THE IMPACT OF CLIMATE CHANGE ON HYDRO-ELECTRICITY GENERATION

Report for CEATI INTERNATIONAL INC. 1010 Sherbrooke Street West, Suite 2500 Montreal, Quebec, Canada H3A 2R7 Website: www.ceati.com WATER MANAGEMENT INTEREST GROUP (WMIG) CEATI REPORT No. T072700-0409 THE IMPACT OF CLIMATE CHANGE ON HYDRO-ELECTRICITY GENERATION Prepared by OURANOS Montreal, Quebec, Canada BC Hydro Generation Bonneville Power Administration Brookfield Renewable Power Électricité de France (EDF) ESKOM Fortum Generation AB Hydro-Quebec Hydro Tasmania Manitoba Hydro National Grid New Brunswick Power Generation Corporation Principal Investigators Biljana Music, Ph.D., Env. Sciences André Musy, Ph.D., Hydrology René Roy, Ph.D., Hydrology Sponsored by New York Power Authority Newfoundland and Labrador Hydro Ontario Power Generation Rio Tinto Alcan Sacramento Municipal Utility District SaskPower Seattle City Light Tennessee Valley Authority U.S. Army Corps of Engineers U.S. Bureau of Reclamation Technology Coordinator Robert P. Metcalfe, P.Eng. August 2008

NOTICE This report was prepared by the CONTRACTOR and administered by CEATI International Inc. ( CEATI ) for the ultimate benefit of CONSORTIUM MEMBERS (hereinafter called SPONSORS ), who do not necessarily agree with the opinions expressed herein. Neither the SPONSORS, nor CEATI, nor the CONTRACTOR, nor any other person acting on their behalf makes any warranty, expressed or implied, or assumes any legal responsibility for the accuracy of any information or for the completeness or usefulness of any apparatus, product or process disclosed, or accept liability for the use, or damages resulting from the use, thereof. Neither do they represent that their use would not infringe upon privately owned rights. Furthermore, the SPONSORS, CEATI and the CONTRACTOR HEREBY DISCLAIM ANY AND ALL WARRANTIES, EXPRESSED OR IMPLIED, INCLUDING THE WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, WHETHER ARISING BY LAW, CUSTOM, OR CONDUCT, WITH RESPECT TO ANY OF THE INFORMATION CONTAINED IN THIS REPORT. In no event shall the SPONSORS, CEATI or the CONTRACTOR be liable for incidental or consequential damages because of use or any information contained in this report. Any reference in this report to any specific commercial product, process or service by tradename, trademark, manufacturer or otherwise does not necessarily constitute or imply its endorsement or recommendation by the CONTRACTOR, the SPONSORS or CEATI. Copyright 2008 CEATI International Inc. All rights reserved. ii

ABSTRACT Hydropower is the leading source of electrical production in many countries. It is a clean and renewable source and certainly will continue to play an important role in the future energy supply. However, the effects of climate change on this valuable resource remain questionable. In order to identify the potential initiatives that the hydropower industry may undertake, it is important to determine the current state of knowledge of the impacts of climate change on hydrological variables at regional and local scales. Usually, the following steps are taken: First, general circulation models (GCMs) are used to simulate future climate under assumed greenhouse gas emission scenarios. Then, different techniques (statistical downscaling/regional climate models) are applied to downscale the GCM outputs to the appropriate scales of hydrological models. Finally, hydrologic models are employed to simulate the effects of climate change at regional and local scales. Outputs from these models serve as inputs to water management models that give more details about hydropower production. This report presents a critical review of the methods used to determine impact of climate change on water resources and hydropower generation. The major results from recent studies worldwide are reported, and future scientific actions are recommended to provide a better understanding of climate change impacts on the hydrological regime. This study is expected to provide direction to the hydropower industry for mitigating the impacts of climate change. Keywords: Climate change, water resources, hydropower, climate models, hydrological models, statistical downscaling, dynamical downscaling. iii

ACKNOWLEDGEMENTS This report was prepared under CEATI International Inc. Agreement No. T072700-0409 with the sponsorship of the following participants of CEATI s Water Management Interest Group (WMIG): BC Hydro Generation BC Canada Bonneville Power Administration OR USA Brookfield Renewable Power ON Canada Electricite de France (EDF) France ESKOM South Africa Hydro Tasmania Australia Fortum Generation AB Sweden Hydro-Quebec QC Canada Manitoba Hydro MB Canada National Grid NY USA New Brunswick Power Generation Corporation NB Canada New York Power Authority NY USA Newfoundland and Labrador Hydro NL Canada Ontario Power Generation ON Canada Rio Tinto Alcan QC Canada Sacramento Municipal Utility District CA USA SaskPower SK Canada Seattle City Light WA Canada Tennessee Valley Authority TN USA U.S. Army Corps of Engineers USA U.S. Bureau of Reclamation USA The investigators are grateful to CEATI for the opportunity to work on this interesting issue. The constant support and guidance by the CEATI Technology Coordinator Robert Metcalfe, as well as Project Monitors Denis Aelbrecht of Électricité de France (EDF), Joan Frain of Ontario Power Generation, David Raff of U.S. Bureau of Reclamation, Richard St-Jean of Brookfield Renewable Power and Efrem Teklemariam of Manitoba Hydro, was greatly appreciated by the investigators. iv

EXECUTIVE SUMMARY This report focuses on methods used to determine impact of climate change on water resources and hydropower generation and provides an overview of recent impact assessment studies. The impact assessment studies are based on numerical modeling that is now a commonly accepted approach for addressing complex environmental and water-resource problems. The following modeling tools/techniques are normally used: Global Climate Models (GCMs), Regional Climate Models (RCMs), Variable-Resolution/Stretched-Grid Global Models (VRGM), statistical downscaling techniques and hydrological models. General circulation models (GCMs) provide simulations of future climate under assumed greenhouse gas emission scenarios. As GCMs have a coarse spatial resolution, application of downscaling techniques is required to provide information at regional and local scale. Downscaling of GCM large-scale variables can be made by using RCMs or by applying statistical downscaling techniques to GCM outputs. Studies show that both approaches have advantages and shortcomings. Dynamical methods (RCMs) are physically based and take into account non-linear effects and other dynamical features of the atmospheric circulation, whereas statistical methods often lack the full range of true variability. While dynamical downscaling has problems with systematic biases, statistical downscaling avoids this problem, as it is a data based method. Hydrological models are useful in assessing the impacts of GCM/RCM projected climate change on hydrological regimes at regional or local scales. Outputs from these models serve as input to water management models that look into more details related to hydropower production. There are two approaches in transferring climate change signal to hydrological models: delta change approach and direct approach. Use of direct approach is recommended where extreme events are evaluated, because extremes resulting from delta approach are simply the extremes from present climate observations that have either been enhanced or dampened according to the delta factors. Direct method is usually associated with modification to RCM outputs to correct for biases before transfer to hydrological models. Although such methods also have limitations, they are more consistent with the RCMs, compared to the delta approach. Review of recent climate impact studies clearly indicates that water resources and, hence, hydropower potential in many regions of the globe, can be strongly affected by climate change. Climate models simulations for the 21st century are consistent in projecting increases of annual mean precipitation in high latitudes (e.g. most of Canada; most of northern European countries). Precipitation is expected to decrease in southwestern regions of North America, southern Europe and Mediterranean region, southern South America, southwest of South Africa, as well as along the south coast of Australia. For some regions, direction of precipitation change is uncertain (e.g. southern part of Mississippi River basin, regions at latitudes near 40ºN in North America, regions at latitudes near 50ºN in Europe, southeast of South Africa, central parts of Australia). Changes in annual runoff (and, hence, water availability and related hydropower potential) vary regionally and, in general, follow projected precipitation changes. Over North America, runoff is projected to increase in northern regions and to decrease in western US, as well as in some regions of southern Canada and over the Okanagan Valley (BC). In Europe, runoff increase is expected over most of Scandinavia, and for regions north of 50 N, while over central, Mediterranean and some regions of Eastern Europe, runoff reduction is projected. Over South America, runoff is projected v

to decline in northeastern countries (northeastern Brazil, Venezuela, Guyana, Suriname), as well as in some southwestern regions of the continent (South Argentina, Chile). An increase is expected over some southeastern and northwestern regions of the continent. Over Australia, runoff is projected to decrease for some south coast regions, but a large uncertainty is present over most of the continent. Studies addressing sensitivity of glacier-fed rivers to climate change indicate that temperature increase causes an initial increase in total glacial runoff by the middle of the 21 st century, followed by significant reduction as the glaciers retreat. Increasing temperature and precipitation extremes is projected to enhance the risk of flooding and drought in many areas. Many studies carried out at catchment-scale indicate possible changes in streamflow annual cycle. These changes have practical implications for the design and operation of local hydroelectric power plants. In regions where snowfall makes a large proportion of winter precipitation, it is likely that a rise in temperature will cause more precipitation to fall as rain and, hence, winter runoff increases and spring snowmelt decreases. The uncertainties involved in quantitative projections of climate change impacts on regional water resources have great implications in renewing decaying hydropower plants and building new hydropower structures. In the years to come, an optimal use of available climate information, together with adaptive management techniques under an uncertain and changing climate, will be major challenges for water managers. vi

TABLE OF CONTENTS ABSTRACT... iii ACKNOWLEDGEMENTS...iv EXECUTIVE SUMMARY...v 1.0 INTRODUCTION... 1-1 1.1 Greenhouse Gas Scenarios... 1-1 1.2 Hydroelectric power: run-of-river and storage facilities...1-2 2.0 CLIMATE MODELING, DOWNSCALING TECHNIQUE AND HYDROLOGICAL MODELLING...2-1 2.1 General Circulation Models...2-1 2.2 Regional Climate Modelling: Dynamical downscaling...2-5 2.2.1 Variable- resolution stretched-grid global models (VRGMs)...2-6 2.2.2 Regional Climate Models...2-6 2.3 Statistical Downscaling...2-9 2.3.1 Transfer function...2-9 2.3.2 Weather classification schemes (Weather typing)...2-10 2.3.3 Stochastic weather generators...2-10 2.4 Hydrological Modelling... 2-11 3.0 METHODS FOR EVALUATION OF HYDROLOGICAL RESPONCES TO GLOBAL CLIMATE CHANGE...3-1 3.1 Direct use of catchment-scale hydrological model (CHM) Evaluation of sensitivity of any specific watershed...3-1 3.2 Direct use of GCM hydrological output...3-2 3.3 Coupling GCMs and macroscale land surface hydrological models (MLS-HMs)...3-2 3.4 Direct use of RCM hydrological output...3-3 3.5 Coupling RCM and MLS-HM...3-4 3.6 Transferring RCM climate change signal to catchment-scale hydrological model (CHM) or macroscale water balance hydrological model (MWB-HM)...3-4 3.7 Statistical downscaling (STAT) of the GCM climate output for use in hydrological model...3-5 4.0 OVERVIEW OF MOST RECENT STUDIES OF IMPACT OF CLIMATE CHANGE ON HYDROLOGICAL REGIME WORLDWIDE...4-1 4.1 North America...4-1 4.1.1 Observed changes of hydroclimatic variables...4-1 4.1.2 Projected changes of hydroclimatic variables...4-3 4.2 Europe...4-8 4.2.1 Observed changes of hydroclimatic variables...4-8 4.2.2 Projected changes of hydroclimatic variables...4-10 4.2.3 Projected changes of hydroclimatic variables over Scandinavia...4-17 4.3 South America...4-22 4.3.1 Observed changes of hydroclimatic variables... 4-22 4.3.2 Projected changes of hydroclimatic variables... 4-25 4.4 Australia...4-29 Page vii

4.4.1 Observed changes of hydroclimatic variables... 4-29 4.4.2 Projected changes of hydroclimatic variables...4-31 4.5 South Africa...4-37 4.5.1 Observed changes of hydroclimatic variables... 4-37 4.5.2 Projected changes of hydroclimatic variables... 4-38 5.0 FUTURE SCIENTIFIC ACTION TO BETTER UNDERSTAND CLIMATE CHANGE IMPACT ON HYDROLOGICAL REGIME...5-1 5.1 Seasonal and Inter-annual Variability...5-1 5.2 Hydrological extremes...5-2 5.2 General comments on uncertainty...5-2 6.0 REFERENCES...6-1 viii