Introduction

The Rialto-Colton groundwater subbasin (California Department of Water Resources, 2018; hereafter referred to as “Rialto-Colton subbasin”) is in southern California about 60 miles (mi) east of Los Angeles (fig. 1). The basin is bounded by the San Jacinto fault (SJF) on the northeast, the Rialto-Colton fault (RCF) on the southwest, the San Gabriel Mountains on the northwest, and the Box Springs Mountains on the southeast (Dutcher and Garrett, 1963; Catchings and others, 2008). The semi-arid region has a mild winter and warm summer climate (Dutcher and Garrett, 1963). Historically, irrigated and non-irrigated agriculture was the dominant land-use type, but by the late 1950s, urban development sharply increased and eventually replaced agriculture as the dominant land use. In 1993, about two-thirds of the basin consisted of residential, industrial, and commercial land use (Woolfenden and Koczot, 2001). Local purveyors receive State Water Project water to supplement native groundwater extraction, primarily through artificial recharge (San Bernardino Valley Municipal Water District, 2020).

In 1991, the San Bernardino Valley Municipal Water District and the U.S. Geological Survey entered into a cooperative agreement to characterize the groundwater-flow system of the subbasin including the construction of a groundwater-flow model (Woolfenden and Koczot, 2001). The purpose and scope of this report is to map structural controls of groundwater flow in and near the Rialto-Colton subbasin using Interferometric Synthetic Aperture Radar (InSAR) methods for 1992–2010. Structural control information derived from InSAR results is used to improve the understanding of groundwater flow. Spatially detailed InSAR results also can be incorporated into future conceptual and hydrologic models of the Rialto-Colton subbasin and help water managers direct limited resources to areas and maximize the efficacy of future water resources management strategies.

Geohydrologic Setting

The lithology of the Rialto-Colton subbasin consists of Quaternary unconsolidated alluvial deposits, Tertiary to Quaternary partly consolidated continental deposits, Tertiary consolidated continental deposits, and a pre-Tertiary basement complex (Dutcher and Garrett, 1963). The unconsolidated and partly consolidated deposits primarily consist of sand and gravel with interbedded silt and clay, the consolidated deposits consist of clay and compacted sand, and the basement complex consists of igneous and metamorphic rock (Woolfenden and Kadhim, 1997). The region has several northwest-southeast trending faults. Two prominent faults in the region that partially define the areal extent of the Rialto-Colton subbasin are the SJF and the RCF. The SJF is a right-lateral strike-slip fault that has a vertical component (Dutcher and Garrett, 1963); the RCF likely is an abandoned trace of the SJF that vertically offsets the basement complex, where the basement is generally deeper in the Rialto-Colton subbasin than in the neighboring Chino groundwater subbasin (Woolfenden and Kadhim, 1997; California Department of Water Resources, 2018; fig. 1).

The unconfined to partly confined aquifer system consists of four hydraulically connected water-bearing units described by Dutcher and Garrett (1963) and Woolfenden and Kadhim (1997): unconsolidated alluvial deposits and upper, middle, and lower water-bearing zones in the partly consolidated continental deposits. Underlying the water-bearing units are the consolidated Tertiary continental deposits and basement complex which are not considered part of the aquifer system. The complex lithologic and structural framework of the Rialto-Colton subbasin controls groundwater flow in the basin and across boundaries to nearby basins.

Interferometric Synthetic Aperture Radar

Synthetic Aperture Radar (SAR) imagery is produced by reflecting radar signals off a target area and measuring the two-way travel time back to the satellite. The SAR interferometry technique uses two SAR images of the same area acquired at different times and “interferes” (differences) them, resulting in maps called interferograms that show line-of-sight ground-surface displacement (range change) between the two periods. The generation of an interferogram produces amplitude and phase components. Amplitude is the measure of the radar signal intensity returned to the satellite. Amplitude depends on the variable reflective properties of the landscape and shows features such as buildings, roads, mountains, and other reflective features. Phase is proportional to line-of-sight distance from the ground to the satellite (range) and is used to measure land-surface deformation (uplift or subsidence). For example, if the ground has moved away from the satellite (subsidence), an increase in phase is measured (fig. 2). Conversely, if the ground has moved closer to the satellite (uplift), a decrease in phase is measured. Interferograms contain phase values that repeat from 0 to 2 pi (a sine wave) and are termed “wrapped” interferograms because each time the data reach 2 pi, the values “wrap” back to zero. The wrapped interferograms in this study consist of bands of repeating color cycles where one complete cycle of color equates to approximately 1.1 inches (in.) of relative range (phase) change (fig. 2).

Interferograms in this study achieved a spatial resolution of approximately 100 feet (ft) because a 100-ft resolution digital elevation model was used for georeferencing and terrain corrections during processing. The SAR satellites used in this report “look” to the side at an average angle of about 23 degrees from vertical, which means that values of range change are about 92 percent of vertical elevation change. Interferograms in this study are presented in relative line-of-site range change because (1) the additional processing necessary to convert to vertical elevation change (“unwrapping”) can introduce spatially dependent errors that can negatively affect the identification of lineaments and (2) vertical magnitudes are not required to map structural control. Interferogram quality is dependent on topography, satellite orbit geometry, land-use practices, and other factors. For example, interferograms in areas of high topographic relief can produce excessive topographic effects that can mask land-surface deformation, and less than optimal satellite orbit geometry can exaggerate this error. Additionally, densely forested areas tend to absorb or reflect the radar signals from various depths in the tree canopy, resulting in random signal noise in those locations. In contrast, areas with low topographic relief and consistently uniform surfaces like the largely urban Rialto-Colton subbasin, or sparsely vegetated areas generally produce higher signal quality because the previously discussed error sources are minimized. Sneed and Brandt (2020) describe InSAR methodology in more detail.

For this study, SAR data from the European Space Agency’s (ESA; https://www.esa.int/) European Remote-Sensing 1 and 2 (ERS1 and ERS2) satellites (1992–2000) and from ESA’s ENVIronmental SATellite (ENVISAT) satellite (2003–10) were acquired from the University NAVSTAR Consortium (2013) for InSAR analysis. There were 31 ERS (ERS1 and ERS2 combined) and 35 ENVISAT interferograms that were of suitable quality to be used in the analysis. Interferograms spanned various periods during September 1992 and May 2010, with time spans ranging from about 1 month to about 4 years. Four representative interferograms are presented in this report (two ERS and two ENVISAT).

Mapping Structural Control of Groundwater Flow

In unconsolidated or partly consolidated basins like the Rialto-Colton subbasin, faults commonly are barriers to groundwater flow caused by the juxtaposition of deposits of differing compressibilities (such as consolidated rocks against unconsolidated deposits) or from the displacement of preferential flow paths in water-bearing deposits (Galloway and others, 1999). The locations of groundwater-flow barriers can be inferred where groundwater pumping occurs in compressible deposits and causes the aquifer system to compact or expand at differing amounts across the barrier. The high spatial resolution of interferograms can be used to infer the locations of buried faults not readily evident on the surface by identifying lineaments in the interferogram (Sneed and others, 2014). In the context of this report, lineaments are linear or partially linear features in interferograms that truncate land-surface deformation caused by the differential compaction or expansion of the aquifer system. Greater deformation rates from groundwater pumping facilitate a more robust identification of lineaments because the differential subsidence or uplift on either side of the groundwater-flow barrier is more pronounced. In the Rialto-Colton subbasin, periods of groundwater-level decline generally correlate with subsidence (aquifer-system compaction), and periods of groundwater-level recovery generally correlate with uplift (aquifer-system expansion; figs. 3, 4, 5).

Identification of Lineaments in Interferograms

InSAR results have identified three areas in and near the Rialto-Colton subbasin where lineaments are interpreted as buried faults acting as groundwater-flow barriers: (1) the northwestern area about 3 mi northwest of the City of Rialto, (2) the San Jacinto fault area west of the City of San Bernardino, and (3) the southeastern area about 2 mi southeast of the City of Colton. These areas where lineaments were identified are generally suitable for InSAR analysis because the areas where deformation was detected are fairly flat and largely urban. Some of the areas where lineaments were identified also are proximal to known groundwater production wells (figs. 1, 3, 4), and each area generally is near previously mapped faults. The names of lineaments identified in this report are generally based on the names of nearby barriers or faults identified in previous reports. For example, Lineament Hw is west of Barrier H (Woolfenden and Kadhim, 1997) and Lineament SJF is near the San Jacinto fault (Morton and Miller, 2006).

Differentiating Pumping-Induced Deformation from Tectonic Deformation

Pumping-induced deformation (aquifer-system compaction) and tectonic-induced deformation is differentiated qualitatively. The regional deformation associated with long-term tectonic plate movement and earthquakes can be differentiated from the localized deformation caused by pumping because tectonic motions cause deformation over a relatively large area (Massonnet and others, 1993) compared with the relatively localized deformation typically associated with pumping (Bawden and others, 2003).

The continuous Global Positioning System (CGPS) site P612 (University NAVSTAR Consortium, 2014) is in the foothills of the San Bernardino Mountains (fig. 1) and was used to help differentiate vertical deformation caused by long-term tectonic plate movement from displacements caused by localized pumping. Day-to-day CGPS vertical deformation measurements varied by up to 2 in. or more, which likely is the result of variable atmospheric moisture, electronic noise, and other effects not directly related to land-surface-deformation (Zerbini and others, 2001; Williams and others, 2004; Langbein, 2008). To minimize this high-frequency noise, a 31-day moving average was applied to the CGPS vertical deformation measurements. The removal of day-to-day variations in the CGPS measurements did not remove seasonal or long-term deformation trends and permitted more meaningful land-surface deformation measurements (fig. 6).

During 2005–10, CGPS site P612 did not record substantial, longer-term (annual to multi-annual) vertical land-surface deformation (fig. 6). Shen and others (2011) identified up to about 0.04 inches per year (in/yr) of tectonic uplift along the Transverse Ranges during 1970–2004 where the San Bernardino Mountains are located. However, Shen and others (2011) did not indicate substantial longer-term, vertical land-surface deformation during 1970–2004 in the area immediately southwest of the San Andreas fault zone near the location of site P612 (fig. 4). Therefore, tectonic uplift in the Transverse Ranges did not contribute substantially to CGPS-measured vertical land-surface deformation and is unlikely to adversely affect InSAR measurements in the study area.

Data from the U.S. Geological Survey Earthquake Hazards Program (U.S. Geological Survey, 2014a) indicated about 100 earthquakes ranging from Magnitude (M) 4.5 to M 7.3 occurred less than 100 mi from the study area during 1992–2010. Seven of these earthquakes, ranging from M 4.5 to M 5.5, occurred during the operational history of site P612 (2005–10) and had no appreciable effect on vertical deformation measured at the site (fig. 6). Therefore, earthquakes of up to M 5.5 likely did not adversely affect InSAR analyses in this study. Six interferograms used in this study included either the January 17, 1994, Northridge (M 6.7) earthquake or the October 16, 1999, Hector Mine (M 7.1) earthquake (fig. 4B). Although these larger earthquakes predate the installation of site P612, the observed deformation patterns and lineaments in these interferograms are consistent with other interferograms used in this study that do not span large earthquakes.

Correlation of Changes in Pumping, Groundwater Levels, and Deformation

In the northwestern area, data from selected production and observation wells indicated that changes in pumping correspond, in part, with changes in groundwater levels in nearby wells, and these changes in groundwater levels correlated with observed deformation patterns. For example, pumping records for production well 1S/5W-5A3 during 1996–98 indicate a nearly 1,000 acre-feet (acre-ft) reduction in pumping volume, which corresponded to about 10 ft of groundwater-level increases in nearby observation well 1S/5W-4D2 and about 1 in. of uplift for the same period (figs. 3, 4A). During 1999–2000, pumping was increased by about 1,000 acre-ft at production well 1S/5W-5A3, which corresponded to about 2 ft of groundwater-level decline in the nearby observation well (1S/5W-4D2) and about 0.5 in. of subsidence for the same period (figs. 3, 4B). However, from 2003 to 2010, changes in pumping are not as well correlated with changes in nearby groundwater levels. Production well 1N/5W-33N1 showed an increase of nearly 350 acre-ft in pumping volume, but nearby observation well 1S/5W-4D2 showed nearly 10 ft of groundwater-level decline. Furthermore, nearby production wells 1S/5W-5A4 and 1S/5W-5A5 each indicated about 200 acre-ft of reduced pumping volume (fig. 3B). Land-surface deformation was not observed in this area during 2003–10 despite the increase in total pumping (fig. 3B). The two production wells (1N/5W-33N1 and 1S/5W-5A3) that appear to partially influence groundwater levels in observation well 1S/5W-4D2 are northeast of Barrier H, whereas production wells (1N/5W-5A4 and 1S/5W-5A5) with the least influence on observation well 1S/5W-4D2 are southwest of Barrier H (fig. 1). As noted earlier in this report, the three production wells being used during the second period of this study (2003–10; 1N/5W-5A4, 1S/5W-5A5, and 1N/5W-33N1) penetrate deeper into the aquifer system than the single production well used during the first period of this study (1992–2000). The increase in total groundwater pumping and lack of observed subsidence in this area during the second period (2003–10) indicated that the location (relative to mapped groundwater barriers or lineaments) or the depth of pumping during this time was not sufficient to create conditions in which lineaments can be readily identified in interferograms and indicated that pumping during this later time may have occurred in less compressible deposits.

In the San Jacinto fault area, pumping data were not available; however, changes in groundwater levels in observation wells generally correlated with observed deformation patterns. Groundwater levels at well 1N/4W-31P3 decreased between 12 and 45 feet during 1996–98 and 1999–2000, where between 0.5 and 0.75 in. of subsidence was observed, respectively. Data from observation well 1N/4W-31P3 also indicated groundwater levels increased about 75 ft during 2005, whereas interferograms that also spanned 2005 show about 0.75 in. of uplift (figs. 3A, 5A). Groundwater levels at well 1S/4W-6H1 were relatively stable during 1996–2001 (compared to groundwater levels at the nearby well 1N/4W-31P3), and substantial deformation did not occur at this location during a similar period (figs. 3A, 4).

In the southeastern area, groundwater-pumping data were not available, and the correlation between changes in available groundwater levels and observed land-surface deformation was less clear. Groundwater levels at well 1S/4W-27M1 increased about 9 ft during 2004–05, and about 0.75 in. of uplift was observed (figs. 3A, 5A). However, land-surface deformation observed during 2004–05 (fig. 5A) was distinct from land-surface deformation observed in most of the other interferograms (1992–2010) used in this study (fig. 5B). Therefore, observed land-surface deformation patterns were not strongly influenced by changes in groundwater levels at observation well 1S/4W-27M1.

Summary

The U.S. Geological Survey, in cooperation with the San Bernardino Valley Municipal Water District, analyzed structural control of groundwater flow in the Rialto-Colton groundwater subbasin using Interferometric Synthetic Aperture Radar (InSAR) methods. The high spatial resolution of InSAR imagery revealed three areas in and near the Rialto-Colton groundwater subbasin where buried faults are interpreted as groundwater-flow barriers where groundwater pumping occurs. Groundwater-flow barriers were detected in InSAR imagery for the San Jacinto fault area and the southeastern area during 1992–2010. Groundwater-flow barriers were detected in InSAR imagery for the northwestern area during 1992–2000 and absent in InSAR imagery during 2003–10 when deformation was not detected. Available pumping records at selected wells in the northwestern area indicated the amount of groundwater extracted in this area increased greatly after 2002 and was being pumped from a deeper part of the aquifer. The absence of InSAR-detected lineaments (and the deformation that facilitated their identification) in this area during 2003–10 indicated that the location (relative to mapped groundwater barriers or lineaments) or the depth of pumping during this time was not sufficient to create conditions in which lineaments can be readily identified in interferograms and indicated that pumping during this later period may have occurred in less-compressible deposits.

The locations of many faults in and near the Rialto-Colton groundwater subbasin are not precisely known. Traditional geologic mapping techniques and modeling studies have identified (or inferred) numerous faults acting as groundwater barriers; however, traditional methods rely in part on the spatial density of existing lithologic and hydrologic data to infer the locations of faults acting as groundwater barriers. The high spatial resolution of InSAR data combined with the knowledge gained from previous studies were used to more precisely define the location and extent of faults acting as groundwater barriers. This increase in precision will inform development of better conceptual and hydrologic models of the Rialto-Colton groundwater subbasin.