1. California Water Demand Scenarios David Groves Pardee RAND Graduate School Scott Matyac and Tom Hawkins Department of Water Resources * * * California Water Plan Update Advisory Committee Meeting January 20, 2005

2. 2 - Groves- Jan 20, 05 Purpose of Analysis Provide preliminary quantitative estimates of 2004 Water Plan narrative water demand scenarios (not FORMAL estimates) Advance conceptual thinking on using a scenario approach for future Water Plans Create a scenario estimator for other non- Water Plan related analyses

3. 3 - Groves- Jan 20, 05 History of Collaboration Water Plan Staff and Advisory Committee embraced scenario approach for Water Plan phased work plan Concurrently, RAND began a scenario-based look at long-term water resources planning in California RAND and Water Plan staff now collaborating to create water demand scenario generator and quantify narrative scenarios RAND expects to continue to use and develop this model for further analyses

4. 4 - Groves- Jan 20, 05 Approach 1) Create model to generate plausible average 2030 water demand scenarios –Simple, understandable, fast running, and easily modifiable –Ability to mimic/incorporate results of detailed models –Specify scenarios through unique parameter values 2) Select parameter values congruent to narrative descriptions 3) Quantify, evaluate, and interpret projected water demand –Test for plausibility –Identify aspects needing further study –Gain insight 4) Follow-up with more detailed analysis

5. 5 - Groves- Jan 20, 05 A Few Words About Scenarios Scenarios are NOT predictions –No one scenario is expected to predict what will occur –Instead, they reflect multiple plausible views of the future Scenarios are useful when –Ability to predict is low due to large uncertainties –Individual outcomes are important Desire to avoid low probability, negative events Scenarios help analysts and decision-makers: –Evaluate uncertain potential outcomes –Generate new ideas for successful policies Scenarios should be evaluated together as a package

6. 6 - Groves- Jan 20, 05 The Water Demand Scenario Estimator Three Modules –Urban –Agricultural –Environmental Annual time step from 2000 to 2030 Disaggregated by Hydrologic Region

7. 7 - Groves- Jan 20, 05 Urban Water Demand Estimates based upon projections of water use by: –Households –Economic activity (based on employment) –Public activities (based on population) –Losses and intentional groundwater recharge Water use per demand unit varies –Water price, income, household size, naturally occurring conservation, and efficiency adoption Initialized using year 2000 data Can use detailed models (e.g. IWR-MAIN) for calibration

8. 8 - Groves- Jan 20, 05 Urban Water Demand Details Demand Units –Households Single- and multi-family Interior and exterior –Commercial Employees –Industrial Employees –Institutional Use (per capita)

9. 9 - Groves- Jan 20, 05 Population Changes Drive Housing and Employment SF and MF houses a function of: –Population –Fraction of population housed –Share of SF houses –Household Size Commercial & Industrial employees a function of: –Population –Employment rate –Share of commercial versus industrial jobs

10. 10 - Groves- Jan 20, 05 Agricultural Water Demand Estimates based on: –Projected future agricultural land use –Changes in crop water needs –Changes in water application technology and practices Uses results of other models for calibration –ETAW (initializing data) –CALAG (when ready)

11. 11 - Groves- Jan 20, 05 Irrigation Demand Calculated by Estimating Crop Demand IU = State-wide irrigation water use ICA = Irrigated crop area Irrigated Land Area + Multi-cropped Area AW = Required applied water per area by crop

12. 12 - Groves- Jan 20, 05 Irrigation demand changes over time IU changes in response to changes in: –Irrigated land area (ILA) –Multi-cropped area (MA) –Applied Water (AW) – improved varieties of crops, better irrigation methods or technology, change in weather –Cropping pattern – reflected in changing ICA by crop

13. 13 - Groves- Jan 20, 05 Rule-based Procedure for Agricultural Land Use Changes 1)Estimate state-wide changes Irrigated land area Multi-cropped area Irrigated cropped area 2)Apportion state-wide changes to hydrologic regions Some regions more apt to absorb changes than others 1998 CWP land-use forecasts for low change regions 3)Estimate crop mix changes for each hydrologic region Low value crops reduced more than high value crops Ratio of area multi-cropped over area with multi- cropping potential must remain within a specified range

14. 14 - Groves- Jan 20, 05 Environmental Demand Very rudimentary procedure –Based on year 2000 unmet needs (Environmental Defense) More complete treatment would incorporate variable hydrology

15. 15 - Groves- Jan 20, 05 Water Plan Narrative Scenarios Chapter 3, Table 3-1

16. 16 - Groves- Jan 20, 05 No single method for choosing numerical values for parameters There is no “correct” scenario Other modeling studies inform quantification –Ex: DOF demographic projections Important to quantify drivers independently of scenario results Check intermediate results for plausibility

17. 17 - Groves- Jan 20, 05 Scenario Specification (Urban 1)

18. 18 - Groves- Jan 20, 05 Scenario Specification (Urban 2)

19. 19 - Groves- Jan 20, 05 Scenario Specification (Agriculture)

20. 20 - Groves- Jan 20, 05 Statewide Results Increase in urban demand, decrease in agricultural demand Net demand differs by scenario Cannot offset state-wide urban increases with state-wide agricultural decreases!!! Ag water use reduction is largest in Current Trends due to specification of agricultural land use in narratives.

21. 21 - Groves- Jan 20, 05 Results (Illustrative and Preliminary) by Hydrologic Region Model initially developed to evaluate state-wide trends Disaggregating by HR difficult due to unique regional characteristics –Unique land use thresholds –Different economic forces Ideally, each HR would have own set of rules to constrain scenarios …. work in progress

22. 22 - Groves- Jan 20, 05 North Coast and North Lahontan NC: Ag demand reduction due to efficiency improvements NL: Ag increases due to increased irrigated land area

23. 23 - Groves- Jan 20, 05 San Francisco and Central Coast SF: Urban water demand increase [4% -> 32%] CC: Ag water use reductions due to reduction in irrigated land area. No changes in multi-cropping

24. 24 - Groves- Jan 20, 05 Sacramento River Urban water use increases greater than reductions in agricultural water use Large increase in multi-cropping leads to increases in ICA under all scenarios

25. 25 - Groves- Jan 20, 05 San Joaquin River Large agricultural demand reductions in Current Trends and Resource Efficient scenarios Large urban demand increases

26. 26 - Groves- Jan 20, 05 Tulare Lake Similar to San Joaquin without environmental water demand increases

27. 27 - Groves- Jan 20, 05 South Coast Agricultural land reduction drives reductions in agricultural water use Urban water demand increases overwhelm agricultural demand decreases except in Resource Efficient scenario

28. 28 - Groves- Jan 20, 05 Colorado River Large decreases in agriculture offset urban increases in Current Trends and Resource Intensive scenarios Increase in multi-cropping ranges from 20%-26%

29. 29 - Groves- Jan 20, 05 South Lahontan Urban demand increases are greater than agricultural demand reductions

30. 30 - Groves- Jan 20, 05 Closing Remarks Interpretation of Scenarios –NOT forecasts –Will be evaluated using more detailed analyses and models for 2008 Water Plan Additional scenarios can easily be generated and evaluated (stay tuned…) Improvements to scenario generator –Introduce interannual variability due to weather –Couple to water supply and management scenarios Reflect annual cycle Interannual hydrologic variability