Water and Climate: What's Changing, and Does It Matter to Water Managers? Dennis P. Lettenmaier Department of Civil and Environmental Engineering University of Washington for 2009 AAAS Annual Meeting Session on 21st Century Water: Friend or Foe? Chicago February 14, 2009

What are the “grand challenges” in hydrology? From Science (2006) 125th Anniversary issue (of eight in Environmental Sciences): Hydrologic forecasting – floods, droughts, and contamination From the CUAHSI Science and Implementation Plan (2007): … a more comprehensive and … systematic understanding of continental water dynamics … From the USGCRP Water Cycle Study Group, 2001 (Hornberger Report): [understanding] the causes of water cycle variations on global and regional scales, to what extent [they] are predictable, [and] how … water and nutrient cycles [are] linked?

Important problems all, but I will argue instead (in addition) that understanding hydrologic change should rise to the level of a grand challenge to the community.

From Stewart et al, 2005

Magnitude and Consistency of Model-Projected Changes in Annual Runoff by Water Resources Region, 2041-2060 Median change in annual runoff from 24 numerical experiments (color scale) and fraction of 24 experiments producing common direction of change (inset numerical values). +25% +10% +5% +2% -2% -5% -10% -25% Decrease Increase (After Milly, P.C.D., K.A. Dunne, A.V. Vecchia, Global pattern of trends in streamflow and water availability in a changing climate, Nature, 438, 347-350, 2005.) 96% 75% 67% 62% 87% 71% 58% 100% Slide 1 shows a subset of the information on slide 2. The notes are the same for both slides. The Water Resources Regions are colored according to the projected percent change in mean annual runoff for the period 2041-2060, relative to the period 1901-1970. These projections are model-estimated changes associated with hypothetical ("SRES A1B" scenario) changes in climate forcing. The printed value inside each region of the map (slide 2 only) is the majority percentage of the 24 experiments that are in agreement on the direction (increase or decrease) of change; for example, the value of 67 in the Texas Gulf region indicates that 67% (i.e., 16) of the 24 experiments projected a decrease in runoff. Actual future changes in runoff can be expected to differ from these projections, primarily because of departures of actual forcing from the SRES A1B scenario, errors in the models' representation of runoff response to climate forcing, and unforced variability ("randomness") of the climate system. The majority percentages (slide 2 only) should not be read as probabilities, but rather as a combined measure of two factors: the degree of agreement among models and the modeled strength of the forced runoff change relative to modeled internal variability of the climate system. In Water Resources Regions for which a strong majority of experiments agree on the direction of change (Alaska, Upper Colorado, Lower Colorado, and Great Basin), the models suggest that the forced runoff change by 2041-2060 will be large compared to unforced runoff variability. In other Regions, either (1) the forced runoff change will not be large relative to the model estimate of unforced variability, or (2) the forced runoff change will be large relative to unforced variability, but this fact is obscured by substantial differences in model errors from one model to the next. The color of a Region is determined only by changes in runoff produced inside the Region. However, where a downstream Region (e.g., the Lower Colorado or the Lower Mississippi) receives streamflow from one or more upstream Regions, the streamflow through the downstream Region will be affected by runoff changes in both the downstream and the upstream Regions. Thus, increasing runoff in the Upper Mississippi, Ohio, and Tennessee Regions implies increasing flow of the Mississippi River through the Lower Mississippi Region, even though the projected runoff change in the Lower Mississippi Region is small. The figure is based on figure 4 of Milly et al. (2005); that reference documents the computations in detail. The computational differences from the published figures are (1) the geographic scope here is limited to the United States; (2) instead of depicting changes in point values of runoff, this figure depicts only changes in areal averages of runoff over Water Resources Regions of the U.S. Water Resources Council; (3) the composite across experiments is formed from the median instead of the mean. The projected changes are median values over a set of 24 climate-model experiments conducted on 12 climate models. The number of experiments exceeds the number of models, because the experiment was run more than once on some models. The 12 models used were the subset of 23 (IPCC AR4) candidate models that best reproduced the global pattern of observed time-mean streamflow during the 20th Century. Reference: Milly, P.C.D., K.A. Dunne, and A.V. Vecchia, 2005, Global pattern of trends in streamflow and water availability in a changing climate, Nature, v. 438, p. 347-350.

Timeseries Annual Average PCM Projected Colorado R. Temperature Timeseries Annual Average ctrl. avg. hist. avg. Period 1 2010-2039 Period 2 2040-2069 Period 3 2070-2098

Timeseries Annual Average PCM Projected Colorado R. Precipitation Timeseries Annual Average hist. avg. ctrl. avg. Period 1 2010-2039 Period 2 2040-2069 Period 3 2070-2098

Annual Average Hydrograph Simulated Historic (1950-1999) Period 1 (2010-2039) Control (static 1995 climate) Period 2 (2040-2069) Period 3 (2070-2098)

Natural Flow at Lee Ferry, AZ allocated 20.3 BCM Currently used 16.3 BCM

Total Basin Storage

Annual Releases to the Lower Basin target release

Annual Releases to Mexico target release

Annual Hydropower Production

Case study 1: Yakima River Basin Irrigated crops largest agriculture value in the state Precipitation (fall-winter), growing season (spring-summer) Five USBR reservoirs with storage capacity of ~1 million acre-ft, ~30% unregulated annual runoff Snowpack sixth reservoir Water-short years impact water entitlements Elevation 8184 ft to 340 ft Historic Temp and precip 22-76F, 80 in-140 in at 2300 ft, 27-90F, 0 in-10 in at 350 ft 61% -81% precip in October-March This water is used in many ways… -TWSA: multiple regression analysis uses correlations with precipitation, streamflow, snow measures, forecasts are made for precipitation levels of 50, 100 and %150 normal. Water supply during growing season in lower basin primarily from snowmelt, depends on reservoirs for storage Six USBR reservoirs with storage capacity of ~1 million acre-ft, ~25% unregulated runoff Managed system vulnerable to drought with increasing water use and changing snowpack DETAILS ON TWSA: Max, Min, and Avg all from 1981-2000 and 2007 values of: (1) System Unregulated Flow Volume (system of reservoirs, sum of inflows), (2) Observed Flow Volume (system of reservoirs, sum of outflows), (3) Parker Unregulated Flow Volume (Yakima River NR Parker mean daily natural discharge), (4) Parker Observed Flow Volume (Yakima River NR Parker mean daily regulated discharge) (5) Yakima System Diversions (5 major irrigation diversions above Parker) (6) Yakima System Storage (Reservoir system storage, mean daily reservoir volume) REFERENCES: (USBR, 2002) http://www.wrcc.dri.edu/cgi-bin/cliMAIN.pl?wa4154: KENNEWICK, WASHINGTON (454154) http://www.wrcc.dri.edu/cgi-bin/cliMAIN.pl?wa4406: LAKE KACHESS, WASHINGTON (454406) Precipitation varies considerably across the basin throughout the year. Mean-annual precipitation ranges from 140 inches in the higher mountains of the northwestern part of the basin to less than 10 inches throughout the lower Yakima Valley. The amount of precipitation that occurs during the period of Oct ober to March period, both the arid and alpine parts of the basin ranges from 61-81% of the annual precipitation. The variation in annual precipitation can be large. The geographic variability of mean annual precip 1951-1980 in high mts between 80-140 inches in lower between 0-10 inches (USBR 2002, pg 2-2) Water quality, generally high in upper basin. Degrades downstream. Many reaches in Federal Clean Water Act 303(d) list. Issues of turbidity, pesticides, low dissolved oxygen, elevated temperatures, metals, fecal coliform, low flows, and pH. Air temps, inversely related to altitude. Min and Max occur in Jan and July. Values above from LAKE KACHESS (1971-2000 Monthly Climate Summary, Average Min in Jan and Average Max in July) -Reserviors been in place since 1930s -Irrigation project serves 465,000 acres -Basin drains about 4 million acres (Basin drains about 6150 sq miles) -Yakima River flows about 215 miles -Estimated unregulated runoff of Yakima Basin (1961-1990) is 3.97 million acre-ft FROM POSTER, AGU 2006 -The USBR Yakima Project supports approximately 464,000 irrigated acres (via four irrigation districts -- Roza, Yakima-Tieton, Sunnyside Valley and Kittitas -- and the Wapato Division). -Most of the water in the Yakima River comes from snowmelt, and is caught in a series of reservoirs to ensure sufficient water supply throughout the season.

Yakima River Basin 2080s 2020s historical Basin shifts from snow to more rain dominant Water prorating, junior water users receive 75% of allocation Junior irrigators less than 75% prorating (current operations): 14% historically 32% in 2020s A1B (15% to 54% range of ensemble members) 36% in 2040s A1B 77% in 2080s A1B

Crop Model - Apple Yields There are similar impacts for cherry producers. Note the difference between yiled with and without CO2 fertilization effects Yields decline from historic by 20% to 25% (2020s) and 40% to 50% (2080s)

PCM Business-as-Usual scenarios California (Basin Average) BAU 3-run average historical (1950-99) control (2000-2048) important point(s): we’re modeling most of the US at 1/8 degree now with the VIC model, but we are performing this forecasting exercise in the Columbia River basin. The plusses show the grid of the numerical weather prediction (forecasting) model that we used (GSM), and the ¼ degree hydrology model resolution can just be discerned in the figure. 24 climate model grid points were used, and 1,668 VIC model cells. We’ve aggregate the VIC model to ¼ degree from 1/8 degree in the Columbia River basin to speed the forecast runs.

PCM Snowpack Changes Business-as-Usual Scenarios California April 1 SWE important point(s): we’re modeling most of the US at 1/8 degree now with the VIC model, but we are performing this forecasting exercise in the Columbia River basin. The plusses show the grid of the numerical weather prediction (forecasting) model that we used (GSM), and the ¼ degree hydrology model resolution can just be discerned in the figure. 24 climate model grid points were used, and 1,668 VIC model cells. We’ve aggregate the VIC model to ¼ degree from 1/8 degree in the Columbia River basin to speed the forecast runs.

Current Climate vs. Projected Climate Storage Decreases Sacramento Range: 5 - 10 % Mean: 8 % San Joaquin Range: 7 - 14 % Mean: 11 %

Current Climate vs. Projected Climate Hydropower Losses Central Valley Range: 3 - 18 % Mean: 9 % Sacramento System Range: 3 – 19 % Mean: 9% San Joaquin System Range: 16 – 63 % Mean: 28%

Stationarity—the idea that natural systems fluctuate within an unchanging envelope of variability—is a foundational concept that permeates training and practice in water-resource engineering. In view of the magnitude and ubiquity of the hydroclimatic change apparently now under way, however, we assert that stationarity is dead and should no longer serve as a central, default assumption in water-resource risk assessment and planning.

How can the water management community respond? Central methodological problem: While water managers are used to dealing with risk, they mostly use methods that are heavily linked to the historical record

“Synthetic hydrology” c. 1970 Figure adapted from Mandelbrot and Wallis (1969)

Ensembles of Colorado River (Lees Ferry) temperature, precipitation, and discharge for IPCC A2 and B1 scenarios (left), and 50-year segments of tree ring reconstructions of Colorado Discharge (from Woodhouse et al, 2006)

Hybrid Climate Change Perturbations New time series value = 19000 Objective: Combine the time series behavior of an observed precipitation, temperature, or streamflow record with changes in probability distributions associated with climate change. Hybrid perturbation schemes largely preserve important historic time series behavior while introducing potentially complex changes in the seasonal (e.g. monthly) probability distributions of temperature, precipitation, or streamflow. Traditional delta method approaches show the effects only of changes in the mean. This more sophisticated approach explicitly incorporates changes in the full probabability distribution. The quantile mapping procedure facilitates a completely non parametric transformation of the observed time series. Value from observed time series = 10000

Observed and Climate Change Adjusted Naturalized Streamflow Time Series for the Snake River at Ice Harbor KAF KAF Blue = Observed time series Red = Climate change time series

Other implications of nonstationarity Hydrologic network design (station discontinuance algorithms won’t work) Need for stability in the evolution of climate scenarios (while recognizing that they will almost certainly change over time)

Another complication: Water resources research has died in the U.S. No federal agency has a competitive research program dedicated to water resources research (e.g., equivalent to the old OWRT) As a result, very few Ph.D. students (and hence young faculty) have entered the area And in turn, the research that would identify alternatives to classic stationarity assumptions is not being done See Lettenmaier, “Have we dropped the ball on water resources”, ASCE JWRPM editorial, to appear Nov., 2008

Conclusions Ample evidence that stationarity assumption is no longer defensible for water planning (especially in the western U.S.) What to replace it with remains an open question A key element though will have to be weaning practitioners from critical period analysis, to risk based approaches (not a new idea!!) Support for the basic research needed to develop alternative methods (a new Harvard Water Program?) is lacking