CIVIL JUSTICE EDUCATION 6 ENERGY AND ENVIRONMENT HEALTH AND HEALTH

This document and trademark(s) contained herein are protected by law as indicated in a notice appearing later in this work. This electronic representation of RAND intellectual property is provided for non-commercial use only. Unauthorized posting of RAND PDFs to a non-RAND Web site is prohibited. RAND PDFs are protected under copyright law. Permission is required from RAND to reproduce, or reuse in another form, any of our research documents for commercial use. For information on reprint and linking permissions, please see RAND Permissions.

Limited Electronic Distribution Rights

This PDF document was made available from www.rand.org as a public

service of the RAND Corporation.

6Jump down to document

THE ARTS

CHILD POLICY

CIVIL JUSTICE

EDUCATION

ENERGY AND ENVIRONMENT

HEALTH AND HEALTH CARE

INTERNATIONAL AFFAIRS

NATIONAL SECURITY

POPULATION AND AGING

PUBLIC SAFETY

SCIENCE AND TECHNOLOGY

SUBSTANCE ABUSE

TERRORISM AND HOMELAND SECURITY

TRANSPORTATION ANDINFRASTRUCTURE

WORKFORCE AND WORKPLACE

The RAND Corporation is a nonprofit research organization providing objective analysis and effective solutions that address the challenges facing the public and private sectors around the world.

Visit RAND at www.rand.org

Explore RAND Environment, Energy, and

Economic Development Program

View document details

For More Information

A RAND INFRASTRUCTURE, SAFETY, AND ENVIRONMENT PROGRAM

Environment, Energy, and Economic Development

Purchase this document

Browse Books & Publications

Make a charitable contribution

Support RAND

http://www.rand.org/pdfrd/publications/permissions.html

http://www.rand.org/pdfrd/

http://www.rand.org/pdfrd/research_areas/arts/

http://www.rand.org/pdfrd/research_areas/children/

http://www.rand.org/pdfrd/research_areas/civil_justice/

http://www.rand.org/pdfrd/research_areas/education/

http://www.rand.org/pdfrd/research_areas/energy_environment/

http://www.rand.org/pdfrd/research_areas/health/

http://www.rand.org/pdfrd/research_areas/international_affairs/

http://www.rand.org/pdfrd/research_areas/national_security/

http://www.rand.org/pdfrd/research_areas/population/

http://www.rand.org/pdfrd/research_areas/public_safety/

http://www.rand.org/pdfrd/research_areas/science_technology/

http://www.rand.org/pdfrd/research_areas/substance_abuse/

http://www.rand.org/pdfrd/research_areas/terrorism/

http://www.rand.org/pdfrd/research_areas/terrorism/

http://www.rand.org/pdfrd/research_areas/infrastructure/

http://www.rand.org/pdfrd/research_areas/infrastructure/

http://www.rand.org/pdfrd/research_areas/workforce/

http://www.rand.org/pdfrd/

http://www.rand.org/pdfrd/ise/environ/

http://www.rand.org/pdfrd/pubs/documented_briefings/DB550/

http://www.rand.org/pdfrd/ise/environ/

http://www.rand.org/pdfrd/pubs/documented_briefings/DB550/

http://www.rand.org/pdfrd/pubs/online/

http://www.rand.org/pdfrd/giving/contribute.html

This product is part of the RAND Corporation documented briefing series. RAND

documented briefings are based on research briefed to a client, sponsor, or targeted au-

dience and provide additional information on a specific topic. Although documented

briefings have been peer reviewed, they are not expected to be comprehensive and may

present preliminary findings.

A RAND INFRASTRUCTURE, SAFETY, AND ENVIRONMENT PROGRAM

Environment, Energy, and Economic Development

Preparing for an Uncertain Future Climate in the Inland EmpireIdentifying Robust Water-Management Strategies

David G. Groves, Robert J. Lempert,

Debra Knopman, Sandra H. Berry

Sponsored by the National Science Foundation

The RAND Corporation is a nonprofit research organization providing objective analysis and effective solutions that address the challenges facing the public and private sectors around the world. RAND’s publications do not necessarily ref lect the opinions of its research clients and sponsors.

R® is a registered trademark.

© Copyright 2008 RAND Corporation

All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from RAND.

Published 2008 by the RAND Corporation1776 Main Street, P.O. Box 2138, Santa Monica, CA 90407-2138

1200 South Hayes Street, Arlington, VA 22202-50504570 Fifth Avenue, Suite 600, Pittsburgh, PA 15213-2665

RAND URL: http://www.rand.orgTo order RAND documents or to obtain additional information, contact

Distribution Services: Telephone: (310) 451-7002; Fax: (310) 451-6915; Email: [email protected]

The research described in this report was sponsored by the National Science Foundation and was conducted under the auspices of the Environment, Energy, and Economic Development Program (EEED) within RAND Infrastructure, Safety, and Environment (ISE), in partnership with the RAND Frederick S. Pardee Center for Longer Range Global Policy and the Future Human Condition.

http://www.rand.org

mailto:[email protected]

iii

Preface

This documented briefing reports on the fourth in a series of workshops, which was held on September 17, 2007, at the Chino, California, headquarters of the Inland Empire Utilities Agency (IEUA). It presents the slides from the workshop, along with supporting information. The primary purpose of this fourth workshop was to present follow-up analysis of water-man-agement options for the IEUA service area to accommodate uncertain but potentially signifi-cant climate change and other management uncertainties.

The findings of the first three workshops, held in the fall of 2006, are described in a RAND Corporation report, Presenting Uncertainty About Climate Change to Water-Resource Managers: A Summary of Workshops with the Inland Empire Utilities Agency (Groves, Knopman, et al., 2008).

The work reported here represents one part of a larger project, “Improving Decisions in a Complex and Changing World,” a multiyear effort funded by the National Science Foundation (NSF) as part of the climate-change decisionmaking under uncertainty (DMUU) component of the agency’s human and social dynamics (HSD) priority area, that aims to improve methods of supporting decisions under deep uncertainty. As part of this project, we are working with water agencies in California to help them better understand how climate change might affect their systems and what actions, if any, they need to take to address this challenge. For this project, RAND partnered with IEUA to (1) evaluate different methods for evaluating and pre-senting water-management uncertainty, (2) examine how the IEUA region’s plan and enhance-ments would perform under different future conditions, and (3) suggest management strategies that would be robust to uncertainties about the future.

The audience for this documented briefing includes the workshop participants from IEUA, other water managers and civic leaders from the IEUA service area, California’s regional and state water managers, and water managers and decisionmakers elsewhere with an inter-est in developing strategies for adapting to uncertain climate change and other uncertainties affecting water supply, reliability, and quality.

The RAND Environment, Energy, and Economic Development Program

This research was conducted under the auspices of the Environment, Energy, and Economic Development Program (EEED) within RAND Infrastructure, Safety, and Environment (ISE), in partnership with the RAND Frederick S. Pardee Center for Longer Range Global Policy and the Future Human Condition. The mission of ISE is to improve the development, operation, use, and protection of society’s essential physical assets and natural resources and to enhance

iv Preparing for an Uncertain Future Climate in the Inland Empire

the related social assets of safety and security of individuals in transit and in their workplaces and communities. The EEED research portfolio addresses environmental quality and regula-tion, energy resources and systems, water resources and systems, climate, natural hazards and disasters, and economic development—both domestically and internationally. EEED research is conducted for government, foundations, and the private sector. The mission of the Pardee Center is to explore trends and potential developments in our world 35 to 200 years from now, to develop and apply means of evaluating their effect on individuals and society, and to enhance the quality of human life over that span by improving policies aimed at coping with those trends and developments. Pardee Center projects address such areas as global warming, genetic engineering, the Internet, nuclear-waste disposal, potable water, population growth, and the sustainable use of natural resources.

Questions or comments about this briefing should be sent to the project leader, Robert Lempert ([email protected]). Information about the Environment, Energy, and Eco-nomic Development Program is available online (http://www.rand.org/ise/environ), as is infor-mation about the Pardee Center (http://www.rand.org/international_programs/pardee/) and the DMUU project (http://www.rand.org/ise/projects/improvingdecisions/). Inquiries about EEED projects should be sent to the following address:

Michael Toman, DirectorEnvironment, Energy, and Economic Development Program, ISERAND Corporation1200 South Hayes StreetArlington, VA 22202-5050703-413-1100, [email protected]

Inquiries about the Pardee Center and its research should be sent to this address:

Robert Lempert, DirectorFrederick S. Pardee Center for Longer Range Global Policy and Future Human Condition1776 Main StreetP.O. Box 2138Santa Monica, CA 90407-2138310-393-0411, [email protected]


http://www.rand.org/ise/environ

http://www.rand.org/international_programs/pardee/

http://www.rand.org/ise/projects/improvingdecisions/



v

Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiiSummary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiAcknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiAbbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

CHAPTER ONE

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

CHAPTER TWO

The Climate-Change Challenge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

CHAPTER THREE

Evaluating Adaptive Management Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17Supplies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25Demands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26Chino Basin Groundwater . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26Recycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26Shortages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38Chino Basin Groundwater (Direct Use) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39Desalted Groundwater . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39Imported Supply . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39Groundwater Replenishment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39Recycled Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40Water Saved Through Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

CHAPTER FOUR

Comparison of Options. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45Major Implications and Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

CHAPTER FIVE

Survey Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

vii

Summary

Water-resource managers have long strived to meet their goals of system reliability and envi-ronmental protection in the face of many uncertainties, including demographic and economic forecasts, intrinsic weather variability, and short-term climate change induced by El Niño and other naturally occurring cycles. Now water managers face a new uncertainty—the potential for longer-term and more persistent climate change, which, in coming years, may significantly affect the availability of supply and patterns of water demand. Information about the future effects of climate change is deeply uncertain and likely to remain so for the foreseeable future. Thus, the scientific community is debating how to most usefully characterize this important yet uncertain information for decisionmakers.

The RAND Corporation is conducting a large, multiyear study under a grant from the National Science Foundation (NSF) on climate-change decisionmaking under uncertainty (see ISE, 2007). As part of that project, we are working with water agencies in California to help them better understand how climate change might affect their systems and what actions, if any, they need to take to address this challenge. As a key component of this effort, we have conducted four workshops in cooperation with the Inland Empire Utilities Agency (IEUA), whose service area overlies Southern California’s Chino Groundwater Basin. We chose to work with IEUA because of its forward-leaning management team and articulation of strategies intended to improve the long-term water outlook for its service area. The first three workshops are described in detail in Groves, Knopman, et al. (2008). This document describes the analysis developed for the fourth workshop, held in September 2007.

The workshop described here was developed to help water managers, technical staff, elected officials, and other planners from the IEUA region consider the significance of potential climate change relative to other key planning uncertainties and evaluate options for reducing their vulnerability to supply shortfalls under a wide range of plausible future conditions. We also administered surveys to the participants before and after the workshop to assess how the participants’ views about the climate-change challenge and their possible responses changed, as well as how well various aspects of our presentation informed them about these issues.

We used an integrated decision framework to evaluate the performance of IEUA plans over a wide range of uncertain planning assumptions. The framework centers around a water-management model developed using the Water Evaluation and Planning (WEAP) modeling environment (see SEI, undated). The WEAP model incorporates temperature and precipitation time series reflective of historical weather and plausible ranges of climate change. Each set of assumptions and specifications of IEUA plans is evaluated using a set of outcome measures, including the cost of providing supply to IEUA’s retail customers and the cost of incurring shortages.

viii Preparing for an Uncertain Future Climate in the Inland Empire

We considered nine adaptive management strategies. Each assumed that the manage-ment actions described in IEUA’s 2005 Regional Urban Water Management Plan (IEUA, 2005) would be enacted. Some strategies included near-term commitment to implementing addi-tional water-management strategies. Others required adaptation by implementing additional management actions if conditions deteriorated over time. The specific augmenting actions were chosen in consultation with IEUA staff members. They included actions that either are under active consideration or have been identified as potential strategies for promoting water-supply reliability and addressing climate change.

We evaluated these various plans under a wide range of planning assumptions and pos-sible weather sequences. Following the scenario-discovery procedure outlined in Groves and Lempert (2007) and Lempert, Groves, et al. (2006), we then use a cluster-finding algorithm called the patient rule induction method (PRIM) (Friedman and Fisher, 1999) to find and characterize clusters in the database of simulations that represent management conditions under which the plans perform poorly.

In some climate-change scenarios, we found that the actions described in the 2005 Regional Urban Water Management Plan (IEUA, 2005) would lead to significant shortages and high-cost outcomes. We identified the key risk factors leading to costs that are above a specific cost-impact threshold—namely, declining precipitation, strong effects of climate change on imports, and greater than anticipated declines in Chino basin percolation. When the 2005 Regional Urban Water Management Plan (the 2005 UWMP) is specified to adapt to chang-ing conditions over time, there are fewer high-cost outcomes across the various futures, and those futures that lead to high-cost outcomes are characterized by slightly different factors. We evaluated other strategies as well. In general, strategies that increased the development of local resources in the near term lead to lower overall costs, regardless of the scenario for which the plan was evaluated. The key risk factors are different for the different plans.

We presented the ensembles of results using two types of visualizations—histograms of performance and scatter plots that disaggregate supply-provisioning costs and shortage costs. The scatter plots, which show the results of the individual simulations as single points on a graph, provide more useful information but are more complex and difficult to understand. These scatter plots revealed that not all high-cost cases can be explained by the key risk factors identified and that some cases classified by the risk factors do not result in costs that exceed the threshold. For the former case, one may also seek to develop an additional characterization of results that exceeds the cost threshold but does not arise from the first set of key risk factors.

In summary, the findings suggest that the IEUA region can reduce its exposure to adverse climate-change effects by taking advantage of the favorable economics of local resource devel-opment in the region. As long as the local resource-development activities (including increas-ing efficiency) are less costly than imports and precipitation trends under climate change are flat to strongly decreasing, there is little risk in overinvestment. The results also indicate that, if current IEUA regional leadership expects that future water managers and city planners will respond to decreasing reliability, some actions may be deferred. However, if the region were to implement only the 2005 UWMP now and wait before augmenting its plans, the region would remain vulnerable to scenarios in which there are significant precipitation declines, reductions in basin percolation, and strong reductions in imports due to climate change.

The results of the surveys that were administered before and after the workshop suggested that the survey respondents slightly increased their support for additional near-term water-management actions. They reported that they felt that robust decisionmaking (RDM) was a

Summary ix

useful methodology for presenting climate-related risks and presenting information in a useful way that was helpful for planning. Most notably, the comparison of the before- and after-workshop responses suggests that the RDM analysis provided information that increased the participants’ concerns about the effect of climate change—both the likelihood of significant effects and the severity of the effects. At the same time, water planners increased their belief that they could mitigate or manage these effects. This is a noteworthy finding, as a primary objective of an RDM analysis is to illustrate possible threats and then identify strategies that can be taken to address these threats without having a clear understanding of exactly how the future may evolve. In this case, the survey results suggest that this objective was met.

xi

Acknowledgments

We would like to thank the National Science Foundation for the support it has provided for this research under its grant SES-0345925. David Yates of the National Center for Atmospheric Research continued to work with us to develop improved climate data for the IEUA region. We are also indebted to IEUA General Manager and CEO Richard Atwater and Executive Manager for Policy Development Martha Davis for their encouragement and substantive and logistical support of the workshops. We would also like to thank Ryan Shaw and Andy Camp-bell of IEUA for their help in compiling data for the water-management model and Sarah Olmstead, a fellow at the Frederick S. Pardee RAND Graduate School, for her assistance with some of the modeling. This work would not have been possible without positive participation of the 18 workshop attendees who provided us with their views on the methods and results before, during, and after the workshop. Finally, as this work builds extensively on the work documented in an earlier report (Groves, Knopman, et al., 2008), we extend our appreciation once again to those recognized in the earlier report.

xiii

Abbreviations

af acre-feet

AOGCM atmosphere-ocean general circulation model

CDF cumulative distribution function

DMUU decisionmaking under uncertainty

DYY dry-year yield

EEED Environment, Energy, and Economic Development Program

gpd gallons per day

HSD human and social dynamics

IEUA Inland Empire Utilities Agency

IPCC Intergovernmental Panel on Climate Change

ISE RAND Infrastructure, Safety, and Environment

Metropolitan Metropolitan Water District of Southern California

NCAR National Center for Atmospheric Research

NSF National Science Foundation

PRIM patient rule induction method

PV present value

RDM robust decisionmaking

RUWMP IEUA’s 2005 Regional Urban Water Management Plan

SWP State Water Project

taf thousand acre-feet

UWMP urban water-management plan

WEAP Water Evaluation and Planning

WMM water-management model

1

CHAPTER ONE

Introduction

2 Preparing for an Uncertain Future Climate in the Inland Empire

This documented briefing reports on the fourth in a series of workshops conducted with the Inland Empire Utilities Agency (IEUA), headquartered in Chino, California, about 40 miles east of Los Angeles. In the previous three workshops, described in detail in Groves, Knopman, et al. (2008), we developed a simulation model that evaluated the performance of IEUA’s plans over a wide range of futures, including potential effects of climate change; used this simulation to examine the vulnerabilities of IEUA’s current plans; and explored how different character-izations of uncertainty affect water managers’ understanding and willingness to act on this information. In this fourth workshop, we built on the results of the previous three to exam-ine a larger set of management options available to IEUA. In particular, this study compared strategies that implement a wider range of actions now with strategies that defer some actions to a later date, when more reliable information may become available on the effects of climate change on the IEUA region.

We chose to work with IEUA because of its forward-leaning management team and articulation of strategies intended to improve the long-term water outlook for its service area. IEUA’s management team was interested in working with RAND because it wished to further explore the adequacy of its existing plans under a wide range of possible climate conditions.

Note that this documented briefing, in some cases, presents slightly modified versions of the slides shown to workshop participants. In particular, some slides were eliminated to improve the flow of the briefing.

Introduction 3

For nearly all types of planning decisions related to future resource use, decisionmakers face varying degrees of uncertainty about changes in the physical environment, land use, popu-lation dynamics, and governance. The prospect of climate change, from local and regional to global scale, adds a further dimension of complexity to planning. Lempert, Popper, and Bankes (2003) and Popper, Lempert, and Bankes (2005) described uncertainty as “deep” when theory and past observations are insufficient to provide well-defined probability distributions for important future events on which all participants in a decision can agree and that they can accept as accurate. The effects of climate change certainly contribute such deep uncertainty to IEUA’s planning.

RAND is conducting a large multiyear study, “Improving Decisions in a Complex and Changing World,” under a grant from the National Science Foundation (NSF) as part of the climate-change decisionmaking under uncertainty (DMUU) component of the agency’s human and social dynamics (HSD) priority area. This larger project aims to conduct basic research to improve computer-based tools that support decisionmaking under conditions of deep uncertainty; examine the best means to represent uncertain scientific information to individuals and groups so that they can act on it more effectively; and strengthen the scientific foundations of robust decisionmaking (RDM), a new approach to decision support under deep uncertainty. As part of this project, we are working with water agencies in California to help them better understand how climate change might affect their systems and what actions, if any, they need to take to address this challenge. For this project, RAND partnered with IEUA to (1) evaluate different methods for evaluating and presenting water-management uncertainty, (2) examine how the IEUA region’s plan and enhancements would perform under different future conditions, and (3) suggest management strategies that would be robust against uncer-tainties about the future.


This fourth workshop built on three previous workshops, held September 28, October 20, and November 3, 2006, and described in detail in Groves, Knopman, et al. (2008). We helped IEUA managers, technical staff, other water managers and planners from the IEUA region, and other participants involved in California water planning to consider the significance of potential climate change relative to a few other key uncertainties and how planners might respond by reducing their vulnerability to supply shortfalls under various scenarios.

We prepared for the workshops by first developing a computer simulation model to esti-mate the performance of IEUA’s 2005 Regional Urban Water Management Plan (the 2005 UWMP) (IEUA, 2005) and potential modifications to the plan over a wide variety of scenar-ios. These scenarios included different future climates as well as different levels of achievement of critical recycling and replenishment goals in the 2005 UWMP.

Future climate scenarios were developed from two approaches: extrapolations of past con-ditions in the IEUA service area and statistically generated temperature and precipitation time series using results derived from a number of global-scale models that capture greenhouse gas–driven climate change. The other uncertainties considered in the scenarios include two key features of the 2005 UWMP: recycling and replenishment. Recycling refers to the collec-tion, treatment, and reuse of water for outdoor landscaping and selected indoor uses (e.g., toilet flushing). Replenishment refers to moving surface-water supplies from run-off or other sources into the Chino Groundwater Basin, an aquifer underlying the IEUA service area that meets more than a third of the IEUA’s current demand.

The RAND project team presented three characterizations of uncertainty and admin-istered surveys to workshop participants before, during, and after each of the workshops to record their views about the effectiveness and implications of the different presentations. The first workshop characterized what is known about future climate change and then demon-

Introduction 5

strated differences in the performance of the 2005 UWMP and the plans in place prior to this, based on assumptions that the current climate would continue into the future. In this first workshop, the RAND team also presented climate and other uncertainties using a traditional scenario approach, in which planners examine a small set of future conditions without assign-ing any likelihood or probability to their occurrence.

In the second workshop, we presented state-of-the-art probabilistic scenarios of climate change and then used these probability distributions to estimate the expected performance of the 2005 UWMP.

Finally, in the third workshop, we presented policy-relevant scenarios derived from one key step in RDM, a quantitative decision-analysis approach for supporting decisions under deep uncertainty. These scenarios, in contrast to the hand-crafted scenarios, are analytically derived by assessing the 2005 UWMP over a very wide range of plausible future conditions, noting those futures in which the 2005 UWMP fails to perform adequately and using statisti-cal approaches to identify those few key factors most important in predicting whether a future scenario might cause the 2005 UWMP to fail. Unlike the probabilistic scenarios, probabilities are not assigned to these scenarios. We intended these policy-relevant scenarios to help IEUA consider ways in which it might augment the RUWMP to make it more robust against poten-tially stressful future conditions.

The traditional scenario analysis demonstrated that current plans will perform well if the future climate is benign—that is, wetter than historic conditions—even with incomplete implementation of IEUA’s recycling and replenishment goals. If the future climate is adverse—that is, drier and warmer than historic conditions—then one option to ensure sufficient supply to meet demand would be for IEUA to fully meet its recycling and replenishment goals, as well as invest in more efficiency, and possibly allow more groundwater replenishment with recycled-water supplies.

The probability-weighted scenarios suggest that, if one believed the best-available proba-bilistic information about both the future climate and the IEUA region’s ability to meet the agency’s recycling and replenishment goals, then the current RUWMP would ensure that the chance of a shortage over the next 25 years will not exceed 7 percent.

The policy-relevant scenarios derived from the RDM method identified two sets of conditions most threatening to the success of IEUA region’s water-management plans—a Dry, Flashy, Low-Recycling scenario and a Wet, Effective-Recycling scenario. Under the Dry, Flashy, Low-Recycling scenario, the current plan fails to prevent frequent and significant short-ages. These shortages would impose significant economic and social costs to water users. Under the Wet, Effective-Recycling scenario, IEUA’s RUWMP generates significantly more available water supply than the agency needs to meet demand, could indicate overinvestment in supply enhancements, and could lead to higher costs than necessary to water users to pay for the unneeded water investments. Note that these policy-relevant scenarios highlight two funda-mentally different types of risks: the risk arising from physical shortage of supplies and the risk arising from overinvestment when supplies are adequate.

In the specific case of IEUA, there are likely other benefits of supply enhancements (e.g., less dependency on imported supplies and the ability to sell additional supply to other regions) that we did not include in this first phase of the analysis that would likely reduce or even elimi-nate the negative impact of this scenario. The analysis developed for the workshop presented here reflects the benefit of displacing expensive imports with less expensive local resources or through demand reductions.


Additional efficiency and groundwater-management strategies improve performance under the Dry, Flashy, Low-Recycling scenario but may also generate excess surpluses if this adverse scenario does not come to pass. The analysis, described in Groves, Knopman, et al. (2008), suggested that, if managers from the IEUA region believe that future conditions are more than 25 percent likely to be consistent with the Dry, Flashy, Low-Recycling scenario, then investments in greater efficiency or more use of recycled water for groundwater replenish-ment than what is specified by the RUWMP would be prudent.

Regarding the different presentations of uncertainty, the traditional scenarios provided a simple description of a range of future conditions relevant to IEUA. But many such scenario analyses fall short of decisionmakers’ needs, because the choice of scenarios can appear arbi-trary and there is not a direct way to infer policy choices from the scenarios. Probabilistic sce-narios can provide a concise ranking of the desirability of alternative IEUA plans but can lead to errors of omission in planning by downplaying the potential importance of possible futures that deviate from “most likely” conditions. Generating policy-relevant scenarios was found to provide useful information but can also prove to be a nontrivial analytical exercise and one more complicated to explain to decisionmakers than the other approaches are.

Introduction 7

Following the fall 2006 workshops, IEUA discussed with us ways to improve the representa-tion of the management options that the IEUA service area faced and evaluate scenarios. As a result, we decided to improve the model, develop a revised RDM analysis, and present findings in a fourth workshop. In particular, we expanded the model to provide the capacity to develop water-management plans that can adapt to changing conditions over time. We also worked with IEUA to include monetary estimates of providing water supplies to the region’s end users and estimates of costs of incurring shortages. These cost estimates were generated for each model simulation. We could thus better compare performance of the various management plans across specific future conditions. Finally, with researchers from the National Center for Atmospheric Research (NCAR), we developed an improved set of monthly weather sequences (temperature, precipitation, and wind speed) reflective of plausible climate change.

Using this improved model, the RAND team addressed uncertainty and identified prom-ising water-management plans for the IEUA region. In contrast to the first three exercises, this workshop used only RDM methods and did not compare alternative means of characterizing the uncertainty. In response to the third workshop’s survey results, we attempted to improve the presentation of the RDM-derived results.

9

CHAPTER TWO

The Climate-Change Challenge

This documented briefing is organized in six parts. The preceding chapter set the stage for the fourth workshop by reviewing the design and findings of the first three workshops. This second chapter describes the global and more local-scale challenge presented by changing climatic conditions, serving as a review of a more extended discussion on climate change presented in the earlier workshops. The third chapter describes the nature of the adaptation decision that water managers face: Should they enhance the RUWMP now, or should they wait until some point in the future when they may know more about climate trends? The fourth chapter dem-onstrates the implications of different strategic options under a wide range of possible futures. In the fifth chapter, we summarize the results of the survey administered to participants before and after the workshop, which was intended to help us better understand the influence of the analytical approach on perceptions and preferences for changes in current plans. An appendix, available online, presents the surveys administered to the participants before and after the pre-sentation (Groves, Lempert, et al., 2008).


An expanding body of scientific evidence suggests that the historic climate records that IEUA currently uses to support its planning will no longer prove a reliable guide to the future (Solo-mon, 2007). Temperatures in the region will very likely increase, and precipitation may increase or decrease. These climate trends could have significant effects on the IEUA region, increasing demand for irrigation; generating more intense and, possibly, less frequent storms; diminishing the Sierra snow pack from which the region draws much of its imported water; and decreasing the local supplies available for recharging groundwater (California Climate Change Center and California Energy Commission, 2006).

For some geophysical parameters, such as temperature, the general trends are clear. For others, such as precipitation or storm frequency, they are not. For both, the details at the geo-graphic scale generally required for IEUA planning remain deeply uncertain. Temperature and precipitation forecasts derive from atmosphere-ocean general circulation models (AOGCMs). More than 20 such models exist worldwide, and most give quantitatively different forecasts for Southern California (Tebaldi, 2006). In addition, none of these models gives a satisfactory rep-resentation of the potential for abrupt climate changes. The possibility for such rapid change has increasingly engaged the scientific community as it has increasingly observed changes in the earth’s climate system occurring more quickly than expected (Hansen et al., 2007).

The Climate-Change Challenge 11

The Intergovernmental Panel on Climate Change (IPCC) fourth assessment report provided estimates of the scientific community’s confidence in many different types of climate projec-tions, largely derived from these AOGCMs (Solomon, 2007). For instance, the report indicated that scientists give a probability of higher than 90 percent that the earth’s atmosphere will con-tinue to warm during the 21st century. Scientists give a smaller but still significant probability of 66 percent or greater that the areas affected by drought will increase in the future.

While the IPCC does not directly address the issue of uncertainty about the structure of climate models or missing phenomenology that might significantly lessen or worsen the sever-ity of its projections (Oppenheimer et al., 2007), these probability estimates provide a rough gauge of scientists’ confidence that their models correctly project some of the factors that might most affect IEUA.


Assuming that the AOGCMs are valid representations of future climate dynamics, one can look at the range of predicted temperature and precipitation changes over Southern California. According to 21 AOGCMs, the likely range (± one standard deviation of the mean projected increase) of average summertime temperature increase is projected to be between 0.1 degrees Celsius and 2.1 degrees Celsius. The likely range of average wintertime precipitation trends is between a 19 percent decrease and an 8 percent increase (Tebaldi, 2006).


Water planners in the IEUA service area also face other deep uncertainties that may affect the future performance of their system. The reliability of imported supplies from the State Water Project (SWP) via the Metropolitan Water District of Southern California (Metro-politan) depends on difficult-to-predict legal decisions and potentially catastrophic events. For instance, in December 2007, a U.S. District Court decision restricted the amount of water that can be pumped from the San Francisco Bay Delta in order to protect the delta smelt, a threat-ened species (NRDC v. Kempthorne, 2007). Because of this decision, pumping in 2008 will be reduced by up to 30 percent (DWR, 2007). As another example, an earthquake could cause catastrophic levee failures in the delta, resulting in the elimination or significant decrease in water exports to Southern California for months to years (Lund et al., 2007).

IEUA also faces uncertainty about its future water needs. Projections of future water demand can be developed by estimating changes in the number of water users or activities and changes in the amount of water used by each user or activity. Any such estimates, however, are very uncertain. Although it is widely expected that rapid urbanization of the IEUA service area will continue through about 2030 and lead to increases in the number of households and commercial businesses and thus the number of water users (Husing, 2006), economic factors largely unrelated to water management—such as the recent mortgage crisis and rise in mort-gage defaults—can occur unexpectedly and have a major effect on development patterns. Per-household water demands are also expected to decline, but by how much is difficult to predict. Trends in new residential developments in the IEUA service area suggest significantly lower water-use rates (180 gallons per day [gpd] versus 270 gpd) than are typically considered in water-demand forecasts (Davis, 2007).

Acknowledging these uncertainties, the California Water Plan Update 2005 (DWR, 2005) developed three scenarios of future water demand in which the growth of the number


of households and the increase in naturally occurring conservation (or declines in water-use intensity due to the natural replacement and modernization of water-using devices) varied by significant amounts. Wilkinson and Groves (2006) expanded on the water-plan analysis and looked at different demand projections for Southern California.

The effectiveness of water-management programs and initiatives in the region is also uncertain. In our first workshop, in 2006, we surveyed the 31 participants about their beliefs about the region achieving its ambitious recycling goals. Only 67 percent believed that the goals would be achieved by 2025 (Groves, Knopman, et al., 2008). A variety of reasons were discussed in subsequent workshops and conversations with IEUA staff members about the rea-sons for this. Significant factors included the continued support of the cities to purchase and distribute recycled water to large water users, the ability to raise the required state and federal funding to partially fund the capital improvements necessary to distribute recycled-water sup-plies throughout the service area, and the sustained interest of the region’s cities and agencies to continue to cooperate and plan effectively.


Decisionmakers often find it difficult to manage such uncertainties. Their training, experience, and analytic tools derive from an analytic tradition that requires one to characterize uncertain-ties with a single set of well-defined probability-density functions to then rank the desirability of alternative decision options. Such predict-then-act approaches can be extraordinarily useful for many decisions but can prove inadequate under conditions of deep uncertainty. When decisionmakers attempt to apply predict-then-act approaches under such conditions, they may face pressures to underestimate the uncertainties, find it hard to make decisions because com-peting uncertainty estimates suggest very different conclusions, and leave themselves vulnera-ble to surprise by neglecting valuable information that is not useful for improving probabilistic forecasts (Lempert, Nakicenovic, et al., 2004; Lempert, Popper, and Bankes, 2003).

17

CHAPTER THREE

Evaluating Adaptive Management Strategies

In this chapter, we first describe the nature of the adaptation decision that IEUA managers—and water managers throughout the western United States and elsewhere—face as they con-sider their systems’ vulnerability to climate change and other uncertainties that may dimin-ish system reliability or add to costs. Next, we describe the four components of the modeling framework to evaluate both nonadaptive and adaptive management strategies in the IEUA service area. The components of the modeling framework are

the simulation model usedthe specific management strategies evaluatedthe uncertainties considered as a basis of the scenariosthe metrics used to evaluate the performance of each management strategy.


Faced with uncertainty, decisionmakers often must choose between early action and a wait-and-see approach. For instance, the skipper of the trailing boat on this slide faces a choice: Tack now and stay close behind the leaders, or hold course, diverge from the others, and perhaps gain the lead with a more favorable wind. Each choice offers uncertain costs and benefits.

In this chapter, we describe an analytical approach to assist IEUA managers in assessing the costs and benefits of their alternative early-action and wait-and-see options.

Evaluating Adaptive Management Strategies 19

As a base for our analysis, we considered the IEUA’s 2005 Regional Urban Water Manage-ment Plan (IEUA, 2005). As described in Groves, Knopman, et al. (2008), the plan’s pri-mary approach to addressing new water-supply needs includes increasing the use of the Chino Groundwater Basin by about 75 percent through increased surface-water replenishment (some-times referred to as replenishment of the groundwater) and developing an extensive recycled water system to deliver up to 69,000 acre-feet (af) of supply by 2025.


In addition to implementation of the various water-management strategies described in the RUWMP, the IEUA region is both pursuing and considering other options.

IEUA anticipates continued reductions in the intensity of water use in new construction as the region’s cities develop new landscape ordinances by 2010 to comply with California Assembly Bill 1881 (signed by the governor in September 2006). Landscape ordinances have the potential to dramatically limit outdoor water use through the restriction of landscaped area and mandates to use water-efficient irrigation techniques and practices. IEUA could also achieve greater demand reductions through the expansion of programs designed to retrofit existing homes and commercial buildings with water-saving devices, such as upgraded toilets and urinals (Gleick et al., 2003). Finally, the region could follow the lead of such cities as Las Vegas, which has begun to pay water users to remove highly water-using landscape (such as turf).

IEUA is also considering a more aggressive roll-out of its recycled-water–use system. Spe-cifically, in the summer of 2007, IEUA developed a business plan to increase the annual avail-able supply to 49 taf by 2010, which represents a speed-up of the RUWMP expansion schedule by five years.

Finally, there are several options for increasing use of the Chino Groundwater Basin supply. First, IEUA is involved in negotiations with Metropolitan to increase the size of the dry-year yield (DYY) program from 33 taf annual yield to 50 taf annual yield. The DYY pro-gram enables IEUA to forgo the use of groundwater during periods of increased availability of imported supplies and then increase groundwater use during periods of low imported-supply availability. Next, the Chino Basin Watermaster is working with the California Department of Health and Safety to change the Chino Groundwater Basin replenishment-blending require-ment to allow a higher percentage of recycled water to be percolated into the groundwater


basin. Currently, only about one-third of the replenishment water may be recycled water; the rest must be made up of local storm water or imported surface supplies. It may be possible to increase this percentage from 33 percent to 50 percent. Another option for increasing the effec-tive yield of the Chino Groundwater Basin is to increase the capture of storm water. Effective yield is the amount of water that can be withdrawn annually from a groundwater basin with-out causing significant drawdown of overall volume in storage. Increased storm-water capture would allow for more replenishment of groundwater, including increased capture of recycled water used in outdoor landscaping.


Long-term water management is inherently an adaptive process. Water planners typically review their long-term plans periodically and adjust them in response to changing and unpre-dictable conditions. When evaluating different policy or management choices, planners com-monly compare how fixed actions will perform in the future and thus determine which actions to commit to implementing in the near term. This approach can work well when the future is predictable. But under deep uncertainty, conditions are unlikely to play out as expected, and, thus, planners need to make adjustments to their earlier strategies. An analysis that fails to consider how decisionmakers will attempt to adjust to changing conditions can exaggerate the poor performance of previously established policies or neglect opportunities for near-term actions that can have significant and beneficial effects on future choices.

In the context of IEUA, we modeled adaptive strategies to evaluate which actions should be undertaken now and which can be put off for possible later action. We describe the details of the model’s implementation of these actions in the remainder of this chapter.


The RAND team developed an integrated decisionmaking framework to evaluate the perfor-mance of IEUA plans over a range of uncertain planning assumptions. The framework centers on a water-management model (WMM) developed using the Water Evaluation and Planning (WEAP) modeling environment (see SEI, undated). Each set of assumptions and specifica-tions of IEUA plans is evaluated using a set of outcome measures. Each of these components is described next.


The WEAP simulation model evaluates the performance of IEUA plans under specific scenarios, each of which reflects plausible trends in climate change and other planning assumptions.


The WEAP model simulates water supply and demand using a stylized representation of the system’s major water flows on a monthly time scale from 2005 to 2040. Building on the ver-sion used in our previous workshops (Groves, Knopman, et al., 2008), the model represents IEUA’s supply and demand relationships by using a set of nodes corresponding to discrete water-management elements, such as catchments, indoor-demand sectors, surface supplies, and groundwater basins. Rivers, conveyance facilities, and other pathways (such as percolation flows) link these elements together. Time series of monthly weather parameters are used to drive the system’s hydrology, using a soil-moisture model. The model specification was based on IEUA’s RUWMP, Chino Basin Optimum Management Program (Wildermuth Environ-mental, Inc., 1999), and other documents provided by IEUA, such as Husing (2006), and the model continues to be refined and developed as a part of this ongoing project. This slide shows the basic model schematic.

The overall model was calibrated and validated by comparing (1) the results for the simu-lation that corresponds to historical temperature and precipitation trends and management conditions and actions described in the RUWMP (e.g., the 2005 UWMP) with (2) the projec-tions of average water demand and supply provided by the RUWMP. Results from the calibra-tion, along with additional model details, are provided in Groves, Knopman, et al. (2008).

Key features of the WMM are described in the following sections.

Supplies

Major water sources to the region included the following:


precipitation over catchments: drives local supplies, Chino Groundwater Basin replenish-ment, and storm water for replenishment imports from Metropolitan: direct use (both “commitment” supplies based on existing agreements and additional amounts as needed and available) and replenishment (both for direct replenishment and in-lieu replenishment via the DYY program)non–Chino basin groundwater: fixed representation of non–Chino Groundwater Basin use by IEUA members.

Demands

Major water demands include the following:

Urban indoor: Household demand is based on projections of the number of households and per-household water use. Note that new development is specified to use less water than existing houses do. Indoor commercial and industrial demand is based on projec-tions of the number of employees and per-employee water use.Urban outdoor: Outdoor demand is estimated by WEAP’s rainfall-runoff model, which considers (1) the projected area of urban development, (2) the fraction of urban land that is irrigated landscaping, (3) monthly temperature and precipitation, and (4) parameters that define the irrigation needs of landscaping and infiltration of precipitation and irri-gation out of the root zone (Groves, Knopman, et al., 2008; Yates, Purkey, et al., 2005; Yates, Sieber, et al., 2005).Agricultural demand: Demand is assumed to come from irrigation based on similar parameters to those used to model urban outdoor demand. The WMM assumes a signifi-cant decline in agricultural area and water demand consistent with the 2005 UWMP.

Chino Basin Groundwater

The WMM also simulates Chino Groundwater Basin inflow and outflow. The WMM shows direct-use groundwater, desalted groundwater, the Chino Groundwater Basin replenishment program (using imported Metropolitan, recycled, and storm water), and the DYY program (modeled as a separate groundwater basin).

Recycling

The WMM reflects IEUA’s recycling program for both direct use and Chino Groundwater Basin replenishment. Recycled supply comes from treated wastewater from direct indoor water use. However, the amount of available recycled supply is specified exogenously in the model; that is, it is not directly linked to treated wastewater from indoor use. This simplification is immaterial, as the total amount of recycled supply made available is always less than the direct indoor water use.


The WEAP model incorporates temperature and precipitation time series reflective of histori-cal weather and plausible ranges of climate change. Historical weather sequences were devel-oped using historical daily observations of temperature, precipitation, and dew-point tempera-ture from 1980 to 2003 and monthly climatological wind-speed data for the region. Data were obtained for three locations representative of IEUA’s agricultural region, urban region, and upper watershed from the Daymet U.S. Data Center (NTSG, undated). We supplemented this information with monthly climatological wind-speed data for the region.

Claudia Tebaldi, a researcher at NCAR, provided us with probabilistic temperature and precipitation forecasts for the region of Southern California that covered the IEUA service area. These forecasts were based on 21 AOGCMs and a global carbon-emission scenario iden-tified as A1B in the IPCC Special Report on Emissions Scenarios (Nakicenovic, 2000). Tebaldi’s methodology (Tebaldi, Mearns, et al., 2004; Tebaldi, Smith, et al., 2005) uses a fully Bayesian approach to combine the predictions of the individual AOGCM forecasts into a single proba-bilistic forecast using the criteria of minimizing bias and maximizing convergence (i.e., the reli-ability ensemble average criteria from Giorgi and Mearns, 2002). The plots in this slide show the probability-density functions for summertime temperature and wintertime precipitation. The numbers on the curves represent the values corresponding to nine deciles of the corre-sponding cumulative distribution function (see Groves, Knopman, et al., 2008). The triangles indicate the actual trends that each AOGCM forecast reported that were used as the basis for the probability-density function.


AOGCMs generate forecasts at scales consistent with the 500 km2 “cells” of their model grids, which are considerably larger than the IEUA service area. Thus, the Tebaldi estimates were at too coarse a scale to be useful in the modeling of IEUA’s hydrology. For this reason, we needed to add another step in generating climate scenarios for IEUA, through a process known as “down-scaling,” bringing regional-scale estimates down to the more local scale of IEUA. Another collaborator, David Yates of NCAR, applied a climate-downscaling procedure called K-nn (Yates, Gangopadhyay, et al., 2003), to project the regional-scale climate changes on the local weather patterns as characterized by the historical data. The K-nn procedure resamples historical daily weather data so that the resulting sequences retain the same basic statistical character of the historical data but include a specified exogenous trend. For this analysis, the K-nn method was used to create 25 realizations for each of the nine deciles of precipitation and temperature change, producing a total of 450 time series.

The Bayesian climate-ensemble method does not provide information about the relation-ship between temperature and precipitation trends (known as covariance). A simple regression of historical annual temperature on precipitation suggests a statistically insignificant negative relationship. Looking forward, the evidence is mixed. Most climate models suggest that strong warming will deepen persistent drought in the subtropical region—IEUA is at the northern edge of this boundary. This effect suggests a strong negative relationship between tempera-ture and precipitation. However, if El Niño conditions are disproportionately strengthened as warming trends continue, then the region could experience more warm and tropical storms, which would be consistent with a positive relationship. So we assumed that the driest (wettest) precipitation decile corresponds to the warmest (coolest) temperature decile. This assumption is a conservative assumption, in that it will lead to the most inclusive set of plausible climate scenarios as a basis of testing the water-management plans of the IEUA region.


The K-nn resampling method does not guarantee that sequences with the appropriate precipitation trend will correspond to the desired temperature trend. We therefore adjusted the monthly temperature trends (using a growing linear additive factor) to meet the 2030-appro-priate trend specified by the cumulative distribution function (CDF).

The figures on this slide show three specific temperature (top) and precipitation (bottom) sequences drawn from the ensemble of series. The red sequence reflects one that exhibits strong warming and drying trends. The green reflects a moderate warming trend and no precipitation trend. Finally, the blue line reflects only slight warming and a slight precipitation increase.


The monthly temperature and precipitation time series affect estimates of outdoor demand, via a soil-moisture model. As described in Groves, Knopman, et al. (2008), the soil-moisture model estimates the irrigation requirements and the return flow for any precipitation and applied irri-gation water (e.g., evapotranspiration, surface runoff, subsurface runoff, deep percolation). For the urban catchment and agricultural catchment, the following parameters are specified: irri-gated land area, crop coefficient, soil water capacity, leaf-area index (parameter that controls the surface runoff response), root-zone conductivity, preferred flow direction (parameter that specifies the portion of water that infiltrates past the root-zone flows horizontally to adjacent rivers or vertically to the underlying aquifer), lower and upper irrigation threshold (parameters that control how frequently irrigation is used), and the initial soil-moisture amount.

This slide shows the historical annual precipitation from 1980 to 2003 projected forward (applying 1980 to 2005, 1981 to 2006, and so on) and corresponding urban outdoor and agricultural water demand. Note that this precipitation sequence is derived from the same historical sequence shown in the previous slide. The decline in agricultural water demand and corresponding increase in urban outdoor water demand are due to land conversion from agri-cultural to urban uses. The interannual variability in demand is directly related to annual pre-cipitation and temperature (not shown). During wet years, demand is low. During dry years, demand is higher.


The WMM simulates annual deliveries of the California SWP system as delivered by Metro-politan. The anticipated annual available supply is specified by the water-management strat-egy. For example, the 2005 UWMP projects a supply of 60.2 taf in 2005 to increase to 82.5 taf by 2025. The actual delivered supply in any simulation is affected by (1) the previous October–March precipitation totals for the IUEA region (dry years lead to reductions in Met-ropolitan imports as specified in the 2005 UWMP, and wet years permit additional deliveries); (2) the extent of the climate-change signal (as defined by the climate decile, explained below); and (3) an uncertain model parameter that specifies how much the Metropolitan supplies are reduced under climate change. Note that this scheme assumes a perfect correlation between precipitation in the IEUA service area and that in Northern California—the source area of the SWP. An improved representation would relax this assumption. The default sensitivity of Metropolitan supplies to climate change ranges between a 10 percent and 50 percent reduc-tion by 2030 (depending on the severity of the climate-change signal—treated as an exogenous parameter in the WMM). The top line in the bottom graph shows how annual deliveries of imports are affected under historical weather conditions. Note that two dry years (2024 and 2027) trigger dry-year import reductions (by an amount specified in the 2005 UWMP).

Local supplies, primarily surface runoff from the San Gabriel Mountains, comprise a relatively small portion (about 5 percent) of the total IEUA supply. This supply responds to the weather pattern specified for the Upper Watershed. The model is tuned so that, on average, 18.7 taf of local supply is available. During dry years, this amount decreases approximately proportionally to the percentage decrease in precipitation during those years. The lower line in the bottom graph shows the local supply response to a simulation using historical weather.


The WMM adjusts the Chino basin “safe yield” to achieve a balance between the inflows and outflows. Safe yield is the amount of water than can be withdrawn from a groundwater basin over several years without exceeding its ability to be replenished naturally during that time and is based on the Peace 1 Agreement described in Wildermuth Environmental (1999). The inflows to the Chino basin are percolation from the urban outdoor, agriculture, and native vegetation catchments and replenishment from local storm water, recycled water, and imports. The outflows include extractions of low-quality water that is desalted prior to delivery to end users in the region and direct-use pumping by IEUA members and a few outside agencies.

At 2010 and every five years thereafter, the WMM evaluates the amount of water in storage in the Chino basin (at the beginning of the simulation, storage is 5 million af). If the storage has dropped below this initial level, then the safe yield is reduced by one-half the amount of the previous deficit for the next five years. The safe yield is then recalibrated to reflect the new groundwater condition.


For the fourth workshop, we used the WMM to evaluate nine specific management plans for the IEUA region. The first five plans are nonadaptive, meaning that the plans do not change over time (though actions may change as specified by the plan). The next four are adaptive, meaning that they update as described in the slide.

The first plan, UWMP forever, is designed to represent implementation of the RUWMP without any updates over time. This is also the base strategy evaluated for the first three work-shops, as described in Groves, Knopman, et al. (2008).

The UWMP + DYY with recycling plan includes two additional management actions:1 (1) increase the size of the DYY program to 50 taf yield by 2010 and 75 taf yield by 2020, and (2) expand the recycled-water system to deliver 49 taf per year of supply to end users by 2010. This represents a speed-up of the expansion schedule in the 2005 UWMP by five years.

The UWMP + replenishment plan includes two additional management actions to the UWMP forever plan: (1) increase the allowable recycled-water content of Chino Groundwater Basin replenishment water from 33 percent to 50 percent, and (2) increase storm-water capture from 12 taf to 35 taf (an increase of 23 taf) by 2020.

The UWMP + efficiency plan increases the efficiency targets of the UWMP forever plan by 10 percent for indoor efficiency and 20 percent for outdoor efficiency by 2030.

The UWMP + all enhancements plan includes all the UWMP forever actions plus the additional actions included in the other three nonadaptive management plans.

1 When referring to plans, a plus sign (+) indicates that the change is implemented now; with indicates that the change will be implemented in the future.


The next four plans (UWMP with updates, UWMP + DYY and recycling with updates, UWMP + replenishment with updates, and UWMP + efficiency with updates) are similar to the first four plans except that they are adaptive.

Referring back to slide 16, we can now explain how the WMM handles adaptive strate-gies. The first step in developing adaptive strategies is to determine whether any near-term deci-sions should be made to augment the RUWMP. If the choice is no, then only the RUWMP actions are implemented (upper branch of diagram). In 2015 and every five years thereafter, the WMM then evaluates whether supply has been sufficient to meet demand over the preceding five years. If the average surplus is 10 taf lower than the anticipated surplus in the RUWMP, then adaptivity is triggered within the model. If the average surplus is sufficient, then imple-mentation of the RUWMP action continues as planned. Alternatively, the IEUA region may choose to implement some additional action now and then evaluate it every five years as before (lower branch of diagram in slide 16).

If adaptivity is triggered, then the model specifies that water-use efficiency will increase and that facilities to capture storm water for groundwater replenishment will be developed. The implementation rate is set to the rate it would be if the policy were enacted in the first period and completed by the end of the simulation. Thus, if adaptivity is triggered late in the simulation period, these actions will not be fully implemented by the end of the simulation. The adaptive strategies remain in effect even if conditions improve. We make this assumption because, in general, it is likely that it will be financially beneficial to continue these actions once they have been initiated.

In many jurisdictions, conditions could change such that development of local resources would lead to growing water surpluses. In these cases, adaptivity would trigger a scale-back of actions. In the IEUA case, however, preliminary analysis suggested that the RUWMP would unlikely be viewed as too aggressive because any local resource development would enable the region to reduce the use of its highest-cost water source—imports supplied by Metropolitan. More discussion of the costs of the various water-management actions can be found on slide 29.


The WMM evaluates each plan under specific assumptions. For this workshop, we consid-ered the eight assumptions listed in this slide. The ranges for these assumptions were derived from a variety of sources. The precipitation- and temperature-trend ranges reflect the ranges of trends specified in the weather-sequence generation procedure described in slide 21. The recy-cling-delay and replenishment-achievement ranges were derived from the participant survey results from the first workshop, as described in Groves, Knopman, et al. (2008). The range of new-development water intensity is derived from information received from Davis (2007). The effect of climate change on Metropolitan supplies derives from an internal review of sev-eral recent modeling studies that estimated the reliability of SWP deliveries under different climate-change scenarios (DWR, 2006; Vicuña, 2006; Zhu, Jenkins, and Lund, 2006). The range of basinwide percolation changes was specified by the authors and chosen to represent a modest range of values for purposes of the scenario analysis. The range of projected increase in the costs of imported supplies were specified by varying the assumed escalation rate (4.5 per-cent) used by Sunding (2007). Finally, the escalation in costs of improving efficiency is set to range between one-half and double the nominal increase of 4.5 percent per year.


The WMM produces a wide variety of outputs. In the first three workshops, we considered the probability of a shortage in any particular year and the average annual surplus as the two key metrics. We evaluated the performance of a plan played out in a particular future (set of planning assumptions). Although these metrics can be informative—clearly, plans that lead to few shortages are preferred over those that lead to large shortages—it is not clear how large surpluses ought to be before they are viewed as excessive. Based on the workshops and sub-sequent discussions with IEUA planners, it was determined that a better metric would be the discounted economic cost of providing water supplies and coping with any shortages under each future, as simulated by the WMM.

For the fourth workshop, we evaluated each future by annual outcome measures related to demand, supply, cost of provisioning supply, and financial effect of shortages from 2005 to 2040. We summarized the outputs by evaluating the present value (PV) of the costs of provid-ing supply from the perspective of the region’s water agencies to the end users and of incurring shortages.


As shown in this slide, the total cost for water in each simulation is the sum of the PV of the annual water-supply cost and the PV of the annual water-shortage cost. The water-supply cost is the volume of water used multiplied by its average cost. The water-shortage cost is the volume of unmet supply multiplied by the cost of shortages. For the analysis presented here, we use a discount rate of 4.5 percent (the current risk-free long-term interest rate) to match that of a recent study regarding the net benefits of alternative Chino basin management programs by Sunding (2007).


We worked with IEUA staff to estimate the cost of provisioning supply to end users. The model considers levelized costs from the perspective of the region’s water agencies. Levelized cost is the total cost to an agency of providing water divided by the total amount of water pro-vided to end users. The use of levelized costs assumes constant marginal costs within the range of use across the simulations. The costs of not having available supplies to provide are included as costs of shortages. The costs of reducing water demand through new efficiency programs or mandates are also specified. Each cost increases over time at a specific rate.

This slide presents the levelized cost in dollars per af for 2005 (dark bars) and 2015 (light bars) of shortages (orange bar), of each supply (green bars), and of demand reduction (blue bar) for base-case cost assumptions. Note that the cost increases for the different supplies are not uniform. Costs of imported supplies, in particular, increase at a much faster rate than the others. By 2015, the cost of tier 2 imported supply exceeds all others, including desalted groundwater.

Next, we describe the basis for each estimate of current costs and an escalation rate. Note that the WEAP model does not currently enable an analysis using marginal costs.

Shortages

The effective cost of incurring a supply shortage in the IEUA region is roughly equivalent to the recently established rate set by Metropolitan for delivering supply to an agency beyond its allocated amount during the current drought conditions. This rate is set to be twice the treated


tier 2 water rate, or roughly $1,200 per af.2 This effective cost is projected to rise 4.5 percent per year.

Chino Basin Groundwater (Direct Use)

We used an estimate of the cost of pumping Chino basin groundwater for direct use by munic-ipalities and agriculture to be about $130 per af (Campbell, 2007). This cost also includes facil-ity maintenance and escalates at 4.5 percent per year.

Desalted Groundwater

The cost of water supplied by the Chino basin desalters is $550 per af, which includes costs of filtration of $50 per af, according to the Chino Basin Desalter Authority (Atwater, 2007). These costs are expected to rise by about 4.5 percent per year.

Imported Supply

Costs of imported supply (provided by Metropolitan) are based on current 2008 imported-water rates ($341 per af for tier 1 and $437 per af for tier 2). The cost of these supplies is held constant until 2008, at which point they rise about $100 per af per year, about the rate increase expected in Metropolitan’s forthcoming updated long-range financial plan (Atwater, 2007):

2009: $450 per af (tier 1) and $560 per af (tier 2)2010: $550 per af (tier 1) and $670 per af (tier 2)2011: $650 per af (tier 1) and $785 per af (tier 2).

After 2011, costs rise between 2.5 percent and 8 percent per year, as specified by an uncer-tain parameter in the model.

Groundwater Replenishment

Costs of replenishment are disaggregated into costs for the replenishment facilities, the replen-ishment water, and the cost of spreading. Costs for the facilities are estimated to be about $94 per af (Campbell, 2007). Costs for spreading range from $15 to $50 per af (depending on hydrology); we used $30 per af as an average cost. Costs for replenishment supplies are as follows:

imported by Metropolitan: $238 per af plus $13 per af IEUA administrative feerecycled: $63 per aflocal storm water: $0 per af.

2 Metropolitan charges its customers two different rates for direct-use imported supply. The tier 1 rate applies to a specific amount of water delivered (59.7 taf per year for IEUA). Deliveries above the tier 1 amount are charged a higher tier 2 rate.


Recycled Water

The cost of recycled supply is specified to be the sales price of the water by IEUA to its member agencies, which then distribute it to end users: $60 per af now, increasing to $76 per af (2012) and $153 per af (2036) (Shaw, 2007). This relatively low cost for recycled water reflects signifi-cant state and federal cost sharing and grants for capital costs.

Water Saved Through Efficiency

Formal estimates for the cost of saving water through efficiency in the region were not avail-able. We worked with IEUA staff members to develop estimates that were consistent with the other cost estimates and also reflective of the costs that the region’s agencies faced. We did not estimate a total social cost of implementing efficiency. Instead, we developed estimates of the costs that the agencies within the region would bear to induce a unit of water savings. There would be other, potentially significant, costs borne by the region’s developers and water users to meet efficiency mandates or program requirements that are not included in the model. Using program budget information and estimates of efficiency savings provided by IEUA, we estimated a base-case cost of efficiency to be $43 per af. These costs increase over time as a function of the amount of efficiency savings induced (having more aggressive saving targets induces greater costs) and an uncertainty parameter. For example, efficiency costs in 2040 for a 15 percent savings by 2040 range between $200 per af and $400 per af.


The next several slides demonstrate the basic results from several specific model simulations. In the upper left chart on each slide, we show the annual precipitation used for each simulation. In the lower left, we show the computed supply mix for the simulation. The supply categories are (from bottom to top): Chino groundwater, non-Chino groundwater, recycled water, local supply, imports, DYY, unneeded supply, and shortage). The right graph shows the correspond-ing total PV of discounted costs for the simulation (including shortage costs). Note that this measure reflects both the relative use of the different supplies as shown in the supply-mix graph and the per-unit cost of the water as shown in the previous slide.

This slide shows the projected performance of the UWMP forever plan under historical climate conditions and base-case planning assumptions. Note that there are no shortages or costs associated with shortages and that a sizable surplus (unneeded supply) is projected during most years (as suggested by the 2005 UWMP). The significant expansion of the recycled-supply program limits the amount of imported supply required in this simulation. The total cost of this strategy and the climate scenario as simulated is about $3 billion.


This slide shows the projected performance of the UWMP forever plan under a set of climate conditions in which precipitation is lower than average and declining (adverse conditions). All other planning assumptions remain the same as in the previous slide. In this case, significant shortages occur in 2030 and beyond. Also, the amount of Chino groundwater is adjusted downward several times—first, very slightly in 2015, again more dramatically in 2020, and again in 2030. Note that, during the early years, there are sizable surpluses, and the amount of imported water use is low (as recycled supplies increase). Once groundwater use is curtailed due to lower infiltration and lower replenishment, use of imported supplies rises to the maxi-mum available amount. The DYY program is used around 2020 and again in 2030, although it runs out of supply by 2034, as conditions continue to deteriorate. Finally, notice that the total demand, the top of the surplus or shortage area, is greater under this simulation than in the previous one; this is due to the hotter and drier conditions in this case. The cost of this strat-egy and the climate scenario as simulated exceeds $4 billion, largely due to the costs associated with shortages.


This slide shows exactly the same conditions as those in the previous slide but for the UWMP with updates plan. In this case, adaptation dramatically improves the performance of the UWMP. Specifically, because conditions are dry at the beginning of the simulation, adaptiv-ity is triggered early (in 2015). This then initiates significant improvements in efficiency and storm-water capture such that, by 2030 (when shortages were experienced in the previous sim-ulation), demand is reduced considerably and groundwater use is less curtailed due to increased replenishment. The total cost of this adaptive strategy and the climate scenario is much smaller than in the previous case, because IEUA’s costs of efficiency and additional storm-water cap-ture are estimated to be low.

45

CHAPTER FOUR

Comparison of Options

We now use the WMM to compare how different water-management options (or plans) per-form across the wide range of possible future conditions, to better understand which actions would be most desirable for the agencies in the IEUA region to take now.


As discussed earlier, traditional analytic decisionmaking approaches would evaluate IEUA’s options by first characterizing the uncertainty the agency faces as a prelude to suggesting its best course of action. This slide shows the basic structure of such an analysis, common in water planning. A standard decisionmaking analysis begins by structuring the problem, by develop-ing a computer model that estimates how particular decisions (or strategies) will perform. If there is uncertainty about how a strategy will perform, these uncertainties are next character-ized with probability distributions over the inputs to the model. Each option is then evaluated and ranked according to one or more decision criteria. The top-ranked option (assuming a weighting for multiple-criteria problems) is often interpreted as optimum. Finally, a sensitivity analysis is performed to assess how the optimum strategy would perform under different com-binations of uncertain conditions. Such predict-then-act analyses are most useful when water planners can characterize the uncertainty with sufficient reliability that the sensitivity analysis does not significantly change the optimum strategy. Under conditions of deep uncertainty, the uncertainty is not so reliably characterized, and it may thus be impossible to confidently choose a single optimum strategy.

Comparison of Options 47

RDM is a new approach to supporting decisions under such conditions of deep uncertainty. It seeks to identify strategies that perform reasonably well, compared with the alternatives, over a wide range of plausible futures (Groves and Lempert, 2007; Lempert and Collins, 2007; Lem-pert, Popper, and Bankes, 2003). We have found in our previous work that RDM often char-acterizes uncertainty and options for managing it in ways that decisionmakers find credible, because they often naturally seek robust strategies. RDM can also contribute to contentious political debates, because parties to a decision can agree on near-term actions without having to agree on expectations about the future.


An RDM analysis begins, similarly to a standard probabilistic analysis, by structuring the prob-lem. But instead of next characterizing the uncertainties as a prelude to ranking the decision-making strategies, an RDM analysis first proposes alternative strategies and then characterizes the uncertainties according to their effect on the choice among these options. To accomplish this, the RDM analysis begins with a set of alternative strategies and then assesses each over a wide range of plausible futures. Each future is defined by a particular combination of values for the uncertain input parameters to the model. The analysis then uses statistical procedures to determine those combinations of uncertainty parameters most important to the choice among the strategies. For instance, we might find that, if future precipitation levels drop below a cer-tain threshold in the IEUA region, the agency might consider augmenting its current UWMP. (As described later in this chapter, what we actually find is a bit more complicated.) Once the RDM analysis has isolated the combinations of uncertainties most important to comparing the policy options, it identifies the key trade-offs among the most promising strategies. At this stage, RDM may employ probabilistic information to illuminate such trade-offs. For instance, the analysis might suggest that decisionmakers choose one strategy over another as long as future precipitation was sufficiently likely to remain above some threshold, helping the deci-sionmakers decide whether this is a risk they are willing to take.

Ideally, an RDM analysis helps identify a robust strategy, one that performs reasonably well compared with the alternatives over a very wide range of uncertainties. If none of the strategies initially considered emerges as robust, the RDM process shown in this slide can be repeated, using the information about key uncertainties and trade-offs to suggest new and potentially more robust strategies.

In many cases, this type of analysis can be very helpful, because it allows decisionmakers to identify and agree on a robust strategy, even if the uncertainties are very large. However,


the success of an RDM analysis often depends on the ability to find strategies that have this robustness property. If no such strategies exist,1 the results of an RDM analysis may not prove particularly useful or satisfying to decisionmakers. But even when no robust strategy exists, the RDM analysis may provide some guidance to decisionmakers on where they need to invest in research to reduce key uncertainties or find ways to expand their options.

1 Lempert and Collins (2007) described some conditions under which robust strategies do not exist.


We began this phase of our analysis by evaluating the performance of the UWMP forever plan under the entire set of weather sequences derived from the AOGCM data (225 runs = 25 runs times 9 deciles).

This slide shows the average annual unmet demand over the 35-year simulation period for each climate-change decile; the leftmost bar is associated with strongly decreasing precipi-tation trends. As expected, the average annual unmet demand for the UWMP forever plan is greatest for the weather sequences that exhibit the largest precipitation decline. There are no shortages at all for the top third of the climate sequences (those with increasing precipitation trends, on average).

These results suggest that subsequent analysis of vulnerabilities of the region’s plans is best focused on those weather sequences that exhibit decreasing precipitation trends. As increasing precipitation trends will only increase local and imported supplies while suppressing demand for outdoor irrigation in the model used for the analysis, there would be no vulnerabilities to probe for such scenarios.


We next evaluated the WMM and various plans under a wide range of planning assumptions and possible weather sequences (with a focus on those futures with declining precipitation—deciles 1–4). For the next several slides, we present histograms of the PV of total cost, in bil-lions of dollars, for 200 individual runs of the WMM. We use a Latin hypercube experimental design (i.e., a quasiuniform design) across the model parameters associated with the assump-tions described in slide 26 (except that we restrict the range of weather sequences to only those from deciles 1 to 4). Note that the PV measure includes both the cost of providing supply to the end users and the cost of incurring any shortages.

This slide shows a histogram of PV total costs for the UWMP forever plan. For reference, we also indicate the result for the UWMP forever plan under historical weather conditions and nominal planning assumptions. The PV total costs for most of the 200 simulations are greater than those under historical weather and nominal planning assumptions (to the right of the arrow in the graph).


We next compare alternative plans to the UWMP forever plan. This slide shows the UWMP forever plan (in blue) as in the previous slide but also overlays a histogram representing the results for the UWMP + DYY and recycling plan (in gold). Note that there are fewer simula-tions in which PV total costs is $4.5 billion or more for the UWMP + DYY and recycling plan. These results indicate that adding these actions now reduces the effect of unfavorable climate and planning assumptions on the UWMP forever plan.


This slide compares the UWMP forever (blue) plan with the UWMP + replenishment (gold) plan for 500 futures. The UWMP + replenishment plan reduces even more high-cost outcomes than the UWMP forever or the UWMP + DYY and recycling plans.


This slide compares the UWMP forever (blue) plan with the UWMP + efficiency (gold) plan for 200 futures. The UWMP + efficiency plan reduces even more high-cost outcomes than have the other plans evaluated so far.


This graph compares the UWMP forever (blue) plan with the UWMP + all enhancements (gold) plan for 200 futures. The UWMP + all enhancements plan reduces the most high-cost outcomes than have the other nonadaptive plans evaluated so far.


We showed that augmenting the UWMP forever plan with all enhancements reduces the number of high-cost outcomes. Although the PV total cost metrics include what IEUA believes to be the major financial costs of implementing these programs, there are other costs and bar-riers not captured by the model. For example, a significant amount of coordination across the cities and agencies will be required to speed up the expansion of the recycling program and increase the amount of storm water that can be captured and used for Chino basin replenish-ment. Planners in the IEUA region are thus keenly interested in what specific risks are reduced by implementing additional management programs.

Following the RDM approach, we specify a threshold output value above which the PV total cost shows poor performance. In this case, the value is set at $3.75 billion in PV total costs (about 20 percent higher than would be incurred under the UWMP forever plan and nominal planning assumptions and also the upper level of costs for the UWMP + all enhance-ments plan). We then ask what factors cause the UWMP forever plan to incur higher costs than these.

Following the scenario-discovery procedure outlined in Groves and Lempert (2007) and Lempert, Groves, et al. (2006), we use a cluster-finding algorithm called the patient rule induc-tion method (PRIM) (Friedman and Fisher, 1999) to find and characterize clusters in the data-base of simulations that represent management conditions in which the UWMP forever plan performs poorly.

PRIM is a data-mining algorithm designed to generate a set of low-dimensional “boxes” of high-dimensional data containing regions in which the value of a particular function is large (or small) compared with its value outside these boxes. PRIM is particularly useful for identifying scenarios, because it aims to optimize both the classification accuracy of the boxes (the percentage of large- or small-function values they contain) and the interpretability of the


boxes (the simplicity of the rules needed to define them). We implement PRIM using publicly available software that inputs a data set (which can be the output of a model run over many combinations of input values) and a criterion for cases of interest. The algorithm yields as out-puts the descriptions of several alternative low-dimensional regions, or boxes, that contain a high density of and span a high proportion of the interesting cases.

The PRIM algorithm yielded three key factors that lead the UWMP forever plan to per-form poorly: (1) large declines in precipitation (deciles 1 and 2), (2) strong climate effect on imports (greater than a 22 percent decline in average annual imports by 2040), and (3) strong reductions in Chino basin percolation (greater than 3.8 percent decline). Simulations in which these conditions are simultaneously achieved explain 53 of the 116 (46 percent) high-cost out-comes for the UWMP forever plan. Note that 95 percent of the outcomes that include these three key factors are high cost.


In summary, we found that enhancing the UWMP with all measures now reduces significant risks due to declining precipitation, strong effects of climate change on imports, and greater-than-anticipated declines in Chino basin percolation. Note that these are similar to the factors found in the analysis in the third workshop (Groves, Knopman, et al., 2008)—strong precipi-tation declines, low achievement of recycling goals, and greater-than-anticipated declines in Chino basin percolation.

These results suggest the following question: “Could the region wait until later to act?” To address this question, we evaluate a series of adaptive plans in the same manner as we evaluated the nonadaptive plans.


In this slide, we evaluate the risks of the UWMP with updates plan versus the UWMP + all enhancements plan. When we add adaptivity to the UWMP forever plan, there are fewer high-cost outcomes across the various futures, and those futures that lead to costs that are above the $3.75 billion threshold are characterized by slightly different factors. The UWMP with updates plan is still vulnerable to large declines in precipitation (decile 1 or 2) and strong climate effects on imports (declines greater than 25 percent by 2040), but, instead of being vul-nerable to declines in Chino basin percolation, it is more vulnerable to higher water demand (water intensity in new developments being less than 25 percent lower than that of current development). The UWMP with updates plan is less vulnerable to declines in groundwater infiltration because of the increases in groundwater replenishment that occur as an adaptive action in response to deteriorating conditions. High-cost outcomes, then, are more frequently caused by futures in which demand is higher than expected. Note that these risk factors explain 23 of the 33 high-cost outcomes (67 percent); 35 percent of all outcomes described by these risk factors are high-cost outcomes.


In this slide, we evaluate the risks of the UWMP + DYY and recycling with updates plan versus the UWMP + all enhancements plan. The results are very similar to the UWMP with updates plan, though the previous vulnerability to low water demand is replaced by vulner-ability to high costs of imports. Specifically, UWMP + DYY and recycling with updates plan is vulnerable to large declines in precipitation (decile 1 or 2) and strong climate effects on imports (declines greater than 33 percent by 2040), but, instead of being vulnerable to declines in Chino basin percolation or water demand, it is more vulnerable to higher costs of imports (via Metropolitan). The fact that a key factor is now related not to the occurrence of shortages but simply to higher costs of delivering water suggests that vulnerabilities to this plan are now related to situations in which investments in more local resources (as is done in the UWMP + all enhancements plan) pay off because they allow the region to reduce expensive imported sup-plies. A key vulnerability of a less aggressive plan (e.g., the UWMP + DYY and recycling with updates plan) is that costs of imports increase even faster than anticipated, leading to higher total costs. Note that these risk factors explain eight of the 23 high-cost outcomes (35 percent) and that 24 percent of all outcomes described by the risk factors are high-cost outcomes.


In this slide, we evaluate the risks of the UWMP + efficiency with updates plan versus the UWMP + all enhancements plan. Both of these plans do well, suggesting that increasing effi-ciency now can avoid most of the high-shortage cases in later years.


In summary, these results suggest that, if current IEUA regional leadership expects that future water managers and city planners will respond to decreasing reliability, some actions may be deferred. For example, if the region implements the RUWMP with increased efficiency now and increases storm-water replenishment if surpluses decline too much (e.g., implements the UWMP + efficiency with updates plan), it will face similar risks to those it would face if it implemented all actions now. However, if the region were to implement the RUWMP and rely on updates, it would still be vulnerable to significant precipitation decline, reductions in basin percolation, and reduction in imports due to climate change.


The results shown so far have aggregated the 200 model runs. We now show comparisons of different plans using scatter plots, which show the results of the individual simulations as single points on a graph. Such scatter plots may provide more useful information to decisionmakers but at a cost of increased complexity. The x-axis indicates the PV of the shortage costs, and the y-axis indicates the PV of the supply costs. Results in the lower-left corner, for example, have low shortage costs and low supply costs. Results in the upper-right corner have high shortage and supply costs. The sum of the PV shortage costs and PV supply costs are the PV total costs shown in the histograms in the earlier slides.

This graph shows the results of the UWMP forever plan in two scenarios. Scenario A exhibits modest warming and minimal precipitation decrease, and scenario B exhibits signifi-cant warming and precipitation decrease. Note that the UWMP forever plan leads to no short-age costs in scenario A but PV shortage costs of about $2 billion in scenario B. Scenario B also leads to higher supply costs.


This graph is the same as the graph on the previous slide except that it includes the results for the UWMP forever plan in 198 other scenarios. Note that the histogram on slide 38 summa-rizes these same results. For scenarios that experience no shortages (e.g., PV shortage costs are 0), the PV supply costs range from around $2.75 billion to $3.5 billion. Shortage costs, on the other hand, range from $0 to more than $3 billion for a few extreme cases.


We now show comparisons of the UWMP forever and the UWMP + DYY and recycling with updates plans using the scatter plots. We overlay a horizontal and vertical line to help emphasize how the patterns of results differ for the two plans. For example, 63 percent of the scenarios lead to more than $500 million in PV shortage costs for the UWMP forever plan, whereas only 16 percent of the scenarios lead to more than $500 million in PV shortage costs. About 67 percent of the scenarios lead the former plan to have PV supply costs greater than $3 billion, compared to only 38 percent for the latter plan. Thus, including future updates and implementing the DYY + recycling plan’s additional actions reduces both shortage and supply costs.


This slide shows the results from the previous slide (upper two graphs) as well as two more plans: the UWMP + efficiency with updates plan and the UWMP + all enhancements plan. These two new plans significantly reduce the risk of adverse climate and other planning condi-tions. It is interesting to note that there is still significant variability in supply costs (as seen by the vertical spread of the points along the y-axis). This is expected as, during dry years, under either plan, the region is likely to rely on expensive imported supplies. During less dry years, the additional local resource development leads to lower supply costs as the region substitutes away from imports toward local resources, such as recycled supply, storm water, and efficiency savings.


The previous slide suggests that early action reduces costs due to shortages and supply-delivery costs overall. A common concern with water-resource development is that, under some condi-tions, the additional development will be unneeded or go unused. These so-called stranded assets then lead to higher costs than would have been experienced under plans with less-aggressive infrastructure investment.

To evaluate this concern, we calculate the difference in PV total cost between the UWMP forever plan and the best-performing strategy for one-fourth of the 200 runs evaluated pre-viously. The graph shows these results plotted against the total shortage under the UWMP forever plan. The symbol shape indicates which plan is the best performer. Note that the best-performing plan is either the UWMP + efficiency with updates or the UWMP + all enhance-ments. The dots aligned along the y-axis of the graph correspond to simulations in which there are no shortages in the UWMP forever plan. Even under these cases, an enhanced plan would perform better. For these cases, the best-performing plan would reduce PV total costs by between about $150 million and more than $400 million. In summary, overinvesting is unlikely to be a significant concern to the IEUA region, provided that precipitation trends decrease or stay the same over the next 35 years. More analysis is needed to ensure that overin-vestment would not occur in scenarios in which precipitation trends increase.


We now look at the remaining risk for a selection of the plans. The four plans presented in this slide are the same as those presented in slide 53. The red diagonal line indicates the $3.75 bil-lion threshold shown in slides 55 and 56. Recall that the sum of the x-axis and y-axis values equals the PV total costs. Dots to the right of the threshold line indicate those simulations with high costs.

The plot shows that, similarly to what we saw in the histograms, as adaptivity and addi-tional near-term local resource-development activities are implemented, the number of cases with high costs decreases significantly.


This slide presents the same key risk factors for each plan identified and described in earlier slides. We indicate with red triangles those simulation results that are characterized by the key risk factors. In this view, one can see that not all bad cases are explained by the key risk factors and that some cases classified by the risk factors do not result in costs that exceed the threshold. For the former case, one may also seek to develop an additional characterization of results that exceed the cost threshold but do not arise from the key risk factors identified here.


To summarize our results and frame a discussion of their implications, we graph the number of outcomes that exceed the $3.75 billion cost threshold for each IEUA region plan. We also include the text description of those factors that lead to many of the high-cost outcomes.

The key findings of this study can thus be summarized as follows:

The IEUA region faces risk from decreasing precipitation trends and other adverse plan-ning conditions.Under the UWMP forever plan (e.g., simply implementing the actions described in IEUA’s RUWMP), there are numerous plausible future conditions in which both shortage and supply costs are higher than would be the case with more-aggressive local resource devel-opment. These high-cost outcomes are characterized by cases in which there are large declines in precipitation, strong effects of climate on imports, and moderate reductions in Chino basin percolation over the simulation period (2005–2040).Out of eight alternative plans that increase local resource development in the near term and in response to degrading conditions, the plans that do more lead to greater reductions in high-cost outcomes (the sum of discounted delivery and shortage costs).The key factors generating the risks change modestly as the region implements more local resources. For example, the region’s vulnerability to reductions in Chino basin per-colation is reduced with updates and more wastewater replenishment. Vulnerabilities to higher costs of imports and the underlying water demand in the region become more important as the region increases its reliance on local resources.The UWMP + efficiency with updates and the UWMP + all enhancements plans per-• form similarly well in reducing high-cost outcomes.


Major Implications and Recommendations

These findings suggest that the IEUA region can reduce its exposure to adverse climate-change effects by taking advantage of the favorable economics of local resource development in the region. As long as the local resource-development activities (including increasing efficiency) are less costly than imports and precipitation trends under climate change are flat to strongly decreasing, there is little risk of overinvestment. We have demonstrated that preparing the region to adapt in the future as climate-change effects and other planning uncertainties become more apparent can be very effective and, in some cases, as effective as developing more resources now. One promising well-hedged strategy would be to implement the actions within the RUWMP while increasing the region’s commitment to indoor and outdoor water-use effi-ciency. Implementing these actions, as well as studying the options for increasing storm-water capture for groundwater replenishment in coming decades, would represent a sensible and sig-nificant response to the threat of climate change.

73

CHAPTER FIVE

Survey Results

We administered surveys to the workshop participants at the beginning (before the presenta-tion) and at the end (after the presentation) of the workshop. We compared the responses at the two points in time to assess how participants’ views changed in response to the material presented in the workshop.

Twelve of the participants in this workshop reported that their primary area of cur-rent professional activity was being managers or directors, four reported being planners, two reported being public officials or consumer representatives, and the other six reported being “other.” None was a professor or student.

Nine reported spending all or a lot of their time working on planning for water resources in Southern California. Three spend half their time, eight spend some or a little of their time, and three spend almost none of their time. One respondent declined to answer this question. When asked how familiar they were with the IEUA RUWMP, four reported that they were involved in developing it; two worked with it extensively as a planning tool or read it thor-


oughly; 10 were familiar with it, though not in detail; and eight were slightly or not at all familiar with it. Finally, six of the participants in this workshop had attended all three of the previous workshops, two had attended two of them, three had attended one workshop, and 13 had not attended any of the prior workshops.

The next set of slides presents some findings from the survey results.

Survey Results 75

Surveys before and after the presentation were administered to 24 workshop participants (six of the respondents had attended at least one previous workshop). Because of the small number of individuals sampled, results cannot be viewed as having statistical significance. Instead, we report these data as notional findings, suggestive of hypotheses that we may choose to test under more-rigorous and -controlled conditions at a later date.

Both the pre- and postworkshop surveys asked several questions about how serious cli-mate change may be, how likely it was for the serious effects on the IEUA region, and how well equipped IEUA-region planners were to manage the effects of climate change. Of the 24 par-ticipants, about half thought that their view of the consequences of bad climate change and the likelihood of bad climate change were the same after the workshop as before, but 35 percent thought that consequences of bad climate change were more serious than before and 40 per-cent thought that the likelihood of bad climate change was greater than before. However, 75 percent changed their views in the positive direction about the ability of water planners at the IEUA to plan for and manage the effects of bad climate-change outcomes on the water supply. This seems to indicate that the RDM-based workshop may have made most of the participants more confident about managing the future effects of climate change in their region.

In summary, the comparison of the before- and after-workshop responses suggests that the briefing provided information that increased the participants’ concerns about the effect of climate change—both the likelihood of significant effects and the severity of the effects. At the same time, water planners increased their belief that they could mitigate or manage these effects. This is a noteworthy finding, as a primary objective of an RDM analysis is to illustrate possible threats and then identify strategies that can be taken to address these threats without having a clear understanding of exactly how the future may evolve. In this case, the survey results suggest that this objective was met.


The surveys also asked the participants to rate the importance of implementing various water-management actions in the near term. Most participants believed that most of the actions listed in the survey (see Groves, Lempert, et al., 2008) should be implemented now, including significant proposals, such as retrofitting houses with more-efficient water-using devices and gray-water or dual-use systems. For most proposed actions, participants reported only small changes in their willingness to implement most proposed actions due to the workshop—nearly every participant endorsed taking each action (that is, thought IEUA should be doing it now) at the beginning of the session, and only one or two additional participants thought that was the case by its end.

It is important to note that the workshop was presented at the end of one of the driest years on record for the IEUA region. It is probable that support for implementing many of these actions was due both to the general anxiety throughout the region that the drought would continue and to the longer-term effects of climate change. Also, as mentioned earlier, in the first workshop, in 2006, the surveys revealed significant concern that the region would not meet all its recycling or groundwater-replenishment goals. The analysis suggests that doing conservation now (for example) does hedge against some uncertainties, such as those surround-ing the implementation of the recycling and replenishment program, and this may be behind some of the support for additional near-term water-management actions.

Survey Results 77

The exception was the item “carry out outdoor landscape retrofits, including turf removal.” While only 44 percent before the workshop said to do it now, 85 percent after the workshop endorsed doing it now.

We also asked about how certain the participants were about their ratings of actions to be taken. We asked, “How sure are you that your choice is correct?” and asked them to rate their level of certainty on a 1 to 5 scale, where 1 = completely sure and 5 = not sure at all. The preworkshop means varied from 1.95 to 2.43. The postworkshop means varied from 1.59 to 1.82. Because of small sample size, we cannot say whether this result is statistically significant, but notionally, it seems to indicate greater levels of certainty after the workshop. This would be consistent with the hypothesis that RDM can help groups converge on the actions to be taken and be more able to undertake them.


We asked workshop participants to rate the usefulness of RDM on a variety of dimensions using a 1–4 scale, where 1 = agree strongly and 4 = disagree strongly. More than 60 percent of participants agreed strongly that RDM provided results that could be used in planning (75 percent), enabled comparisons of climate-related risks (60 percent), and presented information in an objective way (61 percent). Fifty-five percent agreed strongly that RDM provided results that support a choice among plans. Nearly everyone agreed strongly or somewhat that RDM was useful for these purposes. Note that, unlike the design of the previous three workshops, in this workshop, RDM was not being compared with another method.

Survey Results 79

However, we found that only 15 percent agreed strongly that RDM was easy to understand, 10 percent agreed strongly that it was easy to explain to stakeholders, and 35 percent agreed strongly that it presented sufficient information about climate change for planning purposes. Nonetheless, again, nearly all participants agreed strongly or agreed somewhat that RDM was useful for these purposes.


We presented three types of graphical displays about the results from RDM. These were histo-grams and scatter plots that summarized results of a large number of scenarios and the charac-terization of key risk factors. We asked participants using the same 4-point “agree strongly” to “disagree strongly” scale whether these were useful ways of presenting data. A majority agreed strongly that all three approaches were useful, and the majority of the rest somewhat agreed that they were useful.

81

References

AB—see California Assembly Bill.

Atwater, Richard, general manager and chief executive officer, Inland Empire Utilities Agency, email to David G. Groves, September 10, 2007.

California Assembly Bill 1881, Water Conservation in Landscaping Act, September 2006.

California Climate Change Center, and California Energy Commission, Our Changing Climate: Assessing the Risks to California, Sacramento, Calif.: California Climate Change Center, 2006. As of January 15, 2008: http://www.energy.ca.gov/2006publications/CEC-500-2006-077/CEC-500-2006-077.PDF

California Department of Water Resources, California Water Plan Update 2005: A Framework for Action, Sacramento, Calif.: State of California, Resources Agency, Department of Water Resources, 2005. As of February 22, 2008: http://www.waterplan.water.ca.gov/previous/cwpu2005/index.cfm

———, Progress on Incorporating Climate Change into Planning and Management of California’s Water Resources, Sacramento, Calif.: California Department of Water Resources, 2006. As of January 15, 2008: http://baydeltaoffice.water.ca.gov/climatechange/DWRClimateChangeJuly06.pdf

———, “DWR Releases Water Delivery Impact Estimates Following Wanger Decision,” advisory, Sacramento, Calif., December 24, 2007. As of January 15, 2008: http://www.water.ca.gov/news/newsreleases/122407wanger.pdf

Campbell, Andy, groundwater recharge coordinator, Inland Empire Utilities Agency, telephone conversation with David G. Groves, July 26, 2007.

Davis, Martha, manager of strategic policy development, Inland Empire Utilities Agency, telephone conversation with David G. Groves, August 22, 2007.

DWR—see California Department of Water Resources.

Friedman, Jerome H., and Nicholas I. Fisher, “Bump Hunting in High-Dimensional Data,” Statistics and Computing, Vol. 9, No. 2, April 1999, pp. 123–143.

Giorgi, Filippo, and Linda O. Mearns, “Calculation of Average, Uncertainty Range, and Reliability of Regional Climate Changes from AOGCM Simulations via the ‘Reliability Ensemble Averaging’ (REA) Method,” Journal of Climate, Vol. 15, No. 10, May 2002, pp. 1141–1158.

Gleick, Peter H., Dana Haasz, Christine Henges-Jeck, Veena Srinivasan, Gary Wolff, Katherine Kao Cushing, and Amardip Mann, Waste Not, Want Not: The Potential for Urban Water Conservation in California, Oakland, Calif.: Pacific Institute for Studies in Development, Environment, and Security, November 2003. As of January 15, 2008: http://www.pacinst.org/reports/urban_usage/waste_not_want_not_full_report.pdf

Groves, David G., Debra Knopman, Robert J. Lempert, Sandra H. Berry, and Lynne Wainfan, Presenting Uncertainty About Climate Change to Water-Resource Managers: A Summary of Workshops with the Inland Empire Utilities Agency, Santa Monica, Calif.: RAND Corporation, 2008. As of January 15, 2008: http://www.rand.org/pubs/technical_reports/TR505/

Groves, David G., and Robert J. Lempert, “A New Analytic Method for Finding Policy-Relevant Scenarios,” Global Environmental Change, Vol. 17, No. 1, February 2007, pp. 73–85.

http://www.energy.ca.gov/2006publications/CEC-500-2006-077/CEC-500-2006-077.PDF

http://www.waterplan.water.ca.gov/previous/cwpu2005/index.cfm

http://baydeltaoffice.water.ca.gov/climatechange/DWRClimateChangeJuly06.pdf

http://www.water.ca.gov/news/newsreleases/122407wanger.pdf

http://www.pacinst.org/reports/urban_usage/waste_not_want_not_full_report.pdf

http://www.rand.org/pubs/technical_reports/TR505/


Groves, David G., Robert J. Lempert, Debra Knopman, and Sandra H. Berry, Preparing for an Uncertain Future Climate in the Inland Empire: Identifying Robust Water-Management Strategies, Appendix: Workshop Questionnaires, Santa Monica, Calif.: RAND Corporation, DB-550-NSF, 2008. As of January 28, 2008: http://www.rand.org/pubs/documented_briefings/DB550/

Hansen, James, Makiko Sato, Pushker Kharecha, Gary Russell, David W. Lea, and Mark Siddall, “Climate Change and Trace Gases,” Philosophical Transactions of the Royal Society of London, Series A: Mathematical and Physical Sciences, Vol. 365, May 18, 2007, pp. 1925–1954. As of January 15, 2008: http://pubs.giss.nasa.gov/docs/2007/2007_Hansen_etal_2.pdf

Husing, John, Inland Empire Utility Agency Demand Drivers: 2004 to Build-Out, Fontana, Calif.: Inland Empire Utilities Agency, June 28, 2006. As of January 15, 2008: http://www.ieua.org/docs/Reports/JohnHusingReport.pdf

IEUA—see Inland Empire Utilities Agency.

Inland Empire Utilities Agency, 2005 Regional Urban Water Management Plan, Fontana, Calif.: Inland Empire Utilities Agency, 2005. As of January 15, 2008: http://www.ieua.org/docs/Reports/WFA_UWMP_2005/WFA_2005.pdf

ISE—see RAND Infrastructure, Safety, and Environment.

Lempert, Robert J., and Myles T. Collins, “Managing the Risk of Uncertain Threshold Responses: Comparison of Robust, Optimum, and Precautionary Approaches,” Risk Analysis, Vol. 27, No. 4, August 2007, pp. 1009–1026.

Lempert, Robert J., David G. Groves, Steven W. Popper, and Steven C. Bankes, “A General, Analytic Method for Generating Robust Strategies and Narrative Scenarios,” Management Science, Vol. 52, No. 4, 2006, pp. 514–528.

Lempert, Robert, Nebojsa Nakicenovic, Daniel Sarewitz, and Michael Schlesinger, “Characterizing Climate-Change Uncertainties for Decision-Makers,” Climatic Change, Vol. 65, No. 1, July 2004, pp. 1–9.

Lempert, Robert J., Steven W. Popper, and Steven C. Bankes, Shaping the Next One Hundred Years: New Methods for Quantitative, Long-Term Policy Analysis, Santa Monica, Calif.: RAND Corporation, MR-1626-RPC, 2003. As of January 15, 2008: http://www.rand.org/pubs/monograph_reports/MR1626/

Lund, Jay R., Ellen Hanak, William Fleenor, Richard Howitt, Jeffrey Mount, and Peter Moyle, Envisioning Futures for the Sacramento–San Joaquin Delta, San Francisco, Calif.: Public Policy Institute of California, 2007. As of January 15, 2008: http://www.ppic.org/main/publication.asp?i=671

Meehl, Gerald A., Warren M. Washington, Benjamin D. Santer, William D. Collins, Julie M. Arblaster, Aixue Hu, David M. Lawrence, Haiyan Teng, Lawrence E. Buja, and Warren G. Strand, “Climate Change Projections for the Twenty-First Century and Climate Change Commitment in the CCSM3,” Journal of Climate, Vol. 19, No. 11, 2006, pp. 2597–2616.

Mystic Seaport, Rosenfeld Collection, “International Crossing Tacks,” September 7, 1941. Image 102650F copyright Mystic Seaport, Rosenfeld Collection, Mystic, Conn.. As of January 25, 2008, via search for image name: http://www.rosenfeldcollection.com

Nakicenovic, Nebojsa, Special Report on Emissions Scenarios, Geneva, Switzerland: Intergovernmental Panel on Climate Change, 2000. As of January 15, 2008: http://www.grida.no/climate/ipcc/emission/

NRDC v. Kempthorne, 2007 U.S. Dist. LEXIS 91968, E.D. Cal., December 14, 2007.

NTSG—see Numerical Terradynamic Simulation Group.

Numerical Terradynamic Simulation Group, University of Montana, “Daymet Daily Surface Weather and Climatological Summaries,” undated Web page. As of January 15, 2008: http://www.daymet.org/

http://www.rand.org/pubs/documented_briefings/DB550/

http://pubs.giss.nasa.gov/docs/2007/2007_Hansen_etal_2.pdf

http://www.ieua.org/docs/Reports/JohnHusingReport.pdf

http://www.ieua.org/docs/Reports/WFA_UWMP_2005/WFA_2005.pdf

http://www.rand.org/pubs/monograph_reports/MR1626/

http://www.ppic.org/main/publication.asp?i=671

http://www.rosenfeldcollection.com

http://www.grida.no/climate/ipcc/emission/

http://www.daymet.org/

References 83

Oppenheimer, Michael, Brian C. O’Neill, Mort Webster, and Shardul Agrawala, “Climate Change: The Limits of Consensus,” Science, Vol. 317, No. 5844, September 14, 2007, pp. 1505–1506. As of January 15, 2008: http://www.sciencemag.org/cgi/content/full/317/5844/1505

Popper, Steven W., Robert J. Lempert, and Steven C. Bankes, “Shaping the Future: Given the Unavoidable Uncertainty of the Long-Term Future, a New Analytical Approach to Setting Policy May Help with Problems Such as Global Warming,” Scientific American, Vol. 292, No. 4, 2005, pp. 66–71.

RAND Infrastructure, Safety, and Environment, “Improving Decisions in a Complex and Changing World,” last modified December 13, 2007. As of November 14, 2007: http://www.rand.org/ise/projects/improvingdecisions/

SEI—see Stockholm Environment Institute.

Shaw, Ryan, email to David G. Groves, May 4, 2007.

Solomon, Susan, Climate Change 2007: The Physical Science Basis: Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge and New York: Cambridge University Press, 2007. As of January 15, 2008: http://www.ipcc.ch/ipccreports/ar4-wg1.htm

Stockholm Environment Institute, “Water Evaluation and Planning System,” undated Web page. As of January 15, 2008: http://www.weap21.org/

Sunding, David, “Report on the Distribution of Benefits to Basin Agencies from the Major Program Elements Encompassed by the Peace Agreement and Non-Binding Term Sheet,” Berkeley, Calif.: Berkeley Economic Consulting, Inc., 2007.

Tebaldi, Claudia, researcher, National Center for Atmospheric Research, email to David G. Groves, May 31, 2006.

Tebaldi, Claudia, Linda O. Mearns, Doug Nychka, and Richard L. Smith, “Regional Probabilities of Precipitation Change: A Bayesian Analysis of Multimodel Simulations,” Geophysical Research Letters, Vol. 31, 2004.

Tebaldi, Claudia, Richard L. Smith, Doug Nychka, and Linda O. Mearns, “Quantifying Uncertainty in Projections of Regional Climate Change: A Bayesian Approach to the Analysis of Multimodel Ensembles,” Journal of Climate, Vol. 18, No. 10, May 2005, pp. 1524–1540.

Vicuña, Sebastian, Predictions of Climate Change Impacts on California Water Resources Using CALSIM II: A Technical Note: White Paper, Sacramento, Calif.: California Energy Commission, 2006. As of January 15, 2008: http://digitalarchive.oclc.org/request?id%3Doclcnum%3A70928680

Wildermuth Environmental, Inc., Optimum Basin Management Program: Draft Phase 1 Report, Chino Basin Watermaster, August 19, 1999. As of January 15, 2008: http://www.cbwm.org/docs/engdocs/obmpphas1rep/Text/OBMP_Ph1_Report.pdf

Wilkinson, Robert, and David G. Groves, Rethinking Water Policy Opportunities in Southern California: An Evaluation of Current Plans, Future Uncertainty, and Local Resource Potential, Santa Barbara, Calif.: University of California, Santa Barbara, Donald Bren School of Environmental Science and Management, June 2006. As of January 15, 2008: http://g.icess.ucsb.edu/rtmime/1168389566-18330/Socal%20Project%20Report%20v07%2024%202006-FINAL.pdf

Yates, David, Subhrendu Gangopadhyay, Balaji Rajagopalan, and Kenneth Strzepek, “A Technique for Generating Regional Climate Scenarios Using a Nearest-Neighbor Algorithm,” Water Resources Research, Vol. 39, No. 7, 2003, p. 1199.

Yates, David, David Purkey, Jack Sieber, Annette Huber-Lee, and Hector Galbraith, “WEAP21—A Demand-, Priority-, and Preference-Driven Water Planning Model, Part 2: Aiding Freshwater Ecosystem Service Evaluation,” Water International, Vol. 30, No. 4, 2005, pp. 501–512.

http://www.sciencemag.org/cgi/content/full/317/5844/1505

http://www.rand.org/ise/projects/improvingdecisions/

http://www.ipcc.ch/ipccreports/ar4-wg1.htm

http://www.weap21.org/

http://digitalarchive.oclc.org/request?id%3Doclcnum%3A70928680

http://www.cbwm.org/docs/engdocs/obmpphas1rep/Text/OBMP_Ph1_Report.pdf

http://g.icess.ucsb.edu/rtmime/1168389566-18330/Socal%20Project%20Report%20v07%2024%202006-FINAL.pdf


Yates, David, Jack Sieber, David Purkey, and Annette Huber-Lee, “NWAP21—A Demand-, Priority-, and Preference-Driven Water Planning Model, Part 1: Model Characteristics,” Water International, Vol. 30, No. 4, 2005, pp. 487–500.

Zhu, Tingju, Marion W. Jenkins, and Jay R. Lund, Estimated Impacts of Climate Warming on California Water Availability Under Twelve Future-Climate Scenarios, Sacramento, Calif.: California Energy Commission, 2006. As of January 15, 2008: http://www.energy.ca.gov/2006publications/CEC-500-2006-040/CEC-500-2006-040.PDF

http://www.energy.ca.gov/2006publications/CEC-500-2006-040/CEC-500-2006-040.PDF

Documents

CIVIL JUSTICE EDUCATION 6 ENERGY AND ENVIRONMENT HEALTH AND HEALTH