Water Management & Climate Change in the Okanagan Basin

Edited by Stewart Cohen & Tanuja Kulkarni

Environment Canada & University of British Columbia

Contributing Authors:

Roxanne Brewer1 Stewart Cohen2,4 Erin Embley3 Stuart Hamilton1 Maggie Julian3 Tanuja Kulkarni2,4 Bill Taylor1 James Tansey4 Rob VanWynsberghe4 Paul Whitfield1

For further information, please contact Stewart Cohen at: [email protected]

Sponsoring Agency

This project was supported by a grant from the Climate Change Action Fund (Project A 206), Adaptation Liaison Office, Natural Resources Canada, 601 Booth Street, Room 388, Ottawa ON, K1A 0E6.

1 Environment Canada, Pacific and Yukon Region 2 Environment Canada, Adaptation and Impacts Research Group 3 University of British Columbia, School of Regional and Community Planning 4 University of British Columbia, Sustainable Development Research Institute

Table of Contents EXECUTIVE SUMMARY… ………………………..….…………………………………………………6

1.0 PURPOSE OF THE STUDY (Stewart Cohen)…………………………………………………….7

2.0 Background (Erin Embley, Ken Hall, Stewart Cohen)………..……………………………….8 2.1 Regional Context…………………..………………………………………………...………….9 2.2 Case Study: Trout Creek ……………………………………………………….…….……….11

3.0 STUDY FRAMEWORK (Stewart Cohen)…………………………………………………………12

4.0 CLIMATE (Roxanne Brewer, Bill Taylor) 4.1 The Climate of the Okanagan Basin…………………………………………………………14 4.2 Global Climate Change………………………………………………………………………..14 4.3 Climate Variability……………………………………………………………………………...14 4.4 Climate Change in the Okanagan Basin: The Past 100 Years……………………………15 4.5 Establishing a Baseline: Climate Normals…………………………………………………..17 4.6 Predicting Climate Change 4.6.1 Global Climate Models……………………………………………………………..…18 4.6.2 Model Intercomparisons………………………………………………………………19 4.6.3 Climate Change Scenarios for the Okanagan Basin–PRISM Maps……………..20 4.7 Effects of Climate Change on Degree Days………………………………………………...21 4.8 Conclusions…………………………………………………………………………………….22

5.0 HYDROLOGY (Stu Hamilton) 5.1 Introduction……..……………………………………………………………………….…….22 5.2 Methods……………………………………………………………………………………..…23 5.2.1 Statistical Modeling (Paul Whitfield)……………………………………………..23 5.2.2 Temporal Analogue Determination………………………………………………..23 5.2.3 Process Modeling……………………………………………………………………24 5.3 Results…………………………………………………………………………………………26 5.3.1 Statistical modeling (Paul Whitfield)……………………………………………..26 5.3.1.1 TEMPERATURE……………………………………………………………..26 5.3.1.2 PRECIPITATION……………………………………………………………..26 5.3.1.3 NATURAL STREAMS………………………………………………………...26 5.3.1.4 REGULATED STREAMS……………………………………………………..26 5.3.1.5 SUMMARY…………………………………………………………………..26 5.3.2 Analogue and Process Modeling………………………...………………………..27 5.4 Discussion……………………………………………………………………………………..29

6.0 DIALOGUE 6.1 Ongoing Regional Consultation Process (Erin Embley)…………………………………29 6.2 Focus Group (Rob VanWynsberghe)……………………………………………………...29 6.2.1 Focus Group: Agenda Script……………………...………………………………….30 6.3 Stakeholder Reaction to Scenarios (James Tansey, Maggie Julian, Tanuja Kulkarni, Stewart Cohen)..………………………………………………………………………...32 6.3.1 Data Analysis…...………………………………………………………..…………...32 6.3.2 Impacts…………………………………………………………………..…………….33 6.3.3 Climate Change Adaptation……………………………………………..…………..35 6.4 Implications of Adaptation Choices (Maggie Julian, Tanuja Kulkarni, James Tansey, Stewart Cohen)…………………………………………………………………..37

7.0 FUTURE WORK (Stewart Cohen)……………………….……………………………………….39

8.0 ACKNOWLEDGEMENTS……… ………………………………………………………………….40

______3 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

9.0 REFERENCES……………………………………………………………………………………….41

10.0 APPENDIX 1: Climate Observations…………………………………………………………….44

11.0 APPENDIX 2: Hydrological Observations………………………………………………….……48

12.0 APPENDIX 3: Focus Group Results………………………………………………………….….69

______4 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

List of Figures

Figure 2-1. The Okanagan Valley Figure 2-2. Okanagan Populations 1976-1999 Figure 3-1. Okanagan Climate and Water Resource Impacts Study Framework Figure 4-1. Trend in Minimum Temperature at Vernon Coldstream Ranch Figure 4-2. Trend in Maximum Temperature at Vernon Coldstream Ranch Figure 4-3. Trend in Frost Free Days at Summerland Figure 4-4. Trend in Total Precipitation at Kelowna Airport Figure 4-5. Temperature and Precipitation Normals (1961-1990) at Kelowna airport Figure A1-1. Seasonal Change in Minimum Temperature (°C) in the Okanagan Basin – 2020s Figure A1-2. Seasonal Change in Maximum Temperature (°C) in the Okanagan Basin – 2020s Figure A1-3. Seasonal Change in Precipitations (%) in the Okanagan Basin - 2020s Figure A1-4. Projections of January Mean Minimum Temperature in the Okanagan Basin Figure A1-5. Projections of July Mean Minimum Temperature in the Okanagan Basin Figure A1-6. Projections of Annual Total Precipitation in the Okanagan Basin Figure A1-7. Projected Changes in Growing Degree Days (>5°C) at Kelowna Airport Figure A1-8. Projected Changes in Heating Degree Days (<18°C) at Kelowna Airport Figure A1-9. Projected Changes in Cooling Degree Days (>18°C) at Kelowna Airport Figure A2-1. Examples of Recent Shifts in Temperature for 11 day time periods in the Okanagan Figure A2-2. Examples of Recent Shifts in Precipitation for 5 day time periods in the Okanagan Figure A2-3. Examples of Recent Shifts in Natural Streamflows for 5 day time periods in the Okanagan Figure A2-4. Examples of Recent Shifts in Altered Streamflows in South-central BC Figure A2-5. Bellevue Creek Figure A2-6. Camp Creek Figure A2-7. Dave’s Creek Figure A2-8. Pearson Creek Figure A2-9. Vaseaux Creek Figure A2-10. Whiteman Creek Figure A2-11. Total Basin Snowpack Figure A2-12. Mean Flow During Spring Season Figure A2-13. Mean Flow During Summer Season Figure A2-14. Mean Flow During Fall Season Figure A2-15. Mean Flow During Winter Season Figure A2-16. Duration of Low Flow Figure A2-17. Date of Low Flow Figure A2-18. Duration of Peak Flow Figure A2-19. Date of Peak Flow Figure A3-1. Summary of Ranked Impacts and Adaptations Figure A3-2. Summary of Impacts and Implications for Health and Recreation Figure A3-3. Summary of Impacts and Implications for Agriculture Figure A3-4. Summary of Impacts and Implications for Forestry Figure A3-5. Summary of Impacts and Implications for Infrastructure Figure A3-6. Summary of Impacts and Implications for Ecosystems Figure A3-7. Summary of Impacts and Implications for Fisheries

List of Tables

Table 2-1. Transect of Climatic Conditions in the Okanagan Valley, B.C. (North to South) Table 6-1. Okanagan Climate Change Impacts Table 6-2. Adaptation to Climate Change Impacts Table 6-3. Implications of Adaptation Table 7-1. Next Steps for Okanagan Water Resources and Climate Change Studies Table A2-1. Description of Hydrometric and Climate Stations Used for Hydrological Modeling

______5 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

EXECUTIVE SUMMARY

Water resources, their management and use, are known to be sensitive to variations in climate, and will be influenced by projected climatic change. This study concerns potential implications of climatic change for the Okanagan region of British Columbia. The study has two main goals:

•= To identify climate change impacts and possible adaptation strategies for the Okanagan region, and •= To test an approach for engaging resource managers and regional stakeholders as collaborators in research and dialogue on climate change impacts and adaptation.

The use of climate scenarios, hydrologic modeling, and dialogue enables the inclusion of issues that may be difficult to model. Dialogue is used for those indicators and processes that include a human component, such as fisheries management or regional institutional arrangements.

Recent trends indicate that the regional climate is becoming warmer and wetter. Minimum temperatures have increased at a greater rate than maximum temperatures. Frost-free days have increased by 3.1 days/decade during the 20th century.

Climatic change scenarios, obtained from simulations from three climate models, indicate a temperature increase of 1.0-2.5°C from the 1961-1990 base period to the 2020s, and 3-5°C by the 2080s. Higher precipitation is projected for winter, but the climate model simulations do not agree on the direction of change for summer.

The recent warming has led to changes in streamflow. Observed changes in unregulated streams include earlier onset of the annual spring peak, lower fall flows and higher early winter flows. Regulated streams are showing decreases in discharge.

The hydrologic scenarios were calculated using a locally developed variant of a Swedish model known as HBV. HBV is sensitive to changes in both climate and landscape. In this study, however, only scenarios of climatic change were assessed. Results for the climate scenarios for six unregulated creeks within the Okanagan Basin indicate earlier onset of spring peak flows, by as much as 4-6 weeks. The peak is generally lower than current peak flows. All areas show loss of snowpack, with the highest elevation creeks showing the smallest loss. Winter flow increases, while summer flow decreases. There is no consensus on scenario changes to total annual flow.

The dialogue process consisted of focus group exercises designed to elicit views on impacts, adaptation, and implications of adaptation choices. Participants were shown the results of the climatic and hydrologic simulations during workshops held in January and March 2001. They identified impacts for forestry, agriculture, fisheries, infrastructure, health and ecosystems. These included changes in operating seasons, risks (e.g. fire, pests) and wildlife habitats.

Subsequent consideration of adaptation options resulted in a preference for structural measures, particularly intervention to prevent impacts (e.g. snow making, dams at high elevations, controls on land use and irrigation). Other popular alternatives were developing alternative uses for resources (e.g. alternative energy, grey water) and changing land use plans (e.g. densification of urban land).

Participants were then asked to consider implications of adaptation choices. Considerable attention was given to water licensing (e.g. sensitivity to increased license rates, need for more bureaucracy, loss of control to downstream users), flow regulation through dams (e.g. change from natural hydrograph, impacts on erosion and downstream fisheries), and potential restrictions on development (e.g. creates conflict over future visions). Many comments indicated a need for additional research and outreach activities (e.g. develop regional archive of water supplies and use, explore options for xeriscaping) or changes in consultative processes associated with a particular option.

______6 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

This was a short-term exercise in the use of scientific models to form the foundation for a regional dialogue on climate change impacts and adaptation. For this effort to be effective over the long term, there needs to be a sustained investment in this process.

1.0 PURPOSE OF THE STUDY

Water resources, their management and use, are known to be sensitive to variations in climate, and will be influenced by projected climatic change. Hydrologic studies of various watersheds throughout the world (Schriner and Street, 1988; Arnell et al., 2001) suggest changes in total annual flows, seasonal aspects of water supply and demand, implications for ecosystems, and challenges for managers in meeting multiple objectives (energy, irrigation, navigation, flood control, etc.).

The Columbia Basin has been the subject of some detailed case studies (Mote et al., 1999; Hamlet and Lettenmaier, 1999; Miles et al., 2000), and there has been an initial attempt to bring in transboundary perspectives (Cohen et al., 2000). This work suggests that a warmer climate would lead to changes in hydrology, including reduced snow pack and earlier snowmelt peaks, with subsequent implications for regional water supplies and fisheries. The earlier peak would lead to increased flow during winter months and an earlier flood season. Less water would be flowing during the summer months when irrigation demand is highest. Low summer flows would also affect hydroelectricity production and salmon habitat.

These findings have contributed to a growing dialogue on water management and climate change on the American side of the Columbia Basin. This needs to be matched by a similar dialogue in Canada, and ultimately, a bi-national one. Although basin-wide hydrologic assessments have been done, detailed hydrologic studies on the Canadian side were not part of this. The Okanagan region is one such area needing attention. The Okanagan is already experiencing rapid population growth and land use changes, with associated stresses on its water resource systems.

Another important aspect is the need to broaden the dialogue on adaptation (Smith et al., 2001). Changes in climate parameters may change opportunities and risks, but climate represents only one of many issues to be considered in resource management and use. In principle, there are many technical and institutional options available, but the implications of selecting any particular options have not been explored in a long-term planning and climate change scenario context. Certain options, such as new water pricing regimes or water banks (Bruce et al., 2000; Miller, 2000) are being considered and tested in areas facing water shortages now (e.g. California). However, it is not clear how these or other options would perform under a particular scenario of regional development (e.g. conversion of pasture to high intensity horticulture, urban growth, tourism growth, demands from the Columbia system) or a different climatic and hydrologic regime (e.g. longer warmer growing season, higher summer demand for electricity, changing winter/spring flood regime, changing fire and disease risks).

Besides the impacts and adaptation challenges being faced by the Okanagan and Columbia watersheds, there is also an important methodological concern – bringing regional aspects of climate change into a global-scale research and policy environment. Analyses within natural and social sciences have long been faced with the tradeoff of choosing between the difficulties of accounting for complex regional detail, and the ease of simpler aggregation accompanied by unrealistic assumptions about natural processes and human behaviour. National and global model-based impacts studies are, by necessity, highly aggregated. These incorporate assumptions about adaptation choices, and their acceptance into practice. Damage costs and adaptation costs and benefits have been estimated for the U.S. using these kinds of assumptions (Mendelsohn and Neumann, 1999). Results can be quite sensitive to what is assumed about

______7 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

levels of shoreline protection, choice of tree species for planting, etc., and how widespread these practices would become during the scenario time period.

If we’re going to estimate the costs of climate change impacts, and the value of any adaptation investments, there needs to be more attention given to the development of adaptation scenarios that could reflect regional opportunities and constraints associated with any options that might be considered. Just because a growing season may become longer doesn’t mean that all decision makers will be able to adjust to this in the same way or at the same pace. What would stakeholders really do in the face of such changes? How would governments, communities and the private sector incorporate uncertain scenarios of climatic change into their planning? This suggests that dialogue with stakeholders needs to be an explicit part of the process of framing research questions and carrying out impact and adaptation assessments.

This study of Okanagan water resources and climate change implications has two main goals:

a) to identify climate change impacts and possible adaptation strategies for the Okanagan region, and b) to test an approach for engaging resource managers and regional stakeholders as collaborators in research and dialogue on climate change impacts and adaptation.

2.0 BACKGROUND

The following section describes the Okanagan regions current water resource situation, including a case study Figure 2-1. The Okanagan Valley of one watershed under stress. The (www.panorama-map.com/z-pointsite/zoka.html) scenarios of climate and hydrology describe the area outside of the context of development pressures and expanding demand on natural resources.

______8 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

2.1 Regional Context

The Okanagan Valley is located in the southern interior of British Columbia, situated around Okanagan Lake (Figure 2-1). The valley is approximately 160 km in length and encompasses approximately 8200 km2 of land surrounding Okanagan Lake and Okanagan River (Okanagan Basin Water Board 1974, in OBWB 2000).

The Okanagan has a dry continental climate as the Valley sits in the rain-shadow of the Coast and Cascade Mountain Ranges. The semi-arid climate receives approximately 30 cm of precipitation per year; of this 85% is lost through evapotranspiration and evaporation from local lakes. The resultant gradient of climatic conditions in the Valley is shown in Table 2-1. The main water sources for the Okanagan are the tributary streams, which generally supply good quality water, are gravity fed and fairly inexpensive. However; there have been outbreaks of both giardiasis and cryptosporidiosis in the water of Valley communities. The Okanagan basin encompasses 13 municipalities, 3 regional districts, 4 First Nation communities and 59 “improvement districts” all of which share responsibility in water delivery. Various water managers oversee more than 45 community watersheds, 6900 water licenses (Okanagan + Shuswap) and over 385,000 acre-feet in water allocations. This has left very little leeway for further water allocations as most streams are listed as either “fully recorded” or “water shortage”.

Table 2-1. Transect of Climatic Conditions in the Okanagan Valley, B.C. (North to South)

Station Armstrong Vernon Kelowna Penticton Oliver Precipitation (mm) 448 387 315 290 305 Snowfall (m) 1.35 1.09 0.89 0.56 0.60 July Temp. (ºC) 19.2 20.2 20.3 20.1 23 Jan. Temp. (ºC) -6.4 -6.0 -3.5 -2.9 -2.0 Frost Free Days 196 226 228 235 257 Precipitation and Snowfall are mean annual values. Temperatures are mean values (Cannings et al., 1987)

Currently dams on Kalamalka, Okanagan, Skaha and Vaseaux Lakes control water flows, in order to provide for both consumptive needs and flood control. There is also secondary management in place to provide for fish, as imposed controls on waterways and lake levels can impede the spawning ability of fish and alter various aquatic habitat niches. Kokanee, rainbow trout and sockeye salmon are the primary species local management initiatives focus on (OBWB, 2000).

The extensive natural resources of the Okanagan Valley contribute to the region’s thriving tourism industry. Over 1 million tourists visit the region each year, bringing approximately $830 million dollars into local communities (HRDC, 2001). As well as the natural resources, recreation opportunities abound throughout the region including snow skiing at local Big White, Apex and Silver Star ski resorts and golf at more than 50 golf courses in the valley. The high season for Okanagan tourism is during the summer when 53% of the years visitors embark on local lakes, parks and other hot spots, while 20% visit during the wine season of autumn, 10% during the winter ski season and 18% arrive in the spring (HRDC, 2001).

Agriculture and farming have been part of the Okanagan since the 1850s. Although many crops including Tobacco and Hops have come and gone in the valley, by the mid 1980’s the Okanagan

______9 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

and Similkameen Valleys had become the most prominent location for 90% of orchards and 95% vineyards in the province (Pidwirny and Gow, 2001). Today, Okanagan agricultural land use consists of 23% cropland, 8% fruit orchards, 1% grapes, 54% improved pasture, and 8% unimproved pasture. As of 1996, there were 2278 farms, with 78,283 ha of farmed land, and 16,151 ha under crops. Between 1976 and 1996, the total area farmed has decreased from 80,428 ha to 78,283 ha, likely due to urbanization (Statistics Canada 1996). The Okanagan has various climatic and topographic features that contribute to its agricultural success. Some of these features include mild winters with high snowfall, which together are able to protect dormant trees and vines. Springtime in the Okanagan is mild; supporting delayed budding, while the long hot summers promote local fruit development (Pidwirny & Gow, 2001). The arid Okanagan summers, however beneficial for fruit development, provide insufficient moisture and irrigation is steadily used to support the growing crops. The valley contains diverse set of microclimates, as shown in Table 2-1, each location supporting a different array of crops. In Penticton and Oliver, warmer climates allow for the growth of soft fruits like cherries, apricots, and peaches, while North of Penticton to Vernon, apples are more common and further north of Vernon, very hardy crops and dairy farming are the norm (Pidwirny and Gow, 2001).

Due to the need for irrigation and irregular terrain, many industrial methods are utilized in the Okanagan in order to achieve high agricultural yields. The local agriculture industry depends on the use of pesticides and fertilizers, which in combination have caused questions about the health risks for residents, local wildlife and various aquatic habitats. These chemicals are potential hazards for lake eutrophication, imposed anoxic conditions (Osoyoos Lake), and algae blooms - all of which could contribute to significant aquatic perturbations. The rapid population growth in the Okanagan (Figure 2-2) has led to significant land use changes, specifically loss of farmland. As local farmland decreases, a larger problem becomes apparent as the Okanagan Valley is relied upon as one of only two significant areas in Canada that can sustain tender fruit and grape production, the other being the Niagara Peninsula. Any changes to the ability of these areas to support the national demand for fruit and other local crops could have serious economic implications.

300000

250000

Central Okanagan

200000 North Okanagan

Okanagan-Similkameen

150000 Population

100000

50000

0 1976 + 1978 1980 1982 1984 1986 + 1988 1990 1992 1994 1996 ‡ 1997e 1999e Year

Figure 1-2. Okanagan Population 1976-1999

______10 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

Forestry and Mining Industries have both played large roles in the formation of Okanagan communities, from the early gold rush days to the present. Mining in the Thompson-Okanagan (Okanagan + Kamloops area) still maintains 19% of the provinces total although many sites have been closed in the valley within the last 10 years and its role appears to be diminishing. Forestry in the Okanagan is allocated based on a fiber supply called the Annual Allowable Cut (AAC). The AAC is the amount of timber allowed by the Ministry of Forests to be cut from a Timber Supply Area (TSA) in one year. This fiber source is best understood as a long term loan, as forestry companies must replant logged areas to a sufficient stocking standard, in a predetermined length of time, prior the responsibility of the area reverting back to the government. Fiber sources in the Okanagan extend beyond TSA’s, as Tree Farm Licenses (TFL’s), Non-Okanagan Crown forest land and private land supply the industry with 37% of the fiber used every year. In the Okanagan approximately 27% of the land base is considered unsuitable for harvesting (British Columbia Forest Service, 2000), and approximately 4,106,000 cubic meters of wood is milled per year (Ministry of Forests, 2001). It is difficult to sum up the total area of the Okanagan watershed impacted by forestry activities over the many years of timber extraction, as most government and data collection boundaries do not correlate with watershed boundaries. A general understanding of the area of the region impacted by forestry activities can be gained through the use of satellite photography. It is also important to be aware of the many factors that are considered and managed for in Okanagan forests. Other than timber harvesting, considerations of practicing foresters have been analyzed in a recent review of the Okanagan TSA which included requirements for biodiversity, visual quality, ungulate winter range, and community watersheds (British Columbia Forest Service, 2001).

2.2. Case Study: Trout Creek Trout Creek flows into Okanagan Lake just south of Summerland and is the second largest watershed in the Okanagan. The Trout Creek watershed provides a variety of resources for the surrounding built and natural environments. The watershed is a source of recreation, irrigation and domestic water supply for local communities. The area provides habitat for various land and water-based flora and fauna that are niche specific to the unique climate and topography found in the south Okanagan. The catchment area of Trout Creek is approximately 759 km2. The mainstream itself is 72 km long and falls from an elevation of 2017 m to 340 m at Okanagan Lake. Approximately 255 km of the Trout Creek system is described as providing ‘fish bearing reaches’ for species including rainbow trout, mountain whitefish, brook trout and potentially spawning kokanee (MOELP 2000). Over time, Trout Creek has created a large delta of alluvial material and these impacts continue several hundred meters into Okanagan Lake (MOELP 2000).

The Trout Creek watershed averages 550 mm precipitation with 60% falling as snow. The average annual discharge is 2.1 m3/s. There are 80 to 85 water licenses issued for irrigation, domestic supply and storage. The licenses allow 18.6 M m3/y for diversion and approximately 13 M m3/y of water to be stored. The Summerland Water District has 15 of the licenses, which represents 97% of the flow volume for diversion (Dobson Engineering, 1998).

The lower reaches of Trout Creek exemplify a trend in the Okanagan of increasing development and intensification of agriculture. The developed land base in this area is steadily increasing and putting pressure on both the natural areas and the sites historically used for agriculture. Due to this pressure and a changing industry, many farms, orchards and vineyards have intensified their production through the use of variations like dwarf species. This change, combined with increasing demands remaining from other developments, has resulted in higher water per acre usage of the local resources.

Forestry and recreation have been greater forces in the upper reaches of the watershed with visual estimates of impacted areas reaching 30% in some areas. In 1998, the Equivalent Cut

______11 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

Area (ECA) was 17% of the watershed with some sub-watersheds up to 32% (e.g. Camp Creek). An environmental assessment of the logging activities in the watershed indicates that there is a moderate peak flow hazard due to logging activities in Isinok, Lost Chain, North Trout, and Upper Trout Creek. There are also high surface erosion hazards for 5 of the 11 major sub-basins in the watershed (Dobson Engineering, 1998). Logging from 1998 to 2003 is expected to result in from a 0% to 5% increase in the areas logged in the sub-watersheds. In light of this data, in combination with the potential incremental effects of climate change, there will likely be effects on water supply and quality for the main water users like the Summerland water district (see Section 6).

Trout Creek is a bellwether watershed that is able to tell many of the land use stories of the region, such as the impacts from development pressures, agriculture and irrigation demands, forestry impacts and growing recreation activity. When these trends are combined with the global stress of climate change, what would the implications be? The many scenarios obtained from climate models, combined with other analysis tools and local knowledge, produce information that can help visualize and personalize what this global trend could mean locally in the Okanagan Basin (see Section 3). In Trout Creek, local industries and resident lifestyles are currently putting substantial pressure on the natural environment. This environmental strain greatly complicates the regions ability to respond to a changing environment and climate.

3.0 STUDY FRAMEWORK

Scenarios and dialogue are the products of this study, so the framework has to enable both of these to develop. Several steps are needed:

a) understand the region’s use of and concerns for water resources, b) generate regional scenarios of climate change from global-scale climate models, c) develop regional scenarios of hydrologic changes based on the climate change scenarios, and d) create dialogue through existing and new processes to obtain stakeholder views on climate change impacts, potential adaptation options, and implications of choosing any of these options.

Figure 3-1 illustrates how these four study components are linked. The relatively short duration of available time for this study (one year) precluded consideration of a major effort at modeling all natural and human processes relevant to climate and water. Rather than developing an all- inclusive model of water resources, in which the connections to climate and management decisions are mathematically expressed, the approach here is to use dialogue to complement mathematical models. This enables the inclusion of issues that may be difficult to model in such terms. Mathematical models are used strategically to generate information on known environmental indicators and processes (temperature, precipitation, snow pack, runoff, streamflow). Dialogue is used for those indicators and processes that include a human component (irrigation, land use, forestry, fisheries, institutional arrangements).

______12 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

Climate change scenarios global regional

Hydrologic change scenarios Dialogue •snow pack •regional impacts •streamflow •adaptation options •annual cycle •implications of adaptation choices

Regional development context •population •water & land use •institutions •local case study – Trout Creek ??

Figure 3-1. Okanagan Climate and Water Resource Impacts Study Framework.

This is not to suggest that the development of climatic and hydrologic scenarios have led to predictions with any known levels of probability or accuracy. There are a number of important uncertainties associated with these modeling efforts (see Sections 4 & 5). The models are based on established physical principles, but additional steps have been necessary to apply these tools to this location and scale. Throughout this report, we refer to model-based results as scenarios, rather than as forecasts with a known probability or accuracy. The full range of possible outcomes is not known.

An important technical challenge is downscaling of global scale climatic simulations to regional scale scenarios. Although sophisticated statistical and dynamical downscaling techniques are being explored, they have not yet been sufficiently developed for this area. In this study, we employ a simple approach using General Circulation Model (GCM) data at its original resolution (see Section 4). The use of single GCM grid cells and the interpolation of surrounding grid cells are simple techniques to apply, but carry with them considerable uncertainty. We justify their application here by noting that the magnitude and direction of projected changes in climate are consistent with other scenarios developed for the Columbia basin of which the Okanagan is a part (Hamlet and Lettenmaier, 1999).

The question marks in the framework represent uncertainty in the post-study impacts of the dialogue initiated by this work. It has been the intention of this component to build on existing regional consultation processes, such as the watershed round tables, stewardship councils, and the Okanagan Basin Water Board to bring regional stakeholders into the study. Two sets of focus group exercises have been organized to initiate dialogue on impacts and adaptation, and to learn about possible connections between climate change and regional development (see Section 6). Once the final report of this study is completed and distributed within the region, it is up to regional actors to decide how to proceed. However, this also presents an opportunity for follow- up research (see Section 7).

______13 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

4.0 CLIMATE CHANGE IN THE OKANAGAN BASIN

4.1 The Climate of the Okanagan Basin

The Okanagan Basin is located in south central British Columbia and is part of the Southern Interior Ecoprovince, lying east of the crest of the Coast and Cascade mountain ranges and west of the Columbia Mountains. Due to the rain shadow effect of these mountain ranges, the region contains some of the warmest and driest areas of the province in the summer. The Okanagan Basin is nested within the Thompson Okanagan Ecoregion, characterized by rolling plateaus and large north south valley systems with rich glacial deposits carved by river paths. The mean annual temperatures of the major valleys is approximately 6°C with a summer mean of 15°C and a winter mean of -3.5°C. The mean annual precipitation ranges between 250-300 mm in the major valleys to 400-600 mm in plateau regions, to over 1,000 mm in sub alpine and alpine areas. Sub alpine areas support lodgepole pine, engelmann spruce and sub alpine fir. At lower elevations, mixed forests of lodgeople pine, trembling aspen, white spruce and douglas fir cover the rolling landscape. Drier valley bottoms support drought-adapted species such as douglas fir and pine grass, or scattered ponderosa pine among bluebunch wheat grass and sagebrush. Varied grasslands, sagebrush, and antelope bush characterize the most arid areas south of Penticton. Many animal species inhabit this unique habitat including bighorn sheep, mule and white-tailed deer, elk, black bear, coyote, bobcat and cougar. Bird species include blue grouse, waterfowl and long-billed curlew. Reptiles, such as rattlesnakes, are found the more arid areas of the region (ESWG, 1996).

4.2 Global Climate Change

Over the past century, the earth’s mean global temperature has increased by 0.6 ±0.2°C (IPCC, 2001). Since the mid 1970s warming has accelerated, suggesting that this increase is not just a cyclic trend but also a response to an increase in atmospheric concentrations of greenhouse gases, such as carbon dioxide (CO2), methane (CH4) nitrous oxide (N2O) and tropospheric ozone (O3). Since pre-industrial times, CO2, the most important greenhouse gas, has increased by 31%, CH4 by 151%, N2O by 17% and O3 by 36% (IPCC, 2001). Collectively, this increase in greenhouse gases has altered the radiative balance of the earth to the point that it is now having a discernible influence on the global climate (IPCC, 1996). As concentrations of greenhouse gases continue to rise, the greatest warming is predicted for higher latitudes and in the interior of continents. Higher latitudes are expected to experience more warming due to the melting of snow and ice which will result in greater absorption of solar energy and enhanced heat transfer from oceans to land. Similarly, the interiors of continents are especially sensitive due to their vulnerability to temperature extremes. In addition, warming is expected to be higher for the northern hemisphere than the southern hemisphere due to more land mass in the former and enhanced deep water ocean circulation in the latter. Consequently, Canada may experience a greater temperature change over the next several decades than most regions of the world. Already a change in climate has been observed in many parts of British Columbia, including the Okanagan Basin.

4.3 Climate Variability

Climate variability occurs due to both internal and external factors influencing the earth’s climate system. Year to year variability is common, and temperature anomalies in British Columbia are especially pronounced during El Niño episodes, which have a 2-7 year return period. On a larger scale, an important phenomenon influencing decadal climate variability in British Columbia is the Pacific Decadal Oscillation (PDO). The cool phase of the PDO is associated with higher

______14 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

than normal atmospheric sea level pressures in the North Pacific between the months of November and March, resulting in warmer than normal sea surface temperatures in the central North Pacific and cooler than normal sea surface temperatures off the west coast of North America. The warm phase of the PDO is the opposite and is characterized by warmer than normal sea surface temperatures off the west coast of North America (Nicholls et al., 1996). The PDO has a 20-30 year cycle whose cool and warm phases have had a discernible influence on the climate of western North America over the past century (Mantua et al., 1997). The 20th century began with cooler annual temperatures from about 1900 to 1925 (cool PDO phase), which were followed by warmer temperatures between 1925-1945 (warm PDO phase). Cooler temperatures again predominated between 1945-1975 (cool PDO phase), followed by a resumption of warmer temperatures in the last quarter of the 20th century. In spite of the decadal scale temperature variability associated with the PDO, temperatures have continued to rise and the rate of increase has accelerated, especially during the last quarter of the 20th century. There is now almost unanimous consensus among scientists that the continued rise in atmospheric concentrations of greenhouse gases are influencing global warming on a scale unprecedented in human history.

4.4 Climate Change in the Okanagan Basin: The Past 100 Years

Climate trends can be determined by analyzing data collected at climate stations over the past century. Minimum daily temperatures are a good indicator of nighttime lows. Figure 4-1 shows the annual minimum temperatures measured at the Vernon Coldstream Ranch climate station from 1901 to 1999. Although decadal oscillations have occurred over the period of record, minimum annual temperatures at Vernon Coldstream Ranch have increased by approximately two degrees over the past century. Daily maximum temperatures have also increased at this station (Figure 4-2), although not as much as minimum temperatures. The greater increase in minimum temperatures is likely due to increased nighttime cloud cover, which minimizes heat loss to the atmosphere. Observed increases in both minimum and maximum temperatures at the Vernon Coldstream Ranch are paralleled by warming trends observed at nearby Kamloops, Summerland and Cranbrook (Brewer and Taylor, draft, 2001), indicating that this is not an isolated occurrence but a general warming trend in south central and eastern British Columbia.

Figure 4-1. Trend in Minimum Temperature at Vernon Coldstream Ranch

______15 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

Decadal oscillations in temperature are an important modifier of climate, as indicated by the lower minimum temperatures between approximately 1945 to 1975 (Figure 4-1) which are linked to low PDO indices associated with the most recent cool phase of the Pacific Decadal Oscillation (Mantua et al., 1997). Conversely, warm cycles, which were most pronounced during the 1940s and in the last two decades of the 20th century, correspond to positive PDO indices associated with the warm phase of the Pacific Decadal Oscillation.

Figure 4-2. Trend in Maximum Temperature at Vernon Coldstream Ranch

The number of frost-free days per year is a good indicator of the growing season in a given region. Figure 4-3 presents the number of frost-free days per year at Summerland between 1907-1993. An increasing trend is evident in the time series with the number of frost-free days increasing by 3.1 days per decade over the period of record, equivalent to 27 fewer days with frost by 1993. This trend is highly statistically significant, reflecting the parallel increase in minimum temperatures at this site. Similar increasing trends of comparable magnitude have been observed at other stations throughout B.C (Brewer and Taylor, draft, 2001).

An increasing trend in precipitation has also been observed in the Okanagan Basin. Figure 4-4 shows annual precipitation values at Kelowna Airport over the 20th century. Decadal scale variability is a prominent feature of the precipitation pattern at Kelowna Airport. The century began with a brief wetter period for the first decade followed by a drier period through the 1920s and 1930s. A wetter period followed in the 1940s followed by a drier period between 1950 and the late 1970s. Wetter conditions returned in the 1980s, coinciding with the resumption of warmer temperatures throughout British Columbia. In spite of the observed decadal variability, precipitation has increased by approximately 125 milliliters over the past one hundred years. A significant correlation was found between spring maximum temperatures and precipitation. This suggests that increased rates of evaporation, resulting from warmer temperatures, have likely increased the amount of rainfall in the area during the spring months. A significant increase in the amount of precipitation of a similar magnitude has also been found at other southern BC locations such as Summerland and Cranbrook (Brewer and Taylor, draft, 2001), reflecting a larger global trend towards an increase in precipitation (IPCC, 2001), believed to be a natural consequence of an intensification of the hydrological cycle in response to rising temperatures.

______16 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

Figure 4-3. Trend in Frost Free Days at Summerland

Figure 4-4. Trend in Total Precipitation at Kelowna Airport

4.5 Establishing a Baseline: Climate Normals

Before we look at climate change in the Okanagan Basin, it is important to establish a baseline against which to measure this change. Climate normals represent the climate at a given station

______17 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

based on average measurements over the most recent 30 year period on record. The most recent climate normals were computed for the period spanning from 1961-1990. Figure 4-5 shows the most recent temperature and precipitation normals measured at Kelowna Airport. As expected, peak minimum and maximum temperatures occurred during July and August while the lowest temperatures occurred in January. Note that average maximum summer temperatures were around 27°C, while minimum winter temperatures averaged around –8°C. Precipitation was most abundant during the summer, from May to October, and during the winter in December and January. In this area, summer precipitation occurs largely due to local convective activity, while winter precipitation is usually linked to weakened synoptic systems (winter storms) originating in the Pacific Ocean. Kelowna Airport received about 366 millimeters of precipitation per year, reflective of the relatively dry climate of this region. This is a relatively low amount compared to central coastal areas of BC receiving over 3,000 mm of precipitation per year.

Max

Min

Figure 4-5. Temperature and Precipitation Normals (1961-1990) at Kelowna Airport

4.6 Predicting Climate Change

4.6.1 Global Climate Models

Global climate models are used to make predictions about the future state of the world’s climate. Coupled general circulation models (GCMs) simulate the climate system using mathematical equations that describe the earth’s radiation budget, its translation into heat and motion and the operation of the water cycle. An example of such a model is the Canadian Global Coupled Model (CGCM1), developed by the Canadian Centre for Climate Modeling and Analysis in Victoria, British Columbia. CGCM1 is composed of four key components: an atmospheric general circulation model, an ocean general circulation model, a simple land surface model and a thermodynamic sea ice model. Within each of these four models, physical processes are simulated according to the best scientific knowledge to date. The atmospheric general circulation model is composed of 10 vertical levels and has a horizontal resolution of approximately 3.7° latitude and longitude (about 400 km). The model contains interactive

______18 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

simulations of atmospheric pressure and circulation, wind stress, cloud processes and precipitation. The ocean general circulation model, with 29 vertical layers and a horizontal resolution of 1.8° x 1.8° (about 200 km), is able to reproduce large-scale features of the ocean circulation as well as variations in temperature and salinity. Physical attributes of continents such as topography and soil properties are included in the simple land surface model, which calculates runoff and soil moisture based on the balance between precipitation, surface evaporation and water holding capacity of the soil. Lastly, the thermodynamic sea ice model allows ice to grow and melt in response to seasonal change. These individual models are coupled together to most accurately represent the interaction between the atmosphere, hydrosphere, geosphere and biosphere.

Using a combination of historical climate data, physical data and expected increases in aerosols and greenhouse gases, CGCM1 is able to generate climate scenarios for three time slices: 2010-2039 (2020s), 2040-2069 (2050s) and 2070-2099 (2080s). These scenarios are generated at spatial resolution of 400 km and a temporal resolution of one month.

Coupled general circulation models similar to the Canadian counterpart have also been developed by the UK, Germany, USA, Australia and Japan. On a large scale, simulations generated with these different coupled models are in relatively good agreement. For example, st all models predict a global temperature rise in response to a doubling of CO2 over the 21 century. They do differ, however, in their sensitivity to this change, as reflected in their prediction of the exact temperature rise. Similarly, the models all predict warming to be greater at high latitudes and in the interior of continents. Where they differ, however, is in their regional projections of the magnitude of warming at the regional scale and in the magnitude and direction of precipitation changes. Discrepancies between models occur largely as a result of the different assumptions built into the various models. Such assumptions include modeled processes of heat and energy transfer between the atmosphere, oceans and land, modeling of cloud parameters and land-atmosphere interactions (e.g. evaporation and evapotranspiration rates).

Another source of variability between models arises from differences in the starting point from which the model is run. Because of the natural variability of climate, the initial starting point will have an effect on the modeled results. To minimize errors relating to internal climate variability, an ensemble of runs is often run from several starting points. The range of results from the different ensemble runs is an indicator of the uncertainty due to climate variability. The results are often averaged together to provide a more robust estimate of a climate scenario (Kattenberg et al., 1996).

Because of the relatively coarse resolution of global climate models, GCMs are limited in their ability to predict changes at a regional scale. This is due primarily to limited computational ability, as fully coupled GCM computer runs at a regional scale are still time prohibitive. Given these uncertainties, potential users of GCM simulations should consider these climate projections as likely scenarios, not absolute predictions.

4.6.2 Model Inter-comparisons

The amount of agreement between different GCMs is considered to be a good indicator of the degree of confidence in a particular climate projection. Comparison among models is achieved by using the same baseline data (1961-1990 normals) and expected CO2 and aerosol increase over a given period of time. The standard IS92a scenario, which employs the observed CO2 levels to 1990 followed by a 1% increase of CO2 and aerosols for each year thereafter (IPCC, 1996), is commonly used for this purpose. Figure A1-1 (Appendix 1) compares the results of Canadian, German and British model simulations of minimum temperature in the Okanagan Basin for the 2020s. The three Canadian simulations indicate different climate scenarios based on three ensemble members, which represent different initial starting conditions. All three models agree that minimum temperatures will increase across the four seasons, although they

______19 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

differ in the exact magnitude of the change. The Canadian model predicts that minimum temperatures will increase relatively evenly across the four seasons. In contrast, the German model predicts minimum temperatures to increase more during the winter and summer, while the British model predicts the largest increases to occur during the winter and spring. In spite of these seasonal differences, the models agree that the minimum temperature will rise and that this rise will be somewhere between 1.0-2.5°C. These projections are in agreement with observed increasing trends in minimum temperatures in south central British Columbia over the 20th century.

Figure A1-2 compares model projections for maximum temperatures in the Okanagan Basin for the 2020s. The Canadian, German and British models are in relatively good agreement, predicting an increase in maximum temperatures of approximately 1.0- 2.5°C. Seasonally, however, there are some differences, with the Canadian model predicting the increase to be most pronounced during the spring months as opposed to the summer months, as the German model predicts.

Climate predictions can also vary substantially from one model to another. This is especially true for precipitation scenarios, where there is a greater degree of uncertainty in the underlying processes. Figure A1-3 compares model projections of seasonal change in total precipitation for the Okanagan Basin for the 2020s. Results are expressed as percent change from the most recent climate normals (1961-1990). The Canadian, German and British models are in relatively good agreement in their projections of winter and spring precipitation, however they differ substantially in projections for the summer and fall seasons. This difference is especially evident for the summer where the German model predicts precipitation to decrease by as much as 18%, which is substantially different than the 1% decrease predicted by the Canadian model or the 3% percent increase predicted by the British model. In terms of projections over the entire year, both the Canadian and British models are predicting a general increase in annual precipitation.

4.6.3 Climate Change Scenarios for the Okanagan Basin – PRISM Maps

Climate maps are a convenient way of illustrating how climatic conditions vary across the landscape. The topography of British Columbia is very rugged, so the variation in the climate can be large over a relatively small area. Generally, conditions become colder and wetter with increases in elevation. Thus, valleys in the Okanagan are generally much warmer and drier than conditions at higher elevations. These features are reflected in the climate maps, which were prepared using the 1961 to 1990 climate normals and based on the PRISM model. The PRISM model is a statistical approach to climate mapping that is premised on the strong variation in climate with elevation. PRISM uses linear regression of temperature and precipitation on elevation in conjunction with a digital elevation model to extrapolate vertically into areas where station data are not available. PRISM fills a significant gap in the climate monitoring network by providing interpolated surfaces of temperature and precipitation that incorporate the influence of topography and coastal proximity on climate. In the absence of a dense climate network, it is not possible to verify the accuracy of these maps, but they are widely accepted as a good estimation of the current climate.

The Spatial Climate Analysis Unit at Oregon State University prepared the PRISM maps. They were constructed at a resolution of 4 kilometers, resulting in highly detailed climate maps. To simulate the climate of the future, baseline climate data were perturbed by simulations derived from the much coarser (400 kilometer resolution) Canadian global climate model (CGCM1) to produce detailed climate scenarios for three time slices: the 2020s, 2050s and 2080s. Naturally, we cannot know the future climate to the degree of precision implied by these PRISM maps. It is important to keep in mind that these maps represent physically plausible scenarios of future climates, rather than forecasts or predictions.

______20 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

Figure A1-4 shows PRISM maps of model projections for January mean minimum temperatures for the Okanagan Basin. According to model simulations, January minimum temperatures will be between 3-4°C warmer in the 2080s compared to the period between 1961-1990. This change is expected to occur gradually with an approximate one-degree change for each time slice. Higher mean minimum temperatures combined with wetter winters would result in a smaller snow pack and earlier spring runoff (see Section 5). Potential impacts of wetter winters are discussed in Section 6.

Figure A1-5 presents July mean maximum temperatures for the Okanagan Basin for the 2020s, 2050s and 2080s. The simulated increase is approximately 1.5°C per time slice, as shown in the PRISM maps. This warming will result in a temperature difference of between 3-5°C higher for the 2080s compared to the most recent temperature normals (1961-1990). Potential impacts of higher temperatures (e.g. fisheries, agriculture) are discussed in Section 6.

CGCM1 projections predict an increase in precipitation of approximately 10-20% by the 2080s for the 400 km square grid area encompassing the Okanagan Basin (CCCma, 2001). PRISM map projections reflect a corresponding increase in precipitation (Figure A1-6); however, as these increases are gradual and relatively small, they are difficult to see at the resolution provided by the PRISM maps. Nevertheless they indicate that the region is expected to experience a wetting trend over the 21st century. This is in agreement with historical records, which show a 30% increase in annual precipitation over the past century in this area. In general, precipitation is expected to increase during the fall and winter and decrease slightly during the summer. A drier summer, coupled with warmer temperatures would result in an extended low flow period in the late summer and early fall and a larger requirement for irrigation and water storage. Conversely, a wetter and warmer winter, spring and fall would result in higher annual stream flows, an earlier spring freshet, greater chance of flooding and higher risk of landslides.

4.7 Effects of Climate Change on Degree Days

Degree days are commonly used as an indicator of climate in a particular region. Growing degree-days are used in agriculture as an index of crop growth. They are defined as the number of degrees that the mean daily temperature is above 5°C (the temperature at which plant growth begins to occur) summed up over the course of a month or a year. According to Canadian model simulations (Figure A1-7), the number of growing degree days in Kelowna will increase from a peak of just over 400 degree days in July (for the 1961-1990 period) to approximately 550 degree days by the 2080s. Over the entire year, this represents a 51% increase over the baseline period. In addition, the growing season will begin earlier and end later. By the 2080s crops may begin sprouting as early as February and last well into the end of October. Such changes could benefit local agriculture by increasing the yearly crop output as well as the type and variety of crops that could be grown in the region. On the negative side, a milder winter could increase pest survival resulting in a larger incidence of crop damage (see Section 6).

Heating degree-days are used to estimate the heating requirements of buildings. They are defined as the cumulative number of degrees that the mean temperature is below 18°C in a given month or year. According to CGCM1 projections, heating degree days for the month of January in Kelowna are expected to decrease from just under 700 degree days to just under 600 degree days by the 2080s (Figure A1-8). Over the entire year this represents a 28% decrease in heating degree days, which could translate into significant savings for space heating. Such savings however would likely be offset by extra costs associated with air conditioning and swimming pool use during the summer months.

Cooling degree days are used to estimate the cooling requirements of buildings. They are defined as the cumulative number of degrees that the mean temperature is greater than 18°C in a given month or year. CGCM1 projections indicate that cooling degree days in Kelowna for the month of July are expected to double by the 2050s (Figure A1-9), compared to the 1961-1990

______21 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

period, and then nearly triple by the 2080s. Over the entire year, cooling degree days are expected to increase by 200% by the 2080s. This compares with an estimated 28% reduction in heating degree days by the 2080s, indicating that any savings derived from the reduction in heating degree days may be exceeded by the additional cost of air conditioning, depending on local responses to this change (see Section 6).

4.8 Conclusion

Analysis of climate data collected over the past 100 years clearly indicates that the climate of the Okanagan Basin has changed over the 20th century. Increasing trends have been observed for both maximum and minimum temperatures, with minimum temperatures showing the most pronounced changes. Similarly, precipitation has increased in the Okanagan Basin, reflecting an intensification of the hydrological cycle in response to increased temperatures. These climatic changes have been observed throughout most of British Columbia and are reflective of larger scale global trends associated with a disturbance in the radiation budget, largely attributed to an increase in greenhouse gases. Global climate model projections for the Okanagan Basin the 21st century indicate a continuing increasing trend in both temperature and precipitation. There is relatively good agreement among different GCMs with respect to future season specific temperature changes, however, there is less certainty regarding future seasonal distribution of precipitation. Climate projections for the Okanagan Basin indicate an increase in growing degree days with possible benefits to regional agriculture. Similarly, projections call for a decrease in heating degree days with associated savings in space heating. However this is likely to be more than offset by a much higher increase in cooling degree days, reflecting air conditioning requirements.

Overall, in spite of the existing uncertainties related to applying coarse GCM projections on a regional scale, the relatively good agreement between observed temperature and precipitation trends with projected trends indicates that a fair amount of confidence can be placed in these projections in spite of uncertainty in the future seasonal distribution of precipitation. Further advances in regional climate modeling and improved downscaling techniques should further refine predictions of future climate states at a regional level.

5. 0 HYDROLOGY

5.1 Introduction

Meteorological variability provides the driving force for hydrological variability in the landscape. However, the landscape has a modifying influence on the distribution of water in time and space in such a way that water quantity, as measured at any single point in the surface water network on the landscape, is not necessarily a direct result of recent meteorological events. The routing of water through the landscape transforms linear meteorological inputs into non-linear hydrological outputs.

The complex interaction between climate and hydrology can be examined using several analytical approaches; including statistical analysis of existing data, temporal and spatial climate analogue techniques, empirical modeling and hydrological process modeling. The use of any one technique in isolation can lead to misinterpretation by over-simplification of complex phenomena. Klemes (1990) argues that modeling of climate change often fails to do more than to reveal the obvious (e.g. increased precipitation will yield increased runoff and increased temperatures will result in increased evapo-transpiration). In this paper three techniques are used to assess the hydrological consequences of climate change on the Okanagan Basin: statistical, temporal analogue, and process modeling.

______22 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

5.2 Methods (This section contains a technical discussion on hydrology methods and tools. Some equations appear in the text.)

Statistical techniques are dependant on existing data. Whereas statistical analysis is of limited value for extrapolation as far beyond the observed data set as is necessary for this study, it is useful to identify whether trends identified in observed data are consistent with changes predicted by the process modeling approach. The statistical modeling approach is not constrained by the assumptions associated with downscaling from the GCM to the hydrological process model. Hence, statistical analysis can be used to identify trends at the time scale of hydrological extremes and thus provide inference about the future prospects for hydrological extremes that cannot be provided from analysis of the process model results.

In this paper, climate change analogues from the period of record are used to establish context for the process model results. The analogues can help in determining whether assumptions implicit in process modeling can be supported with existing data. Furthermore, the climate analogues assist in translating results to a practical level for stakeholder consultation. It is easier for a stakeholder to relate to the hydrology within his/her experience, complete with the knowledge of the real consequences of that year, than it is to relate to an absolute quantification of change in a hydrological variable. The spatial analogue technique (e.g. examining the hydrology of a region with a similar landscape but a climate that matches the predicted climate) is beyond the scope of this study.

5.2.1 Statistical modeling: Statistical Analysis of Changes in Climate and Hydrology of South Central BC 1976-1995

Data presented here were extracted from daily temperature, precipitation, and streamflow series from Environment Canada’s national databases. These daily data sets were smoothed to 5 day (precipitation and streamflow) and 11 day periods (temperature) for the two decades. This results in calculated average conditions during each 5 day or 11 day period for 1976-1985 and 1986- 1995 decades. Statistical significance is determined using the method of Leith and Whitfield (1998) who assessed significance of changes using robust non-parametric statistical test.

The differences between the decades are presented using polar plots to enhance the visualization of changes in magnitude/timing between the two decades. In these figures increases between decades are shown in green, and decreases in yellow. Statistically significant differences are indicated with red arrowheads (increase) or blue arrowheads (decrease). Results from statistical tests were then grouped using the automated clustering procedure described in Whitfield and Cannon (2000a). Further description for each variable is given in section 5.3.

Patterns in natural streams were characterized by an earlier onset to melting of the winter snowpack, followed by lower peak and recession streamflows during the spring, summer, and early fall months. In the western portions the study area early winter streamflows increased, while in the eastern portion winter flows decreased. The difference between these two areas is due to the increase in early winter rains in the western portion of the study area. In streams with existing withdrawals, significant decreases in flows were observed from fall until spring, with an overall decrease in streamflow between the two decades.

5.2.2 Temporal Analogue determination

As an aid to interpretation of process-model simulated results, in the context of inter-annual streamflow variability, an index is used to rank the climatic similarity of observed data with respect to the predicted climate for the 2080 scenario. This index calculated as:

______23 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

+ SIwy BIwy I = 1− (1) wy 2 where Iwy is the index for any given water year wy (October 1 to September 30), SIwy is a shape index for that water year and BIwy is a bias index for that water year. The shape index is given as:

12 2 (Tm,wy - Tm,1970 ) = (2) SI wy 2 m=1 (Tm,2080 - Tm,1970 ) where m is an integer representing each of the 12 months of the water year, Tm,wy is mean temperature for month m in the given water year, Tm,1970 is the mean monthly temperature for month m, for the climate normal period (1961 to 1990), and Tm,2080 is the mean monthly temperature for month m for the 2080 scenario (2071 to 2100). The bias index is given as:

T - T = 2080 wy BIwy (3) T2080 - T1970 where T2080 is the mean temperature for the 2080 scenario, Twy is the mean temperature for the given water year, and T1970 is the mean temperature for the climate normal period.

This index is calculated for each water year for which we have streamflow record. In this way each year of record is compared to the 2080 scenario. Hence, the 2080 scenario has an index score of 1, the climate normal period has a score near zero and any given year will have an index score that is either more similar to the predicted climate (i.e. a positive sign) or less similar (negative sign) to the predicted climate with respect to the climate normal mean.

Interpretation of this analogue must be done with some caution because of the simplicity of the index algorithm, which ignores precipitation. To aid in interpretation the indexed years are also categorized into four groups according to whether they are: warm-wet, warm-dry, cool-wet or cool-dry with respect to the period of record mean precipitation and temperature.

5.2.3 Process modeling

The HBV-EC model, used in this study, is a locally developed variant of the HBV model developed in Sweden by Sten Bergstrom (1995). This model provides several advantages including: minimal data requirements, ease of calibration, and computational efficiency. These advantages are achieved without compromise of the ability of model to represent the hypsometric complexity of the chosen watersheds.

The HBV-EC model is composed of a set of algorithms that represent dominant hydrological processes and the storage and routing of water in the landscape. As the actual physics of these processes cannot be explicitly resolved with available data, calibrated parameters are used to simplify the model algorithms. Optimum parameter values are determined by comparison of model output with observed data. A number of assumptions are implicit with this modeling approach and these assumptions confine the interpretation of results as being reasonable if, and only if, model assumptions are valid for the scenario being simulated.

Beven (2001) discusses several assumptions that are problematic for hydrologists attempting to reproduce hydrological variability within a numerical-modeling environment. Addressing the full context of hydrological assumptions is beyond the scope of this paper, however, within the context of this study, there are two specific modeling assumptions that must be addressed:

1. The assumption that the landscape used for calibration of model parameters is hydrologically similar to the landscape simulated in an altered climate, and

______24 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

2. The assumption that the predicted shifts in the mean monthly temperature and the mean monthly precipitation are uncomplicated by any change in the skewness, or kurtosis of the daily data around the mean monthly values in an altered climate.

In the first instance, one might reasonably assume that land-use change will occur over the time frame between the climate change scenarios. This change may be a result of anthropogenic forcing (e.g. agriculture, forestry, or urbanization) or due to ecosystem forcing (e.g. change in elevation of tree-line as a result of climate warming). In either case, model parameters calibrated to the native landscape will not be valid in the altered landscape. As the future condition of the landscape is not known, the model is not capable of reproducing what will happen in that altered landscape, but only what might happen if the landscape does not change. However, accepting this assumption of hydrologic similarity simplifies the interpretation of results. For example, if both land-use and climate were altered in the model domain, it would be relatively difficult to interpret some of the more subtle effects of climate change that may be masked by land-use change. As such, results based on this assumption are useful for discussion purposes about the climate change issue, but should not be used for specific planning purposes.

The second assumption, that the statistical properties of the future climate will be similar to the existing climate, with the exception of shifts in the monthly mean, cannot be explicitly validated within the context of our current climate modeling capabilities. This simplifying assumption is necessary to downscale from the resolution of the global climate models to a scale of hydrological relevance. At issue is the effect that any unknown changes in the intensity-duration-frequency distribution of precipitation may have on hydrological response characteristics for a given basin. The current generation of global climate models do not have sufficient spatial resolution to solve for convective storms that would be capable of delivering intense rain, and even if that resolution was available, convective parameterization is required that could not be justified for an unknown future landscape. That the assumption of stationarity of monthly climate frequency distributions with respect to the mean cannot be justified essentially invalidates model results at any time scale finer than the monthly means. This precludes analysis of hydrological extremes based on model scenarios, as annual extremes are typically expressed in units less than a month (e.g. maximum daily, maximum instantaneous, 7-day low-flow etc.). However, the violation of this assumption does not affect interpretation of seasonal flow distribution (i.e. time resolutions greater than or equal to monthly time scale).

The model was calibrated using a split sample method whereby the hydrometric record is split into two periods. The first period is used to optimize the model parameters and the second period is used to validate the optimized parameters. The performance measure chosen for model optimization and validation is the Nash-Sutcliffe model efficiency (NSME) calculated as:

é n n ù = 2 2 NSME 1- ê (Qp(i) - Qo(i) ) (Qo(i) - Qp ) (4) ë i=1 i=1

where Qp(i) is predicted discharge at time i, Qo(i) is observed discharge at time i and Qp is mean predicted discharge.

The calibration and validation time periods complete with the NSME scores are provided in Table A2-1. The performance scores show that the model does a good job of estimating the hydrographs for these stations and that there is very little loss in model efficiency when the model is run with verification data. This increases our confidence in the model’s ability to produce reasonable results when run with a perturbed climate. Locations were selected for hydrological modeling based on:

1. The geographic diversity of the Okanagan basin, 2. Un-regulated flow regime,

______25 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

3. At least ten years of hydrometric data, and 4. Proximity to a climate monitoring station with concurrent temperature and precipitation data. This selection process resulted in six Water Survey of Canada hydrometric stations being chosen (Table A2-1), which were associated with five different climate stations.

The global climate model scenarios provide predictions over a coarse spatial domain where each grid point represents approximately 160,000 km2 of land area. However, the hydrological model is sensitive to local climate variability within the area of the watersheds, which range from 34 to 117 km2 in size. Furthermore, in order to apply the calibrated hydrological model parameters to scenario data, the scenario data must be specific to the location of the climate station used for calibration. As there is no physical basis for downscaling GCM grid data to a singular point in the landscape (see Section 4), the monthly climate normal means of temperature and precipitation were perturbed by the magnitude of the change predicted for each scenario for each of the climate models used. The resultant monthly means were then linearly re-distributed to create a daily time-series of temperature and precipitation for each scenario.

5.3 Results

5.3.1 Statistical modeling

5.3.1.1 TEMPERATURE

In South-central BC we examined 20 stations with complete records and one station with some missing data. All of these stations demonstrated the same pattern shift. This pattern is found in cluster #2 of the nine described in Whitfield et al (in prep). This pattern shift is described as showing significant increases in temperature during spring and fall. Four examples of this pattern are shown in Figure A2-1 (Appendix 2).

5.3.1.2 PRECIPITATION

All 19 stations with complete records and 2 stations with a partial record represent cluster #3 from Whitfield et al (in prep) who found 12 clusters across Canada. All stations within the study area show the same pattern: This pattern is described as having significant increases in summer and winter precipitation, and significant decreases in spring and early fall precipitation. Four examples of this pattern are shown in Figure A2-2.

5.3.1.3 NATURAL STREAMS

There are 21 natural streamflow stations in South-central BC with data available for the two decades of interest. Whitfield and Cannon (2000a, 2000b) found that 8 stations fell into cluster #4c and 13 stations into cluster #6. In this classification system the ‘c’ indicates a type of snowmelt hydrology which is consistent across the area. These two groups of stations each occur in spatially coherent regions: The 4c stations in the western portion and the 6c in the central and eastern portion. The 4c stations demonstrate earlier onset to spring, lower fall flows, and higher early winter flows (Similkameen i.e. 08NM134- Figure A2-3). The 6c stations demonstrate an earlier onset to spring, lower fall flows and lower winter flows (Okanagan/Kettle - i.e. 08NM142, 08NM171, 08NM174 - Figure A2-3).

5.3.1.4 REGULATED STREAMS

Whitfield (2001) examined 5 Okanagan streams with modifications of the natural flow regime: (withdrawals, diversions etc.). All the streams in this group showed decreases in discharge that was consistent across all stations. These stations in particular show statistically significant

______26 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

decreases in fall, winter and spring. Figure A2-4 shows four examples of this type of pattern, which only occur in the Okanagan Valley.

5.3.1.5 SUMMARY

Between the decades 1976-1985 and 1986-1995 changes in temperature and precipitation patterns were observed in south-central BC. More statistically significant changes occurred during spring and fall than during summer or winter, with significant increases in temperature and decreases in precipitation noted during these time periods.

5.3.2 Analogue and process modeling

The results are presented as a series of posters (Figures A2-5 to A2-20) to stimulate dialogue in stakeholder workshops. The first six posters are catchment specific and summarize information relevant to each of the six study watersheds. This includes: catchment description, location, land- cover, hypsometry, annual hydrographs, annual mean flow and annual snowpack. The sixth poster summarizes the snowpack simulations for all six of the study areas. Interpretation of the hydrographs is aided by knowledge of these predicted changes in the nature of snow accumulation and melts. The next four posters summarize the seasonal effects of climate change on streamflow. There are two posters summarizing the effects on the low flow season and finally there are two posters summarizing the effects on the high flow season. High and low flow magnitudes are not presented because those statistics are very sensitive to meteorological sequencing at a scale less than the resolution of the GCM’s.

In general, for each of the catchment-specific posters, a shift in timing and magnitude of the annual hydrograph can be seen generally to an earlier, lower peak. The nature of the change is controlled by various catchment characteristics of which basin hypsometry is felt to be most significant. For example, Dave’s creek, which has the smallest catchment area and the lowest maximum elevation, has the greatest net loss of snowpack in the modeled scenarios. As a result, winter streamflow response is, in part, enhanced as a large percentage of the winter precipitation falls below the snowline in a climate change scenario and is therefore immediately available for runoff. The large magnitude of the change in winter flow is due to an earlier snowmelt peak that occurs in late winter rather than early spring.

In graphs showing the scenario predictions in the context of climate analogues, a regression line is shown if the regression is significant at the 95% confidence interval. The presence of a line (e.g. the graph showing Dave’s creek on the poster of timing of peak flow, Figure A2-7) shows that the predicted consequence is consistent with observed effects during years that have similarities with the predicted climate. The actual presence of such a trend could be confirmed by statistical analysis of the existing time-series. However, the value of a process modeling approach is for identifying those instances where the predicted consequence is out-of-range, or near the extremes in the range of observed data, but for which there is no trend with respect to the analogues (e.g. the graph for Dave’s Creek on the poster of duration of peak flow). This pattern indicates that a regime-shift may be imminent which cannot be anticipated by trend analysis. In this instance, one may infer that the basin has a tolerance for inter-annual climate fluctuations which impact seasonal snowpack storage. However, the future climate may push the snow line high enough in the catchment that the basin then becomes sensitive to these inter- annual fluctuations hence causing a dramatic shift in the mean volume of the freshet runoff.

______27 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

The total basin snowpack poster shows bar graphs of the model generated April 1st snowpack for each of the GCM scenarios with respect to the mean snowpack from the period of record. In every instance, the snowpack is predicted to decrease. The geo-spatial context of the decrease is quite complex, and is related to basin hypsometry, latitude and basin aspect. The largest decline in snowpack is predicted to be Dave’s Creek, which is also the lowest elevation station and has a south-facing aspect. Whiteman Creek has a similar maximum elevation but is further north with an east-facing aspect and isn’t predicted to experience the same magnitude of snowpack loss. The stations with the least predicted loss of snowpack are the highest elevation stations (Vaseux and Bellevue Creeks).

The seasonal posters provide interpretation of the hydrograph changes observed on the watershed posters. For example, winter season flow increases in every case, and the trend is significant with respect to the climate analogues. This response is consistent with the notion of warmer winter temperatures resulting in increased occurrence of liquid precipitation events and increased quantities of winter precipitation. The magnitude of the percentage change in winter flow is much greater than the magnitude of change for any other season. The magnitude of this increase is due, at least in part, to the advance of the spring freshet, which is illustrated on the poster showing the timing of peak flow. Enhanced winter flow has important water quality implications because the winter is typically groundwater-discharge dominated, and in an altered climate, the groundwater chemical signature will be diluted by increased surface water runoff.

The signal is less clear for mean flow during the spring season. Only two cases (Vaseux and Pearson) show a significant trend with respect to the climate analogues. In most cases the climate change scenarios predict an increase in flow during the spring, probably due to a re- apportioning of flow due to an earlier freshet. An interesting exception is Dave’s Creek, which shows a decreasing trend in the climate scenarios. The spring freshet is already early for Dave’s Creek and advancing the timing of the freshet moves the event into the winter season, which in turn helps to account for the extreme magnitude of the change in winter streamflow.

Summer season flow is generally predicted to decrease, and in many cases this prediction is consistent with climate analogues (e.g. Whiteman, Camp and Bellevue Creeks). One may infer that this decrease is due to less snow available in the system to augment summer flow and at least in two cases, (Whiteman and Bellevue Creeks, which have a substantial portion of catchment area at low elevation) there is already evidence of that effect. An alternate hypothesis for low summer flow is that enhanced evapo-transpiration caused by warmer temperatures may be responsible. This hypothesis is not supported by the climate analogues, which show that cool years from the period of record are among the lowest of summer flows (e.g. 1973).

The fall season predictions show no significant trends with respect to the climate analogues.

The date of peak flow is generally predicted to advance earlier in the season. However, it should be noted that the use of a 30-year climate scenario average masks important diagnostic information. For example, peak flow occurred during the fall season in 1986 and 1987 for Whiteman Creek and in the year 1979 at Vaseux Creek. Using an average Julian date for peak flow is clearly inappropriate for stations where the peak can be either in the spring freshet or in the fall. These results are presented to illustrate that while an earlier freshet peak can be expected in general, in some cases, (e.g. Whiteman and Vaseux) there will also be an increase in the number of years where the annual peak occurs in the fall.

The duration of high flow is an indirect measure of the volume of the annual peak. This statistic is used because it should be less sensitive to the problem of downscaling from the GCM data to the catchment scale than other peak flow statistics. In all cases, the climate change scenarios for this statistic are at the lower end of the scatter shown for the climate analogues, however only three stations (Whiteman, Camp and Pearson) show a statistically significant trend in the plots. Paradoxically, five of the stations (all except Pearson) show an increasing tendency within the

______28 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

climate change scenarios. The analogue years, which plot closest in magnitude to the scenario peak flow durations, can be cool and dry (e.g. 1979 and 1973), warm and dry (e.g. 1977 or 1987) or warm and wet (e.g. 1988). This lack of predictive capability of the analogue years reduces the diagnostic value of this statistic.

5.4 Discussion

The hydrologic sensitivity of the Okanagan landscape to climate change varies amongst catchment areas. For some streams, changes in seasonal timing and magnitude of flow will follow trends already apparent in the observed record, and the thirty year mean for each of the predicted scenarios may be within the range of inter-annual variability. In these cases, aquatic ecosystems and local economies may have the capacity to adapt to the projected change. In other instances, the results of this investigation indicate that a regime shift, outside of the realm of recent experience for local communities, may be imminent.

6.0 DIALOGUE

6.1 Ongoing Regional Consultation Process

The Okanagan Basin Water Board (OBWB) remains the primary basin wide body addressing issues that transcend district and city boundaries. The OBWB consists of three elected officials from the each of the valley’s Regional Districts and is legally constituted to tax for water management initiatives agreed upon by the three Regional Districts. The mandate of the OBWB stretches beyond water quality and quantity and currently focuses on issues of sewage and aquatic weed control. The OBWB has recently been directed by the Regional Districts to improve inter-jurisdictional consultation on such issues as non-point source pollution from on-site treatment systems, storm water discharges, water conservation and to work toward more uniform valley standards and goals.

In addition to the OBWB, Fisheries and Oceans Canada has funded a ‘Habitat Conservation and Stewardship Program Coordinator” to set up local stewardship roundtables and to work with local volunteers and stakeholders on specific streams and watersheds of local interest. The current stewardship coordinator is in the preliminary stages of initiating a ‘Central Okanagan Stewardship Centre’ in order to develop a central support, resource and coordination centre for volunteer groups and the public at large to access government resource management information for land, water and air, stewardship initiatives (CORS, 2001). There are also other Provincial and Municipal Initiatives in the valley including; the Land and Resource Management Plan containing the “Protected Areas Strategy”, the Okanagan Valley Transportation Plan and Highway 97 Inland Corridor Initiative, Regional Growth Management Strategies as well as the South Okanagan Ecosystem Recovery plan (OBWB 2000). Each of these initiatives although not focused on water specifically, have implications for water use and aquatic habitat in the Okanagan Valley. There remains little direct collaboration between the various districts and city water managers in the basin due to limited communication lines and heavy workloads. Currently some streamlining of water suppliers and water purveyors is taking place in order to improve the general efficiency of water management and delivery in some areas but, a broader, more inclusive, basin wide management approach to water resources is not expected in the near future.

6.2 Focus Groups

The focus group is the most commonly known type of group interview. Conceived by sociologists (Merton et al., 1956) during the 1940s, it is most commonly used for marketing and to a lesser extent,

______29 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

political purposes. The focus group is primarily an exploratory tool, a self-contained means of data collection, or a method that can profitably be used in combination with other techniques. It was all three in these workshops.

The focus groups we employed were usually comprised of about 7-10 people, with a minimum of 4 and maximum of 12. The purpose was to allow for shared insight and diverse understandings of the issues. Most of the focus groups were comprised of participants who knew, or knew of one another, although there were enough strangers to avoid the development of hierarchical relationships and reluctance to discuss contrary ideas (group-think),

Focus groups are distinct from other group interactions because they produce data of specific interest to the researcher. Not intended to develop consensus, they are the best way to express the desire to learn from participants' greater experience. These focus groups provided every individual with the chance to make a contribution. They kept one person or a coalition of individuals from dominating the group and encouraged shy or inhibited group members to participate. We were able to obtain responses from the entire group so as to ensure the fullest possible coverage of the topic.

6.2.1 Focus Group: Agenda Script

The January & March workshop agendas followed a “script” of planned exercises. After a period of time to allow the participants to view posters illustrating the climatic and hydrologic trends and scenarios (Sections 4-5) and to ask questions, the workshops began.

The January gatherings had three objectives, the overarching goal of which was to begin a regional dialogue on climate change implications for water operations.

1. To determine the impacts of climate change on water resources 2. To determine if the impacts scenario makes a difference to stakeholders’ operation and use of water, and 3. To identify potential adaptations to climate change.

Each of the meetings began with the presentation of the models and their first order impacts. Following this step, we asked participants why this information mattered, the corollary of which asks participants why they were in attendance. Pairs of participants created a list of reasons of why they are concerned about what the hydrological model represents for their operation or institution.

The purposes of this exercise were:

1. To explain what the climate and hydrology models suggest before discussing regional impacts. 2. To give participants an opportunity to explain “where they are coming from” to others in attendance.

The next step focused on identifying past, present, and future (scenario) impacts. In pairs, participants brainstormed the possible impacts (present and future) of climate change for their operation in this region of the Okanagan. Again, people worked in pairs for 15 minutes, and this was followed by a round table of 30 minutes. The purposes were:

1. To link model-based scenarios to local observations and perspectives. 2. To list potential impacts of scenario changes in climate and hydrology (Box 6-1). 3. To consider impacts that are already happening in another location and whether these could be replicated here.

In the first of the three January workshops, we had participants cluster the impacts into scenarios, a kind of “where-we-are-heading” detailing of four future scenarios. The clustering of these impacts was designed to act as a precursor to a discussion of future adaptation scenarios. We placed the

______30 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

impacts on a continuum (e.g., cost- benefit and immediate – delayed). We then asked the participants to pick two by vote, with the 2 most diverse being the criteria. After we chose quadrants we then fit impacts into the 4 quadrants. The impacts were “plotted out” into the relevant quadrant. This means that impacts might be closer or farther from the middle where axes intersect.

BOX 6-1. Examples of Climate Change Impacts Suggested by January focus groups

Climate Change Impacts (on water management)

•= Extended growing season •= Increased demand for irrigation •= Management of seasonal shortages/scarcity •= Reduced water quality •= Degradation of fish habitat •= Increased water-use conflicts.

This work proved difficult and the decision was made to advance the March workshop objective of getting concrete expressions of impacts (e.g., specific amounts of monetary costs associated with low water levels etc). The result was that the third breakout session featured the illumination of the impacts using three questions:

1. Are these impacts public, private, or both? 2. Are many people going to be affected or just a few? 3. Who are the players who could respond to this impact?

The March 2001 workshop had three objectives that were directly related to the January ones. These were:

1. To generate adaptation priorities based on the impacts and implications list that was generated in January. 2. To determine the implications of prioritized climate change adaptation options. 3. To gather input on future research and dialogue.

Participants were organized into topical breakout groups (Agriculture, Forestry and Ecosystems, Health and Recreation, Infrastructure/Land Use/Tourism). In the first breakout session, each group was to brainstorm adaptations to each of the climate change impacts in the water regime. We then prioritized the adaptations by having each group brainstorm adaptation options and then shared these with the whole group. The breakout groups then ranked items on the list by choosing the four they thought were most important. Box 6-2 provides some examples (see Section 6.3 for more discussion).

______31 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

BOX 6-2. Examples of adaptation strategies (for water management)

•= Increase the supply of water •= Increase storage facilities •= Seek new sources •= Use reclaimed water for non-potable use •= Reduce the demand for water •= Water conservation programs •= Reduce irrigated acreage •= Adopt agricultural practices that reduce soil moisture loss

A second breakout session examined the implications of adaptations for water operations. Participants were re-assigned into 4 new mixed non-sectoral groups where each member brainstormed with a partner on the possible impacts (present and future) of adaptation choices for their operation or area of interest. These are the intended and unintended consequences of pursuing alternatives such as increasing supply and decreasing demand (e.g., discussion of costs/benefits/risks associated with these strategies such as increasing storage facilities). An example is shown in Box 6-3.

BOX 6-3. Example of Implications of Adaptations for Agriculture Adaptation Costs Benefits Risks Increase water •=monetary investment •=ensures adequate water •=money invested in building new storage •=land required for sitting of supply for crop irrigation reservoirs, but extra water not capacity on reservoirs •=insures against drought immediately required (over- farms •=land use concerns associated (peace of mind) preparation) with location of reservoir •= side benefits: can use water •=environmental concerns for non-agricultural use, associated with building if needed reservoir

We then brainstormed on the dialogue that needs to take place and how future research can facilitate this dialogue. Essentially a ‘next steps’ conversation, we saw the need for three kinds of feedback. The first asked participants how they thought we should continue the climate change dialogue? We also sought insight on the process for dialogue on impacts/adaptation; that is, which groups need to talk further or who should take charge to implement an adaptation. The third question involved the identification of research needs or knowledge gaps. This step replaced an evaluation of the workshop as it constituted an assessment of the whole process.

6.3 Stakeholder Reaction to Scenarios

6.3.1 Data analysis

Dialogue with workshop participants was recorded on flipcharts and in the notes of the participants at the meetings. Following the meetings, the notes taken by each pair of facilitators from each of the workshop groups were merged, resulting in the production of three aggregate documents: impacts, adaptations, implications. Figure A3-1 summarizes the discussion from the workshops: the first box shows the modeled impacts, the second box shows the impact areas that were ranked in the first set of workshops, and the third box shows the ranked adaptation options that were the product of the full-day workshop in March. These three documents were used as the basis of the data analysis. The feedback from each group in these documents was clustered

______32 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

by the sector they described and a coding process was used to do this more systematically. It should be noted that a number of comments from the meetings are not reflected in this analysis since they were not consistent with the approach of the project. The first was a resistance to the notion that we should intervene on any level in response to climate change impacts. The second assertion that reappeared a number of times during the workshops was that the effects of climate change would be swamped by the impacts of future population increases in the region, and associated demands that would be created for water resources and land.

The notes were analysed using Atlas/ti, a qualitative data analysis tool (as was previously done in Cohen et al., 2000). The data analysis process involved the coding of the text in the documents in order to reduce the amount of data and to enable the researchers to make some generalizations from the workshops. The process of coding involves labeling sections of the text with discrete codes with comments attached to each one. The comments elaborate on the meaning of the code and a single code may be allocated to many sections of the text if this represents a recurrent theme. The meaning of the code is developed in the comment as it is assigned to a greater number of segments of the text. Once an internally consistent list of codes has been developed, the codes themselves can also be clustered in order to identify further levels of generalization. The remainder of this section focuses on the interpretation of the data through this process.

Figures A3-2 to A3-7 are flow charts that illustrate the results of the discussions on impacts and implications in the first set of workshops. Six sectors are profiled. The first set of boxes at the left of each figure show the modeled impacts. The second set of boxes show the impacts specific to each sector (health and recreation, agriculture, forestry, infrastructure, ecosystems, fisheries). The third set of boxes are further implications of these impacts that came out during the workshop discussions. For instance in Figure A3-7, Fisheries Impacts and Implications, the figure illustrates how higher temperature will lead to eutrophication and lower oxygen, the implication of which is higher fish kills.

6.3.2 Impacts

The groups at the March workshop were presented with the climate change impacts identified in the January workshops. The notes from these meetings have been edited so that they only include only the primary impacts of the climate change scenarios. Primary impacts refers to the direct impacts of climate change on biogeophysical systems as distinct from impacts that result from adaptation, i.e. increased demand on groundwater. The impacts are summarised in Table 6-1.

______33 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

Table 6-1. Okanagan Climate Change Impacts

Sector Impacts Health and •= Increased risk of water-borne diseases Recreation •= Decreased water quality •= Short winter recreation season •= Decline in summer water quality •= Change in sport fishery •= Summer instream quantity decline •= Drought stressed forests •= Longer growing season Forestry •= Species viability change •= Longer growing season •= Soil stability problems (erosion) •= Change in harvesting seasons •= Changes in disease and pest regimes •= Change in forest fire frequency and intensity Agriculture •= Higher crop yields and variety (higher CO2) •= Negative impact on dry land agriculture •= Longer growing season •= Warmer winter (less winter kill of pests) •= Changes in pests, loss of beneficial pests and new invasive species •= Changes in disease affecting crops and livestock •= Changes in weeds and invasive species •= Where crops grow best will change (species shift) •= Less water available (more evapotranspiration, lower flows in fall) •= Climate extremes limiting agriculture potential Fisheries •= Temperature change •= Aquatic weed growth (eg millfoil) •= Fish species shift: Good for carp, bass, Bad for salmonids •= Fish abundance [and distribution] changes •= Change in aquatic food chain •= Longer growing season •= Eutrophication and lower dissolved oxygen •= Flow change •= Longer low flow periods •= Impacts on fish lifecycle: Delayed spawning, early migration downstream •= Changes in waterfowl lifecycle •= Increased sediment [and turbidity] •= Lower oxygen due to lower turbulence Infrastructure •= Increased fire potential and Urban •= Increased risk to homes Land Use •= Increased water demand •= Domestic •= Extended irrigation season •= New uses (i.e. dust control) •= Increase in water-borne diseases •= Water supply decreased •= Change in hydrology- higher, earlier peaks •= Water quality, erosion increases Recreation and •= Species shift/ ecosystem change Tourism •= Winter recreation impacts •= Shorter season

______34 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

•= Ski hill investment •= Decreased lake levels •= Impacts on sport fish •= Shift in fish habitat •= Decreased soil stability •= Increased nutrients/algae •= Summer in-stream quantity and quality decline Ecosystems •= Shift in biogeoclimatic zones •= Differential impacts by elevation •= Increased pressure on endangered species-birds •= Vegetation and forestry changes included riparian •= Changes in insects species/abundance •= Change in habitats (increase grasslands) •= Infiltration by new insect species from south •= Lower insect mortality during winter •= Amphibian habitat loss •= White pelican migration north •= Bird food supply •= Migration of non-native plant species into the area (millfoil) •= Desertification •= Species migration through corridors

6.3.3 Climate change adaptation

In the March session, the breakout groups focused on adaptations to the range of climate change impacts that had been identified. These findings were the most significant of the day and the analysis of adaptation is in two parts. The first section below maps out the codes, which describe the different general forms of adaptation. These are clustered into structural adaptation, social adaptation and ‘other’. The numbers beside each code describe the number of times the code appears in the notes.

Structural Adaptation Of the total list of responses to impacts, structural measures were most common. The approaches suggested responding to reduce the impacts of climate change, without changing social practices. The most frequent comment was the need to intervene in natural systems to prevent changes in the availability of natural resources. In addition, there was a strong emphasis on the need for defensive expenditures to ameliorate the effects of changed conditions. The mechanisms for mitigating the effects are described in the next three codes, which recommend regulation through pricing, access and land use.

Social Adaptation In contrast to structural measures, this cluster of comments suggests that a response to climate change impacts in to change social practices. These suggestions ranged from changing energy and water supply technologies, through to the need for public education and lifestyle change

Other The final three codes are fairly self explanatory. The group suggested that alternative governance structures needed to be considered, that more research was required to support decision making and that it might be most appropriate to take a ‘hands off’ approach that allows biogeophysical to evolve in response to changing conditions.

______35 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

Table 6-2. Adaptation to Climate Change Impacts

Cluster Code – (occurrences) Comment Structural Intervene-24 Human intervention to prevent impacts, i.e. more snow Adaptation making. Implies no change in use but interventions to maintain status quo e.g. fish species. Help migratory fish move upstream. Build upland dams and regulate flow. Control erosion. Increase fish hatcheries and decrease fish take. Improve riparian areas to offset climatic impacts. Increased retention of stormwater. Plant new genetic varieties and introduce new technologies for pests. Defensive expenditure- Investment in defensive infrastructure i.e. fire 10 protection. Improve warnings and signage. Spend on medical treatment for water borne disease and increase water quality treatment. Buy out water licences. Aerate lakes to compensate for temperature changes. Increase species diversity to increase resilience. Priced regulation-5 Regulate resource use, i.e. water, through pricing and metering. Regulate access-4 Regulate access to ecosystems, increase conservation of species. Regulate land/water Control forms of agricultural land use, land use use-17 ecosystem oriented. Change land use to for change lake and creek levels. Control irrigation. Protect watersheds and ecological corridors. Prevent clearcutting. Reduce fertilzer use. Social Alternative resource Develop alternative infrastructure in response to Adaptation use-16 changes in demand i.e. alternative energy sources. Small scale water storage. Find alternate water sources, use grey water systems. Plant different crops. Change timing of irrigation and harvesting practices. Manage in terms of ecological costs. Keep options open to CC. Educate to change-4 Educate people to reduce risks i.e. building materials. Increase demand management. Educate people about ecosystems. Lifestyle change-8 Change leisure lifestyle, i.e. different winter sports, different fishing. Reduce watering of lawns and water consumption. Change the 'need for green' Land use planning-12 Densification of urban land use, control of development processes, better and appropriate land use planning. No encroachment on fire zones. Other Hands off-5 Reduce the number of dams and allow for natural regulation. Allow species to adapt to Climate change. Reduce chlorination More research-5 Need to know more about flow regimes, snowpack etc. Better understanding of habitats before intervening Improve governance-3 Better integration among agencies. Watershed level management.

______36 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

Having identified this wide range of adaptation options, the workshop groups were asked to prioritise the options. The adaptation options are listed here:

•= Use water more efficiently •= Pay for water •= Restrict development and or population •= Irrigation methods (practices, systems) •= Metering •= Inventory supply and prioritize water users •= Water licensing •= Increased enforcement •= Change in timing •= Buyouts •= Develop more water storage/other sources •= Regulate flow and erosion through dams •= Adapt silviculture practices •= Develop new seedling varieties (forestry) •= Better integration of government agencies •= Include climate change possibilities when making land use decisions •= Consider ecological costs in resource management •= Consider social costs, making adaptation decisions

6.4 Implications of Adaptation Choices

In many ways, thinking through the implications of the adaptation options that were proposed was the hardest part of the workshop. These implications are third order effects of climates, the unintended impacts of adaptation interventions in response to the bio-geophysical impacts of climate change. The notes from the meeting were edited to exclude comments that were not consistent with the definition of implications we provided to the groups. The most common error was to confuse adaptations with implications. The options and their implications were then clustered and edited to reduce duplications. The final list is included in Table 6-3.

This list includes a mixture of predicted implications and expressed needs for additional research or changes in consultative processes associated with a particular option. These needs may become barriers to implementation.

Table 6-3. Implications of Adaptation Options

Adaptation Implication Water Through metering Licensing •= It is difficult for people to give up their own use for downstream users •= economic costs of retrofitting •= political costs of implementation, especially if retroactive. •= Sensitivity to increased licence rates •= External demand for Canadian water Through restrictions on development •= politically unpalatable to concentrate population •= different attitudes by generation •= population control here may attract more migrants Through changes in irrigation methods •= fiscal costs when compensation required •= positive implications for natural forests Through greater enforcement •= keeps everyone equal

______37 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

•= need for more officers and bureaucracy, leading to greater costs •= resistance due to ignorance •= government accountability at the local level By changing timing of releases •= need education for people responsible for water management •= affects agriculture and tourism – need for balance across sectors •= implication for hydro Regulate Less vulnerable to natural fluctuations Flow With dams- not able to return to natural hydrograph through Water storage- good sites already gone (cheap sites) Dams If site is too low- water quality is impacted Need to make sure there will be enough water to fill up the dam! Affects fisheries coming downstream Loss of peak flow a change in stream cross section Dams effect erosion positively (regular, creates habitat) and negatively (change in physical structure such as wetland and habitat loss) Infrastructure costs $ (capital, maintenance) Save $ in flooding costs BC Hydro makes $ Displacement of people, animals etc. Habitat implications of reservoirs: mercury release? Metals? Uranium? Decreased nutrients downstream Better Harmonization of common goals- different levels integration Stewardship groups should give input into policy of government agencies Climate Need to consider land AND water use change and Need better science land use More accurate models are needed to increase confidence Opportunity to let people know (with information age) what is happening Resistance by vested interests to change land use access

Population Politicians need to play a role- they tend to support growth control and Creates conflict in communities over change, growth, use restricting Development Adapt Species may not survive for 100 years (ones being planted now) silviculture Practices Cost Societal costs of increasing water efficiency/management. Who pays? Is cost implications recoverable? of Equity implications of water management (low income burden) Adaptations Development restrictions will inhibit ability to respond (tax base) Socio-economic vulnerability increases Globalization intensifies this challenge Valuation is inherently divisive when cooperation is needed Resistance because of changes in wealth distribution Concerns around demands for compensation

______38 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

7.0 FUTURE WORK

This study has brought together different forms of knowledge to initiate a dialogue on water resource management and climate change in the Okanagan region. Some insights have been obtained from this exercise, but challenges remain. Can a basin-wide scenario of hydrologic impacts be constructed? Can different scenarios of adaptation be explored to see their impacts on water and land use? Can an integrated assessment for the region be produced, incorporating results of simultaneous assessments of land and water use, using common scenarios?

During the March workshop in Kelowna, participants were asked to recommend next steps for this process. These are summarized in Table 7-1.

Table 7-1. Next Steps for Okanagan Water Resources and Climate Change Studies.

Climate Water Supply Water Use Other Institutional •=Centralized water •=Groundwater authority with legislation technical/research capabilities Modelling •=Regional •=Effects of dams •=Land & water use scenarios and land use on scenarios water temperature Monitoring •=Snowpack •=Shifts in Okanagan •=Trends in other •= Flora and and related hydrologic regime regions fauna indicators •=Trends in other regions Outreach •=Options for dryland •=Regional landscaping water-related success stories Research •= Inventory/archive •=Inventory/archive Support of surface water & of surface water & groundwater groundwater use •= Data collection •=Data collection standards standards

Workshop participants also raised a number of questions regarding research process. How could a multi-stakeholder dialogue be sustained? How could values and individual priorities be accounted for in research (e.g. determining an ‘appropriate balance’ in water use among various users)? Are there analogues of future Okanagan climate/water scenarios already visible in other regions (e.g. eastern Washington state), and if so, how could these contribute to the dialogue in the Okanagan?

This study showed that regional scenarios of changing climate and hydrology could be constructed in a short time, and be communicated to regional decision makers and stakeholders as part of a dialogue on the future of water in the Okanagan. In order for this effort to be effective over the long term, research, monitoring and dialogue need to be sustained. That is the challenge that faces anyone concerned about the regional implications of climate change.

The multi-disciplinary character of climate change means that many are interested but few make direct investments. Climate change science has a natural home within the atmospheric sciences and related disciplines (e.g. oceanography, ecology), but climate change impacts/adaptation has difficulty finding room at the Inn, or in knowing which Inn to go to. Regional circumstances are

______39 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

unique, and local knowledge has an important role to play in impacts/adaptation research, monitoring, dialogue, and ultimately, some form of response. It is hoped that this case study experience will lead to the evolution of a regional home for climate change dialogue in the Okanagan.

8.0 ACKNOWLEDGEMENTS

The research team consisted of the following:

• Team Leader Stewart Cohen • Climate Scenarios Bill Taylor, Roxanne Brewer, Eric Taylor • Hydrologic Scenarios Stuart Hamilton, Dave Hutchinson, John Martin, Jacek Scibek, Paul Whitfield • Trout Creek Case Study Erin Embley, Ken Hall, Les Lavkulich • Stakeholder Dialogue Rob VanWynsberghe, Wendy Avis, Jeff Carmichael, Maggie Julian, Tanuja Kulkarni, Pamela Robinson, James Tansey

This study was funded by the Climate Change Action Fund, Govt. of Canada (Project No. A-206). Additional in-kind support was provided by Environment Canada (Adaptation & Impacts Research Group, and Pacific & Yukon Region) and the University of British Columbia (Sustainable Development Research Institute, and Institute for Resources and Environment).

We would like to express our appreciation to Greg Armour (Okanagan Basin Water Board), Michelle Broshard (Okanagan Similkameen Boundary Regional Stewardship Coordinator), Ted van der Gulik (B.C Ministry of Agriculture), Dawn Machin (Okanagan First Nations Alliance), Peter Schwarzhoff (Environment Canada) and Kim Hyatt (Department of Fisheries and Oceans) for their assistance in identifying regional contacts and publicizing this study.

We would like to thank those who participated in the January and March workshops; Kevin Ade, Jack Allingham, Greg Armour, Jane Bender, Heidi Bennett, Scott Benzer, Janette Bigney, Mary Bowcock, Steve Bryson, Ron Cannan, Barrie Clark, Renee Clark, Stan Combs, Jeff, Curtis, Lorne Davies, Les Deuling, Peter Dill, Mike Dorion, Paul Doyle, Joe Falkoski, Ted Fuller, Steve Gormley, Tracy Gow, Sylviane Guernier, Leah Hartley, Web Hallaver, Juergen Hansen, Robert A. Hawes, Markus Heiunrichs, Greg Henderson, Laura Hooke, Brian Jantz, Ted Jeffery, Ray Jubb, Walter Koch, Darwyn Kutney, Shayla Lawrence, Terry MacDonald, Don McKee, Bruce Mohen, James Moller, Jules Morris, Lisa Mose, Denise Neilson, Smiley Nelson, Harrold Norris, Brian Nuttall, Janine Nyvell, John Parsons, Allan Patton, John Peshbridge, Toby Pike, Denis Potter, Peter Schwarzhoff, Elaine Sharp, Jim Shaver, Dave Smith, Scott Smith, Carmen Stanek, Judie Steeves, Bruce Sumner, Andrew Swetlishoff, Jake Theissen, Ian Walker, Lena Warrington, Mike Watkins, Mark Watt, Dean Watts, Allen Weins, Roger Wysocki

We would also like to thank Troy Gauthier of the Sustainable Development Research Institute for his computer support.

______40 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

9.0 REFERENCES

Arnell, N., C. Liu, R. Compagnucci, L. da Cunha, C. Howe, K. Hanaki, G. Mailu, I. Shiklomanov and E. Stakhiv, 2001. Hydrology and Water Resources. In: Climate Change 2001: Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change [McCarthy, J., O. Canziani, N. Leary, D. Dokken, and K. White (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.

Beven, K., 2001. How far can we go in distributed hydrological modeling? Hydrology and Earth System Sciences, 5(1), 1-12.

Bergstrom, S., 1995. The HBV model. In Computer models of watershed hydrology, V.P. Singh, (ed.)., Water Resources Publications, Highlands Ranch, Colo., 443-476.

Brewer, R. and B. Taylor, Draft, 2001. Climate Indicators for British Columbia and the Yukon Region. Environment Canada, Environmental Conservation Branch. Aquatic and Atmospheric Sciences Division, Vancouver, British Columbia

Bruce, J., I. Burton, H. Martin, B. Mills and L. Mortsch, 2000. Water Sector: Vulnerability and Adaptation to Climate Change, Final Report. Global Change Strategies International Inc. and Atmospheric Environment Service, Environment Canada, Ottawa, 141 pp. British Columbia Forest Service, May 1, 2001, Okanagan Timber Supply Area, Public Discussion Paper, July 2000. Available at http://www.for.gov.bc.ca/tsb/tsr2/tsa/tsa27/pdp/22pdp006.htm

Canadian Centre for Climate Modelling and Analysis, 2001. Environment Canada, Victoria, British Columbia. Available at http://www.ccma.bc.ec.gc.ca. Cannings, R.A., R.J. Cannings and S.G. Cannings. 1987. Birds of the Okanagan Valley, British Columbia. Royal British Columbia Museum, Victoria, B.C. 420p Central Okanagan Resource Stewards (CORS), 2001. “Concept Paper for Partners in the Kelowna Community Stewardship Centre Draft” Unpublished. 11 pages. March.

Cohen, S.J., K.A. Miller, A.F. Hamlet and W. Avis, 2000. Climate change and resource management in the Columbia River Basin. Water International, 25(2): 253-272. Dobson Engineering Ltd, 1998. Interior Watershed Assessment Procedure for the Trout Creek Watershed. 27 p. + appendices.

Ecological Stratification Working Group (ESWG), 1996. A National Ecological Framework for Canada. Ottawa: Agriculture and Agri-Food Canada, Research Branch, Centre for Land and Biological Resources Research, Environment Canada, State of the Environment Directorate, Ecozone Analysis Branch.

Hamlet, A. F. and D. P. Lettenmaier, 1999. Effects of climate change on hydrology and water resources in the Columbia River Basin. Journal of the American Water Resources Association 35 (6): 1597-1624. Human Resources and Development Canada, 2001. “Okanagan Perspective: Hospitality - Industry Outlook”. www.bc.hrdc-drhc.gc.ca/okanagan/common/optourism.shtml, June.

Intergovernmental Panel on Climate Change (IPCC), 1996. Climate Change 1995. The Science of Climate Change. Contribution of Working Group I to the Second Assessment Report of the Intergovernmental Panel on Climate Change. [Houghton J.T, L. G. Meira Filho, B.A. Callander,

______41 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

N. Harris, A. Kattenberg and K. Maskell (eds.)]. Cambridge University Press. ISBN 0 521 564433.

Intergovernmental Panel on Climate Change (IPCC), 2001. Summary for Policymakers: A report of Working Group I of the Intergovernmental Panel on Climate Change, Eighth Session of Working Group I, Shanghai, 17-20 January, 2001. Available at http://www.usgcrp.gov/ipcc/wg1spm.pdf Kattenberg, A., Giorgi, F., Grassl, H., Meehl, G.A., Mitchell, J. F. B., Stouffer, R. J., Tokioka, T., Weaver, A. J. and Wigley, T. M. L. 1996. Climate Models – Projections of Future Climate. In Climate Change 1995. The Science of Climate Change. Contribution of Working Group I to the Second Assessment Report of the Intergovernmental Panel on Climate Change. [Houghton J.T., L. G. Meira Filho, B.A. Callander, N. Harris, A. Kattenberg and K. Maskell (eds.)]. Cambridge University Press. ISBN 0 521 564433.

Klemes, V., 1990. Sensitivity of Water Resource Systems to Climatic Variability. In Proc. of the Canadian Water Resources Association 43rd Annual Conference, Penticton BC., pp. 233-242. Leith, R.M., and P.H. Whitfield, 1998. Evidence of climate change effects on the hydrology of streams in south-central B.C. Canadian Water Resources Journal (23):219-230

Mantua, N. J. , Hare, S. R., Zhang, Y., Wallace, J. M. and Francis, R. C., 1997. A Pacific interdecadal climate oscillation with impacts on salmon production. Bulletin of the American Meteorological Society 78(6):1069-1079.

Mendelsohn, R. and J.E. Neumann, (eds.). 1999. The Impact of Climate Change on the United States Economy. Cambridge University Press, Cambridge, 331 pp.

Merton, R. K., Fiske, M., and Kendall P., 1956. The focused interview: A manual of problems and procedures. New York: Free Press

Miles, E.L., A.F. Hamlet, A.K. Snover, B. Callahan, and D. Fluharty, 2000. Pacific Northwest regional assessment: the impacts of climate variability and climate change on the water resources of the Columbia River Basin. Journal of the American Water Resources Association, 36(2): 399-420.

Miller, K.A., 2000. Managing supply variability: the use of water banks in the western United States. In Drought: A Global Assessment, Volume II [Wilhite, D.A., (ed.)], Routledge, London, 70- 86. Ministry of Environment, Lands and Parks (MOELP), 2000. Profile of a Candidate Sensitive Stream Under the Fish Protection Act: Trout Creek, British Columbia. Unpublished. 8 pages. November.

Mote, P., and 18 others, 1999. Impacts of Climate Variability and Change - Pacific Northwest. A report of the Pacific Northwest regional assessment group for the U.S. Global Change Research Program. The JISAO/SMA Climate Impacts Group, JISAO Contribution #715, University of Washington Seattle, 109 pp.

Nicholls, N., Gruza, G.V., Jouze, J., Karl, T.R., Ogallo, L.A. and Parker, D.E., 1996. Observed Climate Variability and Change. In Climate Change 1995. The Science of Climate Change. Contribution of Working Group I to the Second Assessment Report of the Intergovernmental Panel on Climate Change. [Houghton J.T., L. G. Meira Filho, B.A. Callander, N. Harris, A. Kattenberg and K. Maskell (eds.)]. Cambridge University Press. ISBN 0 521 564433. Okanagan Basin Water Board (OBWB), 2000. Okanagan Basin Briefing Notes. Unpublished. 5 pages, April.

______42 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001

Pidwirny, M. and T. Gow, 2001. “Living Landscapes – Agriculture” Royal British Columbia Museum & Okanagan University College, June 5. http://royal.okanagan.bc.ca/mpidwirn/agriculture/agriculture.html

Schriner, D. and R. Street, 1998. North America. In The Regional Impacts of Climate Change: An Assessment of Vulnerability. A Special Report of IPCC Working Group II Report of the Intergovernmental Panel on Climate Change. [Watson, R., M. Zinwoyera and R. Moss (eds.)]. Cambridge University Press, Cambridge: 253-330.

Smit, B., O. Pilifosova, I. Burton, B. Challenger, S. Huq, R. Klein and G. Yohe, 2001. Adaptation to Climate Change in the Context of Sustainable Development and Equity. In Climate Change 2001: Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change [McCarthy, J., O. Canziani, N. Leary, D. Dokken, and K. White (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.

Statistics Canada, 1996. Census of Agriculture, British Columbia Highlights. http://www.statcan.ca/english/censusag/bc.htm. Trout Creek Flow Monitoring: Year 1 Report. 10 p. + appendices.

Whitfield, P.H., and A. Cannon, 2000a. Recent Variations in Climate and Hydrology in Canada. Canadian Water Resources Journal 25:19-65

Whitfield, P.H., and A. Cannon, 2000b. Recent Variations in Climate and Hydrology of British Columbia and Yukon. International Hydrological Programme. Contributions to IHP-V by Canadian Experts. Technical Documents in Hydrology No. 33. Unesco Paris. pp 1-21.

Whitfield, P.H., and A.J. Cannon, 2000c. Polar Plotting of Seasonal Hydrological and Climatic Data. Northwest Science. (74):76-80.

Whitfield, P. H., 2001. Linked Hydrologic and Climate Variations in British Columbia and Yukon. Environmental Monitoring And Assessment (67):217-238.

Whitfield, P.H., Karin Bodtker, and A.J. Cannon, (in prep). Recent Climate Variations In Canada - Decadal Shifts In Temperature And Precipitation 1976-1995.

Z-Point Graphics. 2001. http://www.panorama-map.com/z-pointsite/zoka.html

______43 Water Management and Climate Change in the Okanagan Basin Cohen and Kulkarni (eds.) 2001 APPENDIX 1: Climate Observations

Figure A1-1. Seasonal Change in Minimum Temperature (°C) in the Okanagan Basin – 2020s

Figure A1-2. Seasonal Change in Maximum Temperature (°C) in the Okanagan Basin – 2020s

Figure A1-3. Seasonal Change in Precipitation (%) in the Okanagan Basin – 2020s

Figure A1-4. Projections of January Mean Minimum Temperature in the Okanagan Basin 45

Figure A1-5. Projections of July Mean Maximum Temperature in the Okanagan Basin

Figure A1-6. Projections of Annual Total Precipitation in the Okanagan Basin

° Figure A1-7. Projected Changes in Growing Degree Days (> 5 C) at Kelowna Airport

Figure A1-8. Projected Changes in Heating Degree Days (< 18°C) at Kelowna Airport

Figure A1-9. Projected Changes in Cooling Degree Days (> 18°C) at Kelowna Airport

48 APPENDIX 2: Hydrological Observations

Figure A2-1. Examples of recent shifts in temperature for 11 day time periods in the Okanagan Valley. Green indicates an increase between the two decades and yellow a decrease. Arrowheads indicates statistical significance.

49 Figure A2-2. Examples of recent shifts in precipitation for 5 day time periods in the Okanagan Valley. Green indicates an increase between the two decades and yellow a decrease. Arrowheads indicates statistical significance.

50 Figure A2-3. Examples of recent shifts in natural streamflows for 5 day time periods in South- central BC. Green indicates an increase between the two decades and yellow a decrease. Arrowheads indicates statistical significance.

Figure A2- 4. Examples of recent shifts in altered streamflows in South-central BC. Green indicates an increase between the two decades and yellow a decrease. Arrowheads indicates statistical significance.

52 Table A2-1 Description of hydrometric and climate stations used for hydrological modeling

Bellevue Creek near Camp Creek Daves Pearson Vaseux Whiteman Station Okanagan near mouth Creek near Creek near Creek above Creek above Name Mission near Thirsk Rutland the mouth Solco Creek Bouleau Creek ID 08NM035 08NM134 08NM137 08NM172 08NM171 08NM174 Area (km2) 77.2 34 34.1 72.8 117 114 Elevation (m) 1550 1480 1350 1550 1700 1300 Start of record 1920 1965 1965 1970 1970 1971 End of record 1986 1999 1986 1987 1999 1999 Years of Record 29 35 22 18 30 29 Climate data Peachland Vernon climate Kelowna Brenda Joe Rich Joe Rich Coldstream station Airport Mines Creek Creek Oliver STP Ranch CS ID 1123970 1126077 1123750 1123750 1125766 1128580 CS elev. (m) 430 1520 875 875 297 482 Calibration Period Start 10/01/69 10/01/77 7/01/70 10/01/70 10/15/70 10/01/71 End 9/30/79 9/30/87 9/30/76 9/30/77 9/01/81 9/30/81 NSME 0.81 0.8 0.57 0.77 0.78 0.76 Qo 0.38 0.148 0.135 1.15 0.901 0.617 Qp 0.378 0.142 0.165 1.053 0.892 0.546 Verification Period Start 10/01/79 10/01/69 10/01/76 10/01/77 10/01/86 10/01/81 End 7/03/86 9/30/75 1/31/83 1/31/83 5/31/99 9/30/96 NSME 0.8 0.79 0.55 0.74 0.73 0.77 Qo 0.502 0.182 0.133 1.115 0.915 0.623 Qp 0.443 0.169 0.116 0.974 0.951 0.587

53 Bellevue Creek

Biogeoclimatic Zones Hypsometric Curves Alpine Tundra 2400 Mountain Spruce 2200 Interior Cedar Engleman Spruce / Subalpine Fir & Hemlock Catchment description 2000 Engleman Spruce / Subalpine Fir 1800 Bellevue Creek near Okanagan Mission Mountain Spruce Interior Douglas Fir (WSC station id 08NM035) is a tributary 1600 Interior Cedar & Hemlock to Okanagan Lake South of Kelowna. The 1400

gauged catchment area is 77.2 km2. This Vaseux Creek 1200 Interior Elevation (m) Whiteman Creek is a very steep catchment with a range of Douglas Fir 1000 Bellevue Creek elevation from 700 to 2200m. Pearson Creek Ponderosa Pine 800 Camp Creek / Bunchgrass Hydrometric Station Dave’s Creek 08nm135 600 0 20 40 60 80 100 120 N Predicted Changes in Annual Cumulative Catchment Area ( km2) Hydrographs

1 km ECHAM4 Model Total Annual Flow 50 2050 POR 40 2020 2000 Landuse 30 1950 Hadley GCM (UK) 20 Old Forest ECHAM4 (German) 1900 10 POR Young Forest CGCM (Canadian) 1850 (1969-1985) 0 3 Discharge (m / s) Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Vernon Selectively Logged 1800 k ee Recently Logged Whiteman Cr Hadley Model 1750 2020 2050 2080 Barren 50 POR Model Scenario (predicted year) 40 2020 Simulated snowpack 30 2050 250 2080 e N k 20 POR a 200 (1969-1985) L k r. e C

e 10 n r o rs

C a

e .

n r P C 0 150 a Kelowna s n e io i iss g v M a Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep a D n a k 100 B O e l l CGCM Model e v u 50 Peachland e C 50 C a POR m r.

p 40 2020 C r . 2050 30 storage (mm water equivalent) Snowpack 0 T ro 2080 ut 2020 2050 2080 C reek 20 Model Scenario (predicted year)

0 Penticton 10 km Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

k a Skaha Lake n a g a

R i

e r Camp Creek Biogeoclimatic Hypsometric Curves Zones Alpine Tundra 2400 CATCHMENT 1 km 2200 C AREA a Catchment description Engleman Spruce (km2) Mountain Spruce m / Subalpine Fir 2000 p Camp Creek (WSC station id: 08NM134) is a small (34 km2) Ponderosa Pine C tributary to Trout Creek; one of the main tributaries feeding 1800 / Bunchgrass r Mountain Spruce e Okanagan Lake. The basin is very steep, with elevations ranging e Engleman Spruce k from 1100 to 2000 metres. 1600 / Subalpine Fir Interior Cedar & Hemlock N 1400 Vaseux Creek 1200 Interior Elevation (m) Whiteman Creek Douglas Fir 1000 Bellevue Creek Pearson Creek Ponderosa Pine 800 Camp Creek Hydrometric Station / Bunchgrass 08nm134 Dave’s Creek 600 0 20 40 60 80 100 120 Cumulative Catchment Area ( km2) Landuse Predicted Changes in Annual C a

Old Forest m Hydrographs p Total Annual Flow Young Forest C ECHAM4 Model 940 r e Selectively Logged e 12 920 k POR Recently Logged 10 900 2020 Hadley GCM (UK) 8 880 ECHAM4 (German) 860 6 N CGCM (Canadian) 840 4 820 POR 3 2 Discharge (m / s) (1969-1987) 800 * 1976 missing Vernon 0 780 Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep 202 205 208 k ee Model Scenario (predicted year) Whiteman Cr 10 km Hadley Model Simulated snowpack 12 250 POR POR 10 (1969-1987) 2020 * 1976 missing 8 200 N 2050 e 6

a 2080 150

L k r. 4 e C

e n r o rs C a 2

e .

n r P 100 C

a Kelowna s n e io 0 i iss g v M a a D n Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep a 50 k B O e l l e v CGCM Model u e 0

Peachland 12 Snowpack storage (mm water equivalent) C C a 2020 2050 2080 m r. p 10 POR

C Model Scenario (predicted year) r . 2020 T 8 ro ut 2050 Cre 6 ek 2080 4

2 10 km Penticton 0 Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

k a Skaha Lake n a g a

R i

e r Dave’s Creek

N 1 km Biogeoclimatic Zones Hypsometric Curves Alpine Tundra 2400 Mountain Spruce Catchment description 2200 Engleman Spruce Interior Cedar & Hemlock / Subalpine Fir Dave's Creek (WSC station id: 08NM137) is a small (31 2000 Engleman Spruce / Subalpine Fir km2) tributrary to Mission Creek; one of the main 1800 tributaries feeding Okanagan Lake. The basin is relatively Mountain Spruce

steep compared to other basins selected in this study. 1600 Elevations range from 900 to 1700 metres with the Interior Cedar & Hemlock median elevation being approximately 1300 metres. 1400

Dominant biogeoclimatic zones are Interior Cedar Vaseux Creek 1200 Hemlock, Englemann Spruce Subalpine Fir, and Montaine Interior Elevation (m) Whiteman Creek Spruce. The basin has been extensively logged, with Douglas Fir 1000 Bellevue Creek only small stands of old growth remaining. Pearson Creek Ponderosa Pine 800 Camp Creek / Bunchgrass Dave’s Creek 600 0 20 40 60 80 100 120 Predicted Changes in Annual Cumulative Catchment Area ( km2) Hydrometric Station Hydrographs 08nm137

ECHAM4 Model 12 Total Annual Flow 10 POR 2020 540 POR 8 (1970-1982) 530 Landuse 6 520 4 Hadley GCM (UK)

2 510 Old Forest ECHAM4 (German) 0 CGCM (Canadian) 500 Young Forest Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

3 490 Selectively Logged Discharge (m / s) Hadley Model 480 Recently Logged 12 470 10 POR 2020 2050 2080 2020 Model Scenario (predicted year) 8 2050 Vernon 6 N 2080 Simulated snowpack ek 4 Whiteman Cre 160 POR 2 (1970-1982) 140 0 Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep 120

100

k CGCM Model 80 a 12 L k r. C e 60 e n r o POR rs 10

C a

e .

n r P 40 C 2020

a Kelowna s n 8 e io i iss g v M a 2050 20 a D n 6 a k 2080 0 B 4 O e Snowpack storage (mm water equivalent) l l 2020 2050 2080 e v u 2 Peachland e Model Scenario (predicted year) C C a m r. 0 p 10 km

C Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep r . T ro ut C reek Pearson Creek

Biogeoclimatic Zones Hypsometric Curves Alpine Tundra 2400

2200 Interior Cedar & Hemlock Catchment description Engleman Spruce / Subalpine Fir 2000 Engleman Spruce / Subalpine Fir Pearson Creek (WSC station id: 08NM172) is a moderate-sized (73 2 1800 km ) tributary to Mission Creek; one of the main tributaries feeding Mountain Spruce Alpine Tundra N Okanagan Lake. The mean annual discharge is 0.85 m3/s, and the period of record extends from 1970 to 1982. 1600 Interior Douglas Fir Interior Cedar & Hemlock 1400 1 km Vaseux Creek 1200 Interior Elevation (m) Whiteman Creek Douglas Fir 1000 Bellevue Creek Pearson Creek Ponderosa Pine Hydrometric Station 800 Camp Creek 08nm172 / Bunchgrass Dave’s Creek 600 0 20 40 60 80 100 120 2 Predicted Changes in Annual Cumulative Catchment Area ( km ) Hydrographs

ECHAM4 Model Total Annual Flow 60 3600

50 POR 3550 2020 Landuse 40 3500 Hadley GCM (UK) POR 30 (1970-1982) 3450 20 ECHAM4 (German) Old Forest 3400 10 CGCM (Canadian) Young Forest 3350 3

0 Discharge (m / s) Recently Logged Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep 3300

Wetlands Hadley Model 3250 2020 2050 2080 60 POR Model Scenario (predicted year) 50 2020 Vernon 40 2050 Simulated snowpack N 2080 ek Whiteman Cre 30 450 20 POR (1970-1982) 400 10 350 0 300 Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

e 250

a CGCM Model 200 L k r. C e 60 e n 150 r o rs

C a

. r 50

n P POR 100 C

a Kelowna s n e sio 2020 i is g v M a 40 50 a D n 2050 a 0 k 30 2080

B Snowpack storage (mm water equivalent) O e 2020 2050 2080 l l e 20 v u e Model Scenario (predicted year) Peachland C 10 C a m r.

p 10 km 0 C r . Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep T ro ut C reek Vaseux Creek Biogeoclimatic Zones Landuse Hypsometric Curves Alpine Tundra 2400

Mountain Spruce Old Forest 2200 Catchment description Engleman Spruce / Subalpine Fir Young Forest 2000 Engleman Spruce / Subalpine Fir Vaseux Creek above Solco Creek (WSC station id Interior Douglas Fir Selectively Logged 08NM171) is a tributary to the Okanagan River 1800 downstream of Okanagan Falls. The gauged Mountain Spruce Recently Logged catchment area is 117 km2. This is a high elevation 1600 Interior Cedar station with a median elevation of 1700 m, with a & Hemlock range from 1200 to 2400 m. The station has been in 1400 N Vaseux Creek operation for 30 years from 1970 to 1999. 1200 1 km Interior Elevation (m) Whiteman Creek Douglas Fir 1000 Bellevue Creek Pearson Creek Ponderosa Pine 800 Camp Creek / Bunchgrass Dave’s Creek 600 0 20 40 60 80 100 120 Predicted Changes in Annual Cumulative Catchment Area ( km2) Hydrographs

ECHAM4 Model HydrometricVernon Station 80 Total Annual Flow 08nm171 70 k POR ee 4400 Whiteman Cr 60 2020 50 4350 40 4300 30 Hadley GCM (UK) 4250 4200 20 ECHAM4 (German) POR 10 4150 (1971-1984) e CGCM (Canadian) k 0 4100

L Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep k .

r 3 4050 C e Discharge (m / s)

e n r so r 4000 C a

e .

n r P C Hadley Model 3950 a Kelowna s n e io i iss g v M a 80 3900 a D n 2020 2050 2080 a 70 k POR B Model Scenario (predicted year) O e 60 l l e 2020 v u 50 e Simulated snowpack Peachland 2050 C C 40 a N m r. p 30 2080

C 200 POR r . 20 (1971-1984) T 180 ro 10 ut C 160 reek 0 Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep 140 120 100 Penticton CGCM Model 80 80 60 O 70 k POR 40 a Skaha Lake n 60 a 2020 20 g 50 a Okanagan n 2050 0 Falls 40 30 2080 storage (mm water equivalent) Snowpack 2020 2050 2080 R k i Va ree

v seux C 20 Model Scenario (predicted year) e

r 10 10 km 0 Oliver Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Whiteman Creek

Biogeoclimatic Zones Hypsometric Curves Alpine Tundra 2400

Catchment description 2200 Engleman Spruce Mountain Spruce / Subalpine Fir This is the second largest catchment of the six used in this 2000 Interior Cedar presentation, and is similar in size to Vaseux Creek catchment. & Hemlock 1800 The period of record extends from 1971 to 1996... Mountain Spruce

Engleman Spruce 1600 / Subalpine Fir N Interior Cedar & Hemlock Interior 1400 Douglas Fir Vaseux Creek 1200 Interior Elevation (m) Whiteman Creek 1 km Douglas Fir 1000 Bellevue Creek Pearson Creek Ponderosa Pine 800 Camp Creek / Bunchgrass Dave’s Creek 600 Hydrometric Station 0 20 40 60 80 100 120 2 08nm174 Predicted Changes in Annual Cumulative Catchment Area ( km ) Landuse Hydrographs Total Annual Flow ECHAM4 Model Old Forest 100 6000 Young Forest 80 POR 5800 2020 Selectively Logged 60 Hadley GCM (UK) 5600 Recently Logged ECHAM4 (German) 40 5400 CGCM (Canadian) Wetlands 20 5200

Barren 0 5000 POR

3 (1971-1996) Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Discharge (m / s) 4800

Hadley Model 4600 100 2020 2050 2080 Model Scenario (predicted year) 80 POR Vernon 2020 Simulated snowpack N 60 ek 2050 Whiteman Cre 40 300 POR 2080 (1971-1996) 20 250

0 200 Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

e 150

k a CGCM Model L k r. C e 100 100 e n r o rs

C a

e . n r P 80 POR C 50 a Kelowna s n e io i iss g v M 2020 a a D 60 n a 2050 0 k B Snowpack storage (mm water equivalent) O e 40 2020 2050 2080 l l e 2080 v u Model Scenario (predicted year) Peachland e 20 C C a m r.

p 10 km 0 C r . Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep T ro ut C reek Predictions bythethreeGlobalClimateModelsforselectedcreeks Okanagan Region:Total BasinSnowpack (April1) during thesummerlow-flowperiod. means lessavailablerunoff toaugmentrunoff The reductionin April 1snowpack storage by 30-60%(compared totheperiodofrecord). April 1snowpack storagemaybereduced the 2050and2080scenariossuggestthat is predictedforthe2020scenario.However, storage (compared totheperiodofrecord) respectively. Littlechangein April 1snowpack to 2070,and20712100periods predicted climatesforthe2011 to2040,2041 scenarios: 2020,2050,2080whichrepresent of record(POR)andthethreeclimate April 1snowpack storagebetweentheperiod equivalent. All creeksexhibitadecreasein stored withinthebasinmeasuredasadepth- storage, whichisthetotal volumeofsnow The graphsillustratechangesinsnowpack ECHAM4 (German),andCGCM(Canadian). Global ClimateModels:Hadley(UK), vary undertheclimatepredictionsforthree April 1snowpack forselectedcreeksmay following graphsillustratehowthemodelled predictor oftotal availablespringrunoff. The snow measurementisoften usedasa spring andsummermonths. The "April1" the landscapethatsustains runoff during integration ofwinterprecipitation storedon Okanagan region.Snowpacks representan source ofrunoff formostcreekswithinthe High elevationsnowpacks arethemain Overview: CGCM (Canadian) ECHAM4 (German) Hadley GCM(UK)

Snowpack storage (mm water equivalent) compared totheperiodofrecord. 1 snowpacksarepredictedtodecreasebyup40% storage underthethreeGCMscenarios.By2080, April Camp Creekshowsadecreasein April 1snowpack Snowpack storage (mm water equivalent) 250 100 150 200 50 100 150 200 250 300 to theperiodofrecord. snowpacks arepredictedtodecreasebyup40%compared storage underthethreeGCMscenarios.By2080, April 1 Whiteman Creekshowsadecreasein April 1snowpack 0 50 0 2020 Model Scenario(predictedyear) 2020 Model Scenario(predictedyear) Camp Creek 2050 2050 Whiteman Creek 2080 2080 (1969-1987) POR

(1971-1996) POR

m C r .

Snowpack storage (mm water equivalent) p

r Peachland

o u

100 120 140 160 180 200 t 20 40 60 80

C r

k W

10 km

h i

2020

t e

Model Scenario(predicted year) O

k m a

an n ag

an C

O r k

a e

Oliver n

a Lake e

g k

a n

2050 R

Penticton

Kelowna

e Falls Okanagan

l e

V u

a Vernon

u x Da v 2080 e’s Cr.

r M e

e i k s

s (1971-1984) POR i

Creek

Pe ar s o n C r . of record. compared totheperiod for the2080scenario, of approximately30% scenario andadecrease storage forthe2020 change insnowpack Vaseux Creekshowsno Vaseux Creek Dave's Creek the catchment. to therelativelylowelevationof snowpack storagemayberelated of record. The large reductionin to 75%comparedtheperiod scenario maybereducedbyup 1snowpack storageforthe2080 2080s. Predicted April to theperiodofrecord,by decrease byupto30%,compared snowpack storageispredictedto the 2020scenario.However, change insnowpackstoragefor Pearson Creekshowslittle Pearson Creek

Snowpack storage (mm water equivalent) 100 120 140 160 20 40 60 80 0 to theperiodofrecord. April 1snowpackstorageforthe2080scenario,ascompared Bellevue Creekshowsadecreaseofapproximately30%in Bellevue Creek

Snowpack storage (mm water equivalent) Snowpack storage (mm water equivalent) 2020 Model Scenario(predictedyear) 100 150 200 250 50 0 100 150 200 250 300 350 400 450 50 0 2020 Model Scenario(predicted year) 2020 2050 Model Scenario(predictedyear) 2050 2080 2050 (1970-1982) POR 2080 2080 (1969-1985) POR (1970-1982) POR Okanagan Region: Mean Flow During Spring Season (Mar - May) 100% Dave's Creek 80% 1974 Predictions by the Canadian Global Climate Model for selected creeks 1972 60% 1978 There is no statistically detectable trend in spring 40% discharge for Daves Creek. 20% Whiteman Creek This is because it is a 1969 1967 Overview: relatively low elevation 0% POR 2020 station. As a result, the The low apparent sensitivity of Whiteman Creek to climate -20% 1981 2050 These graphs show how these streams may effects on spring flow is, in part, a result of the fact increases snowmelt freshet is 1979 2080 typically finished by the end -40% respond to the climate as predicted by the in flow in March and April will be offset by decreases in flow 1977 in May with little net change averaged over the season. Vernon of May. There will likely be -60% Canadian Global Climate Model (GCM) for fewer high flow years like ek 1980 Whiteman Cre -80% 1971, 1972, 1974 and from POR Percent Difference the 2080 scenario (which covers the period 60% 1983 2080 from 2071 to 2100). Individual years from 1993 1978. -100% -1.0 -0.5 0.0 0.5 1.0 the period of record are plotted according 40% 1996 1988 2050 to an index of similarity to the 2080 scenario. Less similar More similar Years which plot to the right hand side of 20% 2020 Index of Climatic Similarity to 2080 Scenario the graph indicate that the climate in those 1972 years is more similar to the predicted climate 0% POR e

1994 k

change than those years which plot to the a

1982 L k r. left hand side. The years are also colour C -20% 1979 e e n r so coded according to whether they are warm- r C a

1992 e .

n r P dry, warm-wet, cool-dry or cool-wet with -40% 1977 C 80%

a Kelowna s 1975 1973 ’ sion e is respect to the reference period of 1961- g v M Percent Difference from POR Percent Difference a a Pearson Creek D 2080 1990. In general, warm wet years are more n 60% -60% a -1.0 -0.5 0.0 0.5 1.0 k A significant upward trend in 2050 similar to the predicted climate and cool- B O e 40% 1971 l spring flow is evident for l dry years are least similar. However, e 1981 2020 Less similar More similar v u Pearson Creek. There are 1972 hydrological response is extremely sensitive Index of Climatic Similarity to 2080 Scenario Peachland e however several years (e.g. 20% 1974 C C to seasonal patterns resulting in a wide a m r. 1971, 1972, 1974, 1981) that

scatter scatter around the mean for the p also show very high spring 0% 1979 POR C r 1977 period of record (POR). The mean for the . flows either because of high T snowpack (1971, 1972)or -20% ro 2020 scenario (from 2011 to 2040) and the u t because of earlier melt(1974, 1982 Cr 2050 scenario (from 2041 to 2060) are also eek 1981) -40% 1975 shown as reference points. If a regression from POR Percent Difference line is shown, then the trend indicated by -60% the line is significant at the 95% confidence -1.0 -0.5 0.0 0.5 1.0 level. Camp Creek Penticton Less similar More similar Index of Climatic Similarity to 2080 Scenario Increases are predicted for Spring flow at Camp Creek, Streamflow during the spring period is N O but the trend is not significant, and the mean predicted k sensitive to antecedant snowpack. If there a spring flow will be less than observed in 1989, which was n a will be less snow accumulation over the a warm-dry year with very early snowmelt. g a Okanagan winter months, spring flow volumes will be n Falls reduced. However, earlier melt means that 50% R more snow will melt in the spring and less 40% 1989 i Va reek v seux C in the summer, which may result in an e 30% r Bellevue Creek 1978 10 km increase in spring flow. 2050 20% 1985 2080 The low apparent sensitivity of Bellevue Creek to climate effects 1983 2020 on spring flow is, in part, a result of the fact increases in flow in Oliver 10% March and April will be offset by decreases in flow in May with 1990 Year label Type 0% 1986 POR little net change averaged over the season. Vaseux Creek 1979 cool wet years -10% 1981 150% 80% 1979 There is a strong upward trend 1979 warm wet years -20% 1992 2080 in spring flow in Vaseux creek. 60% 1972 1983 -30% 1971 1979 warm dry years 1984 100% This trend is a result of earlier 2050 Percent Difference from POR Percent Difference 2080 -40% snowmelt runoff, similar to the 40% 1971 1979 cool dry years 2050 1981 -1.0 -0.5 0.0 0.5 1.0 1983 year 1971 where the peak flow 2020 50% 2020 occurred in May. 1979 predicted by model 20% Less similar More similar 1980 1981 0% POR Index of Climatic Similarity to 2080 Scenario 0% POR 1979 1972 1975 -20% 1976 -50% 1977 1979 1984 -40% Percent Difference from POR Percent Difference

Percent Difference from POR Percent Difference 1973 -100% -60% -1.0 -0.5 0.0 0.5 1.0 -1.0 -0.5 0.0 0.5 1.0 Less similar More similar Less similar More similar Index of Climatic Similarity to 2080 Scenario Index of Climatic Similarity to 2080 Scenario Okanagan Region: Mean Flow During Summer Season (Jun - Aug) 200%

1972 Predictions by the Canadian Global Climate Model for selected creeks 150% 1982

100% 1981 1971 50% Overview: Dave's Creek POR 250% 0% Summer flows are predicted to 1975 2020 1996 These graphs show how these streams may be lower than normal, but the 2080 200% Whiteman Creek -50% 2050 respond to the climate as predicted by the 1972 Vernon average predicted for each of 1969 1977

the scenarios is within the from POR Percent Difference 1979 Canadian Global Climate Model (GCM) for 150% ek 1973 Whitem re the 2080 scenario (which covers the period an C range of summer flows -100% 100% 1982 observed during the period of -1.0 -0.5 0.0 0.5 1.0 from 2071 to 2100). Individual years from record. the period of record are plotted according to 50% Less similar More similar 1990 Index of Climatic Similarity to 2080 Scenario an index of similarity to the 2080 scenario. 0% POR Years which plot to the right hand side of the 2020 graph indicate that the climate in those years -50% 1979 2050 2080 e is more similar to the predicted climate k -100% 1973 1987

a change than those years which plot to the from POR Percent Difference L k r. C left hand side. The years are also colour -150% e e n 120% r so -1.0 -0.5 0.0 0.5 1.0 r C a

coded according to whether they are warm- e . 1972 n r P 100% Less similar More similar C dry, warm-wet, cool-dry or cool-wet with a Kelowna s ’ sion e is g v M 80% respect to the reference period of 1961-1990. Index of Climatic Similarity to 2080 Scenario a Pearson Creek a D n In general, warm wet years are more similar a 60% 1971 k Summer flows are Summer flows are predicted to be lower B 1978 to the predicted climate and cool-dry years O e 40% l predicted to be lower than than normal, but the average predicted l e 1974 are least similar. However, hydrological v normal, but the average for each of the scenarios is within the u 20% response is extremely sensitive to seasonal Peachland e predicted for each of the 1981 range of summer flows observed during C C 0% POR a r scenarios is within the patterns resulting in a wide scatter scatter the period of record. m . p 1975 range of summer flows -20% 2020 around the mean for the period of record C r observed during the . -40% 2050 (POR). The mean for the 2020 scenario (from T period of record. ro 1979 u -60% 2011 to 2040) and the 2050 scenario (from t from POR Percent Difference C 1980 1977 2080 reek 2041 to 2060) are also shown as reference -80% points. If a regression line is shown, then the -1.0 -0.5 0.0 0.5 1.0 trend indicated by the line is significant at Less similar More similar the 95% confidence level. Index of Climatic Similarity to 2080 Scenario Camp Creek Penticton Summer streamflow is not expected to A negative tendency is apparent for N O change too dramatically from the period of k summer flow for Camp Creek a record reference period. The water balance n a during this period will be a function of the 80% g a Okanagan changes predicted for precipitation and to a 1984 n 60% Falls Bellevue Creek lesser degree, changes in the amount of 1986 R 40% i Va reek evaporation due to warmer tempertatures. v seux C Summer flows are predicted to be lower than normal, e

r but the average predicted for each of the scenarios 20% 1990 10 km is within the range of summer flows observed during the period of record. 0% POR Year label Type Oliver -20% 200% 1979 cool wet years 2020 -40% 1980 150% 1979 Vaseux Creek 1972 1979 warm wet years 2050 150% 1986 -60% 1987 2080 Summer flows are Percent Difference from POR Percent Difference 1982 1979 warm dry years 1988 100% predicted to be lower -80% 100% than normal, but the 1979 cool dry years -1.0 -0.5 0.0 0.5 1.0 1974 average predicted for 1975 1982 Less similar More similar 50% each of the scenarios 50% 1979 predicted by model 1984 1974 Index of Climatic Similarity to 2080 Scenario is within the range of 1972 POR POR summer flows 0% 1981 0% observed during the 2020 period of record. 2020 1975 2050 -50% 1979 2050 -50% 2080 2080 Percent Difference from POR Percent Difference 1977 1979 1977 1973

Percent Difference from POR Percent Difference 1973 -100% -100% -1.0 -0.5 0.0 0.5 1.0 -1.0 -0.5 0.0 0.5 1.0 Less similar More similar Less similar More similar Index of Climatic Similarity to 2080 Scenario Index of Climatic Similarity to 2080 Scenario Okanagan Region: Mean Flow During Fall Season (Sep - Nov) 150% Predictions by the Canadian Global Climate Model for selected creeks 1978 100% 1968 1966 1974 Dave's Creek 2080 50% A slight increase in Fall 2050 Overview: discharge is predicted, 1979 2020 Whiteman Creek within the range of 0% POR These graphs show how these streams may observed Fall Discharge A slight increase in Fall discharge is 1972 1967 respond to the climate as predicted by the -50% Canadian Global Climate Model (GCM) for predicted, within the range of observed Fall Discharge 1975 1977 the 2080 scenario (which covers the period from POR Percent Difference -100% from 2071 to 2100). Individual years from 250% Vernon 1974 -1.0 -0.5 0.0 0.5 1.0 the period of record are plotted according to ek 200% W re an index of similarity to the 2080 scenario. hiteman C Less similar More similar Years which plot to the right hand side of the 150% 1986 Index of Climatic Similarity to 2080 Scenario graph indicate that the climate in those years 2080 100% 1987 is more similar to the predicted climate 1996 2050 change than those years which plot to the 50% 1972 left hand side. The years are also colour 2020 0% POR coded according to whether they are warm- e 1979 k Pearson Creek dry, warm-wet, cool-dry or cool-wet with -50% a

L 1990 k r. C respect to the reference period of 1961-1990. e A slight increase in Fall discharge is predicted, within 1975 1995 1992 e n r o -100% rs In general, warm wet years are more similar C a the range of observed Fall Discharge

Percent Difference from POR Percent Difference e .

n r P to the predicted climate and cool-dry years -150% C 250% a Kelowna s ’ sion e is g v M are least similar. However, hydrological -1.0 -0.5 0.0 0.5 1.0 a a D n 200% 1978 response is extremely sensitive to seasonal Less similar More similar a k patterns resulting in a wide scatter scatter B Index of Climatic Similarity to 2080 Scenario O e l 150% l e around the mean for the period of record v u e (POR). The mean for the 2020 scenario (from Peachland 100% C C a 1982 2011 to 2040) and the 2050 scenario (from m r. p 50% 1974 2080

2041 to 2060) are also shown as reference C r 1979 points. If a regression line is shown, then the . 2050 T 2020 ro 0% POR trend indicated by the line is significant at ut C 1972 the 95% confidence level. reek -50% 1977 1981

Percent Difference from POR Percent Difference 1975 Streamflow during the Fall can be sensitive -100% to whether proecipitation falls as rain or as -1.0 -0.5 0.0 0.5 1.0 Penticton snow. In a warmer climate, more of the Camp Creek Less similar More similar precipitation will fall as rain and run off Index of Climatic Similarity to 2080 Scenario immediately. The magnitude of the effect is O A slight increase in Fall discharge N k a largely a function of how much of the is predicted, within the range of n observed Fall Discharge a precipitation during this season falls as snow g a Okanagan during the period of record. Higher elevation n Falls stations are more likely to experience a 150% R significant change in Fall Streamflow i Va reek 2080 v seux C 1986 e

100% r Bellevue Creek 10 km 1982 2050 A slight increase in Fall discharge is predicted, 50% within the range of observed Fall Discharge Year label Type 1985 Oliver 2020 150% 1979 cool wet years Vaseux Creek 0% POR 200% 1979 1982 1979 warm wet years 1982 An increasing trend is 100% 1978 1981 150% evident for Fall 1980 1983 1979 warm dry years -50% discharge. In 1982 ( a 2080 1992 cool-wet year) Fall

Percent Difference from POR Percent Difference 100% 50% 1979 cool dry years 1988 rains produced 2050 -100% 2020 1983 2080 discharge in excess of 1979 predicted by model -1.0 -0.5 0.0 0.5 1.0 POR 50% 1979 2050 the averages predicted 2020 0% 1979 Less similar More similar in the climate change scenarios. Index of Climatic Similarity to 2080 Scenario 0% POR 1974 -50% 1972 1981 1977 -50% 1972 1973 1975 1980 from POR Percent Difference 1975 -100% Percent Difference from POR Percent Difference -100% -1.0 -0.5 0.0 0.5 1.0 -1.0 -0.5 0.0 0.5 1.0 Less similar More similar Less similar More similar Index of Climatic Similarity to 2080 Scenario Index of Climatic Similarity to 2080 Scenario Okanagan Region: Mean Flow During Winter Season (Dec - Feb)

Predictions by the Canadian Global Climate Model for selected creeks 2000% 2080

1500% Dave's Creek Overview: Whiteman Creek winter Whiteman Creek streamflow shows a 1000% Dave's Creek shows a very significant upward trend with significant predicted These graphs show how these streams may predicted flow expected to increase in winter flow. The 1968 2050 respond to the climate as predicted by the be 840% of normal. 1988 is Vernon 500% highest winter flow on record 1981 Canadian Global Climate Model (GCM) for 1000% a useful analogue for 2080 ek is 1968 which was 400% 2020 comparison purposes. Whit re the 2080 scenario (which covers the period eman C above normal which is still 0% 1979 POR Winter flow in 1988 was from 2071 to 2100). Individual years from 800% much less than the 1650% 1972 1977

290% of normal in response from POR Percent Difference increase predicted by the the period of record are plotted according to to a warm, wet year. 2080 scenario. -500% an index of similarity to the 2080 scenario. 600% -1.0 -0.5 0.0 0.5 1.0 Years which plot to the right hand side of the Less similar More similar graph indicate that the climate in those years 400% 1988 2050 Index of Climatic Similarity to 2080 Scenario is more similar to the predicted climate change e 200% k than those years which plot to the left hand a

L k r. 1993 2020 C side. The years are also colour coded e e n 1972 r o 0% POR rs according to whether they are warm-dry, 1987 C a

. r Percent Difference from POR Percent Difference 1992 n P warm-wet, cool-dry or cool-wet with respect 1980 C

a Kelowna s ’ sion 3000% -200% e is to the reference period of 1961-1990. In g v M Pearson Creek a a -1.0 -0.5 0.0 0.5 1.0 D general, warm wet years are more similar to n a 2500% Less similar More similar k Pearson Creek shows a very the predicted climate and cool-dry years are B 2080 O e significant increase in flow l l Index of Climatic Similarity to 2080 Scenario e 2000% least similar. However, hydrological response v predicted during the winter u is extremely sensitive to seasonal patterns Peachland e months. Most of this increase is

C C 1500% a in the months of December and resulting in a wide scatter scatter around the m r.

p January. There are no analogues mean for the period of record (POR). The C r in the period of record for flow 1000% 2050 mean for the 2020 scenario (from 2011 to . T increases of this magnitude. ro 2040) and the 2050 scenario (from 2041 to ut 500% C 1971 1981 2060) are also shown as reference points. If reek 2020 POR a regression line is shown, then the trend 0% 1972 1975 indicated by the line is significant at the 95% from POR Percent Difference 1980 1977 -500% confidence level. -1.0 -0.5 0.0 0.5 1.0 Penticton These graphs show that streamflow is Less similar More similar N Index of Climatic Similarity to 2080 Scenario predicted to increase through the winter O k a months in response to shorter, warmer, wetter n a winters. The magnitude of the increase g a Okanagan depends on how much of a change there will n Camp Creek Falls be from a hydrolgical regime where most of R the precipitation falls as snow and is therefore A 1200% increase in winter flow is predicted for i Va reek v seux C Camp Creek. The highest flow observed during e

unavailable for runoff - to a regime where the 10 km r Bellevue Creek precipitation falls as rain and is therefore the period of record was 1991 which was a cool- wet year with a winter flow which was 190% Bellevue Creek winter streamflow will increase to about available immediately for runoff. The predicted above normal. Oliver 960% above normal by the 2080's. The highest winter change in flow can be quite dramatic as a streamflow on record was recorded in 1970 which was percentage of 'normal'. 1400% a warm-dry year during which the winter discharge was Vaseux Creek about 200% above normal. 1200% 1400% 2080 2080 2080 Vaseux Creek shows a 1200% 1000% Year label Type 1300% significant trend towards higher streamflow, with 800% 1000% 1979 cool wet years 400% predicted streamflow for 2080 being aprroximately 600% 800% 1979 warm wet years 2050 300% 500% higher than recorded during the period of record. 400% 1982 600% 1979 warm dry years 1982 and 1983 were both 1991 200% 2050 200% 2020 high flow years, 1982 was 1979 cool dry years 1986 1983 400% 1990 cool-wet whereas 1983 100% 1971 2050 0% 1979 POR 1981 2020 was warm-wet. 1970 1979 predicted by model from POR Percent Difference 1992 1988 200% 0% POR 1981 -200% 1972 1979 1986 2020 Percent Difference from POR Percent Difference 1979 -1.0 -0.5 0.0 0.5 1.0 1975 1977 0% POR -100% 1977 Less similar More similar from POR Percent Difference 1972 -1.0 -0.5 0.0 0.5 1.0 Index of Climatic Similarity to 2080 Scenario -200% Less similar More similar -1.0 -0.5 0.0 0.5 1.0 Index of Climatic Similarity to 2080 Scenario Less similar More similar Index of Climatic Similarity to 2080 Scenario Okanagan Region: Duration of Low Flow 200 Predictions by the Canadian Global Climate Model for selected creeks 180 160 200 140 Overview: 180 1992 120 These graphs show how these streams may respond 160 100 1969 1975 to the climate as predicted by the Canadian Global Days Climate Model (GCM) for the 2080 scenario (which 140 80 1970 covers the period from 2071 to 2100). Individual Vernon 120 1975 Dave's Creek 60 1972 years from the period of record are plotted according ek POR Whiteman Cre 1981 to an index of similarity to the 2080 scenario. Years 100 Dave's Creek will experience more 40 which plot to the right hand side of the graph indicate Days 1979 1987 winters like 1968, 1974 and 1971 with 1971 1974 2050 that the climate in those years is more similar to the 80 short duration low flow. 20 1968 2020 2080 predicted climate change than those years which 0 plot to the left hand side. The years are also colour 60 POR 1990 -1.0 -0.5 0.0 0.5 1.0 coded according to whether they are: warm-dry, 40 1972 1994 2080 warm-wet, cool-dry or cool-wet with respect to the Less similar More similar 2020 2050 e

20 k reference period of 1961-1990. In general, warm wet 1993 Index of Climatic Similarity to 2080 Scenario

1981 a 1976 years are more similar to the predicted climate and L . 0 k r e C cool-dry years are least similar. However, hydrological e n r so -1.0 -0.5 0.0 0.5 1.0 r response is extremely sensitive to seasonal patterns C a e

. 200 n r P resulting in a wide scatter around the mean for the Less similar More similar C a Kelowna s ’ sion e is g v M 180 period of record (POR). The mean for the 2020 Index of Climatic Similarity to 2080 Scenario a a D scenario (from 2011 to 2040) and the 2050 scenario n Whiteman Creek a 160 k (from 2041 to 2060) are also shown as reference B O e l points. If a regression line is shown, then the trend l e 140 v u indicated by the line is significant at the 95% There will not be any significant change in the e Peachland 120 C Pearson Creek 1975 confidence level. duration of low flow for Whiteman creek. There C a m r. is however a tendency to have fewer years p Pearson Creek will experience 100 like 1992 which had an extended low flow C Low flow occurances are a result of extended absences r Days during a warm dry year. . more winters like 1971 and of inputs from snowmelt or rainfall. Extended periods T 80 1972 1977 ro 1974 with very short periods of sub-freezing weather intercept the supply of ut of low flow and fewer winters C 60 POR atmospheric water in the snowpack. A decrease in the reek like 1975 with an extended low 1981 duration of low flow is likely as the winters become flow period. 40 shorter, warmer and wetter. 1974 20 1971 2020 200 N Penticton 0 2050 2080 180 -1.0 -0.5 0.0 0.5 1.0

160 k a Less similar More similar n 140 a Index of Climatic Similarity to 2080 Scenario g a Okanagan 120 n Falls 1988 10 km R 100 1979 k 200 Days i Va ree

1985 1987 v seux C e

80 1992 r 180 60 POR 160 40 1984 Oliver 140 20 1986 2020 120 Year label Type 2050 1991 1981 Vaseux Creek 1972 1971 0 2080 100 1979 cool wet years 200 Days -1.0 -0.5 0.0 0.5 1.0 Vaseux Creek will experience more 1973 180 80 1979 warm wet years winters like the winter of 1983, with Less similar More similar POR only a very short duration of low flow. 60 1979 warm dry years Index of Climatic Similarity to 2080 Scenario 160 1979 40 1986 1979 cool dry years 140 120 20 1983 2020 1979 predicted by model Camp Creek 2080 2050 100 1973 0 Days The change at Camp creek will be quite profound the period 1977 -1.0 -0.5 0.0 0.5 1.0 of low flow will be quite short and can be contrasted to 1988, 80 1972 POR when the creek was in low flow conditions for almost 1/3 of Less similar More similar 1981 the year. 60 Index of Climatic Similarity to 2080 Scenario 40 1982 2020 Bellevue Creek 1983 20 2080 2050 Similar to Camp Creek the change in Bellevue Creek 0 will be quite profound in response to shorter, warmer, -1.0 -0.5 0.0 0.5 1.0 wetter winters. 1981 and 1983 are both good analogues for the future regime. Less similar More similar Index of Climatic Similarity to 2080 Scenario Okanagan Region: Date of Low Flow Dave's Creek Dave's creek has been rain-dominated in some years(1967,1973,1974), but has most often been snow- Predictions by the Canadian Global Climate Model for selected creeks dominated. The predictions show that in the future there will be many more late-summer occurences of the annual low flow That this stream has a history of shifting between regimes likely means that the aquatic ecosystem has the Overview: capacity to adapt to the regime shift 350 These graphs show how these streams may respond 350 to the climate as predicted by the Canadian Global 300 1975 1983 300 Climate Model (GCM) for the 2080 scenario (which 1992 Vernon 1974 covers the period from 2071 to 2100). Individual 2080 250 ek 250 1967 1979 1986 2050 Whiteman re C 1973 years from the period of record are plotted according 2080 200 2020 200 to an index of similarity to the 2080 scenario. Years POR which plot to the right hand side of the graph indicate that the climate in those years is more similar to the 150 150 2050 predicted climate change than those years which Julian Day 100 Julian Day 100 2020 plot to the left hand side. The years are also colour 1972 POR

e coded according to whether they are: warm-dry, 50 k 50 1980 a 1972 1977 L warm-wet, cool-dry or cool-wet with respect to the k r. C 1971 e 1981 e n 1968 r o reference period of 1961-1990. In general, warm wet 0 rs 0

C a

e . n r P -1.0 -0.5 0.0 0.5 1.0

years are more similar to the predicted climate and -1.0 -0.5 0.0 0.5 1.0 C

a Kelowna s ’ sion e is g v M cool-dry years are least similar. However, hydrological Less similar More similar a Less similar More similar a D n response is extremely sensitive to seasonal patterns Index of Climatic Similarity to 2080 Scenario a Index of Climatic Similarity to 2080 Scenario k B Pearson Creek resulting in a wide scatter around the mean for the O e l l e period of record (POR). The mean for the 2020 v u Whiteman Creek Peachland e Pearson Creek is predicted to show a pronounced scenario (from 2011 to 2040) and the 2050 scenario C C shift in seasonality from an snow dominated a m r.

(from 2041 to 2060) are also shown as reference p regime to a rain-dominated regime. Whiteman creek shows a distinctly bi-modal pattern C r points. If a regression line is shown, then the trend in the timing of low flow. Cool-Wet years such as . T indicated by the line is significant at the 95% 1980 and 1982 almost always experience the ro annual low flow in the spring whereas in most other ut confidence level. Cre years the annual low flow occurs in the late summer. ek 300 The models show that in the future, almost all of the low flow episodes will occur in the late summer. The date of low flow is typically in late winter or early 250 spring for snow-dominated streams and in late summer for rain- dominated streams. In general, one would Penticton 2080 expect an earlier date of low flow for snow-dominated 350 1981 N O streams in a warmer climate. However, for several k a 2050 streams, we see a shift in seasonality of low flow from 300 n a 1988 Julian Day g 2020 late winter to late summer. A shift from one regime to 2080 a Okanagan another could result in profound ecosystem effects that 250 n Falls 2050 R cannot be accurately predicted. k 200 i Va ree

v seux C 1981

e r 150 10 km 0.0 0.5 1.0 2020 Less similar More similar Julian Day 100 Oliver Index of Climatic Similarity to 2080 Scenario 1986 POR Year label Type 50 1979 1979 cool wet years 1984 1987 0 Vaseux Creek 1983 150 1979 warm wet years -1.0 -0.5 0.0 0.5 1.0 Vaseux Creek will continue to experience the 1975 1979 warm dry years Less similar More similar annual low flow in the late winter season. Julian Day 100 1972 Index of Climatic Similarity to 2080 Scenario 1979 cool dry years POR 2020 50 1979 1983 2050 1979 predicted by model Camp Creek 2080 0 Camp creek has been rain-dominated in some years(1981,1988), but has most often been snow- 150 -1.0 -0.5 0.0 0.5 1.0 dominated. The predictions show that in the future there 1981 Less similar More similar Julian Day 1975 will more years in which the annual low flow occurs in 100 2020 late summer. POR 2080 Index of Climatic Similarity to 2080 Scenario 1972 1979 2050 Bellevue Creek 50 1983 Bellevue Creek will remain snow-dominated 0 but the timing of low-flow will be earlier as -1.0 -0.5 0.0 0.5 1.0 shorter winters result in an earlier onset of snowmelt to end the winter recession. Less similar More similar Index of Climatic Similarity to 2080 Scenario Okanagan Region: Duration of High Flow 120

Predictions by the Canadian Global Climate Model for selected creeks 100 Dave's Creek 1972 1978 80 1974 The year with the lowest duration of Overview: 1971 1967 high flow (1980) was cool and wet 60 POR and hence not a very useful analogue 1981 Whiteman Creek Days These graphs show how these streams may respond for the predicted effect. There is no 40 1977 to the climate as predicted by the Canadian Global 90 trend in the scatter of data, indicating 1979 Climate Model (GCM) for the 2080 scenario (which 1972 that there will be a relatively abrupt 2080 80 1986 Vernon 1980 2020 covers the period from 2071 to 2100). Individual 1974 shift to the new regime. The effect will 20 2050 be apparent by the 2020's years from the period of record are plotted according 70 ek 1988 Whiteman Cre to an index of similarity to the 2080 scenario. Years 0 60 1981 which plot to the right hand side of the graph indicate POR -1.0 -0.5 0.0 0.5 1.0 Less similar More similar that the climate in those years is more similar to the 50 2080 predicted climate change than those years which 2050 Index of Climatic Similarity to 2080 Scenario 40 plot to the left hand side. The years are also colour Days 1979 2020 coded according to whether they are: warm-dry, 30 warm-wet, cool-dry or cool-wet with respect to the 1987 20 1973 e reference period of 1961-1990. In general, warm wet k 10 a 120 years are more similar to the predicted climate and L k r. C e 1978 cool-dry years are least similar. However, hydrological e Pearson Creek 0 n r o rs response is extremely sensitive to seasonal patterns -1.0 -0.5 0.0 0.5 1.0 C a

e 100 . n r P 1974 resulting in a wide scatter around the mean for the Less similar More similar C The year with the lowest a Kelowna s ’ sion e is g v M duration of high flow (1980) 1981 period of record (POR). The mean for the 2020 Index of Climatic Similarity to 2080 Scenario a a 80 D is still much greater than the scenario (from 2011 to 2040) and the 2050 scenario n a 1972 POR k predicted high flow regime. (from 2041 to 2060) are also shown as reference B There is a significant negative trend O e That year was cool and wet 60 l l points. If a regression line is shown, then the trend e and hence not a very useful v

to the duration of peak flow which Days u 1977 indicated by the line is significant at the 95% e analogue for the predicted 1979 is supported by the climate change Peachland 40 C confidence level. C effect. There is no trend in 1980 modeling. This is a result of lower a m r. snowpacks which are needed to p the scatter of data, indicating 2020 2050

C that there will be a relatively 20 The duration of high flow for snowmelt-dominated sustain high water for a long period r . abrupt shift to the new 2080 streams is an indication of the total volume of snow of time. High water will occur mor T ro frequently in the fall months driven u regime. The effect will be that is available to melt. In general, the less snow t 0 C apparent by the 2020's there is, the shorter the duration of high flow. A similar by rain event which tend to be much reek -1.0 -0.5 0.0 0.5 1.0 shorter in duration than snowmelt Less similar More similar pattern is evident in the graphs for all stations where events. the predicted duration of high flow is less than almost Index of Climatic Similarity to 2080 Scenario any of the years for which we have data. This change will occur by the 2020's and then there is little N Penticton additional change in subsequent scenarios. This is

significant in that it signals a shift from a snowmelt O dominated regime to a bimodal regime, where there k a n are an increasing number of days of high flow due a to rain but a decreasing number of days of high flow g a Okanagan due to snowmelt. The fact that the predicted scenarios Camp Creek n Falls Bellevue Creek are all at the low end of the scatter of data indicate 10 km R k that there are few or no analogues in the present i Va ree Only 1973 is roughly similar to the predicted Only 1979 is roughly similar to the predicted scenarios. v seux C

e scenarios. This year had low snow pack combined record for the future hydrological regime. All stations This year had low snow pack combined with cool r shown are currently snowmelt dominated and all will temperatures, and as such is not a very good analogue with cool temperatures, and as such is not a very for the warm conditions predicted. There is no trend in good analogue for the warm conditions predicted. become increasingly bi-modal as a result of climate There is no trend in the scatter of data, indicating change. the scatter of data, indicating that there will be a relatively Oliver abrupt shift to the new regime. The effect will be apparent that there will be a relatively abrupt shift to the new by the 2020's regime. The effect will be apparent by the 2020's Year label Type 120 120 1979 cool wet years Vaseux Creek 1978 120 100 100 1979 warm wet years 1990 Only two years (1973 & 1985 1982 100 1983 1979) are roughly similar 1979 warm dry years 80 1982 to the predicted scenarios. 80 POR 1974 1971 1974 These years had low snow 1972 1979 cool dry years 1987 80 1981 60 1981 pack combined with cool 60 POR 1979 temperatures, and as such predicted by model Days 1988 POR 1975 Days 1979 60 are not very good 40 2050 2080 1972 2020 1977 analogues for the warm 40 1977

Days 1979 conditions predicted. There 2080 40 20 is no trend in the scatter of 20 1973 2020 2050 1973 2080 data, indicating that there 20 1979 2050 will be a relatively abrupt 0 2020 shift to the new regime. The 0 -1.0 -0.5 0.0 0.5 1.0 effect will be apparent by -1.0 -0.5 0.0 0.5 1.0 Less similar More similar 0 the 2020's Less similar More similar -1.0 -0.5 0.0 0.5 1.0 Index of Climatic Similarity to 2080 Scenario Index of Climatic Similarity to 2080 Scenario Less similar More similar Index of Climatic Similarity to 2080 Scenario Okanagan Region: Date of Peak Flow 275

Oct Dave's Creek 255 Predictions by the Canadian Global Climate Model for selected creeks Daves Creek is predicted to peak in Sep 235 mid-March by the end of the century compared to a peak which typically 215 occurs at the end of April. Overview: Aug 195 Jul These graphs show how these streams may 175 275 1986 1987 Oct respond to the climate as predicted by the Jun 155 Canadian Global Climate Model (GCM) for the 255 1972 Vernon Julian Day 135 1967 Sep 2080 scenario (which covers the period from 2071 235 May ek POR re 1981 Whiteman C 115 2020 to 2100). Individual years from the period of 215 1969 2050 Aug

record are plotted according to an index of Apr 95 1966 1977 195 2080 similarity to the 2080 scenario. Years which plot 1982

Jul 75 175 1972 to the right hand side of the graph indicate that Mar -1.0 -0.5 0.0 0.5 1.0 1990 2080 the climate in those years is more similar to the Less similar More similar

Jun 155 POR 2020 predicted climate change than those years which Index of Climatic Similarity to 2080 Scenario

e Julian Day 135 2050 Whiteman 1992 k

plot to the left hand side. The years are also May

1979 a 115 Creek L colour coded according to whether they are: warm- k r. 1988 C e

e n r o dry, warm-wet, cool-dry or cool-wet with respect 95 rs Apr C a

e .

n r P to the reference period of 1961-1990. In general, 75 C a Kelowna s ’ sion e is warm wet years are more similar to the predicted g v M Mar -1.0 -0.5 0.0 0.5 1.0 a a D n climate and cool-dry years are least similar. Less similar More similar 275 Pearson Creek a Oct k B However, hydrological response is extremely O e l 255 l Whiteman Creek shows little e Pearson Creek shows little sensitive to seasonal patterns resulting in a wide v u Sep sensitivity to the effects of climate apparent sensitivity to the effects Peachland e 235

scatter around the mean for the period of record C change on the timing of peak flow. C of climate change on the timing of a r m . 215 p

(POR). The mean for the 2020 scenario (from peak flow. The years, 1986 and Aug C 1987, were years in which the r 2011 to 2040) and the 2050 scenario (from 2041 . 195

annual peak occurred in the fall. T Jul 1982 to 2060) are also shown as reference points. If a ro ut 175 1972 C regression line is shown, then the trend indicated reek

Jun 155 2020 by the line is significant at the 95% confidence POR 1981

Julian Day 2050 level. 135 2080

May 1977 115 1979 The timing of peak flow in snowmelt-dominated Penticton

Apr 95 streams is a function of how much energy is N O 75 available to melt snow after snow accumulation k a Mar peaks. In general, warmer temperatures will mean n -1.0 -0.5 0.0 0.5 1.0 a g Less similar More similar that snowmelt will occur earlier because there is a Okanagan more heat in the early spring to melt the snowpack. n Falls Index of Climatic Similarity to 2080 Scenario R However, the timing effect is most pronounced k i Va ree v seux C for low elevation streams because the period of Camp Creek e 275 10 km r

snow-cover will be much shorter at those Oct Camp Creek is predicted to peak in elevations. 255 mid-March by the end of the century Oliver Sep compared to a peak which typically 235 occurs in early May Bellevue Creek Year label Type 215 Aug 275 275 Oct Oct Vaseux Creek Bellevue Creek is predicted 1979 cool wet years 195 255 255 to peak in late April Jul 1979 Vaseux Creek shows little

Sep compared to a peak which Sep 1979 warm wet years 175 235 sensitivity to the effects of 235 typically occurs in mid- climate change on the timing March Jun 1986 1979 warm dry years 155 1990 215 215 of peak flow. The date of Aug Aug

Julian Day peak flow for 1979 occurred 1979 cool dry years 135 1979 POR 1992 195 195

May during the fall.

2020 Jul Jul 1982 115 1987 2050 1979 predicted by model 1991 175 175 1988 2020 1975

Apr 2080 1972 95 Jun Jun 155 POR 155 2080 1981 1972 1981

75 2050 Julian Day Julian Day 135 135 POR 2020 May May Mar -1.0 -0.5 0.0 0.5 1.0 1977 1977 115 115 1979 2050 Less similar More similar 1980 1980 2080 Apr Index of Climatic Similarity to 2080 Scenario Apr 95 95 75 75 Mar Mar -1.0 -0.5 0.0 0.5 1.0 -1.0 -0.5 0.0 0.5 1.0

Less similar More similar Less similar More similar Index of Climatic Similarity to 2080 Scenario Index of Climatic Similarity to 2080 Scenario APPENDIX 3: Focus Group Results RANKED IMPACT AREAS

MODELED IMPACTS •= Fisheries • = Tourism •= Decreased water quality •= Changing Climate Impacts of uncertainty •= •= Higher temperature Changes to water budget and water supply •= •= Decreased snowpack Disparity b/w high and low elevations •= Low flows in fall (timing and quantity) •= Species shift and ecosystem changes •= Air quality – health impacts •= Increased demand for agricultural land and irrigation •= Changes in pests •= Changes in fire regime Changing Hydrology •= Changes in land use •= Earlier peaks in spring freshet •= Aquatic ecosystem •= More precipitation projected, but •= Increased demand for water; groundwater; domestic; irrigation seasonal breakdown is inconsistent •= No consensus on total water flow changes in the system

•= Use water more efficiently: Pay for water, Restrict development and or population, Irrigation methods (practices, systems), Metering •= Inventory supply and prioritize water users •= Water licensing: Increased enforcement, Change in timing, Buyouts •= Develop more water storage/other sources •= Regulate flow and erosion through dams RANKED ADAPTATION OPTIONS •= Adapt silviculture practices; Develop new seedling varieties •= Better integration of government agencies •= Include climate change possibilities when making land use decisions •= Consider ecological costs in resource management •= Consider social costs, making adaptation decisions

Figure A3-1. Summary of Ranked Impacts and Adaptations

Longer growing season IMPACTSHEALTH AND Increased AND RECREATIO Decreased air quality IMPLICATIONN particulates Increase in number of

S temperature inversions Increased energy demand Increase in Decreased water quality climate increases summer Higher temperature Harvesting in warmer Increased risk of water- sedimentation risk borne diseases Increased nutrients, algae, plankton Taste, odour Earlier peak in spring Impact on sport fishery Decline in summer freshet water quality

Less snow for recreation Ski hill investment Shorter winter recreation season Increased demand on groundwater Drought stressed forests Socio-economic vulnerability

Decreased snowpack Summer in stream quantity decline Greater scarcity with higher More precipitation Increased conflict between water users population growth projected but seasonal (95% of watershed breakdown is is licensed) inconsistent across the Increasing summer water demand scenarios meets decreasing supply Decreased supply Lower lake levels of high Decreased water quantity elevation lakes in summer Recreational use changes No consensus on total water flow changes in system Increased uncertainty Figure A3-2. Summary of Impacts and Implications for Health and Recreation Particularly in higher elevations Negative impact on dry land agriculture Limited time for low-impact harvesting in winter months Growing season restricted by water AGRICULTURE Longer growing season availability IMPACTS AND IMPLICATIONS Changes in pests; new invasive species, Loss of beneficial pests less winter kill Habitat may not change fast enough to control invading species

Changes in disease affecting livestock Trees in winter will be less healthy

Where crops grow best will change Species shift Higher temperature Higher altitude harvesting More isolated species- could lead to Climate extremes limiting agriculture potential their extinction

Increased loss of agriculture land Altered timing of seeding Change in harvesting seasons Change in harvesting rates Conflicting water demands Change in timing More difficult to be sensitive- changes in equipment

Increasing demand on surface and groundwater Earlier peak in spring Increased evapotranspiration freshet Lower flows in fall Less water available Less water when it’s needed

Extended irrigation season Decreased Expansion uphill snowpack Increase in irrigation demand Increase demand for agricultural land Availability issues- land quality, pumping costs Soil stability? Roads affected by wetter spring No consensus on total water flow changes in Increased flooding system Figure A3-3. Summary of Impacts and Implications for Agriculture FORESTRY IMPACTS AND IMPLICATIONS Change in insect abundance, diversity More detrimental effects on products Changes in disease and pest regimes Less winter kill of pests Beneficial pests do not survive as well Longer growing season

Higher temperature Change in harvesting seasons Higher elevation harvesting

Species viability change Harder to be flexible (equipment etc)

Harvesting in warmer climate Equilibrium shift

Sedimentation risk increased

Soil stability problems Earlier peak in spring Erosion freshet

Decreased Change in forest fire snowpack frequency and intensity

Figure A3-4. Summary of Impacts and Implications for Forestry

INFRASTRUCTURE IMPACTS AND

IMPLICATIONS

Higher temperature Increased fire potential Increased risk to homes More homes in forested areas Increased energy demand in summer

Increase in water borne disease

Water quality decreased Decreased snowpack

Increased domestic demand Extended irrigation season Increased water demand New uses (e.g. dust control)

Earlier peak in spring Water supply decreased Low flows in fall freshet

More precipitation Local reservoirs, ponds projected but seasonal Impacts on structures in low Dams breakdown is flow or high flow periods

inconsistent across the scenarios Increased erosion

Figure A3-5. Summary of Impacts and Implications for Infrastructure ECOSYSTEMS IMPACTS AND IMPLICATIONS White pelican migration north Bird food supply Impacts on birds Earlier peak in spring Dramatic shifts in fish habitat freshet Impacts on waterfowl Breeding could be (regional/global not local) affected by timing •= Lifecycle timing and spawning of water flow changes •= Decreased fish habitat Decreased snowpack Impacts on fish Warm water fish kill

Impact on species migration through

corridors Shift in biogeoclimatic zones

Endangered birds Differential impacts by elevation Species at risk

Increased pressure on endangered species Vegetation and forestry changes

included riparian Amphibian habitat loss Change in habitats Increase in grasslands

Change in insects, species and abundance Infiltration by new insect species

areas (milfoil) from south Migration of non-native plant species into the Lower insect mortality during winter

Higher fish kills Desertification Eutrophication Higher temperature Responding to blue-green algae life Low elevation loss of productivity time in lakes span

Figure A3-6. Summary of Impacts and Implications for Ecosystems Good for carp, bass FISHERIES Bad for salmonids

IMPACTS AND Aquatic weed growth Fish species shift Fish sensitive to temp change- Osoyoos Lake IMPLICATIONS Intensified through species transplant by humans f Fish abundance and distribution changes Effect on sport fisheries and revenues

Changes in aquatic food chain Change in biodiversity

Longer growing season High elevation

Higher temperature Longer fishing season Eutrophication and lower dissolved oxygen Higher fish kills

Increased sediment and turbidity Sediment build-up and less flushing with decrease in snowpack Lower oxygen due to lower turbulence or Turbidity may not increase prolonged summer stratification associated with stream and lake Loss of productivity in lakes Changes in riparian area vegetation areas Some fish may have to go deeper for Oxygen rich water

Impacts on fish populations Increases in disease in fish

Lake stocks could increase

Increase in oxygen level Life history change Earlier peak in spring Drier spawning areas so delayed spawning freshet Early migration downstream Impacts on fish lifecycle Less time for juveniles to move out- may not be mature enough so higher morbidity in spring

Flow change Longer low flow periods delay the spawning fish Longer recharge time Possibly below baseflow to maintain fish

Figure A3-7. Summary of Impacts and Implications for Fisheries Low flow in fall affects fisheries