Simulating the sensitivity of evapotranspiration and streamflow to large-scale groundwater depletion

American Association for the Advancement of Science (AAAS) | June 19th, 2019

Groundwater pumping has caused marked aquifer storage declines over the past century. In addition to threatening the viability of groundwater-dependent economic activitie

Groundwater pumping has caused marked aquifer storage declines over the past century. In addition to threatening the viability of groundwater-dependent economic activities, storage losses reshape the hydrologic landscape, shifting groundwater surface water exchanges and surface water availability. A more comprehensive understanding of modern groundwater-depleted systems is needed as we strive for improved simulations and more efficient water resources management. Here, we begin to address this gap by evaluating the impact of 100 years of groundwater declines across the continental United States on simulated watershed behavior. Subsurface storage losses reverberate throughout hydrologic systems, decreasing streamflow and evapotranspiration. Evapotranspiration declines are focused in water-limited periods and shallow groundwater regions. Streamflow losses are widespread and intensify along drainage networks, often occurring far from the point of groundwater abstraction. Our integrated approach illustrates the sensitivity of land surface simulations to groundwater storage levels and a path toward evaluating these connections in large-scale models.

Simulation of Ground-Water / Surface-Water Flow in the Santa Clara–Calleguas Ground-Water Basin, Ventura County, California

U.S. Geological Survey (USGS) | July 17th, 2003

Ground water is the main source of water in the Santa Clara–Calleguas ground-water basin that covers about 310 square miles in Ventura County, California. A steady incr

Ground water is the main source of water in the Santa Clara–Calleguas ground-water basin that covers about 310 square miles in Ventura County, California. A steady increase in the demand for surface- and ground-water resources since the late 1800s has resulted in streamflow depletion and ground-water overdraft. This steady increase in water use has resulted in seawater intrusion, inter-aquifer flow, land subsidence, and ground-water contamination.  The Santa Clara–Calleguas Basin consists of multiple aquifers that are grouped into upper- and lower-aquifer systems. The upper-aquifer system includes the Shallow, Oxnard, and Mugu aquifers. The lower-aquifer system includes the upper and lower Hueneme, Fox Canyon, and Grimes Canyon aquifers. The layered aquifer systems are each bounded below by regional unconformities that are overlain by extensive basal coarse-grained layers that are the major pathways for ground-water production from wells and related seawater intrusion. The aquifer systems are bounded below and along mountain fronts by consolidated bedrock that forms a relatively impermeable boundary to ground-water flow. Numerous faults act as additional exterior and interior boundaries to ground-water flow. The aquifer systems extend offshore where they crop out along the edge of the submarine shelf and within the coastal submarine canyons. Submarine canyons have dissected these regional aquifers, providing a hydraulic connection to the ocean through the submarine outcrops of the aquifer systems. Coastal landward flow (seawater intrusion) occurs within both the upper- and lower-aquifer systems. A numerical ground-water flow model of the Santa Clara–Calleguas Basin was developed by the U.S. Geological Survey to better define the geohydrologic framework of the regional ground-water flow system and to help analyze the major problems affecting water-resources management of a typical coastal aquifer system. Construction of the Santa Clara–Calleguas Basin model required the compilation of geographic, geologic, and hydrologic data and estimation of hydraulic properties and flows. The model was calibrated to historical surface-water and ground-water flow for the period 1891–1993. Sources of water to the regional ground-water flow system are natural and artificial recharge, coastal landward flow from the ocean (seawater intrusion), storage in the coarse-grained beds, and water from compaction of fine-grained beds (aquitards). Inflows used in the regional flow model simulation include streamflows routed through the major rivers and tributaries; infiltration of mountain-front runoff and infiltration of precipitation on bedrock outcrops and on valley floors; and artificial ground-water recharge of diverted streamflow, irrigation return flow, and treated sewage effluent.  Most natural recharge occurs through infiltration (losses) of streamflow within the major rivers and tributaries and the numerous arroyos that drain the mountain fronts of the basin. Total simulated natural recharge was about 114,100 acre-feet per year (acre-ft/yr) for 1984–93: 27,800 acre-ft/yr of mountain-front and bedrock recharge, 24,100 acre-ft/yr of valley-floor recharge, and 62,200 acre-ft/yr of net streamflow recharge. Artificial recharge (spreading of diverted streamflow, irrigation return, and sewage effluent) is a major source of ground-water replenishment. During the 1984–93 simulation period, the average rate of artificial recharge at the spreading grounds was about 54,400 acre-ft/yr, 13 percent less than the simulated natural recharge rate for streamflow infiltration within the major rivers and tributaries. Estimated recharge from infiltration of irrigation return flow on the valley floors averaged about 51,000 acre-ft/yr, and treated sewage effluent averaged about 9,000 acre-ft/yr. Artificial recharge as streamflow diversion to the spreading grounds has occurred since 1929, and treated-sewage effluent has been discharged to stream channels since 1930.  Under predevelopment conditions, the largest discharge from the ground-water system was outflow as coastal seaward flow and evapotranspiration. Pumpage of ground water from thousands of water-supply wells has diminished these outflows and is now the largest outflow from the ground-water flow system. The distribution of pumpage for 1984–93 indicates that most of the pumpage occurs in the Oxnard Plain subareas (37 percent) and in the upper Santa Clara River Valley subareas (37 percent). The total average simulated pumpage was about 247,000 acre-ft/yr (59 percent); of which about 146,000 acre-ft/yr was from the Fox Canyon Groundwater Management Agency (FGMA) subareas and 101,000 acre-ft/yr (41 percent) from the non-FGMA subareas. Of the total 1984–93 pumpage, 46 percent was contributed by natural recharge, 22 percent was contributed by artificial recharge from diverted streamflow, 20 percent was contributed by irrigation return flow, 4 percent was contributed from sewage-effluent infiltration, 6 percent was contributed from storage depletion, and 2 percent was contributed from coastal landward flow (seawater intrusion). Seawater intrusion was first suspected in 1931 when water levels were below sea level in a large part of the Oxnard Plain. The simulation of regional ground-water flow indicated that coastal landward flow (seawater intrusion) began in 1927 and continued to the end of the period of simulation (1993). During wet periods or periods of reduced demand for ground water, the direction of coastal flow in the upper-aquifer system reverses from landward to seaward. During the 1984–93 period, the simulated total net seaward flow was 9,500 acre-feet in the upper-aquifer system, which is considerably less than that simulated for predevelopment conditions. During the same period, total simulated landward flow in the lower-aquifer system was 64,200 acre-feet. Water-level declines in the basin have induced land subsidence that was first measured in 1939 and have resulted in as much as 2.7 feet land subsidence in the southern part of the Oxnard Plain. The model simulated a total of 3 feet of land subsidence in the southern part of the Oxnard Plain and as much as 5 feet in the Las Posas Valley subbasins. Model simulations indicate that most of the land subsidence occurred after the drought of the late 1920s and during the agricultural expansion of the 1950s and 1960s. The results also indicate that subsidence occurred primarily in the upper-aquifer system prior to 1959, but in the lower-aquifer system between 1959–93 owing to an increase in pumpage from the lower-aquifer system. The calibrated ground-water flow model was used to assess future ground-water conditions based on proposed water-supply projects in the existing management plan for the Santa Clara–Calleguas ground-water basin. All the projections of the proposed water-supply projects in the existing management plan have reduced pumpage in the FGMA areas which resulted in a reduction but not an elimination of storage depletion and related coastal landward flow (seawater intrusion) and subsidence, a reduction in streamflow recharge, and an increase in coastal seaward flow and underflow to adjacent subareas from the Oxnard Plain. A comparison of management simulations based on historical inflows and a spectral estimate of inflows shows increased coastal landward flow (seawater intrusion), storage depletion, and increased land subsidence due to a drought projected earlier in the spectral estimate of inflows than in the historical inflows. The spectral estimate probably provides a smoother and more realistic transition between historical and future climatic conditions. The model also was used to simulate potential alternative water-supply projects in the Santa Clara–Calleguas ground-water basin. These seven alternative water-supply projects were proposed to help manage the effects of increasing demand and variable supply on seawater intrusion, subsidence, increased withdrawal from storage, and vertical and lateral flow between subareas and aquifers systems. Stopping pumpage primarily in the lower-aquifer system in the South Oxnard Plain subarea had the largest effect on reducing coastal landward flow (seawater intrusion) of all the potential cases evaluated. Shifting pumpage from the lower- to the upper-aquifer system in the South Oxnard Plain subarea yielded the largest combined effect on coastal flow with a reduction of coastal landward flow in the lower-aquifer system and coastal seaward flow from the upper-aquifer system. A seawater-barrier injection project stopped coastal landward flow (seawater intrusion) in the upper-aquifer system but also resulted in large quantities of coastal seaward flow. The recharge of water in Happy Camp Canyon resulted in water-level rises that were above land surface (not feasible) in the East Las Posas Valley subarea but in no significant changes in hydrologic conditions in other parts of the basin. 

Simulation of Ground-Water Flow and Land Subsidence in the Antelope Valley Ground-Water Basin, California

U.S. Geological Survey (USGS) | June 30th, 2003

Antelope Valley, California, is a topographically closed basin in the western part of the Mojave Desert, about 50 miles northeast of Los Angeles. The Antelope Valley grou

Antelope Valley, California, is a topographically closed basin in the western part of the Mojave Desert, about 50 miles northeast of Los Angeles. The Antelope Valley ground-water basin is about 940 square miles and is separated from the northern part of Antelope Valley by faults and low-lying hills. Prior to 1972, ground water provided more than 90 percent of the total water supply in the valley; since 1972, it has provided between 50 and 90 percent. Most ground-water pumping in the valley occurs in the Antelope Valley ground-water basin, which includes the rapidly growing cities of Lancaster and Palmdale. Ground-water-level declines of more than 200 feet in some parts of the ground-water basin have resulted in an increase in pumping lifts, reduced well efficiency, and land subsidence of more than 6 feet in some areas. Future urban growth and limits on the supply of imported water may continue to increase reliance on ground water. To better understand the ground-water flow system and to develop a tool to aid in effectively managing the water resources, a numerical model of ground-water flow and land subsidence in the Antelope Valley ground-water basin was developed using old and new geohydrologic information.

Simulation of Ground-water Flow in the Mojave River Basin, California

U.S. Geological Survey (USGS) | July 1st, 2001

Simulation of Ground-Water/Surface-Water Flow in the Santa Clara–Calleguas Ground-Water Basin, Ventura County, California

U.S. Geological Survey (USGS) | July 15th, 2003

Ground-water is the main source of water in the Santa Clara–Calleguas ground-water basin that covers about 310 square miles in Ventura County, California. A steady incr

Ground-water is the main source of water in the Santa Clara–Calleguas ground-water basin that covers about 310 square miles in Ventura County, California. A steady increase in the demand for surface- and ground-water resources since the late 1800s has resulted in streamflow depletion and ground-water overdraft. This steady increase in water use has resulted in seawater intrusion, inter-aquifer flow, land subsidence, and ground-water contamination. 

Simulation of ground-water/surface-water fow in the Santa Clara-Calleguas ground-water basin, Ventura County, California

U.S. Geological Survey (USGS) | July 17th, 2003

Groundwater is the main source of water in the Santa Clara–Calleguas ground-water basin that covers about 310 square miles in Ventura County, California. A steady incre

Groundwater is the main source of water in the Santa Clara–Calleguas ground-water basin that covers about 310 square miles in Ventura County, California. A steady increase in the demand for surface- and ground-water resources since the late 1800s has resulted in streamflow depletion and ground-water overdraft. This steady increase in water use has resulted in seawater intrusion, inter-aquifer flow, land subsidence, and ground-water contamination. The Santa Clara–Calleguas Basin consists of multiple aquifers that are grouped into upper- and lower-aquifer systems. The upper-aquifer system includes the Shallow, Oxnard, and Mugu aquifers. The lower-aquifer system includes the upper and lower Hueneme, Fox Canyon, and Grimes Canyon aquifers. The layered aquifer systems are each bounded below by regional unconformities that are overlain by extensive basal coarse-grained layers that are the major pathways for ground-water production from wells and related seawater intrusion. The aquifer systems are bounded below and along mountain fronts by consolidated bedrock that forms a relatively impermeable boundary to ground-water flow. Numerous faults act as additional exterior and interior boundaries to ground-water flow. The aquifer systems extend offshore where they crop out along the edge of the submarine shelf and within the coastal submarine canyons. Submarine canyons have dissected these regional aquifers, providing a hydraulic connection to the ocean through the submarine outcrops of the aquifer systems. Coastal landward flow (seawater intrusion) occurs within both the upper- and lower-aquifer systems. A numerical ground-water flow model of the Santa Clara–Calleguas Basin was developed by the U.S. Geological Survey to better define the geohydrologic framework of the regional ground-water flow system and to help analyze the major problems affecting water-resources management of a typical coastal aquifer system. Construction of the Santa Clara–Calleguas Basin model required the compilation of geographic, geologic, and hydrologic data and estimation of hydraulic properties and flows. The model was calibrated to historical surface-water and ground-water flow for the period 1891–1993. Sources of water to the regional ground-water flow system are natural and artificial recharge, coastal landward flow from the ocean (seawater intrusion), storage in the coarse-grained beds, and water from compaction of fine-grained beds (aquitards). Inflows used in the regional flow model simulation include streamflows routed through the major rivers and tributaries; infiltration of mountain-front runoff and infiltration of precipitation on bedrock outcrops and on valley floors; and artificial ground-water recharge of diverted streamflow, irrigation return flow, and treated sewage effluent. Most natural recharge occurs through infiltration (losses) of streamflow within the major rivers and tributaries and the numerous arroyos that drain the mountain fronts of the basin. Total simulated natural recharge was about 114,100 acre-feet per year (acre-ft/yr) for 1984–93: 27,800 acre-ft/yr of mountain-front and bedrock recharge, 24,100 acre-ft/yr of valley-floor recharge, and 62,200 acre-ft/yr of net streamflow recharge. Artificial recharge (spreading of diverted streamflow, irrigation return, and sewage effluent) is a major source of ground-water replenishment. During the 1984–93 simulation period, the average rate of artificial recharge at the spreading grounds was about 54,400 acre-ft/yr, 13 percent less than the simulated natural recharge rate for streamflow infiltration within the major rivers and tributaries. Estimated recharge from infiltration of irrigation return flow on the valley floors averaged about 51,000 acre-ft/yr, and treated sewage effluent averaged about 9,000 acre-ft/yr. Artificial recharge as streamflow diversion to the spreading grounds has occurred since 1929, and treated-sewage effluent has been discharged to stream channels since 1930. Under predevelopment conditions, the largest discharge from the ground-water system was outflow as coastal seaward flow and evapotranspiration. Pumpage of ground water from thousands of water-supply wells has diminished these outflows and is now the largest outflow from the ground-water flow system. The distribution of pumpage for 1984–93 indicates that most of the pumpage occurs in the Oxnard Plain subareas (37 percent) and in the upper Santa Clara River Valley subareas (37 percent). The total average simulated pumpage was about 247,000 acre-ft/yr (59 percent); of which about 146,000 acre-ft/yr was from the Fox Canyon Groundwater Management Agency (FGMA) subareas and 101,000 acre-ft/yr (41 percent) from the non-FGMA subareas. Of the total 1984–93 pumpage, 46 percent was contributed by natural recharge, 22 percent was contributed by artificial recharge from diverted streamflow, 20 percent was contributed by irrigation return flow, 4 percent was contributed from sewage-effluent infiltration, 6 percent was contributed from storage depletion, and 2 percent was contributed from coastal landward flow (seawater intrusion). Seawater intrusion was first suspected in 1931 when water levels were below sea level in a large part of the Oxnard Plain. The simulation of regional ground-water flow indicated that coastal landward flow (seawater intrusion) began in 1927 and continued to the end of the period of simulation (1993). During wet periods or periods of reduced demand for ground water, the direction of coastal flow in the upper-aquifer system reverses from landward to seaward. During the 1984–93 period, the simulated total net seaward flow was 9,500 acre-feet in the upper-aquifer system, which is considerably less than that simulated for predevelopment conditions. During the same period, total simulated landward flow in the lower-aquifer system was 64,200 acre-feet. Water-level declines in the basin have induced land subsidence that was first measured in 1939 and have resulted in as much as 2.7 feet land subsidence in the southern part of the Oxnard Plain. The model simulated a total of 3 feet of land subsidence in the southern part of the Oxnard Plain and as much as 5 feet in the Las Posas Valley subbasins. Model simulations indicate that most of the land subsidence occurred after the drought of the late 1920s and during the agricultural expansion of the 1950s and 1960s. The results also indicate that subsidence occurred primarily in the upper-aquifer system prior to 1959, but in the lower-aquifer system between 1959–93 owing to an increase in pumpage from the lower-aquifer system. The calibrated ground-water flow model was used to assess future ground-water conditions based on proposed water-supply projects in the existing management plan for the Santa Clara–Calleguas ground-water basin. All the projections of the proposed water-supply projects in the existing management plan have reduced pumpage in the FGMA areas which resulted in a reduction but not an elimination of storage depletion and related coastal landward flow (seawater intrusion) and subsidence, a reduction in streamflow recharge, and an increase in coastal seaward flow and underflow to adjacent subareas from the Oxnard Plain. A comparison of management simulations based on historical inflows and a spectral estimate of inflows shows increased coastal landward flow (seawater intrusion), storage depletion, and increased land subsidence due to a drought projected earlier in the spectral estimate of inflows than in the historical inflows. The spectral estimate probably provides a smoother and more realistic transition between historical and future climatic conditions. The model also was used to simulate potential alternative water-supply projects in the Santa Clara–Calleguas ground-water basin. These seven alternative water-supply projects were proposed to help manage the effects of increasing demand and variable supply on seawater intrusion, subsidence, increased withdrawal from storage, and vertical and lateral flow between subareas and aquifers systems. Stopping pumpage primarily in the lower-aquifer system in the South Oxnard Plain subarea had the largest effect on reducing coastal landward flow (seawater intrusion) of all the potential cases evaluated. Shifting pumpage from the lower- to the upper-aquifer system in the South Oxnard Plain subarea yielded the largest combined effect on coastal flow with a reduction of coastal landward flow in the lower-aquifer system and coastal seaward flow from the upper-aquifer system. A seawater-barrier injection project stopped coastal landward flow (seawater intrusion) in the upper-aquifer system but also resulted in large quantities of coastal seaward flow. The recharge of water in Happy Camp Canyon resulted in water-level rises that were above land surface (not feasible) in the East Las Posas Valley subarea but in no significant changes in hydrologic conditions in other parts of the basin.

Simulation of groundwater and surface-water resources of the San Antonio Creek Valley watershed, Santa Barbara County, California

U.S. Geological Survey (USGS) | February 1st, 2022

In the San Antonio Creek Valley watershed (SACVW), western Santa Barbara County, California, groundwater is the primary source of water for agricultural irrigation, the t

In the San Antonio Creek Valley watershed (SACVW), western Santa Barbara County, California, groundwater is the primary source of water for agricultural irrigation, the town of Los Alamos, and supplemental water to Vandenberg Space Force Base (VSFB). Groundwater pumpage has increased since the 1970s as non-irrigated agricultural land has been converted to irrigated land and as local pumping for municipal use has increased. This increase in groundwater use has resulted in declining groundwater levels, adjustments in surface-water flows and species habitats, and changes in water quality. Water managers are addressing the challenges of meeting this increased demand while maintaining sustainable groundwater supplies. To address these challenges, Santa Barbara County Water Agency, Vandenberg Space Force Base (VSFB), and the U.S. Geological Survey (USGS) undertook a cooperative study to characterize the integrated hydrologic system of the SACVW and develop tools to better understand and manage the groundwater system. The objectives of this study were to improve the understanding of the integrated hydrologic system and incorporate the understanding into an integrated groundwater and surface-water flow model that can be used to help manage the water resources in the SACVW. The San Antonio Creek integrated model (SACIM) was developed using the USGS coupled groundwater and surface-water flow model to simulate the hydrologic system of the SACVW and provide annual and average water budgets for 1948–2018 water years. Results from the SACIM indicated that between 1948 and 2018, total groundwater from storage (storage depletion) for the period was 453,300 acre-feet (acre-ft). Agricultural pumpage was the largest discharge and accounted for a total of 1,020,000 acre-ft of groundwater. Increased pumpage since the mid-1980s (of which agricultural pumpage is the primary component) is tied to an increased rate of storage depletion and reduced rates of groundwater evapotranspiration and surface leakage (groundwater discharge to the surface and soil zone). The increased pumpage also reduced subsurface inflow to Barka Slough, resulting in a decline in upward flow through the underlying hydrogeologic units and surface leakage. In addition to quantifying historical changes in the integrated hydrologic system, the SACIM is a tool than can be used by water managers to evaluate the effects of different climatic and hydrologic conditions and management strategies.

Simulation of the Impacts of Spring Diversions on Streamflow in the Strawberry Creek Watershed, San Bernardino County, California, Using an Integrated Hydrological Model

U.S. Geological Survey (USGS) | July 27th, 2026

The Strawberry Creek watershed, in the San Bernardino Mountains of southern California, features a group of natural springs known as Arrowhead Springs that have been augm

The Strawberry Creek watershed, in the San Bernardino Mountains of southern California, features a group of natural springs known as Arrowhead Springs that have been augmented with diversions in the form of sub-horizontal borings and tunnels. Understanding the effects of these structures on streamflow through groundwater capture is crucial for managing surface-water resources in this watershed. In this study, we constructed the Strawberry Creek integrated hydrological model to increase this understanding. The Strawberry Creek integrated hydrological model is an integrated surface runoff and groundwater model, GSFLOW to simulate surface runoff and infiltration and groundwater flow. The model has 263 rows and 176 columns in each of its three layers. The model area includes the Strawberry Creek and four adjacent watersheds. The precipitation- runoff modeling system was calibrated using two streamflow gaging stations, and the groundwater and surface-water flow model was calibrated to the reported spring diversion discharge and a sparse number of groundwater- level measurements. The Strawberry Creek integrated hydrological model was run with and without diversions active, and simulated streamflow was compared, indicating that in the headwaters of Strawberry Creek, about 35 percent of the diversion flow was captured from base flow.

Simulation of Water-Management Alternatives in the Mojave River Ground-Water Basin, California

U.S. Geological Survey (USGS) | July 1st, 2002

Simulations of Ground-Water Flow and Particle Pathline Analysis in the Zone of Contribution of a Public-Supply Well in Modesto, Eastern San Joaquin Valley, California

U.S. Geological Survey (USGS) | July 1st, 2008