Donate
Independent, objective, nonpartisan research

Replenishing Groundwater in the San Joaquin Valley, Technical Appendix

Database

This is the content currently stored in the post and postmeta tables.

View live version

object(Timber\Post)#3711 (44) { ["ImageClass"]=> string(12) "Timber\Image" ["PostClass"]=> string(11) "Timber\Post" ["TermClass"]=> string(11) "Timber\Term" ["object_type"]=> string(4) "post" ["custom"]=> array(5) { ["_wp_attached_file"]=> string(20) "0418ehr-appendix.pdf" ["wpmf_size"]=> string(7) "1073349" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(88554) "Replenishing Groundwater in the San Joaquin Valley Technical Appendi ces CONTENTS Appendix A: Update of the San Joaquin Valley’s Water Balance and Estimate of Water Available for Recharge in 2017 Alvar Escriva-Bou and Ellen Hanak Appendix B: PPIC’s Groundwater Recharge Survey Ellen Hanak and Jelena Jezdimirovic with research support fr om Darcy Bostic and Henry McCann Supported with funding from the S. D. Bechtel, Jr. Foundation and Sustainable Conservation PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 2 Appendix A: Update of the San Joaquin Valley’s Water Balance and Estimate of Water Available for Recharge in 2017 Summary This appendix is divided in two main sections. The first updates e ach component of the San Joaquin Water balance (published in the Technical Appendix A of the 2017 PPIC report Water Stress and a Changing San Joaquin Valley ) to include 2016 and 2017 and provides estimates of how groundwater overdraft is distributed across subregions of the v alley. The second estimates how much water was available for recharge in 2017 replicating the approaches presented by two recently released studies (DWR 2017a and 2018a, and Kocis and Dahlke 2017). From the San Joaquin w ater balance update, we find that 2017 was an extraordinarily wet year when compared to the 1986- 2017 series. With high levels of imports and the highest local inflows in more than three decades, water availability was at peak levels. Although San Joaquin River outflow s were also the highest in the series, there was a net increase in storage of 7.8 maf (4.6 maf in surface reservoirs and 3.2 maf in aquifers). Considering the 1.75 maf of overdraft over the long -term, the 3.2 maf net recharge in the aquifers is roughly 5 m af more total recharge than the 32- year average, and 8 maf more than annual total recharge during the 2012- 16 drought. To obtain the subregional differences in overdraft in the v alley we use data from the two hydrological models that have been independentl y applied in the Central Valley to estimate historical groundwater budgets : CVHM (Faunt et al., 2009) and C2VSim ( Brush et al. 2013). We find that the overdraft estimates are much more significant in the Tulare Lake hydrologic region, and especially in Ker n County. Finally we obtain different estimates for how much water was available for recharge in the San Joaquin Valley in 2017. The approaches used by DWR and Kocis and Dahlke result in a wide range of additional volumes of water that could have been capt ured in 2017: from a high of 6.3 maf (DWR’s unadjusted “maximum project estimate” under our assumptions) to 3.7 maf (Kocis and Dahlke), to 0.88 maf (DWR’s unadjusted “best estimate”). These levels are substantially higher than the long -term averages estimated by these two studies (1.2 and 0.55 maf for DWR’s unadjusted and adjusted “maximum project estimate”; 0.46 maf for Kocis and Dahlke’s post -1989 “impaired period,” and 0.43 maf and 0.19 maf for DWR’s unadjusted and adjusted “best estimate”), reflecting t he fact that 2017 was an extraordinary year. Introduction After a multi-year drought, 2017 was one of California ’s wettest years since record -keeping began in the 1890s . In the San Joaquin Valley high runoff, in conjunction with initial work on sustainable groundwater management plans mandated by the 2014 Sustainable Groundwater Management Act (SGMA), triggered an unprecedented interest in groundwater recharge. In this appendix we seek to answer two questions that are crucial to understanding the potential for groundwater recharge in the San Joaquin Valley over the long term: how did 2017 compare to other years? And how much additional water could have been available for recharge? T his appendix provides information on data sources and methods used to 1) assess the annual water balances in the San Joaquin Valley for water years 1986-2017, including estimat es of groundwater recharge and overdraft at the subregional scale, and 2) estimate water available for recharge in 2017 . 1 1 In this technical appendix we always refer to water years: the 12 -month period between October 1st and September 30th of the following year. The water year is designated by the calendar year in which ends and which includes 9 of t he 12 months. PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 3 For the annual water balances, this report provides an update of Technical Appendix A of the 2017 PPIC report Water Stress and a Changing San Joaquin Valley , with an expanded analysis for water years 2016 and 2017. For that reason we only include here the sources of information used to update the numbers and the results . Readers should consult th e earlier document for a more detailed explanation of methods and results . This appendix is divided in to two main sections. The first updates e ach component of the San Joaquin Water balance t o include 2016 and 2017 and provides estimates of how groundwater overdraft is distributed across subregions of the v alley. The second estimates how much water was available for recharge in 2017, replicating the approaches presented by two recently released studies (DWR 2017a and 2018a, and Kocis and Dahlke 2017). Updating the San Joaquin Valley’s Water Balance A water balance is an accounting statement that estimates water inflows (including precipitation and other water flowing into the area), outflows (including net or consumptive water used locally and water flowing out of the area), and changes in water stored in surface reservoirs and aquifers . As with any mass balance, the sum of inflows, outflows, and changes in storage has to be zero every year, as shown in the following equation: ���������������������������� −������������ PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 4 FIGURE A1 San Joaquin River and Tulare Lake w atersheds Inflows Four types of water inflows are considered here: flows into the valley from local watersheds (including the Central and Southern Sierra Nevada and the C oast R ange), water from net precipitation on the v alley floor, direct diversions from the Delta, and water imported from other regions —especially through the Sacramento -San Joaquin Delta, but also much smaller imports through the Folsom South Canal. These inflows can be either used the same year or stored in surface reservoirs or aquifers for later use. Conversely, some of the water used in a given year can be obtained from withdrawals from surface and groundwater storage. Inflows from local watershe ds To assess inflows into the valley floor from local watersheds we used estimates of full natural or “unimpaired” flows for the main rivers and creeks in the region. 2 We obtained full natural flows (monthly volume) for the period of study from California Data Exchange Center ( CDEC ) for the major rivers in the valley . For some minor local watersheds where CDEC data were insufficient or in consistent we used 1986 –2003 data from DWR (2007). For these minor watersheds w e estimated the data 2 Unimpaired flow is the natural runoff of a watershed in the absence of storage regulation and stream diversions. Full natural flow is the natural runoff of a watershed that would have occurred prior to human influences on the watershed, such as storage , d iversions, or land use changes. In this report we used full natural flows from CDEC stations where available, and unimpaired flows from DWR (2007) for the remaining watersheds. PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 5 after 2003 calibrating a simplified rainfall -runoff model that mimics the runoff response to rainfall in the gaged catchments. Figure A 2 shows total annual inflow volumes from each watershed, with 2017 the highest in the period analyzed. 3 Although there are inflows from 15 local watersheds, five rivers account for nearly 70 percent of the average total: Tuolumne (19%), San Joaquin (17%), Stanislaus (11%), Kings (11%) , and Merced (10%). FIGURE A2 Inflows from local watersheds NOTES: The rivers shown in the bar chart are ordered geographically from south to north. The Kern, Tule, Kaweah , and Kings drain into the Tulare Lake Basin, and the remaining rivers drain into San Joaquin River Basin. The Tulare Lake Basin is a closed basin in mo st years, with all inflows remaining within the basin. The exception is very wet years, when excess flows drain into the San Joaquin River throu gh the James bypass (Fresno slough) . Net valley floor precipitation Total monthly precipitation on the v alley floor has been obtained by clipping the gridded datasets from PRISM Climate Group, Oregon State University, using a GIS layer of the study area. We assume that 15 percent of total monthly precipitation becomes net precipitation —the precipitation that does not evaporate and remains in the Valley, where it is used by plants, flows into streams, or percolates into an aquifer. This estimate is based on the water balances from DWR’s California Water Plan (DWR 2013) for the entire Central Valley, and in the Central Valley water balance of the C2VSim model (Brush et al. 2013). Relative to inflow from the Sierra (shown in Figure A2 ), net precipitation on the v alley floor is less variable across dry and wet y ears (Figure A3) and does not always mirror differences in runoff. 3 Water year 1983 was the wettest year on record in the region in terms of both runoff and precipitation. Water year 2017 was the second wettest for runoff (using CDEC data for full natural flows), and the sixth wettest for precipitation (using NOAA data). 0 5,000 10,000 15,000 20,000 25,000 30,000 19861987198819891990199119921993199419951996199719981999200020012002200320042005200620072008200920102011201220132014201520162017 Local Inflows (thousands of acre - fet) SJ Westside Minor SJ Eastside Minor Cosumnes River Mokelumne River Calaveras River Stanislaus River Tuolumne River Merced River Chowchilla River Fresno River San Joaquin River Kings River Kaweah River Tule River Kern River PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 6 FIGURE A3 Net precipitation on the v alley floor The inflows from local watersheds and the precipitation on the v alley floor are the flows that could be captured in surface reservoirs or used for recharging aquifers. Figure A4 demonstrates that most of these flows come from the San Joaquin River hydrologic region (73% of total inflows), whereas nearly a fifth (19%) comes from the Tulare Basin watersheds, and just 8 percent from precipitat ion on the valley floor. This is important because is an indicator of the amount of water available for recharge (as we show later in th is report). FIGURE A4 Inflows into the valley from local rivers and precipitation on the valley floor 0 500 1,000 1,500 2,000 2,500 Net precipitation (thousands of acre -feet) 0 5 10 15 20 25 30 35 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Millions of acre -feet Dry yearsTulare Basin HRSan Joaquin River HRPrecipitationAverage PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 7 Direct Delta diversions Central and southern Delta agricultural lands—which are considered part of the San Joaquin Valley —use water diverted directly from the Delta. Most of this water flows into the Delta, with lesser volumes from the San Joaquin River and other Delta tributaries. To account for this inflow we assume that all agricultural water use in the area of the Delta that is included in the San Joaquin Valley is directly diverted from the Delta and is estimated as the net or consumptive water use of these agricultural lands. The methods to obtain net or consumptive water use are described in a section below. W e use the same methodology to obtain the direct Delta diversions , but just for the area of the Delta that is in the San Joaquin Valley floor. D iversions average roughly 750 thousand acre -feet per year (Figure A5). FIGURE A5 Direct Delta diversions Imports from other regions Water is imported into the valley from Northern California through pumps in the south Delta. A small volume is also imported from the American River through the Folsom South Canal. Delta imports Delta imports are primarily from the Sacramento River with a small share from the San Joaquin River. These sources mix as they enter the Delta. Daily data for Delta imports from State Water Project (SWP) facilities (Banks Pumping Plant or Clifton Court Int ake), the Central Valley Project (CVP) facilities (C.W. “Bill” Jones Pumping Plant at Tracy ), and the Contra Costa Canal are obtained from Dayflow —a program that estimates daily average 0 200 400 600 800 1,000 1,200 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Direct Delta diversions (thousands of acre -feet) PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 8 Delta outflows (DWR, 2016) —for the period 1986-2015. The water years 2016 and 2017 were provided by DWR . 4 Imports through the Folsom South Canal The Folsom South Canal imports a small volume of water from the American River into the northeastern side of the San Joaquin Valley. Data on water imported through the South Folsom Canal is obtained from the US Bureau of Reclamation Central Valley Operations Annual Delivery Reports (Table 21). Figure A6 shows annual imports from the CVP and SWP pumps in the south Delta (Jones and Banks, respectively), the Contra Costa Canal, and Folsom South Canal. Total imports averaged 4.9 million acre -feet/year (53% from SWP and 47% from CVP), but in dry years imports from the Delta fall below 2 million acre-feet, as occurred in 2015. In 2017 the Delta imports were 6.4 maf, only a little less than the 2011 record of 6.6 maf. Imports through the South Folsom Canal averag ed around 26,000 acre-feet/year. FIGURE A6 Imports from other regions Outflows Four types of water outflows are considered: consumptive water use from evapotranspiration (water consumed by plants, and other evaporation to the atmosphere from the v alley floor), San Joaquin Valley outflows to the Delta, exports from San Joaquin River tributaries to Bay Area water users, and exports of imported water that enters the v alley. 5 Consumptive water use Similar to precipitation, monthly consumptive water use was obtained by clipping evapotr anspiration gridded datasets from the operational Simplified Surface Energy Balance model (Senay et al. 2013) using a GIS layer. This dataset provides high -resolution estimates of evapotranspiration, but only covers the period from 2000 to the 4 CVP deliveries under the Friant Division are not included in these totals—this water is diverted from the San Joaquin River at Millerton Lake to the Friant -Kern Canal, which delivers water to users on the east side of the Tulare Basin. For the purposes of this regional water balance, these are considered local flows. Contra Costa Canal imports are included because the pumps are inside the San Joaquin Valley floor, but as the Contra Costa Water District is entirely outside of the valley floor, these imports are later considered as exports to the Bay Area. 5 This may slightly understate total net water use insofar as it does not include water embodied in manufactured goods produced in the valley. 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Delta Imports (thousands of acre - feet) Folsom South CanalContra Costa Canal CVP pumpsSWP pumps PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 9 present. To obtain earlier values, we employed multiple regression analysis, using temperature, precipitation , and potential evapotranspiration from crops as independent variables (see Arnold and Escriva- Bou 2017 for more details ). The values for 2016 and 2017 were 14.6 and 15.3 maf respectively, at the higher end of the distribution— reflecting the greater volume of water available to evaporate from soils and transpire through vegetation in wetter years. To break down the consumptive use by end users (i.e., agriculture, urban, environmental, and natural landscapes) we obtained the average share of consumptive uses by end use from the California Water Plan (DWR 2013) , and then appl ied these shares to the estimates of evapotranspiration for the entire 3 2-year p eriod (Figure A7) . FIGURE A7 Annual evapotranspiration by end use NOTE: The category “other” includes evapotranspiration from natural landscapes not categorized as wetlands. San Joaquin Valley outflows San Joaquin River and other minor Delta tributaries outflows (Figure A8 ) were obtained from Dayflow and CDEC . These flows are reported as Delta inflows from the San Joaquin River (measured at the Vernalis gage) and eastern Delta inflow (including the Cosumnes and Mokelumne rivers and other minor creeks) . 6 The magnitude of the outflows in the San Joaquin Valley are highly influenced by the storage capacity within the basin , whereas minimum outflows in the river and the Delta are required by water quality and other environmental regulations. In 2017, the outf lows were 12.3 maf, the highest level in the past 32 years. 6 San Joaquin Valley outflows that are subsequently recaptured as either Delta imports or direct Delta diversions are counted as inflows to the region, and included in the Delta import measures presented below. 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000 20,000 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Evapotranspiration (thousands of acre -feet) AgricultureUrbanWetlandsOther PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 10 FIGURE A8 San Joaquin Valley outflows Exports from San Joaquin River tributaries The San Joaquin Valley exports several hundred thousand acre -feet of water annually from local sources to the San Francisco Bay Area (Figure A 9). The amount of export water does not vary significantly from year to year. Water from the Tuolumne River is stored in Hetch Hetchy Reservoir and then conveyed to the San Francisco Public Utilities Commission (SFPUC) service area, while water from the Mokelumne River is stored in Pardee and Camanche Reservoir s and conveyed to the East Bay Municipal Utility District (EBMUD) service area. 7 FIGURE A9 Water exports from San Joaquin River tributaries 7 EBMUD also has a contract with USBR to divert 100 million gallons a day at Freeport on the Sacramento Rive r in an emergency; this is not included here. 0 50 100 150 200 250 300 350 400 450 500 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Exports from local sources (thousands of acre -feet) SF BayEl Dorado ID 0 2 4 6 8 10 12 14 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Millions of acre -feet Dry yearsSan Joaquin Valley outflows PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 11 Data from 1998- 2010 was obtained from the California Water Plan (DWR 2013) , using the series “imports to the San Francisco Bay Hydrologic Region .” The remaining years have been estimated with a regression analysis using the unimpaired flows of the rivers for the entire series as an independent variable, and extrapolating the shares of the diverted data with respect to the unimpaired flows for the 1998- 2010 dataset. 8 Average annual exports f or 1986-2015 are 0.38 maf/year. Figure A 9 also includes a small volume of exports to the El Dorado Irrigation District in the Sacramento River hydrologic region. This water is diverted from Jenkinson Lake on Sly Park Creek, a tributary of the Cosumnes River, and averages 18,000 af per year. Diversion data is from CDEC (station CCN). Note the low variability of water exports from the San Joaquin Valley tributaries compared with the high variability of inflows to these tributaries shown in Figure A 2. Exports of Delta imports Some Delta imports that enter the Valley through the CVP and SWP pumps are delivered to the San Francisco Bay area, the Central Coast, and Southern California (Figure A10 ).  Exports to the San Fra ncisco Bay Region: This includes water from two points of diversion: (1) through the South Bay Aqueduct from the South Bay Pumping Plant (data are from the SWP Annual Reports of Operations), and (2) through the Contra Costa Canal (data are from USBR Centra l Valley Project Annual Re ports of Operations, Table 21).  Exports to the Central Coast: This includes water from two points of diversion: (1) through Las Perillas Pumping Plant on the California Aqueduct (from the State W ater Project Annual Reports of Operation s: Table 1 ); and (2) through the Pacheco Tunnel. 9  Exports to Southern California: This includes water delivered through the A.D. Edmonston Pumping Plant on the California Aqueduct (from DWR SWP Annual R eports of Ops: Table 1 Totals). FIGURE A10 Exports of Delta imports 8 Note that for this update we did not get actual data for these exports, so 2016 and 2017 values were estimated using a regression analysis. 9 San Luis Reservoir Operations, f rom DWR SWP Annual Reports of Oper ations: p ost -2000 Reports, Table 15 Annual San Luis Joint -Use Facility Total , and pre -2000 Reports, Table 13 San Luis Reservoir Operations Total Outflow (Pacheco Tunnel). 2015 data obtained from Santa Clara Valley Wa ter District urban water supply data and data for years 2016 and 2017 was obtained directly from DWR. Some water going through the Pacheco Tunnel goes to the Santa Clara Valley Water District and could be included in the exports to the San Francisco Region. As we do not have access to sufficient data to separate the flows that remain in the Central Coast and those that go to the San Francisco Bay hydrologic r egion, we include them as exports to the Central Coast. 0 500 1,000 1,500 2,000 2,500 3,000 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Exports of Imported Water (thousands of acre -feet) Southern CaliforniaCentral CoastSF Bay PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 12 Changes in S torage Two main storage types are considered: surface reservoirs and water stored in aquifers. Data exist for water stored in surface reservoirs. Changes in aquifer storage are estimated as the volume required to c lose the water balance for the v alley. Changes in reservoir storage From CDEC, we obtained data for monthly storage for 13 major reservoirs in the San Joaquin Valley (Figure A11 ). Annual storage change is the water stored at the beginning of the prior water year minus the storage at the beginning of a new water year (October 1) . The reliability of data for the 1986 and 1987 water years is lower because there were more gaps in the series. 10 FIGURE A11 Water stored in the 13 major reservoirs in the San Joaquin Valley NOTE: The 13 major reservoirs are: New Melones (NML), Don Pedro (DNP), Lake McClure (EXC), Pine Flat (PNF), Lake Isabella (ISB), Success Dam (SCC), Kaweah Lake (TRM), Millerton Lake (MIL), Eastman Lake (BUC), Mariposa Reservoir (MAR), Bear Reservoir (BAR), New H ogan Lake (NHG), and Camanch e Reservoir (CMN). TAF is thousand acre -feet. To include water stored in other minor reservoirs, we obtained the change in total water stored from the California Water Plan (DWR 2013) for the two hydrologic regions and extrapolated the other years using a linear relationship between total changes in storage and changes in the storage in the 13 major reservoirs ( see details in Arnold and Escriva -Bou, 2017) . Total net changes in annual surface storage are shown below in Figure A 12B. The long-term average change in surface water stored is roughly zero. Changes in water stored in aquifers Finally, we determine the change in water stored in aquifers as the residual that closes the water balance for the v alley. In short, the net available water supply (from in flow, precipitation, and changes in storage) must equal the 10 As discussed below, the estimates for 1986 appear low relative to precipitation and runoff year, which may in turn result in an overestimate of the level of aquifer storage in that year . 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 Oct-87 Oct-88 Oct-89 Oct-90 Oct-91 Oct-92 Oct-93 Oct-94 Oct-95 Oct-96 Oct-97 Oct-98 Oct-99 Oct-00 Oct-01 Oct-02 Oct-03 Oct-04 Oct-05 Oct-06 Oct-07 Oct-08 Oct-09 Oct-10 Oct-11 Oct-12 Oct-13 Oct-14 Oct-15 Oct-16 Oct-17 Reservoir Storage (thousands of acre- feet) ISBSCCTRMPNFMILBUCMARBAREXCDNPNMLNHGCMN PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 13 net v olume of water used within the v alley (consumptive use) or exported. The mass balance equation can be formulated as: ∆ �������� 11 Inflows from local supplies can be used directly as surface diversions but also indirectly by replenishing aquifers and pumpe d later as sustainable groundwater use. According to C2VSim groundwater budgets at the subregional scale (Brush et al. 2013), rivers in the San Joaquin Valley contri bute to roughly 0.5 maf of groundwater recharge on average for the period 1973 -2009. Also there is water recharged from unlined canals and percolation of excess irrigation water on agricultural lands. 12 Net imports into the valley are the total water imported from other regions minus the water that is exported to other regions. 13 As mentioned above, reservoir data for the years 1986 and 1987 had low reliability. This could explain why in 1986, a wet year, our model is not reporting an increase in reservoir storage, but a n unusually large increase in water stored in aquifers, relative to the volume of runoff and imported water deliveries. If the balance between surface and groundwater storage in 1986 was in line with more recent wet years, average overdraft for the entire period would be closer to 1.9 maf/year. 14 California has approximately 850 maf to 1.3 billion acre -feet of groundwater in storage (DWR, 1994), and about 45 maf of surface storage (PPIC, 2017) 15 More information about the different changes in temporal patterns can be found in Hanak et al., 2017. PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 14 -8 -6 -4 -2 0 2 4 6 8 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 - 1,0 00 2,0 00 3,0 00 4,0 00 5,0 00 6,0 00 7,0 00 8,0 00 9,0 00 10 ,00 0 Millions of acre -feet Dry yearsNet Reservoir WithdrawalNet Reservoir Recharge 0 5 10 15 20 25 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Millions of acre -feet Dry yearsImportsLocal suppliesDirect Delta DiversionsTotal Net Use -8 -6 -4 -2 0 2 4 6 8 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 - 1,0 00 2,0 00 3,0 00 4,0 00 5,0 00 6,0 00 7,0 00 8,0 00 9,0 00 10 ,00 0 Millions of acre -feet Dry yearsNet GW WithdrawalNet Recharge FIGURE A12 Components of the San Joaquin Valley’s annual water balance A) Water supplies and net water use B) Change in reservoir storage C) Change in groundwater storage PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 15 Estimating Groundwater Overdraft at the S ubregional Level in the San Joaquin Valley Two hydrological models have been independently applied in the Central Valley to estimate historical groundwater budgets : CVHM (Faunt et al., 2009) and C2VSim ( Brush et al. 2013). The models use precipitation, surface- water inflows, and surface- water diversion input data that is either similar or identical (Dogrul et al., 2011) to estimate subregional water budgets , with a focus on assessing agricultural groundwater pumping . Using C2VSim and CVHM models we determined the groundwater overdraft for 14 subregions of the San Joaquin Valley by assessing the decrease in groundwater storage over the lo ng term (Figure A13). Table A1 shows the results for both models for the 14 subregions within the San Joaquin Valley for the period 1975- 2003, and also an average of both models. FIGURE A13 Hydrologic regions and m odel subregions used by both C2VSim and C VHM models SOURCE: Brush et al. (2013). NOTE: The numbers refer to analysis subregions commonly used in hydrologic and economic modeling in the Central Valley. The results of Table A 1 show that the models are in fairly close agreement at the regional scale (plus or minus 20%) , but with some significant discrepancies for individual subregions (note for instance the discrepancies in regions 9 and 18). PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 16 TABLE A1 Change in groundwater storage at the subregional scale ( 1975-2003 ) Subregion Change in ground water storage (in taf/year) C2VSim CVHM Average 8* -52 -3 -28 9* -17 96 39 10 -8 1 -4 11 -11 -5 -8 12 5 32 19 13 -97 -134 -116 14 -271 -177 -224 15 -59 -170 -115 16 -14 -140 -77 17 -124 -107 -116 18 156 -408 -126 19 -210 -192 -201 20 -232 -159 -195 21 -411 -255 -333 Total -1,345 -1,622 -1,483 SOURCE: C2VSim results at the subregional scale were obtained directly from model outputs, while CVHM results were provided by Stephen Maples. NOTE S: * The actual change in groundwater storage in regions 8 and 9 is twice the value we show in this table. We made this adjustment because only a part of regions 8 and 9 are in the San Joaquin River hydrological region. T o provide a rough estimate of the distribution of overdraft across different parts of the valley , we aggregated the results, combining the 14 modeling subregions into five (Table A2) . The San Joaquin River hydrologic region includes the northwest (sub region s 9 and 10 ) and northeast ( subregions 8, 11, 12, and 13), and the Tulare Lake hydrologic region includes the southwest (sub regions 14 and 15 ), the southeast ( subregions 16, 17, and 18 ), and the Kern basin ( sub regions 19, 20, and 21) . This aggregation reinforces the discrepancy between the two models in the southeast: whereas the C2VSim model estimates a positive change of groundwater storage over the long - term, CVHM estimates overdraft of more than 6 50 taf/year. This subregion includes the Kings, Kaweah, and Tule basins, all of which DWR consi ders to be “critically overdrafted.” Moreover, parts of these regions are experiencing significant subsidence. 16 It therefore appears that C2VSim is underestimating pumping and/ or overestimating recharge in this part of the valley. Finally, we also made a s imple adjustment to include years after 2003 , which were drier. For the northwest, where the average change in groundwater storage from the two models is positive, we assumed that there is no contribution to overdraft at the subregional scale (although the re could still be local issues). For the areas with overdraft , we multiplied the 1975- 2003 average by a factor that accounts for the additional overdraft for the San Joaquin Valley from our 1986- 2015 water balance. Using this procedure we are implicitly assuming that overdraft increased at a similar pace after 2003 across affected areas. This rough approximation is supported by the widespread decline in groundwater tables since the mid -2000s shown in DWR’s Groundwater Information Center Interactive Map Application (DWR 2018b). 16 See DWR’s identification of Critically Overdrafted Basins and NASA JPL report on Subsidence in the Central Valley PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 17 TABLE A2 Change in groundwater storage for the five aggregated subregions Subregion Change in groundwater storage (in taf/year) C2VSim 1975-2003 CVHM 1975-2003 Average 1975-2003 Adjusted long- term average* Northwest -25 96 16 0 Northeast -155 -110 -133 -156 Southwest -330 -347 -339 -397 Southeast 18 -655 -318 -374 Kern -853 -606 -729 -856 Total -1,345 -1,622 -1,483 -1,783 Total in overdrafted subregions* -1320 -1,718 -1,519 -1,783 NOTES: *The adjusted long -term average allocates the additional overdraft we find in our 1986 -2015 water balance relative to the 1975 -2003 total for the four subregions experiencing average overdraft (1,783 – 1, 519 = 264 acre -feet) in proportion to their overdraft in 1975 -2003. We assume that the northwest is in average long -term balance. FIGURE A14 Groundwater overdraft at the subregional scale (in thousands of acre -feet per year) NOTE: Given the uncertainties of the estimations, this figure shows results rounded to the nearest 10 taf. PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 18 Estimating Water Available for Recharge in 2017 This section explains estimates of how much water was available for recharge in the San Joaquin Valley in 2017. W e first describe two different approaches from studies that estimated how much water is available to replenish aquifers in California. Then we try to reproduce each approach with the available data from the San Joaquin Valley for 2017 and the othe r years examined in our water balance . Finally, we discuss the results. DWR’s Approach In January 2017, t he Department of Water Resources released Water Available for Replenishment Report Draft (DWR 2017 a). 17 The study estimated water available for replenishment for each of the state’s 10 hydrologic regions and 56 planning areas. Basically , the methodology defines a minimum streamflow requirement that accounts for environmental needs and is not available for re charge. The remaining outflow can be diverted for recharge if there is infrastructure capacity for doing so. 18 To account for uncertainties in its estimates, DWR shows a range of values based upon a range of instream flow s and project capacity. These includ e a “best estimate” that uses the instream flow requirement and an estimate of the existing project capacity, a “sensitivity range ,” and a “maximum project estimate” that illustrates the maximum potential diversion for recharge. DWR’s approach caps the max imum amount of water that can be diverted above the minimum instream flow requirement based on the infrastructure needed to divert this water. To do this they define a conceptual “project diversion capacity” which is the total capacity of the diversions th at would be used to recharge groundwater. To obtain the “best estimate” they define as project diversion capacity the “existing project diversion capacity”, which is sized based on the water right with the largest single point of diversion capacity on a given river/stream. The sensitivity ranges are obtained by using tw o and half times the “existing project capacity.” Finally they obtain the “maximum project estimate” by assuming that there is no infrastructure constraint in the amount of water that can be diverted. The methodology uses historical daily gage data (1930 through 2015 for the San Joaquin River) to develop a n initial set of estimates of available water. It then adjusts these using a simulation model (W ater Evaluation and Planning [ WEAP ]) to account for changes in outflows reflecting current water demands and operations . In our analysis, we refer to the results based on the historical gage data as the “unadjusted” estimates, and those adjusted using WEAP as the “adjusted” estimates. 19 For the S an Joaquin Valley, t he WEAP adjustment reduces the volume available for recharge significantly—to 45 percent of the unadjusted volumes . In Figure A1 5 we show a conceptual depiction of DWR ’s “best estimate” and “maximum project estimate.” In this illustration, which uses actual outflow from Vernalis for the water year 2016, a minimum instream flow requirement of 1,000 cfs is defined for any day of the year, and an existing project diversion capacity is 500 cfs. The shaded region in yellow would be the “best estimate” of water available for replenishment. When there is no upper limit assumed for project diversion capacity, we use the “maximum project estimate.” 17 The fi nal report is expected to be released in 2018. Its appendices have already been updated and published, and we use the estimat es in updated appendix A in our analysis (DWR 2018a). 18 DWR’s Water Available for Replenishment Report Draft does not explicitly identify instream flow requirements n or the existing project capacity data, so to replicate that study’s approach we developed rough estimates below that result in comparable levels of water available for recharge. 19 DWR refers to the estimates based on the historical gage data as the “gage d ata outflow” and the estimates using the WEAP adjustment as “water available for replenishment .” WEAP results are only available at a monthly —not daily—timescale. PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 19 FIGURE A15 C onceptual depiction of DWR’s water available for replenishment For the San Joaquin River hydrological region, DWR ’s “best estimate” of water available for replenishment is 190 taf/year and the “maximum project estimate” , which includes also the water captured under the best estimate assumptions, is 550 taf/year. These totals include wate r in the Tulare Lake region (47 taf/year in the “best estimate” and 213 taf/year in the “maximum project estimate”) because excess surface water from the Tulare Lake region flows into the San Joaquin River. The Kocis and D ahlke Approach: High-M agnitude Streamflow Also in 2017, Tiffany N. Kocis and Helen E. Dahlke, two scientists at UC Davis, published a paper estimating the availability of high -magnitude streamflow for recharge in the Central Valley (Kocis and Dahlke 2017). Their approach is different from DWR’s. U sing historic daily streamflow records available from USGS , they assume that the days on which out flows are within the highest 10 percent (or above the 90th percentile) over a 90-year period exceed both environmental flow requirements and current surface water allocations under California water rights, and could therefore be used to recharge aquifers. Using USGS data for the San Joaquin River at Vernalis for the period Oct ober 19 23 through Sep tember 2014, Kocis and Dahlke obtained a 90 th percentile flow of 11,600 cfs. Figure A16 depicts the high-magnitude flows for the p ast 10 years using their approach, which assumes that the volume in excess of the 90 th percentile could be taken on days with higher outflow . 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 Oct-15Nov-15Dec-15Jan-16Feb-16Mar-16Apr-16May-16Jun-16Jul-16Aug-16Sep-16 San Joaquin River flow (cfs) Extra water available for replenishment under "maximum project estimate" Water available for replenishment "best estimate" Daily stream flow Instream flow requirement Project(s) diversion capacity PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 20 FIGURE A16 Water available fo r recharge in the San Joaquin Valley in 2008 -17 using the Kocis and Dahlke approach For the period 1989- 2014—a subset of years included in this analysis that better reflects current water demand and operations in the v alley—Kocis and Dahlke estimate that an average of 1.3 maf would be available for recharge in years with high-magnitude flows . Because these flows only occur in 36 percent of all years (9 out of 25) , this translates to a long- term average across all years of 455 taf/year. 20 Water Available for Recharge in 2017 In the water balance analysis above, we estimate that there was 3.2 maf net groundwater recharge in the valley in water year 2017, roughly 5 maf more recharge than the 32- year average and 8 maf more than annual recharge during the 2012- 16 drought . To estimate how much additional water was available for recharge in 2017, we try to reproduce both approaches using data for 2017. DWR’s approach DWR’s Water Available for Replenishment Report Draft does not explicitly report the underlying assumptions for instream flow requirements to meet environmental needs and downstream water rights, nor the assumptions on existing project capacity to divert additional flows for recharge . Also, the authors used a complex modeling approach at the planning area scale that would be difficult to replicate precisely . In an effort to provide a rough replic ation of DWR’s approach, w e made some simplifying assumptions. We treat ed the entire San Joaquin Valley as a single basin, with a unique outlet: the San Joaquin River at Vernalis . 21 This facilitates a comparison with the results of Kocis and Dahlke, who also used the Vernalis gage to estimate high -magnitude flow s. 22 W e selected values for instream flow requirements and existing pro ject capacity that 20 The study refers to 1989 -2014 as the “post -impairment” period. For the entire 1923 -2014 per iod of analysis, they estimate an average of 1.46 maf is available in years with high magnitude flows, which occur in 47 percent of years, translating to a long- term average across all years of 686 taf/year. 21 The gage near Vernalis is the point commonly u sed to measure flows leaving the San Joaquin River and Tulare Lake regions for the Delta. Vernalis flow does not include the northernmost San Joaquin hydrologic region tributaries: the Calaveras, the Mokelumne, and the Cosumnes Rivers. As a result, these outflow values differ somewhat from the totals shown in the valley- wide water balance, shown in Figure A8. 22 Kocis and Dahlke (2017) also provide estimates for some gages upstream of Vernalis (for interactive maps and graphics, see recharge.ucdavis.edu/starr ). 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 Oct-07Apr-08Oct-08Apr-09Oct-09Apr-10Oct-10Apr-11Oct-11Apr-12Oct-12Apr-13Oct-13Apr-14Oct-14Apr-15Oct-15Apr-16Oct-16Apr-17 San Joaquin River flows (cfs) High-magnitude flows available for recharge San Joaquin River actual outflow 90th percentile flow PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 21 result in a similar long-term average level of water available for recharge for the region under DWR’s best estimate and maximum project estimate scenarios. Because these assumptions may differ from what DWR used, we cannot ensure that t he annual and seasonal hydrographs we obtain are the same as those obtained by DWR , even though our estimates of the long -term average annual volumes available for recharge are similar . We define the minimum instream flow requirement based on Delta conditions, which is also the focus of DWR’s approach for the San Joaquin hydrologic region (DWR 2018a). 23 First , we assume that water is available only when the Delta is in “excess conditions .” 24 Then we define a minimum daily Delta outflow requirement —as a parameter to calibrate —to meet water quality standards and other environmental regulations in the Delta . The water above this threshold is the uncaptured water —flows that exceed the capacity of existing storage and diversion facilities and regulatory demands. 25 Because this water comes from both the Sacramento and the San Joaquin hydrologic regions, we distribute the total amount based on the share of total inflow from each river on a daily basis using Dayflow data . With this approach, and w ithout capping the amount that can be diverted, we are reproducing DWR’s unadjusted “maximum project estimate .” Calibrating the minimum Delta outflow parameter to 10,000 cfs, we determine that the water available for recharge over the long term (using data for 1986 -2015) is 1.216 taf/year, just 0.3 percent lower than DWR’s unadjusted “maximum project estimate.” 26 To reach DWR’s unadjusted “best estimate” we had to cap the amount of water that can be diverted through the “existing project”— i.e., available conveyance infrastructure. Using a cap of 2 ,250 cfs we get a best estimate of 421 taf/year, just 2.1 % lower than DWR’s unadjusted best estimate. Figure A17 shows how this approach performs from 2000 to 2010. The adjusted best estimate and maximum project es timate—adjusting outflows based on the WEAP model simulations —are 190 and 550 taf/year respectively. 23 Water availability in the Central Valley depends on meeting environmental conditions downstream in the Sacramento -San Joaquin Delta. DWR’s approach for the Sacramento River and San Joaquin River hydrologic regions takes this into account (DWR 2018). 24 Exc ess water conditions occur in periods when releases from upstream reservoirs plus unregulated flows exceed Sacramento Valley in- basin uses plus exports (USDOI 2004). 25 See Gartrell et al. (2017) for a detailed explanation of uncaptured water. Note that this method of determining water available for recharge assumes that upstream environmental flow regulations do not require additional water to reach the Delta, beyond that needed to meet regulations in the Delta and downstream water demands. 26 On average, t his calibration results in just a slightly higher estimate of uncaptured outflow (+11%), with a similar temporal distribution , to estimates of uncaptured Delta outflow by Gartrell et al. (2017). PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 22 FIGURE A17 Water available for recharge for 2000 -1 0 using DWR’s unadjusted “best estimate” and “maximum project estimate” with our “instream flow requirement” and “existing project capacity” assumptions NOTES: Both “best estimate” and “maximum proje ct estimate ” are matched to DWR’s unadjusted estimates (“g age data outflow ”) at Vernalis . Using th is same approach f or water year 2017, we obtain an unadjusted best estimate of 882 taf and maximum proj ect estimate of 6.3 maf (Figure A1 8). Using the same WEAP adjustment factor that DWR applied to the long- term averages, the adjusted values are 397 taf and 2.9 maf, respectively. However, the unadjusted values are likely more appropriate for recent years, since actual outflow data already reflect current water use and operations. 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 Oct-00 Apr-01 Oct-01 Apr-02 Oct-02 Apr-03 Oct-03 Apr-04 Oct-04 Apr-05 Oct-05 Apr-06 Oct-06 Apr-07 Oct-07 Apr-08 Oct-08 Apr-09 Oct-09 Apr-10 Flows (taf/day) Extra water available for recharge under "maximum project capacity" Water available for recharge "best estimate" San Joaquin River actual outflow PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 23 FIGURE A18 Water available for recharge for 2017 using DWR’s unadjuste d “best estimate” and “maximum project estimates” (with our own “instream flow requirement” and “existing project capacity” assumptions ) NOTES: B oth “best estimate” and “maximum proje ct estimate” are matched to DWR’s unadjusted estimates (“g age data outfl ow”) at Vernalis . The Kocis and Dahlke approach Once the 90th percentile daily flow is defined, determining high -magnitude flows available for recharge is straightforward. Every flow since January 11, 2017 to July 11, 2017 was above the 90th percentile, so all th e water above this threshold ( 11,600 cfs) would be available: 3.7 maf for the entire water year (Figure A19). 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 90,000 Oct-16 Nov-16 Dec-16 Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Jul-17 Aug-17 Sep-17 San Joaquin River flows (taf/day) Extra water available for recharge under "maximum project capacity" Water available for recharge "best estimate" San Joaquin River actual outflow PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 24 FIGURE A19 Water available for recharge for 2017 using the Kocis and Dahlke approach Discussion of the E stimates of W ater Available for Recharge Comparison of findings for 2017 The approaches used by DWR and Kocis and Dahlke result in a wide range of additional volumes of water that could have be en captured in 2017, from a high of 6. 3 maf (DWR’s unadjusted “maximum project estimate” under our assumptions) to 3.7 maf (Kocis and Dahlke) , to 0.88 maf (DWR’s unadjusted “best estimate”) . 27 These levels are substantially higher than the long -term averages estimated by these two studies ( 1.2 and 0.55 maf for DWR ’s unadjusted and adjusted “maximum project estimate” ; 0.46 maf for Kocis and Dahlke’s post -1989 “impaired period,” and 0.43 maf and 0.19 maf for DWR’s unadjusted and adjusted “best estimate”) , reflecting the fact that 2017 was an extraordinary year . When com paring the different approaches and the actual monthly out flows in 2017 (Figure A2 0), it is noteworthy that DWR ’s maximum project estimate approach takes more flows almost every month than Kocis and Dahlke’s approach . This is because in this extremely wet year, the Delta requirements were already satisfied with flows much lower than the 90 th percentile used as a cut -off by Kocis and Dahlke. 27 As described further below, we do not include the DWR adjusted estimates here because the 2017 gage data already reflect current demands and operations. 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 Oct-16Nov-16Dec-16Jan-17Feb-17Mar-17 Apr-17May-17Jun-17Jul-17Aug-17Sep-17 San Joaquin River flows (cfs) High-magnitude flows available for recharge San Joaquin River actual outflow 90th percentile flow PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 25 FIGURE A20 Water available for recharge under the different approaches compared to San Joaqui n River actual outflow in 2017 NOTE: B oth the “best estimate” and “maximum proje ct estimate” are matched to DWR’s unadjusted estimates (“g age data outflow ”) at Vernalis . Comparison across the past three decades The comparison of total water available for r echarge across the past three decades further highlights the differen ces between the approaches (Figure A2 1). Whereas DWR’s approach identifies additional water for recharge in most years, Kocis and Dahlke’s approach finds additional water mostly in w et years. 28 This again reflects the different threshold s above which streamflow is considered available. Whereas Kocis and Dahlke only consider flows above the 90 th percentile, DWR ’s approach varies with Delta conditions. 28 Under this approach, there can also be high magnitude flows during big storm events in other year types, but this results in a very small share of the total wat er available for recharge: 2 percent of all high magnitude flows at Vernalis for the 1989-2014 “unimpaired period”, and 6 percent of all high magnitude flows for the 1971 -2017 period. 0.0 0.5 1.0 1.5 2.0 Oct-16 Nov-16 Dec-16 Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Jul-17 Aug-17 Sep-17 Water available for recharge vs San Joaquin River Actual Outflow (millions of acre-feet) DWR best estimateDWR maximum project estimateKocis & DahlkeActual outflow PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 26 FIGURE A21 Water available for recharge under the different approaches Reasons for using DWR’s unadjusted estimates for planning purposes It is also worth noting that while the adjustment DWR makes to account for current demands and operations makes sense when using gage data over a very long timeframe, it might be unnecessary for recent years. 29 As the water balance data presented earlier in this appendix demonstrate (Figures A7 and A12), water use in the San Joaquin Valley has not changed much since the mid- 1980s. Nor have there been any significant changes in water infrastructure over this period. For this reason, the unadjusted data may better capture the potential water available for recharge than the adjusted data that DWR presents as final estimates. One caveat is that the method used here to determin e water available for recharge—focusing on conditions in the Delta —assumes that upstream environmental flow regulations do not require additional water to reach the Delta beyond that needed to meet regulations in the Delta and downstream water demands. T he method used here also shows how simple accounting rules can help determine when water i s available for recharge. F urther development of this approach—taking into account the uncaptured water in the Delta as defined in Gartrell et al . (2017) plus some definition of upstream environmental requirements in the San Joaquin River system —would mak e it possible to estimate water available for groundwater recharge in the San Joaquin Valley. To be useful as a management tool , the flows that need to stay in the river and the amount that can be taken out of the river for recharg e purposes would need to be published in a timely manner (maybe weekly) . Near-term 29 Recall that DWR uses daily average Delta outflow from water year 1930 through 2015 to calculate its long- term averages. 0 1 2 3 4 5 6 7 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Water available for recharge (millions of acre- feet) DWR's adjusted best estimateDWR's adjusted maximum project estimate DWR's unadjusted maximum project estimateKocis & Dahlke PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 27 forecasts of storm conditions and river flows could be linked to this information to help managers prepare for recharge opportunities, which often need to be acted on quickly . This would be a way to overcome regulatory hurdles that impede the expansion of recharge practices in the v alley. Capacity constraints on capturing additional flows In 2017, most rivers in the San Joaquin Valley were above flood stage for months, and water managers faced significant capacity constraints in their water storage and conveyance infrastructure. Capturing significant additional flows in wet years like 2017 will likely require new infrastructure investments, in addition to new water rights permits to divert and store the water. As an example, more than half of the 6.3 maf of water available for recharge in 2017 under DWR’s unadjusted “maximum project estimat e” would have needed to be taken in two months (February and March), with diversions totaling more than 30,000 af/day (Figure A18). On seven of these days, diversions would have had to exceed 70,000 af per day. The Kocis and Dahlke approach would also requ ire capturing very high daily flows during the peak period: more than half the total would be extracted in 48 successive 30,000- plus acre-foot days between mid- February and the end of March, with a maximum daily withdrawal of 57,776 af/day on February 23, 2017 (Figure A19) . While such volumes are not excessive relative to the water moved during the valley’s prime irrigation season, they are concentrated in specific locations where conveyance limits are likely to be a challenge. In this context, it is import ant to bear in mind that there is a geographic mismatch between the supply of water available for recharge and the demand for this water: most of the available flows are in the San Joaquin River hydrologic region, while most overdraft is in the Tulare Lake region (see Figure A14). 30 How much additional infrastructure investment is warranted to capture available outflows will depend both on the costs of specific projects and the frequency with which the added capacity can be put to use. As a rough illustration of conveyance capacity requirements, Table A 2 summarizes the share of total water available for recharge that could have been captured over the 1986 -2017 period with different levels of available diversion capacity for DWR’s unadjusted “maximum project estimate” and Kocis’ and Dahlke’s approach. With 5,000 cfs of capacity —equivalent to roughly 10,000 acre -feet/day —it would have been possible to capture just under half of the available flows under each approach. Doubling that capacity to 10,000 cfs would have made it possible to capture another quarter of available flows. As a frame of reference, design capacity on the Friant -Kern canal —the key infrastructure for moving San Joaquin River flows south —is 10.5 taf/day leaving Millerton Lake, dropping to about 8 taf/day in southern Tulare County (US Bureau of Reclamation 2017). Subsidence has reduced capacity in that southern reach by 60 percent, to just 3.5 taf/day (Fitchette 2018). 31 And as we show in the main report, much of the existing capacity is already put to use in wet years. 30 To develop a rough estimate of daily surface water use valley -wide, we used data for 1998–2010 from the California Water Plan Update (DWR 2013). Total applied water use for agricultural and urban uses (including surface and groundwater) is about 18.8 maf/year: 16.5 maf for cropland i rrigation, 1.3 maf/year for urban uses, and 1 maf/year for conveyance losses. Surface water ranges from 40 to 70 percent of this total between dry and wet years, or 8 to 13 maf/year and 25 to 40 taf/day. But since most agricultural water is delivered in the peak irrigation season (April to September), daily deliveries could reach o ver 60 taf/day—leaving sign ificant spare capacity valley -wide in the winter and early spring months. A key challenge, however, is the much more limited capacity for moving water from the wetter San Joaquin River region to the Tulare Lake region, as discussed in the text and the next note. 31 Capacity in the southern part of the California Aqueduct, which could bring San Joaquin River flood flows through the Delta t o the Tulare Lake region, has fallen from 16.5 taf/day to 13 taf/day near Avenal in Kings County (Department of Water Res ources 2017). This limits high flow deliveries to Kern County and Southern California (Fitchette 2017). PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 28 TABLE A2 Conveyance needs to capture water available for recharge under different approaches DWR unadjusted “maximum project estimate” approach Kocis and Dahlke approach Diversion capacity Share captured (%) Total captured (maf /year ) Share captured (%) Total c aptured (maf /year ) 2,500 cfs 37.0 0.45 30.5 0.13 5,000 cfs 49.8 0.67 45.5 0.24 10,000 cfs 71.9 0.97 72.7 0.38 20,000 cfs 92.5 1.24 96.4 0.51 30,000 cfs 99.1 1.33 99.8 0.53 NOTES: Cfs is cubic feet per second. One cfs running over an entire day is approximately 2 acre-feet per day . The table examines flows available for water years 1986 –2017 , which results in slightly different totals than the long -term averages reported in the tex t (1986-2015 for DWR’s approach, and 1989 -2014 for Kocis and Dahlke’s approach) . We do not report values for DWR’s adjusted “maximum project estimate” since we do not have daily estimates for that approach. The role of surface storage also needs to be part of the infrastructure capacity analysis, because the ability to hold water above ground can make it possible to smooth out the peaks of water available for recharge. In addition to considering new investments in storage —including reservoirs such as Temperance Flat and holding ponds in areas without good recharge capacity —it will be important to consider reoperation of surface reservoirs to fill aquifers . Moving more dry -year storage out of reservoirs and into aquifers during the fall months can increase sp ace in reservoirs during the winter , and reduce the volumes that need to be released as flood flows during peak events. Such a strategy should add some capacity for recharging aquifers with existing conveyance and surface storage infrastructure. 32 32 DWR (2017c) has written a white paper and a factsheet o n Flo od-MAR (Managed Aquifer Recharge) that describes the need to increase this strategy. PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 29 REFERENCES Arnold, Brad, and Alvar Escriva -Bou . 2017. “The San Joaquin Valley’s Water Balance.” Technical Appendix A to Water Stress and a Changing San Joaquin Valley . 13 pp. Public Policy Institute of California, San Francisco, CA. Brush, C. F., Dogrul, E. C., and Kadir, T. N. 2013. Development and Calibrati on the California Central Valley Groundwater Surface Water Simulation Model (C2VSIM), California Department of Water Resources. Escriva -Bou, A., J. Mount , and J. Jezdimirovic. 2017. Dams in California . Public Policy Institute of California, DWR (California Department of Water Resources) . 1994. Bulletin 160 -93, California Water Plan Update , October 1994. DWR (California Department of Water Resources) . Bay -Delta Office. 2007. California Central Valley Unimpaired Flow Data, Fourth Edition. DWR (California Department of Water Resources) . 2013. California Water Plan, Update 2013 . DWR (California Department of Water Resources) . 2016 . Estimates of Natural and Unimpaire d Flows for the Central Valley of California: Water Years 1922- 2014. First Edition (Draft) . DWR (California Department of Water Resources) . 2017a . Draft Water Available for Replenishment Report. . DWR (California Department of Water Resources) . 2017b. NASA Report: San Joaquin Valley Land Continues to Sink - Groundwater Pumping Causes Subsidence, Damages Water Infrastructure. Press release. February 8. Department of Water Resources. 2017c. FLOOD-MAR: Using Flood Water for Managed Aquifer Recharge to Support Sustainable Water Resources . November. DWR (California Department of Water Resources) . 2018a. Water Available for Replenishment. Ap pendix A: Water Available for Replenishment Information and Estimates . DWR (California Department of Water Resources). 2018 b. Groundwater Information Center Interactive Map Application . Online resource. Faunt, Claudia C., Randall T. Hanson, and Kenneth Belitz. 2009. "Groundwater Availability of the Central Valley Aquifer." US Geological Survey professional paper 1766. Fitchette, Todd. 2018. “Subsidence Sinks Friant -Kern Canal Capacity by 60 Percent .” Western Farm Press. January 9. Fitchette, Todd. 2017. “How Land Subsidence Could Reduce Surface Water Deliveries in California. ” Western Farm Press. February 9. Gartrell, G., J. Mount, E. Hanak , and B. Gray . 2017. A New Approach to Accounting for Environmental Water. Insights from the Sacramento-San Joaquin Delta. Public Policy Institute of California. Hanak, Ellen, Jay Lund, Brad Arnold, Al var Escriva-Bou, Brian Gray, Sarge Green, Thomas Harter, Richard Howitt, Duncan MacEwan, Josué Medellín -Azuara, Peter Moyle, and Nathaniel Seavy. 2017. Water Stress and a Changing San Joaquin Valley . Public Policy Institute of California. Kocis, T. N., & Dahlke, H. E. 2017. Availability of High -Magnitude Streamflow for G roundwater Banking in the Central Valley, California. Environmental Research Letters, 12(8), 084009. Senay, G. B., Bohms, S., Singh, R. K., Gowda, P. H., Velpuri, N. M., Alemu, H., et al. 2013. “Operational Evapotranspiration M apping U sing Remote Sensing and W eather Datasets: A N ew Parameterization for the SSEB Approach.” Journal of the American Water Resources Association , 1–15. SWRCB (State Water Resources Control Board). 2016. “Recirculated Draft. Substitute Environmental Document in Support of Potential Changes to the Water Quality Control Plan for the San Francisco Bay -Sacramento San Joaquin Delta Estuary. San Joaquin River Flows and Southern Delta Water Quality.” September. US Bureau of Reclamation. 2017. Friant Dam Fact Sheet. December. US Department of Interior . 2004. Long-Term Central Valley Project Operations Criteria and Plan CVP-OCAP . June 30. PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 30 Appendix B : PPIC’s Groundwater Recharge Survey Introduction The appendix describes the design and deployment of our September 2017 survey of San Joaquin Valley water managers, provides information on the sample characteristics of respondents relative to the overall sample of water districts contacted, and describes our valley -wide estimates of groundwater recharge in 2017. The full survey questionnaire is presented at the end of the appendix. Survey Design and Deployment Survey Design The research team developed the survey questions with input from water managers, farmers, and other experts. This included individual conversations and several group discussions, held in Bakersfield, Madera, and via conference call with members of the Valley Agriculture Water Coalition, whose members include irrigation and water districts across the Valley. The i nstrument was developed in Qualtrics, an online survey platform. It included 19 questions, both multiple choice and text entry. The survey took approximately 10 minutes to complete. The survey was distributed through an email containing individualized links to general managers, engineers, or directors of urban water and agriculture water suppliers, on August 30, 2017. Respondents were required to provide the agency name at the beginning of the survey, but were reminded of the confidentiality of individual responses. Several follow up reminder emails were sent before the survey was officially closed on October 3, 2017. The survey asked respondents if they wanted to participate in a focus group to discuss initial survey findings. Those who expressed interest were invited to a meeting on the Fresno State campus on October 30, 2017, which was attended by about two dozen water managers. Sample Design The overall sample was 202 districts, including 151 agricultural water suppliers and 51 urban water suppliers. F or urban suppliers, we included all water suppliers meeting the state’s size threshold for urban water supplier (those required to prepare Urban Water Management Plans, and who were subject to the State Water Board’s water conservation mandate during the r ecent drought). 33 This group generally includes suppliers serving at least 3,000 connections. Given agriculture’s importance in water use in the valley, we cast a wide net to catch as many entities that manage surface and groundwater for agriculture as poss ible. This required consulting multiple data sources, because no agency maintains a comprehensive list of all entities providing water for agriculture. We drew on State Controller’s Office data for a list of special districts that raise revenues by selling water. To find districts that only have access to groundwater, we reviewed special district municipal service reviews and Integrated Regional Water Management (IRWM) plans. We developed a list of private agricultural water suppliers by compiling informati on about local rivers and their water rights-holders. 34 We also relied on 33 For more details on this list, see Mitchell et al. (2017), Buildi ng Drought Resilience in California’s Cities and Suburbs. Public Policy Institute of California. 34 Local managers helped us to eliminate some ditch companies from the sample that only providing ditch tending services, not ir rigation water supply. PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 31 membership lists of the v alley’s large “umbrella” organizations, like Friant Water Authority, Kaweah and St. Johns Rivers Association , Downstream Kaweah and Tule Rivers Association , Kern County Water Agency, Kings River Water Association, San Joaquin River Exchange Contractors , and San Luis Delta Mendota Water Authority. For additional cross- checking, we consulted Groundwater Sustainability Agency formation notices, State Water Projec t and Central Valley Project contractor lists, and US Bureau of Reclamation NEPA documentation. To our knowledge, the list that was created is the most comprehensive list of agricultural water providers in the San Joaquin Valley. Representativeness of Surv ey Responses We received a total of 81 survey responses, out of a sample of 202 agencies, a 40 percent overall response rate. We review sample representativeness through the lens of five different categories 1) supplier type classification (agricultural vs . urban), (2) service area size, (3) surface water availability, (4) existence of recharge basins, and (5) subregion. Response rates were somewhat higher from agricultural than urban suppliers, larger than smaller urban suppliers, suppliers with larger l and areas and more access to surface water , and suppliers with recharge basins. For these reasons, the sample likely over -represents districts that engage in active groundwater recharge, and particularly districts that have large, formal recharge programs. Supplier Type Classification Response rates were somewhat higher from agricultural than urban suppliers (42% versus 33%). Urban and agricultural entities have quite different patterns of recharge activity, so we sometimes present results separately for t hese two groups. Size One way of gauging sample representativeness was by service area acreage —a metric we developed for both urban and agricultural suppliers using information from GIS files and district planning documents. 35 For urban entities, we also looked at population served. We find that larger suppliers of both types are overrepresented in the sample (Figure B1). In the analysis, we do not break the data down along these lines, instead focusing on surface water deliveries, a measure of size that i s more directly relevant for recharge activities. 35 Urban district information was collected for use in the study by Mitchell et al. (2017) (see note 1). PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 32 FIGURE B1 Comparison of overall sample and survey respondents by size (land area and population) Surface Water Availability We anticipated that the availability of surface water would be an important factor in the ability to recharge. To provide a general means of comparison among districts, we sought information on average water deliveries during the 2005- 08 period, a span inc luding a mix of wet, dry, and normal years. We obtained the information from Central Valley Project, State Water Project, and local water agency and river association records. In a few cases we had to use data from earlier years or make judgment calls abou t how local river water is apportioned among users within associations. Response rates were higher for districts with greater access to surface water (Figure B2). Our analysis often distinguishes among districts by volume of surface water, which is related to recharge activity. FIGURE B2 Comparison of overall sample and survey respondents by surface water deliveries Existence of Recharge Basins The existence of a formal recharge program also plays a role in determining how entities responded to the very wet conditions in 2017. As an indicator of a formal recharge program, we collected publicly available information on which districts operate rech arge basins. Although districts can operate recharge programs using other methods, the presence of basins signals dedicated investment in recharge activities. We find that survey response rates were somewhat higher for districts that operate recharge basins (Figure B3). 2451 25 18 41 41 S (0 - 18,509) M (18,510 - 65,766) L (>65,766)Percent Urban population served All urban organizations Urban survey respondents 47 41 10 37 49 12 S (0 - 9,999) M (10,000 - 99,999) L (>99,999) Percent Land area (acres) All organizations Survey respondents 32 2428 16 22 2527 25 None S (0 - 9,999)M (10,000 - 99,999) L (>99,999) Percent Total surface water (AF) All organizations Survey respondents PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 33 FIGURE B3 Comparison of overall sample and survey respondents by existence of recharge basins Subregion Another useful distinction is geography, because both the extent of overdraft and the suitability of lands for recharge vary consider ably across the valley. We consider five subregions that broadly capture these differences: two in the San Joaquin River hydrologic region (northwest and northeast), and three within the Tulare Lake hydrologic region (southwest, southeast, and Kern basin) (Figure B3). 36 FIGURE B4 Subregions used for analysis Survey respondents are well-represented across all five subregions, but with districts in the Kern basin slightly over -represented, and districts in the southeast slightly underrepresented (Figure B4) . 36 Technical Appendix A describes how we formed these sub -regions from analysis regions for hydrologic models (Figure A13 and related discussion). 78 22 72 28 District does not operate a recharge basin District operates a recharge basin Percent All organizations Survey respondents PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 34 FIGURE B5 Comparison of overall sample and survey respondents by subregion Estimates of Valley-wide Recharge Volumes in 2017 To get a sense of how much water was recharged in calendar year 2017, we asked water managers to provide us with estimates of the total volume their district recharged and the methods used. We received valid responses from 66 districts (including 46 that r echarged water and 20 that did not). On -site active recharge reported for 2017 was 4.1 maf. Districts also reported that they banked 0.5 maf of water off -site through partnerships within the valley. To provide a rough estimate of the total volumes of activ e on-site recharge in the valley (6.5maf) and the proportion of this recharge stored for off -site parties (0.9 maf), we applied the results of a regression analysis for districts that supplied volume data to estimate volumes for the 136 districts that eith er did not report recharge volumes, or did not respond to our survey. This section provides details on data and methods used in the analysis. Data, V ariables, and Methods Our analysis was based on the districts that provided volumes of recharge in their survey response, or indicated that they did not actively recharge in 2017. Using that sample, we ran multivariate regressions with two dependent variables: (1) the natural log (ln) of total recharge volume, and (2) the natural log of on-site recharge volume. Off - site banking is calculated as the difference between these two values. Both models controlled for subregion, average surface water deliveries (in natural logs), whether the district is agricultural or urban, and whether the district has recharge basin s—all variables we could obtain for the entire sample, and which are associated with di fferences in recharge activity (Table B1). We set zero levels of recharge and surface water availability to 1 acre- foot for compatibility with the natural log format of these variables. Table B2 reports descriptive statistics for all variables used in the analysis. 1421 1430 20 19 22 1625 19 KR NENW SESW Percent All organizations Survey respondents PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 35 TABLE B1 Independent variables used in regression analysis Variable Description Average surface water supply Natural log of average annual surface water supply available to district from all sources District has recharge basin(s) Binary variable that equals 1 if the district owns and operates recharge basin Type of water supplier Binary variable indicating if district is an urban supplier (1) or agricultural supplier (0) Subregion Binary variables for four of the five subregions: Southeast, Southwest, Northeast, Northwest (with Kern as the omitted category) TABLE B2 Regression sample descriptive statistics Observations Mean Standard deviation Min Max Recharge volume (acre- feet) 66 69,509 129,848 1 570,000 Ln of recharge volume 66 7.0 4.9 0 13.3 On-site recharge volume (acre- feet) 66 61,679 129,897 1 570,000 Ln of on- site recharge volume 66 6.2 5.0 0 13.3 Average surface water supply (acre-feet) 66 72,379 139,944 1 920,332 Ln of average surface water supply 66 8.2 4.4 0 13.7 District has r echarge basin(s) 66 0.3 0.5 0 1 Urban water supplier 66 0.2 0.4 0 1 Kern basin (omitted subregion) 66 0.21 0.41 0 1 Southeast 66 0.21 0.41 0 1 Southwest 66 0.21 0.41 0 1 Northeast 66 0.20 0.40 0 1 Northwest 66 0.17 0.37 0 1 Results Table B3 presents the results of the regression analysis. Higher volumes of recharge are associated with having a recharge basin, being an agricultural supplier, having more access to surface water supply, and being located in regions with good recharge suitability. The equations have a good fit for cross -sectional regr essions (adjusted R 2 of 0.57 for total recharge and 0.51 for on- site recharge). PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 36 TABLE B3 Regression analysis estimating the total volume of recharge, and the volume of on -site recharge Regression estimating total volume of recharge Regression estimating on-site recharge Ln of average surface water supply 0.344** 0.102 (0.119) (0.131) District has recharge basin( s) 3.015** 6.102** (1.137) (1.250) Urban water supplier -4.871** -3.59* (1.263) (1.387) Southeast -1.539 1.209 (1.250) (1.374) Southwest -1.749 1.407 (1.362) (1.497) Northeast -0.597 2.176 (1.365) (1.500) Northwest -7.166** -2.407 (1.499) (1.6469) Constant 5.978** 3.500* (1.459) (1.602) Observations 66 66 Adjusted R-squared 0.57 0.51 NOTES: The table reports regression coefficients for each variable, with standard errors in parentheses. Statistical significance at the 99 and 95 percent levels are represented by ** and *, respectively . The total volume of recharge by district includes off-site recharge by othe r districts within the valley. We applied coefficients from Table B3 to the 136 districts that did not respond to our survey, or did not provide volumes of recharge. Table B4 provides descriptive statistics for that sample. Aggregating the reported recharg e and the estimates for non -respondents, our calculations provide rough estimates of total valley -wide on- site recharge (6.5 maf) and the proportion of this recharge that is banked for off -site parties within the valley (0.9 maf) in 2017. TABLE B4 Extra polation sample descriptive statistics Observations Mean Standard deviation Min Max Average surface water supply 136 37,451 87,892 1 580,958 Ln of average surface water supply 136 6 5 0 13 District has recharge basin(s) 136 0.2 0.4 0 1 Urban water supplier 136 0.3 0.5 0 1 Region 136 1.9 1.2 0 4 Kern basin (omitted subregion) 136 0.11 0.31 0 1 Southeast 136 0.35 0.48 0 1 Southwest 136 0.20 0.40 0 1 Northeast 136 0.21 0.41 0 1 Northwest 136 0.13 0.34 0 1 PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 37 Survey Questionnaire SAN JOAQUIN VALLEY GROUNDWATER RECHARGE SURVEY Thank you for agreeing to participate in this survey, which aims to obtain first -hand input from San Joaquin Valley water managers regarding groundwater recharge challenges, practices, and opportunities. The results will inform a public document that identifies policies, regulations, and funding tools to support groundwater recharge activities in the region. We have developed the ques tions in consultation with water managers from across the Valley. The survey is designed to take about 10 minutes to complete, and it covers the following topics: 1. Current and potential groundwater recharge methods in your service area, 2. Groundwater recharge activities this year, 3. Barriers to groundwater recharge (e.g., infrastructure, regulatory, financial issues), and 4. Priorities for expanding your system’s potential to engage in groundwater recharge. At the end of the survey, we also ask you to indicate if you would be interested in participating in a focus group discussions of preliminary results, to inform conclusions and recommendations in our report. We will maintain confidentiality of individual r esponses, and present results such that no organization- specific identifiers will be publicly available. If you choose to complete the survey on this form instead of completing the online version, please: - Send a scanned copy to jezdimirovic@ppic.org OR - Mail a paper copy to Jelena Jezdimirovic, PPIC; 500 Washington St., Suite 600; San Francisco CA 94111 Before we start, we’d like to confirm your organization’s name: PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 38 CURRENT AND POTENTIAL GROUNDWATER RECHARGE METHODS [Q1] What methods of active groundwater recharge does your organization currently practice—or envisage using or expanding in the future? (Please check all that apply.) Currently used Potential to expand Dedicated recharge basins Injection wells / ASR (aquifer storage and recovery) Recharging via unlined canals In-lieu recharge (i.e., using surface water instead of groundwater in wetter years) Recharge on cropland (e.g., extra irrigation, winter flooding) Recharge on fallowed farmland Recharge on open space lands Banking groundwater for my customers off-site in other districts Banking groundwater within my district on behalf of off-site parties Other (specify): ___________________________________________ Other (specify): ___________________________________________ None [Q1.1] Please feel free to list other recharge methods we've overlooked, and/or elaborate on any of the answers above. PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 39 ACTIVE GROUNDWATER RECHARGE THIS YEAR [Q2] Has your organization actively recharged groundwater this calendar year (2017)? Yes No [Q3] Please provide the estimated volume of water recharged to date and additional recharge expected this calendar year (by the end of 2017): Recharged to date: _________________________________________________ (acre -feet) Additional recharge expected: _________________________________________ ( acre-feet) [Q4] Please provide the approximate percentage of total recharge by type: Dedicated recharge basins % Injection wells / ASR (aquifer storage and recovery) % Recharging via unlined canals % In-lieu recharge (i.e., using surface water instead of groundwater in wetter years) % Recharge on cropland (e.g., extra irrigation, winter flooding) % Recharge on fallowed farmland % Recharge on open space lands % Banking groundwater for my customers off-site in other districts % Other (specify): ___________________________________________ % Other (specify): ___________________________________________ % Total 100% [Q5] How does the total expected volume of recharge this year compare with the amount you were able to recharge in 2011, which was also a wet year? Much more this year About the same Much less this year Unsure Not applicable: We did not recharge at all in 2011 PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 40 [Q5.1] Please feel free to comment on how your recharge activity this year compares with 2011. [Q6] What were the sources of water for recharge this year? CVP water (including Section 215 and Recovered Water Account) SWP water (including Article 21 water) Water from local rivers or streams (including flood flows) Urban stormwater runoff Recycled wastewater Water purchased from another party Other (specify): Other (specify): Please check all that apply [Q6.1] Please feel free to elaborate on any of the answers above regarding water sources. [Q7] Which of the following statements is most accurate for your system this year: We could have recharged more water with our existing recharge capacity We will have used all of our existing recharge capacity. Unsure. PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 41 BARRIERS TO GROUNDWATER RECHARGE THIS YEAR [Q8] Did you encounter any barriers to recharging groundwater this year? Please check all that apply Capacity constraints in system-wide conveyance (e.g., CVP or SW P canals) Capacity constraints in district-level recharge basins Other district-level capacity issues (e.g., conveyance to recharge locations) Irrigation constraints (e.g., inability to spread water on fields that use drip irrigation) Timing of water availability (e.g., too much water available at some times) Regulatory issues related to project construction (e.g., obtaining permits) Water rights or contracts for recharge water (e.g., SWRCB, CVP, SW P approvals) Permitting and approvals to convey recharge water Issues related to groundwater quality (e.g., waste discharge permits) Proposition 218-related difficulties raising funds to support recharge investments Price of recharge water too high District-level concerns about water migrating to neighboring areas Farmer concerns about benefiting adequately from on-farm recharge on their lands Farmer concerns about on-farm recharge because of crop health or yields None - we did not encounter any barriers [Q8.1] Please feel free to list other barriers to recharge we've overlooked, and/or elaborate on any of the answers above. PRIORITIES FOR EXPANDING GROUNDWATER RECHARGE [Q9] In your opinion, what are the top two to three priorities that need to be addressed to expand the potential of your organization to engage in groundwater recharge activities in the future? (You can refer to barriers listed above or other issues.) 1.______________________________________________________________________________ 2.______________________________________________________________________________ 3.______________________________________________________________________________ PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 42 CONTACT INFORMATION Thank you very much for your participation. So that we can contact you for follow ‐up questions or clarifications, please provide your name and contact information. This information is optional and will remain confidential. Name: Position: Phone: Email: Would you be interested in participating in a focus group discussion of preliminary survey results with other water managers? Yes No We welcome your comments on these topics, as well as comments regarding the questionnaire itself or clarifications of your responses. You may include any written comments in the space below. Thank you for taking the time to fill out this survey. We greatly appreciate your input, and will send you a copy of the final report when it is released. Please check below if you would like to subscribe to our publication alerts and blog: I would like to subscribe to PPIC Water Policy Center publication alerts I would like to subscribe to the PPIC Water Policy Center weekly blog The Public Policy Institute of California is dedicated to informing and improving public policy in California through independent, objective, nonparti san research. Public Policy Institute of California 500 Washington Street, Suite 600 San Francisco, CA 94111 T: 415.291.440 F: 415.291.4401 PPIC.ORG/WATER P PIC Sacramento Center Senator Office Building 1121 L Street, Suite 801 Sacramento, CA 95814 T: 916.440.1120 F: 916.440.1121" } ["___content":protected]=> string(172) "

Replenishing Groundwater in the San Joaquin Valley, Technical Appendix

" ["_permalink":protected]=> string(101) "https://www.ppic.org/publication/replenishing-groundwater-in-the-san-joaquin-valley/0418ehr-appendix/" ["_next":protected]=> array(0) { } ["_prev":protected]=> array(0) { } ["_css_class":protected]=> NULL ["id"]=> int(14664) ["ID"]=> int(14664) ["post_author"]=> string(1) "4" ["post_content"]=> string(0) "" ["post_date"]=> string(19) "2018-04-17 14:45:17" ["post_excerpt"]=> string(0) "" ["post_parent"]=> int(14554) ["post_status"]=> string(7) "inherit" ["post_title"]=> string(70) "Replenishing Groundwater in the San Joaquin Valley, Technical Appendix" ["post_type"]=> string(10) "attachment" ["slug"]=> string(16) "0418ehr-appendix" ["__type":protected]=> NULL ["_wp_attached_file"]=> string(20) "0418ehr-appendix.pdf" ["wpmf_size"]=> string(7) "1073349" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(88554) "Replenishing Groundwater in the San Joaquin Valley Technical Appendi ces CONTENTS Appendix A: Update of the San Joaquin Valley’s Water Balance and Estimate of Water Available for Recharge in 2017 Alvar Escriva-Bou and Ellen Hanak Appendix B: PPIC’s Groundwater Recharge Survey Ellen Hanak and Jelena Jezdimirovic with research support fr om Darcy Bostic and Henry McCann Supported with funding from the S. D. Bechtel, Jr. Foundation and Sustainable Conservation PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 2 Appendix A: Update of the San Joaquin Valley’s Water Balance and Estimate of Water Available for Recharge in 2017 Summary This appendix is divided in two main sections. The first updates e ach component of the San Joaquin Water balance (published in the Technical Appendix A of the 2017 PPIC report Water Stress and a Changing San Joaquin Valley ) to include 2016 and 2017 and provides estimates of how groundwater overdraft is distributed across subregions of the v alley. The second estimates how much water was available for recharge in 2017 replicating the approaches presented by two recently released studies (DWR 2017a and 2018a, and Kocis and Dahlke 2017). From the San Joaquin w ater balance update, we find that 2017 was an extraordinarily wet year when compared to the 1986- 2017 series. With high levels of imports and the highest local inflows in more than three decades, water availability was at peak levels. Although San Joaquin River outflow s were also the highest in the series, there was a net increase in storage of 7.8 maf (4.6 maf in surface reservoirs and 3.2 maf in aquifers). Considering the 1.75 maf of overdraft over the long -term, the 3.2 maf net recharge in the aquifers is roughly 5 m af more total recharge than the 32- year average, and 8 maf more than annual total recharge during the 2012- 16 drought. To obtain the subregional differences in overdraft in the v alley we use data from the two hydrological models that have been independentl y applied in the Central Valley to estimate historical groundwater budgets : CVHM (Faunt et al., 2009) and C2VSim ( Brush et al. 2013). We find that the overdraft estimates are much more significant in the Tulare Lake hydrologic region, and especially in Ker n County. Finally we obtain different estimates for how much water was available for recharge in the San Joaquin Valley in 2017. The approaches used by DWR and Kocis and Dahlke result in a wide range of additional volumes of water that could have been capt ured in 2017: from a high of 6.3 maf (DWR’s unadjusted “maximum project estimate” under our assumptions) to 3.7 maf (Kocis and Dahlke), to 0.88 maf (DWR’s unadjusted “best estimate”). These levels are substantially higher than the long -term averages estimated by these two studies (1.2 and 0.55 maf for DWR’s unadjusted and adjusted “maximum project estimate”; 0.46 maf for Kocis and Dahlke’s post -1989 “impaired period,” and 0.43 maf and 0.19 maf for DWR’s unadjusted and adjusted “best estimate”), reflecting t he fact that 2017 was an extraordinary year. Introduction After a multi-year drought, 2017 was one of California ’s wettest years since record -keeping began in the 1890s . In the San Joaquin Valley high runoff, in conjunction with initial work on sustainable groundwater management plans mandated by the 2014 Sustainable Groundwater Management Act (SGMA), triggered an unprecedented interest in groundwater recharge. In this appendix we seek to answer two questions that are crucial to understanding the potential for groundwater recharge in the San Joaquin Valley over the long term: how did 2017 compare to other years? And how much additional water could have been available for recharge? T his appendix provides information on data sources and methods used to 1) assess the annual water balances in the San Joaquin Valley for water years 1986-2017, including estimat es of groundwater recharge and overdraft at the subregional scale, and 2) estimate water available for recharge in 2017 . 1 1 In this technical appendix we always refer to water years: the 12 -month period between October 1st and September 30th of the following year. The water year is designated by the calendar year in which ends and which includes 9 of t he 12 months. PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 3 For the annual water balances, this report provides an update of Technical Appendix A of the 2017 PPIC report Water Stress and a Changing San Joaquin Valley , with an expanded analysis for water years 2016 and 2017. For that reason we only include here the sources of information used to update the numbers and the results . Readers should consult th e earlier document for a more detailed explanation of methods and results . This appendix is divided in to two main sections. The first updates e ach component of the San Joaquin Water balance t o include 2016 and 2017 and provides estimates of how groundwater overdraft is distributed across subregions of the v alley. The second estimates how much water was available for recharge in 2017, replicating the approaches presented by two recently released studies (DWR 2017a and 2018a, and Kocis and Dahlke 2017). Updating the San Joaquin Valley’s Water Balance A water balance is an accounting statement that estimates water inflows (including precipitation and other water flowing into the area), outflows (including net or consumptive water used locally and water flowing out of the area), and changes in water stored in surface reservoirs and aquifers . As with any mass balance, the sum of inflows, outflows, and changes in storage has to be zero every year, as shown in the following equation: ���������������������������� −������������ PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 4 FIGURE A1 San Joaquin River and Tulare Lake w atersheds Inflows Four types of water inflows are considered here: flows into the valley from local watersheds (including the Central and Southern Sierra Nevada and the C oast R ange), water from net precipitation on the v alley floor, direct diversions from the Delta, and water imported from other regions —especially through the Sacramento -San Joaquin Delta, but also much smaller imports through the Folsom South Canal. These inflows can be either used the same year or stored in surface reservoirs or aquifers for later use. Conversely, some of the water used in a given year can be obtained from withdrawals from surface and groundwater storage. Inflows from local watershe ds To assess inflows into the valley floor from local watersheds we used estimates of full natural or “unimpaired” flows for the main rivers and creeks in the region. 2 We obtained full natural flows (monthly volume) for the period of study from California Data Exchange Center ( CDEC ) for the major rivers in the valley . For some minor local watersheds where CDEC data were insufficient or in consistent we used 1986 –2003 data from DWR (2007). For these minor watersheds w e estimated the data 2 Unimpaired flow is the natural runoff of a watershed in the absence of storage regulation and stream diversions. Full natural flow is the natural runoff of a watershed that would have occurred prior to human influences on the watershed, such as storage , d iversions, or land use changes. In this report we used full natural flows from CDEC stations where available, and unimpaired flows from DWR (2007) for the remaining watersheds. PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 5 after 2003 calibrating a simplified rainfall -runoff model that mimics the runoff response to rainfall in the gaged catchments. Figure A 2 shows total annual inflow volumes from each watershed, with 2017 the highest in the period analyzed. 3 Although there are inflows from 15 local watersheds, five rivers account for nearly 70 percent of the average total: Tuolumne (19%), San Joaquin (17%), Stanislaus (11%), Kings (11%) , and Merced (10%). FIGURE A2 Inflows from local watersheds NOTES: The rivers shown in the bar chart are ordered geographically from south to north. The Kern, Tule, Kaweah , and Kings drain into the Tulare Lake Basin, and the remaining rivers drain into San Joaquin River Basin. The Tulare Lake Basin is a closed basin in mo st years, with all inflows remaining within the basin. The exception is very wet years, when excess flows drain into the San Joaquin River throu gh the James bypass (Fresno slough) . Net valley floor precipitation Total monthly precipitation on the v alley floor has been obtained by clipping the gridded datasets from PRISM Climate Group, Oregon State University, using a GIS layer of the study area. We assume that 15 percent of total monthly precipitation becomes net precipitation —the precipitation that does not evaporate and remains in the Valley, where it is used by plants, flows into streams, or percolates into an aquifer. This estimate is based on the water balances from DWR’s California Water Plan (DWR 2013) for the entire Central Valley, and in the Central Valley water balance of the C2VSim model (Brush et al. 2013). Relative to inflow from the Sierra (shown in Figure A2 ), net precipitation on the v alley floor is less variable across dry and wet y ears (Figure A3) and does not always mirror differences in runoff. 3 Water year 1983 was the wettest year on record in the region in terms of both runoff and precipitation. Water year 2017 was the second wettest for runoff (using CDEC data for full natural flows), and the sixth wettest for precipitation (using NOAA data). 0 5,000 10,000 15,000 20,000 25,000 30,000 19861987198819891990199119921993199419951996199719981999200020012002200320042005200620072008200920102011201220132014201520162017 Local Inflows (thousands of acre - fet) SJ Westside Minor SJ Eastside Minor Cosumnes River Mokelumne River Calaveras River Stanislaus River Tuolumne River Merced River Chowchilla River Fresno River San Joaquin River Kings River Kaweah River Tule River Kern River PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 6 FIGURE A3 Net precipitation on the v alley floor The inflows from local watersheds and the precipitation on the v alley floor are the flows that could be captured in surface reservoirs or used for recharging aquifers. Figure A4 demonstrates that most of these flows come from the San Joaquin River hydrologic region (73% of total inflows), whereas nearly a fifth (19%) comes from the Tulare Basin watersheds, and just 8 percent from precipitat ion on the valley floor. This is important because is an indicator of the amount of water available for recharge (as we show later in th is report). FIGURE A4 Inflows into the valley from local rivers and precipitation on the valley floor 0 500 1,000 1,500 2,000 2,500 Net precipitation (thousands of acre -feet) 0 5 10 15 20 25 30 35 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Millions of acre -feet Dry yearsTulare Basin HRSan Joaquin River HRPrecipitationAverage PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 7 Direct Delta diversions Central and southern Delta agricultural lands—which are considered part of the San Joaquin Valley —use water diverted directly from the Delta. Most of this water flows into the Delta, with lesser volumes from the San Joaquin River and other Delta tributaries. To account for this inflow we assume that all agricultural water use in the area of the Delta that is included in the San Joaquin Valley is directly diverted from the Delta and is estimated as the net or consumptive water use of these agricultural lands. The methods to obtain net or consumptive water use are described in a section below. W e use the same methodology to obtain the direct Delta diversions , but just for the area of the Delta that is in the San Joaquin Valley floor. D iversions average roughly 750 thousand acre -feet per year (Figure A5). FIGURE A5 Direct Delta diversions Imports from other regions Water is imported into the valley from Northern California through pumps in the south Delta. A small volume is also imported from the American River through the Folsom South Canal. Delta imports Delta imports are primarily from the Sacramento River with a small share from the San Joaquin River. These sources mix as they enter the Delta. Daily data for Delta imports from State Water Project (SWP) facilities (Banks Pumping Plant or Clifton Court Int ake), the Central Valley Project (CVP) facilities (C.W. “Bill” Jones Pumping Plant at Tracy ), and the Contra Costa Canal are obtained from Dayflow —a program that estimates daily average 0 200 400 600 800 1,000 1,200 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Direct Delta diversions (thousands of acre -feet) PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 8 Delta outflows (DWR, 2016) —for the period 1986-2015. The water years 2016 and 2017 were provided by DWR . 4 Imports through the Folsom South Canal The Folsom South Canal imports a small volume of water from the American River into the northeastern side of the San Joaquin Valley. Data on water imported through the South Folsom Canal is obtained from the US Bureau of Reclamation Central Valley Operations Annual Delivery Reports (Table 21). Figure A6 shows annual imports from the CVP and SWP pumps in the south Delta (Jones and Banks, respectively), the Contra Costa Canal, and Folsom South Canal. Total imports averaged 4.9 million acre -feet/year (53% from SWP and 47% from CVP), but in dry years imports from the Delta fall below 2 million acre-feet, as occurred in 2015. In 2017 the Delta imports were 6.4 maf, only a little less than the 2011 record of 6.6 maf. Imports through the South Folsom Canal averag ed around 26,000 acre-feet/year. FIGURE A6 Imports from other regions Outflows Four types of water outflows are considered: consumptive water use from evapotranspiration (water consumed by plants, and other evaporation to the atmosphere from the v alley floor), San Joaquin Valley outflows to the Delta, exports from San Joaquin River tributaries to Bay Area water users, and exports of imported water that enters the v alley. 5 Consumptive water use Similar to precipitation, monthly consumptive water use was obtained by clipping evapotr anspiration gridded datasets from the operational Simplified Surface Energy Balance model (Senay et al. 2013) using a GIS layer. This dataset provides high -resolution estimates of evapotranspiration, but only covers the period from 2000 to the 4 CVP deliveries under the Friant Division are not included in these totals—this water is diverted from the San Joaquin River at Millerton Lake to the Friant -Kern Canal, which delivers water to users on the east side of the Tulare Basin. For the purposes of this regional water balance, these are considered local flows. Contra Costa Canal imports are included because the pumps are inside the San Joaquin Valley floor, but as the Contra Costa Water District is entirely outside of the valley floor, these imports are later considered as exports to the Bay Area. 5 This may slightly understate total net water use insofar as it does not include water embodied in manufactured goods produced in the valley. 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Delta Imports (thousands of acre - feet) Folsom South CanalContra Costa Canal CVP pumpsSWP pumps PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 9 present. To obtain earlier values, we employed multiple regression analysis, using temperature, precipitation , and potential evapotranspiration from crops as independent variables (see Arnold and Escriva- Bou 2017 for more details ). The values for 2016 and 2017 were 14.6 and 15.3 maf respectively, at the higher end of the distribution— reflecting the greater volume of water available to evaporate from soils and transpire through vegetation in wetter years. To break down the consumptive use by end users (i.e., agriculture, urban, environmental, and natural landscapes) we obtained the average share of consumptive uses by end use from the California Water Plan (DWR 2013) , and then appl ied these shares to the estimates of evapotranspiration for the entire 3 2-year p eriod (Figure A7) . FIGURE A7 Annual evapotranspiration by end use NOTE: The category “other” includes evapotranspiration from natural landscapes not categorized as wetlands. San Joaquin Valley outflows San Joaquin River and other minor Delta tributaries outflows (Figure A8 ) were obtained from Dayflow and CDEC . These flows are reported as Delta inflows from the San Joaquin River (measured at the Vernalis gage) and eastern Delta inflow (including the Cosumnes and Mokelumne rivers and other minor creeks) . 6 The magnitude of the outflows in the San Joaquin Valley are highly influenced by the storage capacity within the basin , whereas minimum outflows in the river and the Delta are required by water quality and other environmental regulations. In 2017, the outf lows were 12.3 maf, the highest level in the past 32 years. 6 San Joaquin Valley outflows that are subsequently recaptured as either Delta imports or direct Delta diversions are counted as inflows to the region, and included in the Delta import measures presented below. 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000 20,000 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Evapotranspiration (thousands of acre -feet) AgricultureUrbanWetlandsOther PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 10 FIGURE A8 San Joaquin Valley outflows Exports from San Joaquin River tributaries The San Joaquin Valley exports several hundred thousand acre -feet of water annually from local sources to the San Francisco Bay Area (Figure A 9). The amount of export water does not vary significantly from year to year. Water from the Tuolumne River is stored in Hetch Hetchy Reservoir and then conveyed to the San Francisco Public Utilities Commission (SFPUC) service area, while water from the Mokelumne River is stored in Pardee and Camanche Reservoir s and conveyed to the East Bay Municipal Utility District (EBMUD) service area. 7 FIGURE A9 Water exports from San Joaquin River tributaries 7 EBMUD also has a contract with USBR to divert 100 million gallons a day at Freeport on the Sacramento Rive r in an emergency; this is not included here. 0 50 100 150 200 250 300 350 400 450 500 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Exports from local sources (thousands of acre -feet) SF BayEl Dorado ID 0 2 4 6 8 10 12 14 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Millions of acre -feet Dry yearsSan Joaquin Valley outflows PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 11 Data from 1998- 2010 was obtained from the California Water Plan (DWR 2013) , using the series “imports to the San Francisco Bay Hydrologic Region .” The remaining years have been estimated with a regression analysis using the unimpaired flows of the rivers for the entire series as an independent variable, and extrapolating the shares of the diverted data with respect to the unimpaired flows for the 1998- 2010 dataset. 8 Average annual exports f or 1986-2015 are 0.38 maf/year. Figure A 9 also includes a small volume of exports to the El Dorado Irrigation District in the Sacramento River hydrologic region. This water is diverted from Jenkinson Lake on Sly Park Creek, a tributary of the Cosumnes River, and averages 18,000 af per year. Diversion data is from CDEC (station CCN). Note the low variability of water exports from the San Joaquin Valley tributaries compared with the high variability of inflows to these tributaries shown in Figure A 2. Exports of Delta imports Some Delta imports that enter the Valley through the CVP and SWP pumps are delivered to the San Francisco Bay area, the Central Coast, and Southern California (Figure A10 ).  Exports to the San Fra ncisco Bay Region: This includes water from two points of diversion: (1) through the South Bay Aqueduct from the South Bay Pumping Plant (data are from the SWP Annual Reports of Operations), and (2) through the Contra Costa Canal (data are from USBR Centra l Valley Project Annual Re ports of Operations, Table 21).  Exports to the Central Coast: This includes water from two points of diversion: (1) through Las Perillas Pumping Plant on the California Aqueduct (from the State W ater Project Annual Reports of Operation s: Table 1 ); and (2) through the Pacheco Tunnel. 9  Exports to Southern California: This includes water delivered through the A.D. Edmonston Pumping Plant on the California Aqueduct (from DWR SWP Annual R eports of Ops: Table 1 Totals). FIGURE A10 Exports of Delta imports 8 Note that for this update we did not get actual data for these exports, so 2016 and 2017 values were estimated using a regression analysis. 9 San Luis Reservoir Operations, f rom DWR SWP Annual Reports of Oper ations: p ost -2000 Reports, Table 15 Annual San Luis Joint -Use Facility Total , and pre -2000 Reports, Table 13 San Luis Reservoir Operations Total Outflow (Pacheco Tunnel). 2015 data obtained from Santa Clara Valley Wa ter District urban water supply data and data for years 2016 and 2017 was obtained directly from DWR. Some water going through the Pacheco Tunnel goes to the Santa Clara Valley Water District and could be included in the exports to the San Francisco Region. As we do not have access to sufficient data to separate the flows that remain in the Central Coast and those that go to the San Francisco Bay hydrologic r egion, we include them as exports to the Central Coast. 0 500 1,000 1,500 2,000 2,500 3,000 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Exports of Imported Water (thousands of acre -feet) Southern CaliforniaCentral CoastSF Bay PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 12 Changes in S torage Two main storage types are considered: surface reservoirs and water stored in aquifers. Data exist for water stored in surface reservoirs. Changes in aquifer storage are estimated as the volume required to c lose the water balance for the v alley. Changes in reservoir storage From CDEC, we obtained data for monthly storage for 13 major reservoirs in the San Joaquin Valley (Figure A11 ). Annual storage change is the water stored at the beginning of the prior water year minus the storage at the beginning of a new water year (October 1) . The reliability of data for the 1986 and 1987 water years is lower because there were more gaps in the series. 10 FIGURE A11 Water stored in the 13 major reservoirs in the San Joaquin Valley NOTE: The 13 major reservoirs are: New Melones (NML), Don Pedro (DNP), Lake McClure (EXC), Pine Flat (PNF), Lake Isabella (ISB), Success Dam (SCC), Kaweah Lake (TRM), Millerton Lake (MIL), Eastman Lake (BUC), Mariposa Reservoir (MAR), Bear Reservoir (BAR), New H ogan Lake (NHG), and Camanch e Reservoir (CMN). TAF is thousand acre -feet. To include water stored in other minor reservoirs, we obtained the change in total water stored from the California Water Plan (DWR 2013) for the two hydrologic regions and extrapolated the other years using a linear relationship between total changes in storage and changes in the storage in the 13 major reservoirs ( see details in Arnold and Escriva -Bou, 2017) . Total net changes in annual surface storage are shown below in Figure A 12B. The long-term average change in surface water stored is roughly zero. Changes in water stored in aquifers Finally, we determine the change in water stored in aquifers as the residual that closes the water balance for the v alley. In short, the net available water supply (from in flow, precipitation, and changes in storage) must equal the 10 As discussed below, the estimates for 1986 appear low relative to precipitation and runoff year, which may in turn result in an overestimate of the level of aquifer storage in that year . 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 Oct-87 Oct-88 Oct-89 Oct-90 Oct-91 Oct-92 Oct-93 Oct-94 Oct-95 Oct-96 Oct-97 Oct-98 Oct-99 Oct-00 Oct-01 Oct-02 Oct-03 Oct-04 Oct-05 Oct-06 Oct-07 Oct-08 Oct-09 Oct-10 Oct-11 Oct-12 Oct-13 Oct-14 Oct-15 Oct-16 Oct-17 Reservoir Storage (thousands of acre- feet) ISBSCCTRMPNFMILBUCMARBAREXCDNPNMLNHGCMN PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 13 net v olume of water used within the v alley (consumptive use) or exported. The mass balance equation can be formulated as: ∆ �������� 11 Inflows from local supplies can be used directly as surface diversions but also indirectly by replenishing aquifers and pumpe d later as sustainable groundwater use. According to C2VSim groundwater budgets at the subregional scale (Brush et al. 2013), rivers in the San Joaquin Valley contri bute to roughly 0.5 maf of groundwater recharge on average for the period 1973 -2009. Also there is water recharged from unlined canals and percolation of excess irrigation water on agricultural lands. 12 Net imports into the valley are the total water imported from other regions minus the water that is exported to other regions. 13 As mentioned above, reservoir data for the years 1986 and 1987 had low reliability. This could explain why in 1986, a wet year, our model is not reporting an increase in reservoir storage, but a n unusually large increase in water stored in aquifers, relative to the volume of runoff and imported water deliveries. If the balance between surface and groundwater storage in 1986 was in line with more recent wet years, average overdraft for the entire period would be closer to 1.9 maf/year. 14 California has approximately 850 maf to 1.3 billion acre -feet of groundwater in storage (DWR, 1994), and about 45 maf of surface storage (PPIC, 2017) 15 More information about the different changes in temporal patterns can be found in Hanak et al., 2017. PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 14 -8 -6 -4 -2 0 2 4 6 8 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 - 1,0 00 2,0 00 3,0 00 4,0 00 5,0 00 6,0 00 7,0 00 8,0 00 9,0 00 10 ,00 0 Millions of acre -feet Dry yearsNet Reservoir WithdrawalNet Reservoir Recharge 0 5 10 15 20 25 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Millions of acre -feet Dry yearsImportsLocal suppliesDirect Delta DiversionsTotal Net Use -8 -6 -4 -2 0 2 4 6 8 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 - 1,0 00 2,0 00 3,0 00 4,0 00 5,0 00 6,0 00 7,0 00 8,0 00 9,0 00 10 ,00 0 Millions of acre -feet Dry yearsNet GW WithdrawalNet Recharge FIGURE A12 Components of the San Joaquin Valley’s annual water balance A) Water supplies and net water use B) Change in reservoir storage C) Change in groundwater storage PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 15 Estimating Groundwater Overdraft at the S ubregional Level in the San Joaquin Valley Two hydrological models have been independently applied in the Central Valley to estimate historical groundwater budgets : CVHM (Faunt et al., 2009) and C2VSim ( Brush et al. 2013). The models use precipitation, surface- water inflows, and surface- water diversion input data that is either similar or identical (Dogrul et al., 2011) to estimate subregional water budgets , with a focus on assessing agricultural groundwater pumping . Using C2VSim and CVHM models we determined the groundwater overdraft for 14 subregions of the San Joaquin Valley by assessing the decrease in groundwater storage over the lo ng term (Figure A13). Table A1 shows the results for both models for the 14 subregions within the San Joaquin Valley for the period 1975- 2003, and also an average of both models. FIGURE A13 Hydrologic regions and m odel subregions used by both C2VSim and C VHM models SOURCE: Brush et al. (2013). NOTE: The numbers refer to analysis subregions commonly used in hydrologic and economic modeling in the Central Valley. The results of Table A 1 show that the models are in fairly close agreement at the regional scale (plus or minus 20%) , but with some significant discrepancies for individual subregions (note for instance the discrepancies in regions 9 and 18). PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 16 TABLE A1 Change in groundwater storage at the subregional scale ( 1975-2003 ) Subregion Change in ground water storage (in taf/year) C2VSim CVHM Average 8* -52 -3 -28 9* -17 96 39 10 -8 1 -4 11 -11 -5 -8 12 5 32 19 13 -97 -134 -116 14 -271 -177 -224 15 -59 -170 -115 16 -14 -140 -77 17 -124 -107 -116 18 156 -408 -126 19 -210 -192 -201 20 -232 -159 -195 21 -411 -255 -333 Total -1,345 -1,622 -1,483 SOURCE: C2VSim results at the subregional scale were obtained directly from model outputs, while CVHM results were provided by Stephen Maples. NOTE S: * The actual change in groundwater storage in regions 8 and 9 is twice the value we show in this table. We made this adjustment because only a part of regions 8 and 9 are in the San Joaquin River hydrological region. T o provide a rough estimate of the distribution of overdraft across different parts of the valley , we aggregated the results, combining the 14 modeling subregions into five (Table A2) . The San Joaquin River hydrologic region includes the northwest (sub region s 9 and 10 ) and northeast ( subregions 8, 11, 12, and 13), and the Tulare Lake hydrologic region includes the southwest (sub regions 14 and 15 ), the southeast ( subregions 16, 17, and 18 ), and the Kern basin ( sub regions 19, 20, and 21) . This aggregation reinforces the discrepancy between the two models in the southeast: whereas the C2VSim model estimates a positive change of groundwater storage over the long - term, CVHM estimates overdraft of more than 6 50 taf/year. This subregion includes the Kings, Kaweah, and Tule basins, all of which DWR consi ders to be “critically overdrafted.” Moreover, parts of these regions are experiencing significant subsidence. 16 It therefore appears that C2VSim is underestimating pumping and/ or overestimating recharge in this part of the valley. Finally, we also made a s imple adjustment to include years after 2003 , which were drier. For the northwest, where the average change in groundwater storage from the two models is positive, we assumed that there is no contribution to overdraft at the subregional scale (although the re could still be local issues). For the areas with overdraft , we multiplied the 1975- 2003 average by a factor that accounts for the additional overdraft for the San Joaquin Valley from our 1986- 2015 water balance. Using this procedure we are implicitly assuming that overdraft increased at a similar pace after 2003 across affected areas. This rough approximation is supported by the widespread decline in groundwater tables since the mid -2000s shown in DWR’s Groundwater Information Center Interactive Map Application (DWR 2018b). 16 See DWR’s identification of Critically Overdrafted Basins and NASA JPL report on Subsidence in the Central Valley PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 17 TABLE A2 Change in groundwater storage for the five aggregated subregions Subregion Change in groundwater storage (in taf/year) C2VSim 1975-2003 CVHM 1975-2003 Average 1975-2003 Adjusted long- term average* Northwest -25 96 16 0 Northeast -155 -110 -133 -156 Southwest -330 -347 -339 -397 Southeast 18 -655 -318 -374 Kern -853 -606 -729 -856 Total -1,345 -1,622 -1,483 -1,783 Total in overdrafted subregions* -1320 -1,718 -1,519 -1,783 NOTES: *The adjusted long -term average allocates the additional overdraft we find in our 1986 -2015 water balance relative to the 1975 -2003 total for the four subregions experiencing average overdraft (1,783 – 1, 519 = 264 acre -feet) in proportion to their overdraft in 1975 -2003. We assume that the northwest is in average long -term balance. FIGURE A14 Groundwater overdraft at the subregional scale (in thousands of acre -feet per year) NOTE: Given the uncertainties of the estimations, this figure shows results rounded to the nearest 10 taf. PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 18 Estimating Water Available for Recharge in 2017 This section explains estimates of how much water was available for recharge in the San Joaquin Valley in 2017. W e first describe two different approaches from studies that estimated how much water is available to replenish aquifers in California. Then we try to reproduce each approach with the available data from the San Joaquin Valley for 2017 and the othe r years examined in our water balance . Finally, we discuss the results. DWR’s Approach In January 2017, t he Department of Water Resources released Water Available for Replenishment Report Draft (DWR 2017 a). 17 The study estimated water available for replenishment for each of the state’s 10 hydrologic regions and 56 planning areas. Basically , the methodology defines a minimum streamflow requirement that accounts for environmental needs and is not available for re charge. The remaining outflow can be diverted for recharge if there is infrastructure capacity for doing so. 18 To account for uncertainties in its estimates, DWR shows a range of values based upon a range of instream flow s and project capacity. These includ e a “best estimate” that uses the instream flow requirement and an estimate of the existing project capacity, a “sensitivity range ,” and a “maximum project estimate” that illustrates the maximum potential diversion for recharge. DWR’s approach caps the max imum amount of water that can be diverted above the minimum instream flow requirement based on the infrastructure needed to divert this water. To do this they define a conceptual “project diversion capacity” which is the total capacity of the diversions th at would be used to recharge groundwater. To obtain the “best estimate” they define as project diversion capacity the “existing project diversion capacity”, which is sized based on the water right with the largest single point of diversion capacity on a given river/stream. The sensitivity ranges are obtained by using tw o and half times the “existing project capacity.” Finally they obtain the “maximum project estimate” by assuming that there is no infrastructure constraint in the amount of water that can be diverted. The methodology uses historical daily gage data (1930 through 2015 for the San Joaquin River) to develop a n initial set of estimates of available water. It then adjusts these using a simulation model (W ater Evaluation and Planning [ WEAP ]) to account for changes in outflows reflecting current water demands and operations . In our analysis, we refer to the results based on the historical gage data as the “unadjusted” estimates, and those adjusted using WEAP as the “adjusted” estimates. 19 For the S an Joaquin Valley, t he WEAP adjustment reduces the volume available for recharge significantly—to 45 percent of the unadjusted volumes . In Figure A1 5 we show a conceptual depiction of DWR ’s “best estimate” and “maximum project estimate.” In this illustration, which uses actual outflow from Vernalis for the water year 2016, a minimum instream flow requirement of 1,000 cfs is defined for any day of the year, and an existing project diversion capacity is 500 cfs. The shaded region in yellow would be the “best estimate” of water available for replenishment. When there is no upper limit assumed for project diversion capacity, we use the “maximum project estimate.” 17 The fi nal report is expected to be released in 2018. Its appendices have already been updated and published, and we use the estimat es in updated appendix A in our analysis (DWR 2018a). 18 DWR’s Water Available for Replenishment Report Draft does not explicitly identify instream flow requirements n or the existing project capacity data, so to replicate that study’s approach we developed rough estimates below that result in comparable levels of water available for recharge. 19 DWR refers to the estimates based on the historical gage data as the “gage d ata outflow” and the estimates using the WEAP adjustment as “water available for replenishment .” WEAP results are only available at a monthly —not daily—timescale. PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 19 FIGURE A15 C onceptual depiction of DWR’s water available for replenishment For the San Joaquin River hydrological region, DWR ’s “best estimate” of water available for replenishment is 190 taf/year and the “maximum project estimate” , which includes also the water captured under the best estimate assumptions, is 550 taf/year. These totals include wate r in the Tulare Lake region (47 taf/year in the “best estimate” and 213 taf/year in the “maximum project estimate”) because excess surface water from the Tulare Lake region flows into the San Joaquin River. The Kocis and D ahlke Approach: High-M agnitude Streamflow Also in 2017, Tiffany N. Kocis and Helen E. Dahlke, two scientists at UC Davis, published a paper estimating the availability of high -magnitude streamflow for recharge in the Central Valley (Kocis and Dahlke 2017). Their approach is different from DWR’s. U sing historic daily streamflow records available from USGS , they assume that the days on which out flows are within the highest 10 percent (or above the 90th percentile) over a 90-year period exceed both environmental flow requirements and current surface water allocations under California water rights, and could therefore be used to recharge aquifers. Using USGS data for the San Joaquin River at Vernalis for the period Oct ober 19 23 through Sep tember 2014, Kocis and Dahlke obtained a 90 th percentile flow of 11,600 cfs. Figure A16 depicts the high-magnitude flows for the p ast 10 years using their approach, which assumes that the volume in excess of the 90 th percentile could be taken on days with higher outflow . 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 Oct-15Nov-15Dec-15Jan-16Feb-16Mar-16Apr-16May-16Jun-16Jul-16Aug-16Sep-16 San Joaquin River flow (cfs) Extra water available for replenishment under "maximum project estimate" Water available for replenishment "best estimate" Daily stream flow Instream flow requirement Project(s) diversion capacity PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 20 FIGURE A16 Water available fo r recharge in the San Joaquin Valley in 2008 -17 using the Kocis and Dahlke approach For the period 1989- 2014—a subset of years included in this analysis that better reflects current water demand and operations in the v alley—Kocis and Dahlke estimate that an average of 1.3 maf would be available for recharge in years with high-magnitude flows . Because these flows only occur in 36 percent of all years (9 out of 25) , this translates to a long- term average across all years of 455 taf/year. 20 Water Available for Recharge in 2017 In the water balance analysis above, we estimate that there was 3.2 maf net groundwater recharge in the valley in water year 2017, roughly 5 maf more recharge than the 32- year average and 8 maf more than annual recharge during the 2012- 16 drought . To estimate how much additional water was available for recharge in 2017, we try to reproduce both approaches using data for 2017. DWR’s approach DWR’s Water Available for Replenishment Report Draft does not explicitly report the underlying assumptions for instream flow requirements to meet environmental needs and downstream water rights, nor the assumptions on existing project capacity to divert additional flows for recharge . Also, the authors used a complex modeling approach at the planning area scale that would be difficult to replicate precisely . In an effort to provide a rough replic ation of DWR’s approach, w e made some simplifying assumptions. We treat ed the entire San Joaquin Valley as a single basin, with a unique outlet: the San Joaquin River at Vernalis . 21 This facilitates a comparison with the results of Kocis and Dahlke, who also used the Vernalis gage to estimate high -magnitude flow s. 22 W e selected values for instream flow requirements and existing pro ject capacity that 20 The study refers to 1989 -2014 as the “post -impairment” period. For the entire 1923 -2014 per iod of analysis, they estimate an average of 1.46 maf is available in years with high magnitude flows, which occur in 47 percent of years, translating to a long- term average across all years of 686 taf/year. 21 The gage near Vernalis is the point commonly u sed to measure flows leaving the San Joaquin River and Tulare Lake regions for the Delta. Vernalis flow does not include the northernmost San Joaquin hydrologic region tributaries: the Calaveras, the Mokelumne, and the Cosumnes Rivers. As a result, these outflow values differ somewhat from the totals shown in the valley- wide water balance, shown in Figure A8. 22 Kocis and Dahlke (2017) also provide estimates for some gages upstream of Vernalis (for interactive maps and graphics, see recharge.ucdavis.edu/starr ). 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 Oct-07Apr-08Oct-08Apr-09Oct-09Apr-10Oct-10Apr-11Oct-11Apr-12Oct-12Apr-13Oct-13Apr-14Oct-14Apr-15Oct-15Apr-16Oct-16Apr-17 San Joaquin River flows (cfs) High-magnitude flows available for recharge San Joaquin River actual outflow 90th percentile flow PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 21 result in a similar long-term average level of water available for recharge for the region under DWR’s best estimate and maximum project estimate scenarios. Because these assumptions may differ from what DWR used, we cannot ensure that t he annual and seasonal hydrographs we obtain are the same as those obtained by DWR , even though our estimates of the long -term average annual volumes available for recharge are similar . We define the minimum instream flow requirement based on Delta conditions, which is also the focus of DWR’s approach for the San Joaquin hydrologic region (DWR 2018a). 23 First , we assume that water is available only when the Delta is in “excess conditions .” 24 Then we define a minimum daily Delta outflow requirement —as a parameter to calibrate —to meet water quality standards and other environmental regulations in the Delta . The water above this threshold is the uncaptured water —flows that exceed the capacity of existing storage and diversion facilities and regulatory demands. 25 Because this water comes from both the Sacramento and the San Joaquin hydrologic regions, we distribute the total amount based on the share of total inflow from each river on a daily basis using Dayflow data . With this approach, and w ithout capping the amount that can be diverted, we are reproducing DWR’s unadjusted “maximum project estimate .” Calibrating the minimum Delta outflow parameter to 10,000 cfs, we determine that the water available for recharge over the long term (using data for 1986 -2015) is 1.216 taf/year, just 0.3 percent lower than DWR’s unadjusted “maximum project estimate.” 26 To reach DWR’s unadjusted “best estimate” we had to cap the amount of water that can be diverted through the “existing project”— i.e., available conveyance infrastructure. Using a cap of 2 ,250 cfs we get a best estimate of 421 taf/year, just 2.1 % lower than DWR’s unadjusted best estimate. Figure A17 shows how this approach performs from 2000 to 2010. The adjusted best estimate and maximum project es timate—adjusting outflows based on the WEAP model simulations —are 190 and 550 taf/year respectively. 23 Water availability in the Central Valley depends on meeting environmental conditions downstream in the Sacramento -San Joaquin Delta. DWR’s approach for the Sacramento River and San Joaquin River hydrologic regions takes this into account (DWR 2018). 24 Exc ess water conditions occur in periods when releases from upstream reservoirs plus unregulated flows exceed Sacramento Valley in- basin uses plus exports (USDOI 2004). 25 See Gartrell et al. (2017) for a detailed explanation of uncaptured water. Note that this method of determining water available for recharge assumes that upstream environmental flow regulations do not require additional water to reach the Delta, beyond that needed to meet regulations in the Delta and downstream water demands. 26 On average, t his calibration results in just a slightly higher estimate of uncaptured outflow (+11%), with a similar temporal distribution , to estimates of uncaptured Delta outflow by Gartrell et al. (2017). PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 22 FIGURE A17 Water available for recharge for 2000 -1 0 using DWR’s unadjusted “best estimate” and “maximum project estimate” with our “instream flow requirement” and “existing project capacity” assumptions NOTES: Both “best estimate” and “maximum proje ct estimate ” are matched to DWR’s unadjusted estimates (“g age data outflow ”) at Vernalis . Using th is same approach f or water year 2017, we obtain an unadjusted best estimate of 882 taf and maximum proj ect estimate of 6.3 maf (Figure A1 8). Using the same WEAP adjustment factor that DWR applied to the long- term averages, the adjusted values are 397 taf and 2.9 maf, respectively. However, the unadjusted values are likely more appropriate for recent years, since actual outflow data already reflect current water use and operations. 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 Oct-00 Apr-01 Oct-01 Apr-02 Oct-02 Apr-03 Oct-03 Apr-04 Oct-04 Apr-05 Oct-05 Apr-06 Oct-06 Apr-07 Oct-07 Apr-08 Oct-08 Apr-09 Oct-09 Apr-10 Flows (taf/day) Extra water available for recharge under "maximum project capacity" Water available for recharge "best estimate" San Joaquin River actual outflow PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 23 FIGURE A18 Water available for recharge for 2017 using DWR’s unadjuste d “best estimate” and “maximum project estimates” (with our own “instream flow requirement” and “existing project capacity” assumptions ) NOTES: B oth “best estimate” and “maximum proje ct estimate” are matched to DWR’s unadjusted estimates (“g age data outfl ow”) at Vernalis . The Kocis and Dahlke approach Once the 90th percentile daily flow is defined, determining high -magnitude flows available for recharge is straightforward. Every flow since January 11, 2017 to July 11, 2017 was above the 90th percentile, so all th e water above this threshold ( 11,600 cfs) would be available: 3.7 maf for the entire water year (Figure A19). 0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 90,000 Oct-16 Nov-16 Dec-16 Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Jul-17 Aug-17 Sep-17 San Joaquin River flows (taf/day) Extra water available for recharge under "maximum project capacity" Water available for recharge "best estimate" San Joaquin River actual outflow PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 24 FIGURE A19 Water available for recharge for 2017 using the Kocis and Dahlke approach Discussion of the E stimates of W ater Available for Recharge Comparison of findings for 2017 The approaches used by DWR and Kocis and Dahlke result in a wide range of additional volumes of water that could have be en captured in 2017, from a high of 6. 3 maf (DWR’s unadjusted “maximum project estimate” under our assumptions) to 3.7 maf (Kocis and Dahlke) , to 0.88 maf (DWR’s unadjusted “best estimate”) . 27 These levels are substantially higher than the long -term averages estimated by these two studies ( 1.2 and 0.55 maf for DWR ’s unadjusted and adjusted “maximum project estimate” ; 0.46 maf for Kocis and Dahlke’s post -1989 “impaired period,” and 0.43 maf and 0.19 maf for DWR’s unadjusted and adjusted “best estimate”) , reflecting the fact that 2017 was an extraordinary year . When com paring the different approaches and the actual monthly out flows in 2017 (Figure A2 0), it is noteworthy that DWR ’s maximum project estimate approach takes more flows almost every month than Kocis and Dahlke’s approach . This is because in this extremely wet year, the Delta requirements were already satisfied with flows much lower than the 90 th percentile used as a cut -off by Kocis and Dahlke. 27 As described further below, we do not include the DWR adjusted estimates here because the 2017 gage data already reflect current demands and operations. 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 Oct-16Nov-16Dec-16Jan-17Feb-17Mar-17 Apr-17May-17Jun-17Jul-17Aug-17Sep-17 San Joaquin River flows (cfs) High-magnitude flows available for recharge San Joaquin River actual outflow 90th percentile flow PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 25 FIGURE A20 Water available for recharge under the different approaches compared to San Joaqui n River actual outflow in 2017 NOTE: B oth the “best estimate” and “maximum proje ct estimate” are matched to DWR’s unadjusted estimates (“g age data outflow ”) at Vernalis . Comparison across the past three decades The comparison of total water available for r echarge across the past three decades further highlights the differen ces between the approaches (Figure A2 1). Whereas DWR’s approach identifies additional water for recharge in most years, Kocis and Dahlke’s approach finds additional water mostly in w et years. 28 This again reflects the different threshold s above which streamflow is considered available. Whereas Kocis and Dahlke only consider flows above the 90 th percentile, DWR ’s approach varies with Delta conditions. 28 Under this approach, there can also be high magnitude flows during big storm events in other year types, but this results in a very small share of the total wat er available for recharge: 2 percent of all high magnitude flows at Vernalis for the 1989-2014 “unimpaired period”, and 6 percent of all high magnitude flows for the 1971 -2017 period. 0.0 0.5 1.0 1.5 2.0 Oct-16 Nov-16 Dec-16 Jan-17 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Jul-17 Aug-17 Sep-17 Water available for recharge vs San Joaquin River Actual Outflow (millions of acre-feet) DWR best estimateDWR maximum project estimateKocis & DahlkeActual outflow PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 26 FIGURE A21 Water available for recharge under the different approaches Reasons for using DWR’s unadjusted estimates for planning purposes It is also worth noting that while the adjustment DWR makes to account for current demands and operations makes sense when using gage data over a very long timeframe, it might be unnecessary for recent years. 29 As the water balance data presented earlier in this appendix demonstrate (Figures A7 and A12), water use in the San Joaquin Valley has not changed much since the mid- 1980s. Nor have there been any significant changes in water infrastructure over this period. For this reason, the unadjusted data may better capture the potential water available for recharge than the adjusted data that DWR presents as final estimates. One caveat is that the method used here to determin e water available for recharge—focusing on conditions in the Delta —assumes that upstream environmental flow regulations do not require additional water to reach the Delta beyond that needed to meet regulations in the Delta and downstream water demands. T he method used here also shows how simple accounting rules can help determine when water i s available for recharge. F urther development of this approach—taking into account the uncaptured water in the Delta as defined in Gartrell et al . (2017) plus some definition of upstream environmental requirements in the San Joaquin River system —would mak e it possible to estimate water available for groundwater recharge in the San Joaquin Valley. To be useful as a management tool , the flows that need to stay in the river and the amount that can be taken out of the river for recharg e purposes would need to be published in a timely manner (maybe weekly) . Near-term 29 Recall that DWR uses daily average Delta outflow from water year 1930 through 2015 to calculate its long- term averages. 0 1 2 3 4 5 6 7 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Water available for recharge (millions of acre- feet) DWR's adjusted best estimateDWR's adjusted maximum project estimate DWR's unadjusted maximum project estimateKocis & Dahlke PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 27 forecasts of storm conditions and river flows could be linked to this information to help managers prepare for recharge opportunities, which often need to be acted on quickly . This would be a way to overcome regulatory hurdles that impede the expansion of recharge practices in the v alley. Capacity constraints on capturing additional flows In 2017, most rivers in the San Joaquin Valley were above flood stage for months, and water managers faced significant capacity constraints in their water storage and conveyance infrastructure. Capturing significant additional flows in wet years like 2017 will likely require new infrastructure investments, in addition to new water rights permits to divert and store the water. As an example, more than half of the 6.3 maf of water available for recharge in 2017 under DWR’s unadjusted “maximum project estimat e” would have needed to be taken in two months (February and March), with diversions totaling more than 30,000 af/day (Figure A18). On seven of these days, diversions would have had to exceed 70,000 af per day. The Kocis and Dahlke approach would also requ ire capturing very high daily flows during the peak period: more than half the total would be extracted in 48 successive 30,000- plus acre-foot days between mid- February and the end of March, with a maximum daily withdrawal of 57,776 af/day on February 23, 2017 (Figure A19) . While such volumes are not excessive relative to the water moved during the valley’s prime irrigation season, they are concentrated in specific locations where conveyance limits are likely to be a challenge. In this context, it is import ant to bear in mind that there is a geographic mismatch between the supply of water available for recharge and the demand for this water: most of the available flows are in the San Joaquin River hydrologic region, while most overdraft is in the Tulare Lake region (see Figure A14). 30 How much additional infrastructure investment is warranted to capture available outflows will depend both on the costs of specific projects and the frequency with which the added capacity can be put to use. As a rough illustration of conveyance capacity requirements, Table A 2 summarizes the share of total water available for recharge that could have been captured over the 1986 -2017 period with different levels of available diversion capacity for DWR’s unadjusted “maximum project estimate” and Kocis’ and Dahlke’s approach. With 5,000 cfs of capacity —equivalent to roughly 10,000 acre -feet/day —it would have been possible to capture just under half of the available flows under each approach. Doubling that capacity to 10,000 cfs would have made it possible to capture another quarter of available flows. As a frame of reference, design capacity on the Friant -Kern canal —the key infrastructure for moving San Joaquin River flows south —is 10.5 taf/day leaving Millerton Lake, dropping to about 8 taf/day in southern Tulare County (US Bureau of Reclamation 2017). Subsidence has reduced capacity in that southern reach by 60 percent, to just 3.5 taf/day (Fitchette 2018). 31 And as we show in the main report, much of the existing capacity is already put to use in wet years. 30 To develop a rough estimate of daily surface water use valley -wide, we used data for 1998–2010 from the California Water Plan Update (DWR 2013). Total applied water use for agricultural and urban uses (including surface and groundwater) is about 18.8 maf/year: 16.5 maf for cropland i rrigation, 1.3 maf/year for urban uses, and 1 maf/year for conveyance losses. Surface water ranges from 40 to 70 percent of this total between dry and wet years, or 8 to 13 maf/year and 25 to 40 taf/day. But since most agricultural water is delivered in the peak irrigation season (April to September), daily deliveries could reach o ver 60 taf/day—leaving sign ificant spare capacity valley -wide in the winter and early spring months. A key challenge, however, is the much more limited capacity for moving water from the wetter San Joaquin River region to the Tulare Lake region, as discussed in the text and the next note. 31 Capacity in the southern part of the California Aqueduct, which could bring San Joaquin River flood flows through the Delta t o the Tulare Lake region, has fallen from 16.5 taf/day to 13 taf/day near Avenal in Kings County (Department of Water Res ources 2017). This limits high flow deliveries to Kern County and Southern California (Fitchette 2017). PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 28 TABLE A2 Conveyance needs to capture water available for recharge under different approaches DWR unadjusted “maximum project estimate” approach Kocis and Dahlke approach Diversion capacity Share captured (%) Total captured (maf /year ) Share captured (%) Total c aptured (maf /year ) 2,500 cfs 37.0 0.45 30.5 0.13 5,000 cfs 49.8 0.67 45.5 0.24 10,000 cfs 71.9 0.97 72.7 0.38 20,000 cfs 92.5 1.24 96.4 0.51 30,000 cfs 99.1 1.33 99.8 0.53 NOTES: Cfs is cubic feet per second. One cfs running over an entire day is approximately 2 acre-feet per day . The table examines flows available for water years 1986 –2017 , which results in slightly different totals than the long -term averages reported in the tex t (1986-2015 for DWR’s approach, and 1989 -2014 for Kocis and Dahlke’s approach) . We do not report values for DWR’s adjusted “maximum project estimate” since we do not have daily estimates for that approach. The role of surface storage also needs to be part of the infrastructure capacity analysis, because the ability to hold water above ground can make it possible to smooth out the peaks of water available for recharge. In addition to considering new investments in storage —including reservoirs such as Temperance Flat and holding ponds in areas without good recharge capacity —it will be important to consider reoperation of surface reservoirs to fill aquifers . Moving more dry -year storage out of reservoirs and into aquifers during the fall months can increase sp ace in reservoirs during the winter , and reduce the volumes that need to be released as flood flows during peak events. Such a strategy should add some capacity for recharging aquifers with existing conveyance and surface storage infrastructure. 32 32 DWR (2017c) has written a white paper and a factsheet o n Flo od-MAR (Managed Aquifer Recharge) that describes the need to increase this strategy. PPIC.ORG/WATER Technical Appendix A Replenishing Groundwater in the San Joaquin Valley 29 REFERENCES Arnold, Brad, and Alvar Escriva -Bou . 2017. “The San Joaquin Valley’s Water Balance.” Technical Appendix A to Water Stress and a Changing San Joaquin Valley . 13 pp. Public Policy Institute of California, San Francisco, CA. Brush, C. F., Dogrul, E. C., and Kadir, T. N. 2013. Development and Calibrati on the California Central Valley Groundwater Surface Water Simulation Model (C2VSIM), California Department of Water Resources. Escriva -Bou, A., J. Mount , and J. Jezdimirovic. 2017. Dams in California . Public Policy Institute of California, DWR (California Department of Water Resources) . 1994. Bulletin 160 -93, California Water Plan Update , October 1994. DWR (California Department of Water Resources) . Bay -Delta Office. 2007. California Central Valley Unimpaired Flow Data, Fourth Edition. DWR (California Department of Water Resources) . 2013. California Water Plan, Update 2013 . DWR (California Department of Water Resources) . 2016 . Estimates of Natural and Unimpaire d Flows for the Central Valley of California: Water Years 1922- 2014. First Edition (Draft) . DWR (California Department of Water Resources) . 2017a . Draft Water Available for Replenishment Report. . DWR (California Department of Water Resources) . 2017b. NASA Report: San Joaquin Valley Land Continues to Sink - Groundwater Pumping Causes Subsidence, Damages Water Infrastructure. Press release. February 8. Department of Water Resources. 2017c. FLOOD-MAR: Using Flood Water for Managed Aquifer Recharge to Support Sustainable Water Resources . November. DWR (California Department of Water Resources) . 2018a. Water Available for Replenishment. Ap pendix A: Water Available for Replenishment Information and Estimates . DWR (California Department of Water Resources). 2018 b. Groundwater Information Center Interactive Map Application . Online resource. Faunt, Claudia C., Randall T. Hanson, and Kenneth Belitz. 2009. "Groundwater Availability of the Central Valley Aquifer." US Geological Survey professional paper 1766. Fitchette, Todd. 2018. “Subsidence Sinks Friant -Kern Canal Capacity by 60 Percent .” Western Farm Press. January 9. Fitchette, Todd. 2017. “How Land Subsidence Could Reduce Surface Water Deliveries in California. ” Western Farm Press. February 9. Gartrell, G., J. Mount, E. Hanak , and B. Gray . 2017. A New Approach to Accounting for Environmental Water. Insights from the Sacramento-San Joaquin Delta. Public Policy Institute of California. Hanak, Ellen, Jay Lund, Brad Arnold, Al var Escriva-Bou, Brian Gray, Sarge Green, Thomas Harter, Richard Howitt, Duncan MacEwan, Josué Medellín -Azuara, Peter Moyle, and Nathaniel Seavy. 2017. Water Stress and a Changing San Joaquin Valley . Public Policy Institute of California. Kocis, T. N., & Dahlke, H. E. 2017. Availability of High -Magnitude Streamflow for G roundwater Banking in the Central Valley, California. Environmental Research Letters, 12(8), 084009. Senay, G. B., Bohms, S., Singh, R. K., Gowda, P. H., Velpuri, N. M., Alemu, H., et al. 2013. “Operational Evapotranspiration M apping U sing Remote Sensing and W eather Datasets: A N ew Parameterization for the SSEB Approach.” Journal of the American Water Resources Association , 1–15. SWRCB (State Water Resources Control Board). 2016. “Recirculated Draft. Substitute Environmental Document in Support of Potential Changes to the Water Quality Control Plan for the San Francisco Bay -Sacramento San Joaquin Delta Estuary. San Joaquin River Flows and Southern Delta Water Quality.” September. US Bureau of Reclamation. 2017. Friant Dam Fact Sheet. December. US Department of Interior . 2004. Long-Term Central Valley Project Operations Criteria and Plan CVP-OCAP . June 30. PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 30 Appendix B : PPIC’s Groundwater Recharge Survey Introduction The appendix describes the design and deployment of our September 2017 survey of San Joaquin Valley water managers, provides information on the sample characteristics of respondents relative to the overall sample of water districts contacted, and describes our valley -wide estimates of groundwater recharge in 2017. The full survey questionnaire is presented at the end of the appendix. Survey Design and Deployment Survey Design The research team developed the survey questions with input from water managers, farmers, and other experts. This included individual conversations and several group discussions, held in Bakersfield, Madera, and via conference call with members of the Valley Agriculture Water Coalition, whose members include irrigation and water districts across the Valley. The i nstrument was developed in Qualtrics, an online survey platform. It included 19 questions, both multiple choice and text entry. The survey took approximately 10 minutes to complete. The survey was distributed through an email containing individualized links to general managers, engineers, or directors of urban water and agriculture water suppliers, on August 30, 2017. Respondents were required to provide the agency name at the beginning of the survey, but were reminded of the confidentiality of individual responses. Several follow up reminder emails were sent before the survey was officially closed on October 3, 2017. The survey asked respondents if they wanted to participate in a focus group to discuss initial survey findings. Those who expressed interest were invited to a meeting on the Fresno State campus on October 30, 2017, which was attended by about two dozen water managers. Sample Design The overall sample was 202 districts, including 151 agricultural water suppliers and 51 urban water suppliers. F or urban suppliers, we included all water suppliers meeting the state’s size threshold for urban water supplier (those required to prepare Urban Water Management Plans, and who were subject to the State Water Board’s water conservation mandate during the r ecent drought). 33 This group generally includes suppliers serving at least 3,000 connections. Given agriculture’s importance in water use in the valley, we cast a wide net to catch as many entities that manage surface and groundwater for agriculture as poss ible. This required consulting multiple data sources, because no agency maintains a comprehensive list of all entities providing water for agriculture. We drew on State Controller’s Office data for a list of special districts that raise revenues by selling water. To find districts that only have access to groundwater, we reviewed special district municipal service reviews and Integrated Regional Water Management (IRWM) plans. We developed a list of private agricultural water suppliers by compiling informati on about local rivers and their water rights-holders. 34 We also relied on 33 For more details on this list, see Mitchell et al. (2017), Buildi ng Drought Resilience in California’s Cities and Suburbs. Public Policy Institute of California. 34 Local managers helped us to eliminate some ditch companies from the sample that only providing ditch tending services, not ir rigation water supply. PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 31 membership lists of the v alley’s large “umbrella” organizations, like Friant Water Authority, Kaweah and St. Johns Rivers Association , Downstream Kaweah and Tule Rivers Association , Kern County Water Agency, Kings River Water Association, San Joaquin River Exchange Contractors , and San Luis Delta Mendota Water Authority. For additional cross- checking, we consulted Groundwater Sustainability Agency formation notices, State Water Projec t and Central Valley Project contractor lists, and US Bureau of Reclamation NEPA documentation. To our knowledge, the list that was created is the most comprehensive list of agricultural water providers in the San Joaquin Valley. Representativeness of Surv ey Responses We received a total of 81 survey responses, out of a sample of 202 agencies, a 40 percent overall response rate. We review sample representativeness through the lens of five different categories 1) supplier type classification (agricultural vs . urban), (2) service area size, (3) surface water availability, (4) existence of recharge basins, and (5) subregion. Response rates were somewhat higher from agricultural than urban suppliers, larger than smaller urban suppliers, suppliers with larger l and areas and more access to surface water , and suppliers with recharge basins. For these reasons, the sample likely over -represents districts that engage in active groundwater recharge, and particularly districts that have large, formal recharge programs. Supplier Type Classification Response rates were somewhat higher from agricultural than urban suppliers (42% versus 33%). Urban and agricultural entities have quite different patterns of recharge activity, so we sometimes present results separately for t hese two groups. Size One way of gauging sample representativeness was by service area acreage —a metric we developed for both urban and agricultural suppliers using information from GIS files and district planning documents. 35 For urban entities, we also looked at population served. We find that larger suppliers of both types are overrepresented in the sample (Figure B1). In the analysis, we do not break the data down along these lines, instead focusing on surface water deliveries, a measure of size that i s more directly relevant for recharge activities. 35 Urban district information was collected for use in the study by Mitchell et al. (2017) (see note 1). PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 32 FIGURE B1 Comparison of overall sample and survey respondents by size (land area and population) Surface Water Availability We anticipated that the availability of surface water would be an important factor in the ability to recharge. To provide a general means of comparison among districts, we sought information on average water deliveries during the 2005- 08 period, a span inc luding a mix of wet, dry, and normal years. We obtained the information from Central Valley Project, State Water Project, and local water agency and river association records. In a few cases we had to use data from earlier years or make judgment calls abou t how local river water is apportioned among users within associations. Response rates were higher for districts with greater access to surface water (Figure B2). Our analysis often distinguishes among districts by volume of surface water, which is related to recharge activity. FIGURE B2 Comparison of overall sample and survey respondents by surface water deliveries Existence of Recharge Basins The existence of a formal recharge program also plays a role in determining how entities responded to the very wet conditions in 2017. As an indicator of a formal recharge program, we collected publicly available information on which districts operate rech arge basins. Although districts can operate recharge programs using other methods, the presence of basins signals dedicated investment in recharge activities. We find that survey response rates were somewhat higher for districts that operate recharge basins (Figure B3). 2451 25 18 41 41 S (0 - 18,509) M (18,510 - 65,766) L (>65,766)Percent Urban population served All urban organizations Urban survey respondents 47 41 10 37 49 12 S (0 - 9,999) M (10,000 - 99,999) L (>99,999) Percent Land area (acres) All organizations Survey respondents 32 2428 16 22 2527 25 None S (0 - 9,999)M (10,000 - 99,999) L (>99,999) Percent Total surface water (AF) All organizations Survey respondents PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 33 FIGURE B3 Comparison of overall sample and survey respondents by existence of recharge basins Subregion Another useful distinction is geography, because both the extent of overdraft and the suitability of lands for recharge vary consider ably across the valley. We consider five subregions that broadly capture these differences: two in the San Joaquin River hydrologic region (northwest and northeast), and three within the Tulare Lake hydrologic region (southwest, southeast, and Kern basin) (Figure B3). 36 FIGURE B4 Subregions used for analysis Survey respondents are well-represented across all five subregions, but with districts in the Kern basin slightly over -represented, and districts in the southeast slightly underrepresented (Figure B4) . 36 Technical Appendix A describes how we formed these sub -regions from analysis regions for hydrologic models (Figure A13 and related discussion). 78 22 72 28 District does not operate a recharge basin District operates a recharge basin Percent All organizations Survey respondents PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 34 FIGURE B5 Comparison of overall sample and survey respondents by subregion Estimates of Valley-wide Recharge Volumes in 2017 To get a sense of how much water was recharged in calendar year 2017, we asked water managers to provide us with estimates of the total volume their district recharged and the methods used. We received valid responses from 66 districts (including 46 that r echarged water and 20 that did not). On -site active recharge reported for 2017 was 4.1 maf. Districts also reported that they banked 0.5 maf of water off -site through partnerships within the valley. To provide a rough estimate of the total volumes of activ e on-site recharge in the valley (6.5maf) and the proportion of this recharge stored for off -site parties (0.9 maf), we applied the results of a regression analysis for districts that supplied volume data to estimate volumes for the 136 districts that eith er did not report recharge volumes, or did not respond to our survey. This section provides details on data and methods used in the analysis. Data, V ariables, and Methods Our analysis was based on the districts that provided volumes of recharge in their survey response, or indicated that they did not actively recharge in 2017. Using that sample, we ran multivariate regressions with two dependent variables: (1) the natural log (ln) of total recharge volume, and (2) the natural log of on-site recharge volume. Off - site banking is calculated as the difference between these two values. Both models controlled for subregion, average surface water deliveries (in natural logs), whether the district is agricultural or urban, and whether the district has recharge basin s—all variables we could obtain for the entire sample, and which are associated with di fferences in recharge activity (Table B1). We set zero levels of recharge and surface water availability to 1 acre- foot for compatibility with the natural log format of these variables. Table B2 reports descriptive statistics for all variables used in the analysis. 1421 1430 20 19 22 1625 19 KR NENW SESW Percent All organizations Survey respondents PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 35 TABLE B1 Independent variables used in regression analysis Variable Description Average surface water supply Natural log of average annual surface water supply available to district from all sources District has recharge basin(s) Binary variable that equals 1 if the district owns and operates recharge basin Type of water supplier Binary variable indicating if district is an urban supplier (1) or agricultural supplier (0) Subregion Binary variables for four of the five subregions: Southeast, Southwest, Northeast, Northwest (with Kern as the omitted category) TABLE B2 Regression sample descriptive statistics Observations Mean Standard deviation Min Max Recharge volume (acre- feet) 66 69,509 129,848 1 570,000 Ln of recharge volume 66 7.0 4.9 0 13.3 On-site recharge volume (acre- feet) 66 61,679 129,897 1 570,000 Ln of on- site recharge volume 66 6.2 5.0 0 13.3 Average surface water supply (acre-feet) 66 72,379 139,944 1 920,332 Ln of average surface water supply 66 8.2 4.4 0 13.7 District has r echarge basin(s) 66 0.3 0.5 0 1 Urban water supplier 66 0.2 0.4 0 1 Kern basin (omitted subregion) 66 0.21 0.41 0 1 Southeast 66 0.21 0.41 0 1 Southwest 66 0.21 0.41 0 1 Northeast 66 0.20 0.40 0 1 Northwest 66 0.17 0.37 0 1 Results Table B3 presents the results of the regression analysis. Higher volumes of recharge are associated with having a recharge basin, being an agricultural supplier, having more access to surface water supply, and being located in regions with good recharge suitability. The equations have a good fit for cross -sectional regr essions (adjusted R 2 of 0.57 for total recharge and 0.51 for on- site recharge). PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 36 TABLE B3 Regression analysis estimating the total volume of recharge, and the volume of on -site recharge Regression estimating total volume of recharge Regression estimating on-site recharge Ln of average surface water supply 0.344** 0.102 (0.119) (0.131) District has recharge basin( s) 3.015** 6.102** (1.137) (1.250) Urban water supplier -4.871** -3.59* (1.263) (1.387) Southeast -1.539 1.209 (1.250) (1.374) Southwest -1.749 1.407 (1.362) (1.497) Northeast -0.597 2.176 (1.365) (1.500) Northwest -7.166** -2.407 (1.499) (1.6469) Constant 5.978** 3.500* (1.459) (1.602) Observations 66 66 Adjusted R-squared 0.57 0.51 NOTES: The table reports regression coefficients for each variable, with standard errors in parentheses. Statistical significance at the 99 and 95 percent levels are represented by ** and *, respectively . The total volume of recharge by district includes off-site recharge by othe r districts within the valley. We applied coefficients from Table B3 to the 136 districts that did not respond to our survey, or did not provide volumes of recharge. Table B4 provides descriptive statistics for that sample. Aggregating the reported recharg e and the estimates for non -respondents, our calculations provide rough estimates of total valley -wide on- site recharge (6.5 maf) and the proportion of this recharge that is banked for off -site parties within the valley (0.9 maf) in 2017. TABLE B4 Extra polation sample descriptive statistics Observations Mean Standard deviation Min Max Average surface water supply 136 37,451 87,892 1 580,958 Ln of average surface water supply 136 6 5 0 13 District has recharge basin(s) 136 0.2 0.4 0 1 Urban water supplier 136 0.3 0.5 0 1 Region 136 1.9 1.2 0 4 Kern basin (omitted subregion) 136 0.11 0.31 0 1 Southeast 136 0.35 0.48 0 1 Southwest 136 0.20 0.40 0 1 Northeast 136 0.21 0.41 0 1 Northwest 136 0.13 0.34 0 1 PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 37 Survey Questionnaire SAN JOAQUIN VALLEY GROUNDWATER RECHARGE SURVEY Thank you for agreeing to participate in this survey, which aims to obtain first -hand input from San Joaquin Valley water managers regarding groundwater recharge challenges, practices, and opportunities. The results will inform a public document that identifies policies, regulations, and funding tools to support groundwater recharge activities in the region. We have developed the ques tions in consultation with water managers from across the Valley. The survey is designed to take about 10 minutes to complete, and it covers the following topics: 1. Current and potential groundwater recharge methods in your service area, 2. Groundwater recharge activities this year, 3. Barriers to groundwater recharge (e.g., infrastructure, regulatory, financial issues), and 4. Priorities for expanding your system’s potential to engage in groundwater recharge. At the end of the survey, we also ask you to indicate if you would be interested in participating in a focus group discussions of preliminary results, to inform conclusions and recommendations in our report. We will maintain confidentiality of individual r esponses, and present results such that no organization- specific identifiers will be publicly available. If you choose to complete the survey on this form instead of completing the online version, please: - Send a scanned copy to jezdimirovic@ppic.org OR - Mail a paper copy to Jelena Jezdimirovic, PPIC; 500 Washington St., Suite 600; San Francisco CA 94111 Before we start, we’d like to confirm your organization’s name: PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 38 CURRENT AND POTENTIAL GROUNDWATER RECHARGE METHODS [Q1] What methods of active groundwater recharge does your organization currently practice—or envisage using or expanding in the future? (Please check all that apply.) Currently used Potential to expand Dedicated recharge basins Injection wells / ASR (aquifer storage and recovery) Recharging via unlined canals In-lieu recharge (i.e., using surface water instead of groundwater in wetter years) Recharge on cropland (e.g., extra irrigation, winter flooding) Recharge on fallowed farmland Recharge on open space lands Banking groundwater for my customers off-site in other districts Banking groundwater within my district on behalf of off-site parties Other (specify): ___________________________________________ Other (specify): ___________________________________________ None [Q1.1] Please feel free to list other recharge methods we've overlooked, and/or elaborate on any of the answers above. PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 39 ACTIVE GROUNDWATER RECHARGE THIS YEAR [Q2] Has your organization actively recharged groundwater this calendar year (2017)? Yes No [Q3] Please provide the estimated volume of water recharged to date and additional recharge expected this calendar year (by the end of 2017): Recharged to date: _________________________________________________ (acre -feet) Additional recharge expected: _________________________________________ ( acre-feet) [Q4] Please provide the approximate percentage of total recharge by type: Dedicated recharge basins % Injection wells / ASR (aquifer storage and recovery) % Recharging via unlined canals % In-lieu recharge (i.e., using surface water instead of groundwater in wetter years) % Recharge on cropland (e.g., extra irrigation, winter flooding) % Recharge on fallowed farmland % Recharge on open space lands % Banking groundwater for my customers off-site in other districts % Other (specify): ___________________________________________ % Other (specify): ___________________________________________ % Total 100% [Q5] How does the total expected volume of recharge this year compare with the amount you were able to recharge in 2011, which was also a wet year? Much more this year About the same Much less this year Unsure Not applicable: We did not recharge at all in 2011 PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 40 [Q5.1] Please feel free to comment on how your recharge activity this year compares with 2011. [Q6] What were the sources of water for recharge this year? CVP water (including Section 215 and Recovered Water Account) SWP water (including Article 21 water) Water from local rivers or streams (including flood flows) Urban stormwater runoff Recycled wastewater Water purchased from another party Other (specify): Other (specify): Please check all that apply [Q6.1] Please feel free to elaborate on any of the answers above regarding water sources. [Q7] Which of the following statements is most accurate for your system this year: We could have recharged more water with our existing recharge capacity We will have used all of our existing recharge capacity. Unsure. PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 41 BARRIERS TO GROUNDWATER RECHARGE THIS YEAR [Q8] Did you encounter any barriers to recharging groundwater this year? Please check all that apply Capacity constraints in system-wide conveyance (e.g., CVP or SW P canals) Capacity constraints in district-level recharge basins Other district-level capacity issues (e.g., conveyance to recharge locations) Irrigation constraints (e.g., inability to spread water on fields that use drip irrigation) Timing of water availability (e.g., too much water available at some times) Regulatory issues related to project construction (e.g., obtaining permits) Water rights or contracts for recharge water (e.g., SWRCB, CVP, SW P approvals) Permitting and approvals to convey recharge water Issues related to groundwater quality (e.g., waste discharge permits) Proposition 218-related difficulties raising funds to support recharge investments Price of recharge water too high District-level concerns about water migrating to neighboring areas Farmer concerns about benefiting adequately from on-farm recharge on their lands Farmer concerns about on-farm recharge because of crop health or yields None - we did not encounter any barriers [Q8.1] Please feel free to list other barriers to recharge we've overlooked, and/or elaborate on any of the answers above. PRIORITIES FOR EXPANDING GROUNDWATER RECHARGE [Q9] In your opinion, what are the top two to three priorities that need to be addressed to expand the potential of your organization to engage in groundwater recharge activities in the future? (You can refer to barriers listed above or other issues.) 1.______________________________________________________________________________ 2.______________________________________________________________________________ 3.______________________________________________________________________________ PPIC.ORG/WATER Technical Appendix B Replenishing Groundwater in the San Joaquin Valley 42 CONTACT INFORMATION Thank you very much for your participation. So that we can contact you for follow ‐up questions or clarifications, please provide your name and contact information. This information is optional and will remain confidential. Name: Position: Phone: Email: Would you be interested in participating in a focus group discussion of preliminary survey results with other water managers? Yes No We welcome your comments on these topics, as well as comments regarding the questionnaire itself or clarifications of your responses. You may include any written comments in the space below. Thank you for taking the time to fill out this survey. We greatly appreciate your input, and will send you a copy of the final report when it is released. Please check below if you would like to subscribe to our publication alerts and blog: I would like to subscribe to PPIC Water Policy Center publication alerts I would like to subscribe to the PPIC Water Policy Center weekly blog The Public Policy Institute of California is dedicated to informing and improving public policy in California through independent, objective, nonparti san research. Public Policy Institute of California 500 Washington Street, Suite 600 San Francisco, CA 94111 T: 415.291.440 F: 415.291.4401 PPIC.ORG/WATER P PIC Sacramento Center Senator Office Building 1121 L Street, Suite 801 Sacramento, CA 95814 T: 916.440.1120 F: 916.440.1121" ["post_date_gmt"]=> string(19) "2018-04-17 21:45:17" ["comment_status"]=> string(4) "open" ["ping_status"]=> string(6) "closed" ["post_password"]=> string(0) "" ["post_name"]=> string(16) "0418ehr-appendix" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2018-04-17 14:50:52" ["post_modified_gmt"]=> string(19) "2018-04-17 21:50:52" ["post_content_filtered"]=> string(0) "" ["guid"]=> string(59) "http://www.ppic.org/wp-content/uploads/0418ehr-appendix.pdf" ["menu_order"]=> int(0) ["post_mime_type"]=> string(15) "application/pdf" ["comment_count"]=> string(1) "0" ["filter"]=> string(3) "raw" ["status"]=> string(7) "inherit" ["attachment_authors"]=> bool(false) }