultancy, are not familiar with the fi ner details of IUMs. Th e fact training, implementation and maintenance costs – that respondents couldn’t identify the names of the which in turn vary depending on for what and for how commercial or open-source models being used called long the model will be used. their responses into question.  External support from consultancies or research in-  While cellular automata (CA) models were identifi ed stitutes appears to be common in the development of as the most common methodological framework, they integrated models as urban planners working for mu- are by no means the only approach. In contrast, many nicipalities or regions typically do not possess detailed simple, yet eff ective IUMs use a rule-based method knowledge of such models. Th is was also observed by that determines the future state of a land parcel based Berglund (2014), who noted that a majority of IUMs are on strategic planning goals. Th ese are probably the best developed in academia and, if successful, are subse- option for planning authorities making an initial in- quently implemented in practice. vestment in IUMs (Berglund, 2014).  Of the ten municipalities that mentioned they have  Th e attributes identifi ed as restrictions and opportu- used an urban model, only four stated that it was im-

NORDREGIO WORKING PAPER 2014:1 27 plemented by their authority on its own. In contrast, tion not only appears to be quite positive, it also intro- six respondents mentioned that model development duces the question of whether IUM projects could be and/or implementation involved cooperation with oth- seen as an opportunity to support new and strength- er municipalities or some sort of shared approach with ened forms of cooperation between municipalities and regional authorities. Th erefore, the degree of coopera- maybe even regional authorities.

28 NORDREGIO WORKING PAPER 2014:1 4 Directions for future work

An underlying objective of this survey, together with model would appear to be the most appropriate next the inventory of IUMs, was to identify a model that step in widening the use of IUMs in the Nordic coun- could be recommended for use in the Nordic countries tries. In principle, these types of models would be rela- – one that might enable a comprehensive, integrated tively simple to implement and the rationales behind approach to the modelling of urban development pos- their fi ndings are easier to comprehend and thus more sibilities. However, the results of the project show, as likely to gain support within the Nordic planning com- expected, that only a few planning authorities current- munity. ly use IUMs. In defence of this, it seems that there are Lastly, the respondents showed signifi cant interest some important barriers to the use of existing models, in the theme of integrated models. Even though most not least in terms of overall lack of knowledge about the of the respondents haven’t used IUMs in practice (and implementation and benefi ts of IUMs. Th is identifi es a many haven’t been introduced to them at all), the vast knowledge gap within the planning community and majority are keen on learning more about the results of between planners and researchers, which indicates the the survey. We also found that a small group of plan- value of providing additional information, especially if ners who have been working with IUMs in practice the benefi ts of using these models can be eff ectively were more than willing to share their experiences in communicated. Th is could provide practitioners with detail. Creating opportunities for them to share their the knowledge they need to make informed decisions experiences of these models could therefore lead to in- about whether to invest in applying these tools to their creased discussion concerning IUMs and their poten- work. Th is could include deeper analysis of issues such tial benefi ts. as costs, data availability and knowledge development. As researchers, we certainly appreciated the fact that But it could also be valuable to learn more about ex- the majority of the respondents seemed to value the op- actly how the models are being used by planners; for portunity to refl ect on processes within their organi- what purposes, by whom and with what end goals or sation such as the lack of knowledge about what tools objectives in mind. other departments were working with, the gap between Much is currently being written on the need to im- planners and IT departments or even the lack of sup- prove the visual outputs of IUMs (cf. Serras, Bosredon, port from above for investments in these tools – despite Herranz, & Batty, 2014) and the results of the survey requests. From this engagement with respondents we also strongly encourage this. For IUMs to be success- are able to see that the limited use of IUMs could also ful, model developers must continue to respect an im- be a sign of poor integration between the sectoral poli- portant design principle – that the outputs of the model cies of municipal and regional authorities. Th is poses should provide a basis for discussion among diverse the question of whether IUMs are not being used due groups of stakeholders about the long-term strategic to a lack of awareness, because they provide a poor rep- development of city-regions. resentation of reality or because of a deeper lack of in- Also, the lack of existing knowledge and use of tegration in urban spatial planning as a whole. IUMs indicates that a relatively simple, rule-based

NORDREGIO WORKING PAPER 2014:1 29 30 NORDREGIO WORKING PAPER 2014:1 5 References

Bai, X., McAllister, R. R. J., Beaty, R. M., & Taylor, B. (2010). Jensen, J. O., & Elle, M. (2007). Exploring the Use of Tools Urban policy and governance in a global environment: for Urban Sustainability in European Cities. Indoor and Built complex systems, scale mismatches and public participa- Environment, 16(3), 235–247. tion. Current Opinion in Environmental Sustainability, 2(3), 129–135. Katz, B., & Bradley, J. (2013). The Metropolitan Revolution: How Cities and Metros Are Fixing Our Broken Politics and Batty, M. (2008). The Size, Scale, and Shape of Cities. Sci- Fragile Economy. Washington D.C.: Brookings Institution ence, 319, 769–771. Press.

Batty, M. (2013). The New Science of Cities. Boston: MIT Serras, J., Bosredon, M., Herranz, R., & Batty, M. (2014). Press. Urban Planning and Big Data - Taking LUTi Models to the Next level? Nordregio News, (1). Berglund, L. (2014). Review of Land-Use Models -Summary and Documentation. Stockholm, Sweden: WSP Strategy and UNEP. (2011). Towards a Green Economy: Pathways to Sus- Analysis. tainable Development and Poverty Eradication. A synthesis for policy makers. United Nations Environment Programme. EUNOIA. (2012). Urban models for transportation and spatial planning. EUNOIA Constortium.

Jensen, J. O., Christensen, T. H., & Hansson, K. G. (2011). Sustainable Urban Development - compact cities or consum- er practices. Landinspektoeren. Tidsskrift for Kortlaegning Og Arealforvaltning., 46(1), 50–64.

NORDREGIO WORKING PAPER 2014:1 31 ISSN 1403-2511 ISBN 978-91-87295-18-8

Nordregio P.O. Box 1658 SE-111 86 Stockholm, Sweden [email protected] www.nordregio.se www.norden.org