California Agriculture
California Agriculture
California Agriculture
University of California
California Agriculture

Archive

Modeling guides groundwater management in a basin with river–aquifer interactions

Share using any of the popular social networks Share by sending an email Print article
Share using any of the popular social networks Share by sending an email Print article

Authors

Laura Foglia, UC Davis
Jakob Neumann, Technical University Darmstadt
Douglas G. Tolley, UC Davis
Steve Orloff, UC Cooperative Extension
Richard L. Snyder, UC Cooperative Extension and UC Davis
Thomas Harter, UC Davis

Publication Information

California Agriculture 72(1):84-95. https://doi.org/10.3733/ca.2018a0011

Published online March 13, 2018

PDF  |  Citation  |  Permissions  |  Cited by 0 articles

Author Affiliations show

Abstract

The Sustainable Groundwater Management Act (SGMA) of 2014 seeks to maintain groundwater discharge to streams to support environmental goals. In Scott Valley, in Siskiyou County, the Scott River and its tributaries are an important salmonid spawning habitat, and about 10% of average annual Scott River stream flow comes from groundwater. The local groundwater advisory committee is developing groundwater management alternatives that would increase summer and early fall stream flows. We developed a model to provide a framework to evaluate those alternatives. We first created a water budget for the Scott Valley groundwater basin and integrated the detailed, spatiotemporally distributed water budget results into a computer model of the basin that simultaneously accounted for groundwater flow, stream flow and landscape water fluxes. Different conceptual representations (using the MODFLOW RIV package and MODFLOW SFR package) of the stream–aquifer boundary provided significantly different results in the seasonal dynamics of groundwater–surface water fluxes. As groundwater sustainability agencies draw up plans to meet SGMA requirements, they must choose and test simulation tools carefully.

Full text

Management of California's water supplies serves diverse goals. Securing the needs of urban and agricultural water customers is a key goal. Meeting environmental health, ecosystem services and stream water quality goals has also been an integral part of many California water management systems. To meet this range of goals, groundwater, soil water and surface water will need to be managed conjunctively, management will likely become more tightly linked with land use and land resources planning and management, and modelling will play a key role in the development of successful and useful management plans.

The 2014 California Sustainable Groundwater Management Act (SGMA) and recent salt- and nitrate-related regulations to protect groundwater quality have put a focus on groundwater resources management, both quality and quantity, particularly in agricultural regions (Harter 2015). They mandate that local agencies pursue groundwater sustainability goals: avoiding long-term groundwater storage depletion, land subsidence, seawater intrusion, groundwater management–related water quality degradation, and deterioration of groundwater–surface water interactions.

The Scott River is an important salmonid spawning habitat that depends on groundwater to maintain stream flow during the summer. A hydrologic model developed by UC researchers can help predict the impact of different groundwater and surface water management scenarios on stream flow.

The Scott River is an important salmonid spawning habitat that depends on groundwater to maintain stream flow during the summer. A hydrologic model developed by UC researchers can help predict the impact of different groundwater and surface water management scenarios on stream flow.

Particularly important under the SGMA regulations is the interaction between groundwater and surface water: how do groundwater management decisions — by individual landowners or by groundwater sustainability agencies (GSAs) — impact not only beneficial users, but also streams (Zume and Tarhule 2011) and ground-water-dependent ecosystems (GDEs) (Boulton and Hancock 2006; Hatton 1998). Prominent California examples of areas where groundwater–surface water interactions are already addressed include the Napa River in Napa County and the Scott River in Siskiyou County. Both feature important salmonid fish habitat and therefore temperature is a critical issue (Brown et al. 1994; Moyle and Israel 2005); and low or decreased late-summer stream flow over the last half-century has impacted the quantity and quality of fish habitat (Kim and Jain 2010; NCRWQCB 2005; Nehlsen et al. 1991). During drought, portions of these rivers may temporarily dry up. In intermontane Scott Valley, dry sections disconnect lower sections of the stream from tributaries in the headwaters. Summer stream temperatures in the Scott River are affected by groundwater discharge into the streambed and by riparian shading and were being addressed under the federal Clean Water Act (NCRWQCB 2005) before SGMA.

Some measurements can be collected in the field to evaluate groundwater–surface water interactions, but computer models are needed to fully understand groundwater basin flow dynamics and assess impacts to stream flow under future groundwater management scenarios. For example, computer models can show the response of integrated water systems to management decisions such as pumping and intentional recharge. They are expected to play a key role in the implementation of SGMA and regulatory efforts.

Various modeling approaches have been developed for groundwater–surface water interactions (Furman 2008; Harter and Seytoux 2013). These range from analytical or spreadsheet tools (Foglia, McNally, Harter 2013) and coupled or iteratively coupled numerical model codes for computer simulations, such as the MODFLOW river (RIV) package (Harbaugh et al. 2000) and the MODFLOW stream flow routing SFR1 package (Prudic et al. 2004) and SFR2 package (Harbaugh 2005; Niswonger and Prudic 2005), to fully coupled models such as ParFlow (Ashby and Falgout 1996; Kollet and Maxwell 2006) and Hydrogeosphere (Brunner and Simmons 2012).

Fully coupled models provide the physically and mathematically most consistent and complete integration of groundwater, surface water and soil water systems. But they are computationally more expensive and require more parameterization (data input) than iteratively coupled models. In coupled or iteratively coupled models, multiple models are coupled such that one model provides input to the other model and vice versa, sometimes iteratively. Full coupling may not always yield better results (Furman 2008). For some applications, statistical models or analytical tools, which are based on highly simplified concepts and therefore have the least data input requirements and are computationally much less demanding, may be appropriate.

In Scott Valley, groundwater–surface water interactions are analyzed as part of an action plan to meet temperature TMDL (Total Maximum Daily Load) requirements for the Scott River. Climate change and groundwater pumping for irrigation in the valley have impacted late-summer and early fall stream flows in the Scott River (Drake et al. 2000). The local groundwater advisory committee is developing potential groundwater management scenarios that would increase summer and early fall stream flows. To evaluate those scenarios, we explored three levels of conceptual complexity at which information can be obtained about groundwater–surface water interactions: a water budget approach, a groundwater model with a conceptually simplified stream model (RIV) and a fully coupled groundwater–surface water model (SFR).

Almost 70% of Scott Valley is used for agricultural production, with a nearly even split between alfalfa/grain and pasture.

Almost 70% of Scott Valley is used for agricultural production, with a nearly even split between alfalfa/grain and pasture.

Scott Valley study area

Our study area was Scott Valley in northern California. Almost 70% of the valley is used for agricultural production, with a nearly even split between alfalfa/grain and pasture.

Geography and climate

Scott Valley is an intermontane 220-square-kilometer agricultural groundwater basin at an elevation of 2,600 to 3,100 feet in Siskiyou County (fig. 1). The Scott River flows from south to north along the east-central and northern portion of the valley. At the valley's northwest corner, the river descends into a gorge before joining the Klamath River several miles below Scott Valley. The Scott River watershed above Scott Valley extends into the surrounding Klamath Mountains to elevations of over 8,500 feet. The river and its tributaries are an important salmonid spawning habitat, home to native populations of the threatened Oncorhynchus kisutch (coho).

The boundaries of the groundwater model study in Scott Valley, and its surface waters. The Scott River and its tributaries are an important salmonid spawning habitat, home to native populations of the threatened coho. Source: Model extent derived from Mack (1958) and Soil Survey Geographic Database (SSURGO) data. Projection: North American Datum 1983, UTM Zone 10.

FIG. 1. The boundaries of the groundwater model study in Scott Valley, and its surface waters. The Scott River and its tributaries are an important salmonid spawning habitat, home to native populations of the threatened coho. Source: Model extent derived from Mack (1958) and Soil Survey Geographic Database (SSURGO) data. Projection: North American Datum 1983, UTM Zone 10.

Scott Valley formed primarily due to movement along an eastward dipping normal fault, with unconsolidated, highly heterogeneous fluvial and alluvial fan deposits forming an alluvial groundwater basin (Mack 1958). Surrounding the valley, the geology is comprised of relatively impermeable bedrock composed of metamorphic and volcanic units, although fractures do yield some water in the form of springs at the margins of the valley and in surrounding upland areas.

Aquifer thickness may be as much as 400 feet in the wide central part of the valley (Mack 1958). However, there is no evidence of sufficiently coarse material to support agricultural groundwater pumping below 250 feet (Foglia, McNally, Harter 2013). The aquifer pinches out at the valley margin.

Climate in the valley is Mediterranean, with 89% of the nearly 500-millimeter average annual precipitation falling between October and April. Daily mean temperatures range from 70°F in July to 32°F in January. Precipitation depths in the surrounding mountains are much higher, and snowmelt is a major source for ephemeral tributaries feeding the Scott River and recharging into the aquifer. Snowmelt dominates Scott River flows through June. During the summer months, flows in the Scott River immediately below the montane valley (USGS gage 11519500 Ft. Jones) can drop to 4 cubic feet per second (cfs), while maximum flows during winter can reach 40,000 cfs. After snowpack storage has been depleted, the Scott River is dependent on discharge from the Scott Valley aquifer to support base flow. In dry years, sections of the Scott River overlying the valley floor become ephemeral.

Land use and irrigation

Land use was surveyed in 2000 (DWR 2000) and further refined using aerial photo analysis and on-the-ground verification through interviews with landowners. A total of 2,119 land use parcels overlie the Scott Valley groundwater basin (fig. 2): 710 parcels (17,400 acres) are alfalfa/grain (an 8-year rotation with, on average, 1 year of grain crop followed by 7 years of alfalfa), 541 parcels (16,600 acres) are pasture, 451 parcels (20,400 acres) belong to land use categories with significant evapotranspiration but no irrigation (e.g., cemeteries, lawns, natural vegetation) and 417 parcels (1,700 acres) represent land uses with no evapotranspiration or irrigation (e.g., residential areas, parking lots, roads, and — most significantly — historic mine tailings).

Land use information and well locations in Scott Valley. ET/no irrigation reflects nonirrigated vegetation, e.g., lawns and riparian vegetation. No ET/no irrigation represents nonvegetated land surfaces including the mine tailings near Callahan. Well location information was obtained from well logs filed with the Department of Water Resources and verified in the field. Source: Model extent derived from Mack (1958) and SSURGO data. Land use polygon data source: DWR (2000). Revised to reflect 2011 land use patterns (GWAC, Groundwater Advisory Committee). Projection: North American Datum 1983, UTM Zone 10.

FIG. 2. Land use information and well locations in Scott Valley. ET/no irrigation reflects nonirrigated vegetation, e.g., lawns and riparian vegetation. No ET/no irrigation represents nonvegetated land surfaces including the mine tailings near Callahan. Well location information was obtained from well logs filed with the Department of Water Resources and verified in the field. Source: Model extent derived from Mack (1958) and SSURGO data. Land use polygon data source: DWR (2000). Revised to reflect 2011 land use patterns (GWAC, Groundwater Advisory Committee). Projection: North American Datum 1983, UTM Zone 10.

The year 2000 land use survey by DWR (DWR 2000) also identified the irrigation type associated with each land parcel. About 6,200 acres of cropland were identified as nonirrigated, dry or subirrigated. In Scott Valley, flood, center-pivot sprinkler and wheel-line sprinkler irrigation are used almost exclusively. Over the past 25 years, significant conversion from wheel-line sprinkler (but also from flood irrigation) to center-pivot sprinkler has occurred. For our study, we mapped the location (extent) and year of such irrigation-type conversions to land parcels by reviewing 1990 to 2011 aerial photos.

The beginning of the irrigation season is determined by soil moisture depletion but also by grower peer behavior. Earliest irrigation dates reported by local growers were March 15, March 24 and April 15 for grains, alfalfa and pasture, respectively. Growers irrigate based on soil moisture data, experience, peer behavior and established irrigation practices. The irrigation season typically ends on July 10, Sept. 1 and Oct. 15 for grain, alfalfa and pasture, respectively.

Water sources (identified for each land parcel by the DWR 2000 land use survey and updated through landowner survey) include groundwater, surface water, subirrigated (shallow groundwater table, not actually irrigated), mixed groundwater–surface water, and nonirrigated (dryland farming). Land parcels are distributed across nine subwatersheds associated with the major tributaries and the main stem Scott River. Discharge on these streams into the Scott Valley defines available maximum diversion rates for surface water irrigations. Where surface water is the only source of irrigation, lack of surface water will terminate the irrigation season. Groundwater pumping for a land parcel is from nearby or on-site irrigation wells. Well locations and type for the study area were obtained from DWR well permit records (fig. 2).

Hydrogeology

Within the alluvial groundwater basin of the Scott Valley, Mack (1958) distinguished six subareas (fig. 3). In our work, we also included the mine tailings at the southern end of the alluvial basin, an important hydrogeologic area consisting almost exclusively of reworked boulders from mine dredging operations (Foglia, McNally, Harter 2013).

Representation of the main characteristic of the modelled area, including boundary conditions, hydraulic conductivity and specific storage as defined by hydrostratigraphic zone, irrigation ditches, stream flow gaging stations and river segments (represented as Riv1, Riv2 and Riv5). Source: Model extent derived from Mack (1958) and Soil Survey Geographic Database (SSURGO) data. Projection: North American Datum 1983, UTM Zone 10.

FIG. 3. Representation of the main characteristic of the modelled area, including boundary conditions, hydraulic conductivity and specific storage as defined by hydrostratigraphic zone, irrigation ditches, stream flow gaging stations and river segments (represented as Riv1, Riv2 and Riv5). Source: Model extent derived from Mack (1958) and Soil Survey Geographic Database (SSURGO) data. Projection: North American Datum 1983, UTM Zone 10.

Aquifer pumping tests were performed to determine hydraulic properties in the main subarea of the valley, along the Scott River corridor. The tests showed that even within hydrogeologic subareas, hydraulic property values vary greatly. Estimates of hydraulic property values were also obtained from literature available for the region (DWR 2000; Mack 1958; SSPA 2012). The ratio of vertical hydraulic conductivity to horizontal hydraulic conductivity was estimated to be 1:10, a relatively high value representing relatively strong vertical connectivity of the coarser sediments.

The aquifer receives recharge from excess rainfall and irrigation but also from streams entering the basin on highly permeable alluvial fans. Groundwater discharge generally occurs through groundwater-dependent wetlands and riparian vegetation, pumping (primarily for irrigation) and discharge to streams, mostly along the valley thalweg.

Within the alluvial groundwater basin of the Scott Valley, there are six subareas. In this work, the authors also included the mine tailings at the southern end of the alluvial basin, an important hydrogeologic area consisting almost exclusively of reworked boulders from mine dredging operations.

Within the alluvial groundwater basin of the Scott Valley, there are six subareas. In this work, the authors also included the mine tailings at the southern end of the alluvial basin, an important hydrogeologic area consisting almost exclusively of reworked boulders from mine dredging operations.

Modeling tools

We developed the Scott Valley Integrated Hydrologic Model (SVIHM) to (1) provide a tool that integrates a diverse set of data and information within a consistent physical, hydrological framework; (2) estimate water budget components and their seasonal and interannual dynamics in the groundwater, stream and landscape–soil system; (3) better understand the relationship between land use, irrigation, groundwater pumping and stream flow; (4) provide a tool to predict potential impacts on stream flow from future groundwater and surface water management scenarios; and (5) provide an educational and decision-making tool for local stakeholders, regulators and policy- and decision-makers engaged in developing solutions to support and protect groundwater-dependent salmon habitat in the Scott Valley watershed.

For the simulation, we considered the period from October 1991 through September 2011, a period that includes the transformation of the Scott Valley landscape from predominantly sprinkler to significant center-pivot irrigation, a series of wet periods (1996 to 1999, 2006) and dry periods (1991, 2001, 2007 to 2009) and a series of years with potentially higher temperature. We developed several distinct model elements, representing the 1991 to 2011 period of the different hydrologic system components at varying levels of complexity that meet the modeling objectives. These were linked together into the SVIHM:

The upper watershed was represented by a statistical regression model to simulate incoming stream flows in the Scott River and its tributaries from the upper watershed to the valley, which are also used for irrigation. The Scott Valley landscape overlying the groundwater basin was represented by a tipping-bucket-type soil water budget model (SWBM) that simulates daily and monthly landscape-related water fluxes at the land parcel scale (see description above), including irrigation from diversions of surface water inflows to the valley and by groundwater pumping, evapotranspiration and groundwater recharge. Valley groundwater and surface water were simulated using a numerical model capable of simulating groundwater flow dynamics and the groundwater–surface water interface at sufficient detail to guide future data collection and simulate future water management scenarios.

Upper watershed stream flows

Surface water inflows to Scott Valley from the upper watershed are an important source of irrigation water. During the summer, incoming low flows may limit or terminate surface water diversions for irrigation. This in turn affects groundwater pumping in some crop parcels equipped for dual irrigation (surface and groundwater). Quantitative estimates of surface water inflows are also an important input to simulation of stream flow dynamics (including tributaries) within the valley, where streams are in direct connection with groundwater (the groundwater–surface water interface).

Since only limited stream gauging data were available on inflowing streams, a stream flow regression model was developed (Foglia, McNally, Hall 2013). Several factors were considered in developing the regression model, including precipitation, precipitation history, snowpack, and stream flows at the valley outlet, where the USGS Ft. Jones gage has provided nearly continuous records since the early 1940s. Foglia, McNally, Hall (2013) showed that the latter was the most critical factor to predict available monthly total incoming stream flow measured near the valley margins.

Soil water budget model, SWBM

In California, no water rights permits are issued for groundwater pumping, and wells, including wells in the study area, are largely unmetered. The primary purpose of the soil water budget model (SWBM) was therefore to estimate spatially and temporally varying recharge and pumping across the groundwater basin. A second goal was to quantify crop evapotranspiration (crop ET) and irrigation water use from surface water and from groundwater, and to understand the role of soil water storage. Conceptually, the soil water budget model encompasses the managed and unmanaged landscape including its vegetation and soil root zone and also the managed components of the surface water system (diversions) and of the groundwater system (well pumping).

SWBM does not account for fluxes at the ground-water–stream interface (stream recharge, groundwater discharge to streams) or for evapotranspiration due to root water uptake directly from groundwater by nonirrigated crops or in natural landscapes with a shallow water table. These processes were instead accounted for by the groundwater–surface water models MODFLOW RIV or MODFLOW SFR.

SWBM provided daily estimates of groundwater pumping, groundwater recharge, and evapotranspiration from Oct. 1, 1991, to Sept. 30, 2011, for each of the 2,115 parcels delineated in the land use survey of Scott Valley. Storage routing and mass balance were calculated for each land parcel as

where θi is the water content at the end of day i; Padji is the precipitation that infiltrates into the soil and is available for recharge or evapotranspiration on day i; AWi is the applied water (irrigation) amount on day i; ETi is the evapotranspiration on day i (computed as the product of the crop coefficient Kc and measured reference ET); Rechargei is deep percolation to the groundwater below the 1.22 meter (4 foot) deep root zone; and WC4i is the soil-dependent water holding capacity of the 1.22 meter (4 foot) root zone (Foglia, McNally, Harter 2013).

SWBM approximated growers' irrigation decisions in a simplified fashion: In the model, daily irrigation depths, AWi, were controlled by crop evapotranspiration depth and effective precipitation, which in turn were computed from daily climate data, using appropriate crop coefficients:

where AE is the water application efficiency, which was assumed to be constant over the growing season. The AE values were based on published values (Canessa et al. 2011) adjusted for local conditions: 90% for center-pivot sprinkler, 75% for wheel-line sprinkler and 70% for flood irrigation. The model accounted for the strong relationship between crop evapotranspiration and irrigation, but it did not represent temporal details of the actual irrigation schedule or alfalfa cuttings, as these have negligible impact on variations in groundwater conditions. The model also did not account for delivery losses.

MODFLOW simulations

A water budget model accounts for water fluxes into and out of a groundwater basin, the associated landscape and streams, and it provides some insight into large-scale, regional groundwater–surface water interactions. But integrated groundwater–surface water computer models, such as the MODFLOW packages, are more useful to fully assess and understand ground-water–surface water dynamics that are also driven by human impacts (e.g., pumping).

We used the MODFLOW-2005 code to build the groundwater–surface water model element of SVIHM (Harbaugh 2005). MODFLOW-2005 is a computer-based groundwater–surface water model that simulates groundwater flows and surface water flows by representing the aquifer basin and overlying stream system through discretized blocks (much like the way pixels on a TV screen are a representation of a continuous image). Aquifer and stream properties were defined for each block, which allowed the model to not only take on the actual shape of a groundwater–surface water system but also to represent the internal variability in aquifer and streambed properties that best reflects that actual system.

At the core, the model code solved the equations governing groundwater flow and stream flow, one time step after another. The entire Scott Valley groundwater basin (fig. 1) was discretized into 50-meter-by-50-meter cells, and it was divided into two vertical layers to better capture vertical fluxes associated with ground-water–surface water interactions. Due to the basin geometry, the bottom layer is not laterally expanding as much as the top layer (see supporting information S1 online).

Figure 3 summarizes the boundary conditions used to develop the groundwater model. The model simulates groundwater–surface water interactions along the Scott River, along major tributary streams (Shackleford, Mill, Kidder, Oro Fino, Moffett, Patterson, Etna, Crystal, Johnson, Clark Miner's and French Creeks) and along two major irrigation ditches (Farmers Ditch Company and Scott Valley Irrigation District). These features were simulated using different combinations of the river, stream flow routing (SFR1) and drain (DRN) packages of MODFLOW.

In our study, we developed two versions of SVIHM to represent two levels of conceptual complexities in the simulation of the groundwater–surface water interface. Both used the same algorithm to determine groundwater–surface water exchanges based on water level differences between the stream and groundwater, and as a function of streambed hydraulic conductivity.

In SVIHM-RIV, using the MODFLOW RIV package (Harbaugh 2005), stream water levels were user assigned and might vary in time and space. The advantage of SVIHM-RIV is that it is computationally much less expensive (has a much lower simulation run time) than SVIHM-SFR, since it does not simulate the stream flow system. The computational efficiency is advantageous in model calibration. In Scott Valley, only sparse data were available on stream water levels. As an initial modeling design step, we chose a simple approximation of stream water levels using a constant, average stream depth uniform across the valley at all times.

In SVIHM-SFR, using the MODFLOW SFR package (Prudic et al. 2004), inflows from the upper watershed (obtained from the statistical model of watershed inflows), after irrigation diversions (obtained from SWBM), were physically routed by simulation through the valley's stream system. The simulation computed stream water level as a function of flow rate, stream slope, streambed morphology and stream roughness (Manning's equation). Detailed streambed morphology was available from two LIDAR surveys (SSPA 2012). With SFR, stream flow varied from stream cell to stream cell due to diversions, tributary inflows or groundwater–surface water exchanges. In this way, MODFLOW SFR tracked stream water depth variations in time and along the stream system. It could also estimate the timing and location of stream sections that fell dry.

The land parcel–based output results of SWBM — agricultural groundwater pumping, groundwater recharge and irrigation — were used as input to the MODFLOW RIV and MODFLOW SFR versions of SVIHM, which simulated the 21-year period using monthly variabl

Return to top

Author notes

In memory of our co-author Steve Orloff and his many contributions to this work.

Funding for our research was provided by the California State Water Resources Control Board contracts 11-189-110 and 14-020-110. We would like to thank the Scott Valley Groundwater Advisory Committee, Sari Sommarstrom, and Bryan McFadin for many helpful discussions during the development of our modeling tools.

California Agriculture thanks Guest Associate Editor Hoori Ajami for her work on this article.

References

Ashby SF, Falgout RD. A parallel multigrid preconditioned conjugate gradient algorithm for groundwater flow simulations. Nucl Sci Eng. 1996. 124:145-59. (Also available as LLNL Technical Report UCRL-JC-122359.) https://doi.org/10.13182/NSE96-A24230

Banta ER. MODFLOW-2000, the U.S 2000. p.127. Geological Survey Modular Ground-Water Model – Documentation of Packages for Simulating Evapotranspiration with a Segmented Function (ETS1) and Drains with Return Flow (DRT1). U.S. Geological Survey Open-File Report 00-466

Boulton AJ, Hancock PJ. Rivers as groundwater-dependent ecosystems: A review of degrees of dependency, riverine processes and management implications. Aust J Bot. 2006. 54:133-44.

Brown LR, Moyle PB, Yoshiyama RM. Historical decline and current status of Coho salmon (Oncorhynchus kisutch) in California. N Am J Fisheries Manage. 1994. 14(2):237-61. https://doi.org/10.1577/1548-8675(1994)014replacecodelt0237

Brunner P, Simmons CT. HydroGeoSphere: A fully integrated, physically based hydrological model. Ground Water. 2012. 50:170-6. https://doi.org/10.1111/j.1745-6584.2011.00882.x

Canessa P, Green S, Zoldoske D. Agricultural Water Use in California: A 2011 Update 2011. p.80. Staff Report, Center for Irrigation Technology, California State University, Fresno, CA. www.waterboards.ca.gov/waterrights/water_issues/programs/hearings/cachuma/exbhts_2012feir/cachuma_feir_mu289.pdf

Drake D, Tate K, Carlson H. Analysis shows climate-caused decreases in Scott River fall flows. Calif Agr. 2000. 54(6):46-9. http://doi.org/10.3733/ca.v054n06p46 https://doi.org/10.3733/ca.v054n06p46

[DWR] California Department of Water Resources. Siskiyou County Land Use Survey 2000, Division of Planning and Local Assistance 2000. www.water.ca.gov/landwateruse/lusrvymain.cfm

Foglia L, McNally A, Hall C, et al. Scott Valley Integrated Hydrologic Model: Data Collection, Analysis, and Water Budget, Final Report, April 2013 2013. p.101. UC Davis. http://groundwater.ucdavis.edu

Foglia L, McNally A, Harter T. Coupling a spatiotemporally distributed soil water budget with stream-depletion functions to inform stakeholder-driven management of groundwater-dependent ecosystems. Water Resour Res. 2013. 49:7292-310. https://doi.org/10.1002/wrcr.20555

Furman A. Modeling coupled surface–subsurface flow processes: A review. Vadose Zone J. 2008. 7(2):741-https://doi.org/10.2136/vzj2007.0065

Harbaugh AW. MODFLOW-2005, The US Geological Survey Modular Ground-Water Model — the Ground-Water Flow Process 2005. p.253. US Geological Survey Techniques and Methods

Harbaugh A, Banta E, Hill M, McDonald M. MODFLOW-2000, The US Geological Survey Modular Ground-Water Model – Users Guide to Modularization Concepts and the Ground-Water Flow Process 2000. US Geological Survey Open-File Report 00-92

Harter T. California's agricultural regions gear up to actively manage groundwater use and protection. Calif Agr. 2015. 69(3):193-201. http://doi.org/10.3733/ca.E.v069n03p193 https://doi.org/10.3733/ca.E.v069n03p193

Harter T, Morel-Seytoux H. Peer Review of the IWFM, MODFLOW and HGS Model Codes: Potential for Water Management Applications in California's Central Valley and Other Irrigated Groundwater Basins 2013. p.121. California Water and Environmental Modeling Forum, Sacramento, August 2013

Hatton TJ. The Basics of Recharge and Discharge, Part 4: Catchment Scale Recharge Modeling. 1998. Collingwood, Victoria, Australia: CSIRO: Commonwealth Scientific and Industrial Research Organization.

Kim JS, Jain S. High-resolution streamflow trend analysis applicable to annual decision calendars: A western United States case study. Climatic Change. 2010. 102(3):699-707. https://doi.org/10.1007/s10584-010-9933-3

Kollet SJ, Maxwell RM. Integrated surface-groundwater flow modeling: A free-surface overland flow boundary condition in a parallel groundwater flow model. Adv Water Res. 2006. 29:945-58. https://doi.org/10.1016/j.advwatres.2005.08.006

Mack S. Geology and Ground-Water Features of Scott Valley Siskiyou County, California 1958. US Geological Survey Water-Supply Paper 1462. Washington DC

Moyle PB, Israel JA. Untested assumptions: Effectiveness of screening diversions for conservation of fish populations. Fisheries. 2005. 30(5):20-8. https://doi.org/10.1577/1548-8446(2005)30[20:UA]2.0.CO;2

NCRWQCB. Staff Report for the Action Plan for the Scott River Watershed Sediment and Temperature Total Maximum Daily Loads 2005. www.waterboards.ca.gov/water_issues/programs/tmdl/records/region_1/2010/ref3872.pdf

Nehlsen W, Williams JE, Lichatowich JA. Pacific salmon at the crossroads: Stocks at risk from California, Oregon, Idaho, and Washington. Fisheries. 1991. 16(2):4-21. https://doi.org/10.1577/1548-8446(1991)016replacecodelt0004:PSATCSreplacecodegt2.0.CO;2

Niswonger RG, Prudic DE. Documentation of the Stream-flow-Routing (SFR2) Package to Include Unsaturated Flow beneath Streams—A modification to SFR1 2005. p.50. US Geological Survey Techniques and Methods 6-A13

Prudic DE, Konikow LF, Banta ER. A New Stream-Flow Routing (SFR1) Package to Simulate Stream-Aquifer Interaction with MODFLOW-2000 2004. p.95. US Geological Survey Open-File Report 2004-1042

[SSPA] SS Papadopulos & Associates. Groundwater Conditions in Scott Valley 2012. Report prepared for the Karuk Tribe, March 2012

Zume JT, Tarhule AA. Modeling the response of an alluvial aquifer to anthropogenic and recharge stresses in the United States Southern Great Plains. J Earth Syst Sci. 2011. 120(4):557-72. https://doi.org/10.1007/s12040-011-0088-z

Modeling guides groundwater management in a basin with river–aquifer interactions

Laura Foglia, Jakob Neumann, Douglas G. Tolley, Steve Orloff, Richard L. Snyder, Thomas Harter
Webmaster Email: wsuckow@ucanr.edu

Modeling guides groundwater management in a basin with river–aquifer interactions

Share using any of the popular social networks Share by sending an email Print article
Share using any of the popular social networks Share by sending an email Print article

Authors

Laura Foglia, UC Davis
Jakob Neumann, Technical University Darmstadt
Douglas G. Tolley, UC Davis
Steve Orloff, UC Cooperative Extension
Richard L. Snyder, UC Cooperative Extension and UC Davis
Thomas Harter, UC Davis

Publication Information

California Agriculture 72(1):84-95. https://doi.org/10.3733/ca.2018a0011

Published online March 13, 2018

PDF  |  Citation  |  Permissions  |  Cited by 0 articles

Author Affiliations show

Abstract

The Sustainable Groundwater Management Act (SGMA) of 2014 seeks to maintain groundwater discharge to streams to support environmental goals. In Scott Valley, in Siskiyou County, the Scott River and its tributaries are an important salmonid spawning habitat, and about 10% of average annual Scott River stream flow comes from groundwater. The local groundwater advisory committee is developing groundwater management alternatives that would increase summer and early fall stream flows. We developed a model to provide a framework to evaluate those alternatives. We first created a water budget for the Scott Valley groundwater basin and integrated the detailed, spatiotemporally distributed water budget results into a computer model of the basin that simultaneously accounted for groundwater flow, stream flow and landscape water fluxes. Different conceptual representations (using the MODFLOW RIV package and MODFLOW SFR package) of the stream–aquifer boundary provided significantly different results in the seasonal dynamics of groundwater–surface water fluxes. As groundwater sustainability agencies draw up plans to meet SGMA requirements, they must choose and test simulation tools carefully.

Full text

Management of California's water supplies serves diverse goals. Securing the needs of urban and agricultural water customers is a key goal. Meeting environmental health, ecosystem services and stream water quality goals has also been an integral part of many California water management systems. To meet this range of goals, groundwater, soil water and surface water will need to be managed conjunctively, management will likely become more tightly linked with land use and land resources planning and management, and modelling will play a key role in the development of successful and useful management plans.

The 2014 California Sustainable Groundwater Management Act (SGMA) and recent salt- and nitrate-related regulations to protect groundwater quality have put a focus on groundwater resources management, both quality and quantity, particularly in agricultural regions (Harter 2015). They mandate that local agencies pursue groundwater sustainability goals: avoiding long-term groundwater storage depletion, land subsidence, seawater intrusion, groundwater management–related water quality degradation, and deterioration of groundwater–surface water interactions.

The Scott River is an important salmonid spawning habitat that depends on groundwater to maintain stream flow during the summer. A hydrologic model developed by UC researchers can help predict the impact of different groundwater and surface water management scenarios on stream flow.

The Scott River is an important salmonid spawning habitat that depends on groundwater to maintain stream flow during the summer. A hydrologic model developed by UC researchers can help predict the impact of different groundwater and surface water management scenarios on stream flow.

Particularly important under the SGMA regulations is the interaction between groundwater and surface water: how do groundwater management decisions — by individual landowners or by groundwater sustainability agencies (GSAs) — impact not only beneficial users, but also streams (Zume and Tarhule 2011) and ground-water-dependent ecosystems (GDEs) (Boulton and Hancock 2006; Hatton 1998). Prominent California examples of areas where groundwater–surface water interactions are already addressed include the Napa River in Napa County and the Scott River in Siskiyou County. Both feature important salmonid fish habitat and therefore temperature is a critical issue (Brown et al. 1994; Moyle and Israel 2005); and low or decreased late-summer stream flow over the last half-century has impacted the quantity and quality of fish habitat (Kim and Jain 2010; NCRWQCB 2005; Nehlsen et al. 1991). During drought, portions of these rivers may temporarily dry up. In intermontane Scott Valley, dry sections disconnect lower sections of the stream from tributaries in the headwaters. Summer stream temperatures in the Scott River are affected by groundwater discharge into the streambed and by riparian shading and were being addressed under the federal Clean Water Act (NCRWQCB 2005) before SGMA.

Some measurements can be collected in the field to evaluate groundwater–surface water interactions, but computer models are needed to fully understand groundwater basin flow dynamics and assess impacts to stream flow under future groundwater management scenarios. For example, computer models can show the response of integrated water systems to management decisions such as pumping and intentional recharge. They are expected to play a key role in the implementation of SGMA and regulatory efforts.

Various modeling approaches have been developed for groundwater–surface water interactions (Furman 2008; Harter and Seytoux 2013). These range from analytical or spreadsheet tools (Foglia, McNally, Harter 2013) and coupled or iteratively coupled numerical model codes for computer simulations, such as the MODFLOW river (RIV) package (Harbaugh et al. 2000) and the MODFLOW stream flow routing SFR1 package (Prudic et al. 2004) and SFR2 package (Harbaugh 2005; Niswonger and Prudic 2005), to fully coupled models such as ParFlow (Ashby and Falgout 1996; Kollet and Maxwell 2006) and Hydrogeosphere (Brunner and Simmons 2012).

Fully coupled models provide the physically and mathematically most consistent and complete integration of groundwater, surface water and soil water systems. But they are computationally more expensive and require more parameterization (data input) than iteratively coupled models. In coupled or iteratively coupled models, multiple models are coupled such that one model provides input to the other model and vice versa, sometimes iteratively. Full coupling may not always yield better results (Furman 2008). For some applications, statistical models or analytical tools, which are based on highly simplified concepts and therefore have the least data input requirements and are computationally much less demanding, may be appropriate.

In Scott Valley, groundwater–surface water interactions are analyzed as part of an action plan to meet temperature TMDL (Total Maximum Daily Load) requirements for the Scott River. Climate change and groundwater pumping for irrigation in the valley have impacted late-summer and early fall stream flows in the Scott River (Drake et al. 2000). The local groundwater advisory committee is developing potential groundwater management scenarios that would increase summer and early fall stream flows. To evaluate those scenarios, we explored three levels of conceptual complexity at which information can be obtained about groundwater–surface water interactions: a water budget approach, a groundwater model with a conceptually simplified stream model (RIV) and a fully coupled groundwater–surface water model (SFR).

Almost 70% of Scott Valley is used for agricultural production, with a nearly even split between alfalfa/grain and pasture.

Almost 70% of Scott Valley is used for agricultural production, with a nearly even split between alfalfa/grain and pasture.

Scott Valley study area

Our study area was Scott Valley in northern California. Almost 70% of the valley is used for agricultural production, with a nearly even split between alfalfa/grain and pasture.

Geography and climate

Scott Valley is an intermontane 220-square-kilometer agricultural groundwater basin at an elevation of 2,600 to 3,100 feet in Siskiyou County (fig. 1). The Scott River flows from south to north along the east-central and northern portion of the valley. At the valley's northwest corner, the river descends into a gorge before joining the Klamath River several miles below Scott Valley. The Scott River watershed above Scott Valley extends into the surrounding Klamath Mountains to elevations of over 8,500 feet. The river and its tributaries are an important salmonid spawning habitat, home to native populations of the threatened Oncorhynchus kisutch (coho).

The boundaries of the groundwater model study in Scott Valley, and its surface waters. The Scott River and its tributaries are an important salmonid spawning habitat, home to native populations of the threatened coho. Source: Model extent derived from Mack (1958) and Soil Survey Geographic Database (SSURGO) data. Projection: North American Datum 1983, UTM Zone 10.

FIG. 1. The boundaries of the groundwater model study in Scott Valley, and its surface waters. The Scott River and its tributaries are an important salmonid spawning habitat, home to native populations of the threatened coho. Source: Model extent derived from Mack (1958) and Soil Survey Geographic Database (SSURGO) data. Projection: North American Datum 1983, UTM Zone 10.

Scott Valley formed primarily due to movement along an eastward dipping normal fault, with unconsolidated, highly heterogeneous fluvial and alluvial fan deposits forming an alluvial groundwater basin (Mack 1958). Surrounding the valley, the geology is comprised of relatively impermeable bedrock composed of metamorphic and volcanic units, although fractures do yield some water in the form of springs at the margins of the valley and in surrounding upland areas.

Aquifer thickness may be as much as 400 feet in the wide central part of the valley (Mack 1958). However, there is no evidence of sufficiently coarse material to support agricultural groundwater pumping below 250 feet (Foglia, McNally, Harter 2013). The aquifer pinches out at the valley margin.

Climate in the valley is Mediterranean, with 89% of the nearly 500-millimeter average annual precipitation falling between October and April. Daily mean temperatures range from 70°F in July to 32°F in January. Precipitation depths in the surrounding mountains are much higher, and snowmelt is a major source for ephemeral tributaries feeding the Scott River and recharging into the aquifer. Snowmelt dominates Scott River flows through June. During the summer months, flows in the Scott River immediately below the montane valley (USGS gage 11519500 Ft. Jones) can drop to 4 cubic feet per second (cfs), while maximum flows during winter can reach 40,000 cfs. After snowpack storage has been depleted, the Scott River is dependent on discharge from the Scott Valley aquifer to support base flow. In dry years, sections of the Scott River overlying the valley floor become ephemeral.

Land use and irrigation

Land use was surveyed in 2000 (DWR 2000) and further refined using aerial photo analysis and on-the-ground verification through interviews with landowners. A total of 2,119 land use parcels overlie the Scott Valley groundwater basin (fig. 2): 710 parcels (17,400 acres) are alfalfa/grain (an 8-year rotation with, on average, 1 year of grain crop followed by 7 years of alfalfa), 541 parcels (16,600 acres) are pasture, 451 parcels (20,400 acres) belong to land use categories with significant evapotranspiration but no irrigation (e.g., cemeteries, lawns, natural vegetation) and 417 parcels (1,700 acres) represent land uses with no evapotranspiration or irrigation (e.g., residential areas, parking lots, roads, and — most significantly — historic mine tailings).

Land use information and well locations in Scott Valley. ET/no irrigation reflects nonirrigated vegetation, e.g., lawns and riparian vegetation. No ET/no irrigation represents nonvegetated land surfaces including the mine tailings near Callahan. Well location information was obtained from well logs filed with the Department of Water Resources and verified in the field. Source: Model extent derived from Mack (1958) and SSURGO data. Land use polygon data source: DWR (2000). Revised to reflect 2011 land use patterns (GWAC, Groundwater Advisory Committee). Projection: North American Datum 1983, UTM Zone 10.

FIG. 2. Land use information and well locations in Scott Valley. ET/no irrigation reflects nonirrigated vegetation, e.g., lawns and riparian vegetation. No ET/no irrigation represents nonvegetated land surfaces including the mine tailings near Callahan. Well location information was obtained from well logs filed with the Department of Water Resources and verified in the field. Source: Model extent derived from Mack (1958) and SSURGO data. Land use polygon data source: DWR (2000). Revised to reflect 2011 land use patterns (GWAC, Groundwater Advisory Committee). Projection: North American Datum 1983, UTM Zone 10.

The year 2000 land use survey by DWR (DWR 2000) also identified the irrigation type associated with each land parcel. About 6,200 acres of cropland were identified as nonirrigated, dry or subirrigated. In Scott Valley, flood, center-pivot sprinkler and wheel-line sprinkler irrigation are used almost exclusively. Over the past 25 years, significant conversion from wheel-line sprinkler (but also from flood irrigation) to center-pivot sprinkler has occurred. For our study, we mapped the location (extent) and year of such irrigation-type conversions to land parcels by reviewing 1990 to 2011 aerial photos.

The beginning of the irrigation season is determined by soil moisture depletion but also by grower peer behavior. Earliest irrigation dates reported by local growers were March 15, March 24 and April 15 for grains, alfalfa and pasture, respectively. Growers irrigate based on soil moisture data, experience, peer behavior and established irrigation practices. The irrigation season typically ends on July 10, Sept. 1 and Oct. 15 for grain, alfalfa and pasture, respectively.

Water sources (identified for each land parcel by the DWR 2000 land use survey and updated through landowner survey) include groundwater, surface water, subirrigated (shallow groundwater table, not actually irrigated), mixed groundwater–surface water, and nonirrigated (dryland farming). Land parcels are distributed across nine subwatersheds associated with the major tributaries and the main stem Scott River. Discharge on these streams into the Scott Valley defines available maximum diversion rates for surface water irrigations. Where surface water is the only source of irrigation, lack of surface water will terminate the irrigation season. Groundwater pumping for a land parcel is from nearby or on-site irrigation wells. Well locations and type for the study area were obtained from DWR well permit records (fig. 2).

Hydrogeology

Within the alluvial groundwater basin of the Scott Valley, Mack (1958) distinguished six subareas (fig. 3). In our work, we also included the mine tailings at the southern end of the alluvial basin, an important hydrogeologic area consisting almost exclusively of reworked boulders from mine dredging operations (Foglia, McNally, Harter 2013).

Representation of the main characteristic of the modelled area, including boundary conditions, hydraulic conductivity and specific storage as defined by hydrostratigraphic zone, irrigation ditches, stream flow gaging stations and river segments (represented as Riv1, Riv2 and Riv5). Source: Model extent derived from Mack (1958) and Soil Survey Geographic Database (SSURGO) data. Projection: North American Datum 1983, UTM Zone 10.

FIG. 3. Representation of the main characteristic of the modelled area, including boundary conditions, hydraulic conductivity and specific storage as defined by hydrostratigraphic zone, irrigation ditches, stream flow gaging stations and river segments (represented as Riv1, Riv2 and Riv5). Source: Model extent derived from Mack (1958) and Soil Survey Geographic Database (SSURGO) data. Projection: North American Datum 1983, UTM Zone 10.

Aquifer pumping tests were performed to determine hydraulic properties in the main subarea of the valley, along the Scott River corridor. The tests showed that even within hydrogeologic subareas, hydraulic property values vary greatly. Estimates of hydraulic property values were also obtained from literature available for the region (DWR 2000; Mack 1958; SSPA 2012). The ratio of vertical hydraulic conductivity to horizontal hydraulic conductivity was estimated to be 1:10, a relatively high value representing relatively strong vertical connectivity of the coarser sediments.

The aquifer receives recharge from excess rainfall and irrigation but also from streams entering the basin on highly permeable alluvial fans. Groundwater discharge generally occurs through groundwater-dependent wetlands and riparian vegetation, pumping (primarily for irrigation) and discharge to streams, mostly along the valley thalweg.

Within the alluvial groundwater basin of the Scott Valley, there are six subareas. In this work, the authors also included the mine tailings at the southern end of the alluvial basin, an important hydrogeologic area consisting almost exclusively of reworked boulders from mine dredging operations.

Within the alluvial groundwater basin of the Scott Valley, there are six subareas. In this work, the authors also included the mine tailings at the southern end of the alluvial basin, an important hydrogeologic area consisting almost exclusively of reworked boulders from mine dredging operations.

Modeling tools

We developed the Scott Valley Integrated Hydrologic Model (SVIHM) to (1) provide a tool that integrates a diverse set of data and information within a consistent physical, hydrological framework; (2) estimate water budget components and their seasonal and interannual dynamics in the groundwater, stream and landscape–soil system; (3) better understand the relationship between land use, irrigation, groundwater pumping and stream flow; (4) provide a tool to predict potential impacts on stream flow from future groundwater and surface water management scenarios; and (5) provide an educational and decision-making tool for local stakeholders, regulators and policy- and decision-makers engaged in developing solutions to support and protect groundwater-dependent salmon habitat in the Scott Valley watershed.

For the simulation, we considered the period from October 1991 through September 2011, a period that includes the transformation of the Scott Valley landscape from predominantly sprinkler to significant center-pivot irrigation, a series of wet periods (1996 to 1999, 2006) and dry periods (1991, 2001, 2007 to 2009) and a series of years with potentially higher temperature. We developed several distinct model elements, representing the 1991 to 2011 period of the different hydrologic system components at varying levels of complexity that meet the modeling objectives. These were linked together into the SVIHM:

The upper watershed was represented by a statistical regression model to simulate incoming stream flows in the Scott River and its tributaries from the upper watershed to the valley, which are also used for irrigation. The Scott Valley landscape overlying the groundwater basin was represented by a tipping-bucket-type soil water budget model (SWBM) that simulates daily and monthly landscape-related water fluxes at the land parcel scale (see description above), including irrigation from diversions of surface water inflows to the valley and by groundwater pumping, evapotranspiration and groundwater recharge. Valley groundwater and surface water were simulated using a numerical model capable of simulating groundwater flow dynamics and the groundwater–surface water interface at sufficient detail to guide future data collection and simulate future water management scenarios.

Upper watershed stream flows

Surface water inflows to Scott Valley from the upper watershed are an important source of irrigation water. During the summer, incoming low flows may limit or terminate surface water diversions for irrigation. This in turn affects groundwater pumping in some crop parcels equipped for dual irrigation (surface and groundwater). Quantitative estimates of surface water inflows are also an important input to simulation of stream flow dynamics (including tributaries) within the valley, where streams are in direct connection with groundwater (the groundwater–surface water interface).

Since only limited stream gauging data were available on inflowing streams, a stream flow regression model was developed (Foglia, McNally, Hall 2013). Several factors were considered in developing the regression model, including precipitation, precipitation history, snowpack, and stream flows at the valley outlet, where the USGS Ft. Jones gage has provided nearly continuous records since the early 1940s. Foglia, McNally, Hall (2013) showed that the latter was the most critical factor to predict available monthly total incoming stream flow measured near the valley margins.

Soil water budget model, SWBM

In California, no water rights permits are issued for groundwater pumping, and wells, including wells in the study area, are largely unmetered. The primary purpose of the soil water budget model (SWBM) was therefore to estimate spatially and temporally varying recharge and pumping across the groundwater basin. A second goal was to quantify crop evapotranspiration (crop ET) and irrigation water use from surface water and from groundwater, and to understand the role of soil water storage. Conceptually, the soil water budget model encompasses the managed and unmanaged landscape including its vegetation and soil root zone and also the managed components of the surface water system (diversions) and of the groundwater system (well pumping).

SWBM does not account for fluxes at the ground-water–stream interface (stream recharge, groundwater discharge to streams) or for evapotranspiration due to root water uptake directly from groundwater by nonirrigated crops or in natural landscapes with a shallow water table. These processes were instead accounted for by the groundwater–surface water models MODFLOW RIV or MODFLOW SFR.

SWBM provided daily estimates of groundwater pumping, groundwater recharge, and evapotranspiration from Oct. 1, 1991, to Sept. 30, 2011, for each of the 2,115 parcels delineated in the land use survey of Scott Valley. Storage routing and mass balance were calculated for each land parcel as

where θi is the water content at the end of day i; Padji is the precipitation that infiltrates into the soil and is available for recharge or evapotranspiration on day i; AWi is the applied water (irrigation) amount on day i; ETi is the evapotranspiration on day i (computed as the product of the crop coefficient Kc and measured reference ET); Rechargei is deep percolation to the groundwater below the 1.22 meter (4 foot) deep root zone; and WC4i is the soil-dependent water holding capacity of the 1.22 meter (4 foot) root zone (Foglia, McNally, Harter 2013).

SWBM approximated growers' irrigation decisions in a simplified fashion: In the model, daily irrigation depths, AWi, were controlled by crop evapotranspiration depth and effective precipitation, which in turn were computed from daily climate data, using appropriate crop coefficients:

where AE is the water application efficiency, which was assumed to be constant over the growing season. The AE values were based on published values (Canessa et al. 2011) adjusted for local conditions: 90% for center-pivot sprinkler, 75% for wheel-line sprinkler and 70% for flood irrigation. The model accounted for the strong relationship between crop evapotranspiration and irrigation, but it did not represent temporal details of the actual irrigation schedule or alfalfa cuttings, as these have negligible impact on variations in groundwater conditions. The model also did not account for delivery losses.

MODFLOW simulations

A water budget model accounts for water fluxes into and out of a groundwater basin, the associated landscape and streams, and it provides some insight into large-scale, regional groundwater–surface water interactions. But integrated groundwater–surface water computer models, such as the MODFLOW packages, are more useful to fully assess and understand ground-water–surface water dynamics that are also driven by human impacts (e.g., pumping).

We used the MODFLOW-2005 code to build the groundwater–surface water model element of SVIHM (Harbaugh 2005). MODFLOW-2005 is a computer-based groundwater–surface water model that simulates groundwater flows and surface water flows by representing the aquifer basin and overlying stream system through discretized blocks (much like the way pixels on a TV screen are a representation of a continuous image). Aquifer and stream properties were defined for each block, which allowed the model to not only take on the actual shape of a groundwater–surface water system but also to represent the internal variability in aquifer and streambed properties that best reflects that actual system.

At the core, the model code solved the equations governing groundwater flow and stream flow, one time step after another. The entire Scott Valley groundwater basin (fig. 1) was discretized into 50-meter-by-50-meter cells, and it was divided into two vertical layers to better capture vertical fluxes associated with ground-water–surface water interactions. Due to the basin geometry, the bottom layer is not laterally expanding as much as the top layer (see supporting information S1 online).

Figure 3 summarizes the boundary conditions used to develop the groundwater model. The model simulates groundwater–surface water interactions along the Scott River, along major tributary streams (Shackleford, Mill, Kidder, Oro Fino, Moffett, Patterson, Etna, Crystal, Johnson, Clark Miner's and French Creeks) and along two major irrigation ditches (Farmers Ditch Company and Scott Valley Irrigation District). These features were simulated using different combinations of the river, stream flow routing (SFR1) and drain (DRN) packages of MODFLOW.

In our study, we developed two versions of SVIHM to represent two levels of conceptual complexities in the simulation of the groundwater–surface water interface. Both used the same algorithm to determine groundwater–surface water exchanges based on water level differences between the stream and groundwater, and as a function of streambed hydraulic conductivity.

In SVIHM-RIV, using the MODFLOW RIV package (Harbaugh 2005), stream water levels were user assigned and might vary in time and space. The advantage of SVIHM-RIV is that it is computationally much less expensive (has a much lower simulation run time) than SVIHM-SFR, since it does not simulate the stream flow system. The computational efficiency is advantageous in model calibration. In Scott Valley, only sparse data were available on stream water levels. As an initial modeling design step, we chose a simple approximation of stream water levels using a constant, average stream depth uniform across the valley at all times.

In SVIHM-SFR, using the MODFLOW SFR package (Prudic et al. 2004), inflows from the upper watershed (obtained from the statistical model of watershed inflows), after irrigation diversions (obtained from SWBM), were physically routed by simulation through the valley's stream system. The simulation computed stream water level as a function of flow rate, stream slope, streambed morphology and stream roughness (Manning's equation). Detailed streambed morphology was available from two LIDAR surveys (SSPA 2012). With SFR, stream flow varied from stream cell to stream cell due to diversions, tributary inflows or groundwater–surface water exchanges. In this way, MODFLOW SFR tracked stream water depth variations in time and along the stream system. It could also estimate the timing and location of stream sections that fell dry.

The land parcel–based output results of SWBM — agricultural groundwater pumping, groundwater recharge and irrigation — were used as input to the MODFLOW RIV and MODFLOW SFR versions of SVIHM, which simulated the 21-year period using monthly variabl

Return to top

Author notes

In memory of our co-author Steve Orloff and his many contributions to this work.

Funding for our research was provided by the California State Water Resources Control Board contracts 11-189-110 and 14-020-110. We would like to thank the Scott Valley Groundwater Advisory Committee, Sari Sommarstrom, and Bryan McFadin for many helpful discussions during the development of our modeling tools.

California Agriculture thanks Guest Associate Editor Hoori Ajami for her work on this article.

References

Ashby SF, Falgout RD. A parallel multigrid preconditioned conjugate gradient algorithm for groundwater flow simulations. Nucl Sci Eng. 1996. 124:145-59. (Also available as LLNL Technical Report UCRL-JC-122359.) https://doi.org/10.13182/NSE96-A24230

Banta ER. MODFLOW-2000, the U.S 2000. p.127. Geological Survey Modular Ground-Water Model – Documentation of Packages for Simulating Evapotranspiration with a Segmented Function (ETS1) and Drains with Return Flow (DRT1). U.S. Geological Survey Open-File Report 00-466

Boulton AJ, Hancock PJ. Rivers as groundwater-dependent ecosystems: A review of degrees of dependency, riverine processes and management implications. Aust J Bot. 2006. 54:133-44.

Brown LR, Moyle PB, Yoshiyama RM. Historical decline and current status of Coho salmon (Oncorhynchus kisutch) in California. N Am J Fisheries Manage. 1994. 14(2):237-61. https://doi.org/10.1577/1548-8675(1994)014replacecodelt0237

Brunner P, Simmons CT. HydroGeoSphere: A fully integrated, physically based hydrological model. Ground Water. 2012. 50:170-6. https://doi.org/10.1111/j.1745-6584.2011.00882.x

Canessa P, Green S, Zoldoske D. Agricultural Water Use in California: A 2011 Update 2011. p.80. Staff Report, Center for Irrigation Technology, California State University, Fresno, CA. www.waterboards.ca.gov/waterrights/water_issues/programs/hearings/cachuma/exbhts_2012feir/cachuma_feir_mu289.pdf

Drake D, Tate K, Carlson H. Analysis shows climate-caused decreases in Scott River fall flows. Calif Agr. 2000. 54(6):46-9. http://doi.org/10.3733/ca.v054n06p46 https://doi.org/10.3733/ca.v054n06p46

[DWR] California Department of Water Resources. Siskiyou County Land Use Survey 2000, Division of Planning and Local Assistance 2000. www.water.ca.gov/landwateruse/lusrvymain.cfm

Foglia L, McNally A, Hall C, et al. Scott Valley Integrated Hydrologic Model: Data Collection, Analysis, and Water Budget, Final Report, April 2013 2013. p.101. UC Davis. http://groundwater.ucdavis.edu

Foglia L, McNally A, Harter T. Coupling a spatiotemporally distributed soil water budget with stream-depletion functions to inform stakeholder-driven management of groundwater-dependent ecosystems. Water Resour Res. 2013. 49:7292-310. https://doi.org/10.1002/wrcr.20555

Furman A. Modeling coupled surface–subsurface flow processes: A review. Vadose Zone J. 2008. 7(2):741-https://doi.org/10.2136/vzj2007.0065

Harbaugh AW. MODFLOW-2005, The US Geological Survey Modular Ground-Water Model — the Ground-Water Flow Process 2005. p.253. US Geological Survey Techniques and Methods

Harbaugh A, Banta E, Hill M, McDonald M. MODFLOW-2000, The US Geological Survey Modular Ground-Water Model – Users Guide to Modularization Concepts and the Ground-Water Flow Process 2000. US Geological Survey Open-File Report 00-92

Harter T. California's agricultural regions gear up to actively manage groundwater use and protection. Calif Agr. 2015. 69(3):193-201. http://doi.org/10.3733/ca.E.v069n03p193 https://doi.org/10.3733/ca.E.v069n03p193

Harter T, Morel-Seytoux H. Peer Review of the IWFM, MODFLOW and HGS Model Codes: Potential for Water Management Applications in California's Central Valley and Other Irrigated Groundwater Basins 2013. p.121. California Water and Environmental Modeling Forum, Sacramento, August 2013

Hatton TJ. The Basics of Recharge and Discharge, Part 4: Catchment Scale Recharge Modeling. 1998. Collingwood, Victoria, Australia: CSIRO: Commonwealth Scientific and Industrial Research Organization.

Kim JS, Jain S. High-resolution streamflow trend analysis applicable to annual decision calendars: A western United States case study. Climatic Change. 2010. 102(3):699-707. https://doi.org/10.1007/s10584-010-9933-3

Kollet SJ, Maxwell RM. Integrated surface-groundwater flow modeling: A free-surface overland flow boundary condition in a parallel groundwater flow model. Adv Water Res. 2006. 29:945-58. https://doi.org/10.1016/j.advwatres.2005.08.006

Mack S. Geology and Ground-Water Features of Scott Valley Siskiyou County, California 1958. US Geological Survey Water-Supply Paper 1462. Washington DC

Moyle PB, Israel JA. Untested assumptions: Effectiveness of screening diversions for conservation of fish populations. Fisheries. 2005. 30(5):20-8. https://doi.org/10.1577/1548-8446(2005)30[20:UA]2.0.CO;2

NCRWQCB. Staff Report for the Action Plan for the Scott River Watershed Sediment and Temperature Total Maximum Daily Loads 2005. www.waterboards.ca.gov/water_issues/programs/tmdl/records/region_1/2010/ref3872.pdf

Nehlsen W, Williams JE, Lichatowich JA. Pacific salmon at the crossroads: Stocks at risk from California, Oregon, Idaho, and Washington. Fisheries. 1991. 16(2):4-21. https://doi.org/10.1577/1548-8446(1991)016replacecodelt0004:PSATCSreplacecodegt2.0.CO;2

Niswonger RG, Prudic DE. Documentation of the Stream-flow-Routing (SFR2) Package to Include Unsaturated Flow beneath Streams—A modification to SFR1 2005. p.50. US Geological Survey Techniques and Methods 6-A13

Prudic DE, Konikow LF, Banta ER. A New Stream-Flow Routing (SFR1) Package to Simulate Stream-Aquifer Interaction with MODFLOW-2000 2004. p.95. US Geological Survey Open-File Report 2004-1042

[SSPA] SS Papadopulos & Associates. Groundwater Conditions in Scott Valley 2012. Report prepared for the Karuk Tribe, March 2012

Zume JT, Tarhule AA. Modeling the response of an alluvial aquifer to anthropogenic and recharge stresses in the United States Southern Great Plains. J Earth Syst Sci. 2011. 120(4):557-72. https://doi.org/10.1007/s12040-011-0088-z


University of California, 2801 Second Street, Room 184, Davis, CA, 95618
Email: calag@ucanr.edu | Phone: (530) 750-1223 | Fax: (510) 665-3427
Website: https://calag.ucanr.edu