Crystalline Basement Rocks

Crystalline rocks crop out around the margins of the Yucaipa subbasin (fig. 5) and underlie all sedimentary materials (figs. 7A, 7C, 8A, 8B). Herein, these rocks are referred to as crystalline-basement rocks because they form a hard foundation for overlying sedimentary materials. The basement rocks can be subdivided into three distinctive “packages,” or region-wide lithologies (figs. 2, 5; Matti and others, 1992b; Matti and Morton, 1993): (1) crystalline-basement rocks of Mojave Desert-type that outcrop in the San Bernardino Mountains north of the San Andreas fault zone, (2) crystalline-basement rocks of San Gabriel Mountains-type that outcrop in the Crafton Hills and Yucaipa hills north of the Banning fault, and (3) crystalline-basement rocks of Peninsular Ranges-type that outcrop south of the Yucaipa subbasin and south of the Banning fault.

Tertiary Sedimentary Rocks

Tertiary sedimentary rocks are sedimentary materials that are consolidated. They do not crop out in the Yucaipa subbasin but are thought to occur in the deeper subsurface in the Western Heights subarea (fig. 7C). Lower Pleistocene and Pliocene sedimentary rocks crop out in The Badlands south of the Yucaipa subbasin, and presumably these occur in the subsurface south of the Banning fault (figs. 5, 7A). Sedimentary rocks older than the San Timoteo Formation may occur in the deep subsurface south of the Banning fault as well as the Western Heights subarea (figs. 7A, C), and these may be as old as late Miocene. Upper Miocene sedimentary rocks crop out north of the Yucaipa subbasin in the San Bernardino Mountains north of the San Andreas fault zone.

San Timoteo Formation

The San Timoteo Formation is a regionally important stratigraphic unit that flanks the northeast margin of the San Jacinto fault. The unit, for the most part, does not crop out within the Yucaipa subbasin, but is well exposed in The Badlands (fig. 5; Morton and Matti, 2001; Matti and others, 2003b, 201561; Morton and Miller, 2006). For much of this outcrop exposure, the San Timoteo Formation is folded into a northwest-trending asymmetric anticline having a southwestern limb that dips steeply toward the San Jacinto Valley and a northeastern limb that dips shallowly toward (and presumably beneath) the Yucaipa subbasin. The other surface exposure of the San Timoteo Formation is at the southeastern part of the Yucaipa subbasin—in the hangingwall (upper plate, indicated by the presence of sawteeth in fig. 5) of the Cherry Valley thrust fault (fig. 5). There, the unit also is folded into an asymmetric anticline having a steeply dipping southwestern limb and a shallower northeastern limb (Matti and others, 2015).

Regionally, the San Timoteo Formation is separated into multiple members and subunits (fig. 5; Morton and Miller, 2006; Matti and others, 2015). The upper member and Reche Canyon member crop out in the northwestern part of The Badlands (Morton and Miller, 2006) and probably have no counterparts within the Yucaipa subbasin, although the upper member crops out adjacent to the westernmost subbasin boundary. The middle and lower members of the San Timoteo Formation crop out mainly south of the Yucaipa subbasin in the El Casco 7.5-minute quadrangle (Matti and others, 2015). The middle member generally consists of light-gray, sheet-like layers of well-consolidated to cemented pebble-cobble conglomerate, with medium to thick intervals of gray-brown fine- to coarse-grained sandstone and minor amounts of siltstone and mudstone intervals (Frick, 1921; Matti and others, 2015). The lower member is separated into several subunits (Matti and others, 2015) that generally are much finer grained than the middle member, including abundant fine- to coarse-grained sandstone and associated intervals of yellowish-gray colored mudrock.

The San Timoteo Formation is shown as “undifferentiated” in the Beaumont 7.5-minute quadrangle southeast of the Yucaipa subbasin (figs. 2, 5). In this report, the formation in this area is adopted as presented by Dibblee and Minch (2003a) and Rewis and others (2006)—those authors did not differentiate between different members of the San Timoteo Formation. Interpreting different members of the formation in this area was beyond the scope of this report. Typical lithologies of the undifferented sediments of the San Timoteo Formation include well-consolidated to cemented, well sorted fine- to coarse-grained sand and sandstone and sheet-like layers of well-consolidated to indurated pebble-cobble gravel and conglomerate (Dibblee and Minch, 2003a; Rewis and others, 2006).

Sediments of the San Timoteo Formation were sourced from crystalline rocks north and northeast of the formation’s depocenter. In large part the sources were rocks of San Gabriel Mountains-type (both upper and lower plates of the Vincent thrust), but in the northwestern part of The Badlands some sediment was sourced from rocks of Mojave Desert-type (the quartzite-bearing conglomerate unit of Matti and others, 2003b, unit QTstcq; see discussion by Morton and Matti, 1993). The depositional transition between lower and middle members of the San Timoteo Formation and younger parts of the formation—especially the quartzite-bearing conglomerate unit (mapped as part of the upper member in fig. 5; Morton and Matti, 2001; Matti and others, 2003b)—marks an important shift in Quaternary paleogeography for the region surrounding the Yucaipa subbasin. This shift presumably was linked to inception of dextral slip on the San Jacinto fault. Motion on that structure, coupled with simultaneous folding and uplift of the lower and middle San Timoteo members (Kendrick and others, 2002) uplifted The Badlands landscape above adjacent lowlands where sediment continued to accumulate—but in depositional patterns different from those that existed when the lower and middle members accumulated. This new paleogeographic setting not only received sediment sourced from the San Bernardino Mountains but also Pelona Schist-bearing sediment derived from outcrops in the southeastern San Gabriel Mountains. Fossils collected from the San Timoteo Formation in The Badlands (Morton and others, 1986; Repenning, 1987; Morton and Matti, 1993; Albright, 1997, 19992; Reynolds and others, 2013) suggest that these paleogeographic changes occurred after about 1.3–1.5 Ma (Albright, 1999). It is not clear that this paleogeographic evolution is recorded in subsurface sedimentary rocks in the Yucaipa subbasin north of the Banning fault.

The location of cross section A-A’ (figs. 5 and 7A) does not allow the section to show geologic units that occur beneath the San Timoteo Formation in the southeastern part of The Badlands. There, the San Timoteo Formation is underlain by the Mount Eden Formation, a Miocene unit of nonmarine sedimentary rock characterized by clasts exclusively of Peninsular Ranges-type (Matti and others, 2015). Stratigraphic and paleogeographic relations between the Mount Eden beds and deep subsurface sedimentary rocks south of the Banning fault, in the hangingwall of the West Salton detachment fault, are not obvious.

Cross section B-B’ (fig. 7C) depicts undifferentiated San Timoteo Formation in the subsurface of the Western Heights subarea (unit QTst). If present, these sedimentary deposits could correlate with younger parts of the San Timoteo Formation, possibly with the upper member as recognized by Morton and Miller (2006) in the northwestern part of The Badlands. This hypothetical correlation with the upper member of the formation is speculative, however, and is based on the inference that upper member of the San Timoteo Formation may not have accumulated in the Yucaipa subbasin while sediment in the lower and middle members accumulated in The Badlands.

Sedimentary Deposits of Live Oak Canyon

The sedimentary deposits of Live Oak Canyon comprise a large part of the principal water-bearing hydrogeologic unit in the Yucaipa subbasin (Cromwell and others, 2022a). For this reason, the geologic unit’s nomenclatural usage and physical stratigraphy is described in detail.

Quaternary² Alluvial Deposits

Middle Pleistocene through Holocene sedimentary materials occur throughout the Yucaipa subbasin. Many of these deposits are thin youthful veneers that mantle stream valleys and other lowlands, but some deposits are relatively thick older fills that underlie dissected geomorphic terraces that rise above stream valleys. The latter represent former valley-filling deposits that have been abandoned and incised by younger streamflows of the modern landscape. The Quaternary alluvial deposits are unconsolidated except for thin horizons that have been consolidated due to pedogenesis and epigenetic processes, and in most cases the deposits likely lie above the water table (fig. 7). In this report, two Quaternary alluvial deposit units are classified in the Yucaipa subbasin: (1) middle and upper Pleistocene alluvial deposits that appear as continuous outcrops along the broad floor of the Yucaipa Valley and Beaumont plain, and (2), latest Quaternary alluvial deposits that are generally found along streams and incised into older geologic units.

San Andreas Fault Zone

The San Andreas fault zone forms the northern boundary of the Yucaipa subbasin and is the most widely recognized structural element in the area. Matti and others (1985, 1992a) recognized four strands of the fault zone in the vicinity Yucaipa subbasin: from northeast to southwest, the Mill Creek, Wilson Creek, Mission Creek, and San Bernardino strands (fig. 2). The San Bernardino strand defines the modern trace of the San Andreas fault zone in the Yucaipa subbasin and is the only San Andreas fault trace that is relevant to the subbasin. See Matti and others (1992a) and Matti and Morton (1993) for a more detailed discussion of the San Andreas fault zone and each of the four fault strands.

The San Bernardino strand of the San Andreas fault zone breaks all but the youngest surficial materials. Estimates of slip-rate of the San Bernardino strand in the Yucaipa subbasin by Harden and Matti (1989) are compatible with gradual inception of the strand by reactivation of the abandoned Mission Creek strand (illustrated in fig. 2) starting perhaps around 125,000 years before present (Matti and others, 1985, 1992a; Matti and Morton, 1993). In the vicinity of the Yucaipa subbasin the San Bernardino strand has a complicated left-stepping geometry (Harden and Matti, 1989; Matti and others, 2003a) and is characterized in the northwestern part of the subbasin by individual en echelon fault segments that have a northwesterly trend averaging about 55 degrees west of north (Matti and others, 2003a). In the northeastern part of the subbasin, the San Bernardino strand consists of fewer individual fault strands that traverse the south margin of the San Bernardino Mountains (Matti and others, 1983, 1992a).

Banning Fault

The structural role of the Banning fault with regard to the Yucaipa subbasin has been the subject of considerable and varied interpretations, and questions remain about the fault’s mapped distribution, geologic age (age of fault movement), and its effect on groundwater in and adjacent to the subbasin (see below). Therefore, this section examines in detail what is known about the Banning fault as well as questions about its hydrogeologic role.

Live Oak Canyon Fault Zone

Matti and others (2003b) applied the name “Live Oak Canyon fault zone” to the series of arcuate faults that traverse the south margin of Smiley Heights (fig. 5). Matti and others (2003b) did not discuss their slip sense or structural role, but in this report we interpret these faults as north-dipping reverse and thrust faults whose movement has led to uplift of the Smiley Heights and Redlands Heights highland. The fault system most likely is late Quaternary because it disrupts a landscape surface estimated to be approximately 300–700 ka (surface Q3 of Kendrick and McFadden, 1996; Kendrick and others, 2002, p. 2,784). This surface caps the Sedimentary deposits of Live Oak Canyon and is comparable to paleosols reported by Matti and others (2003a) in the upper part of that unit on the east margin of Live Oak Canyon (discussed above).

Tectonic interaction between contractional structures of the Live Oak Canyon fault zone and extensional structures of the Crafton Hills fault zone (fig. 5) is not obvious. Matti and others (2003a, b) interpreted structures of the two fault zones to overlap spatially: specifically, their mapping depicts one strand of the Crafton Hills fault zone extending a short distance into the Redlands 7.5-minute quadrangle, where strands of the Live Oak Canyon fault zone also are depicted (fig. 5). In addition, along trend in the adjacent Yucaipa 7.5-minute quadrangle, Matti and others (2003a) showed only strands of the Crafton Hills fault zone, implying that the contractional Live Oak Canyon fault zone does not extend far to the northeast. However, in the southwest corner of the Yucaipa 7.5-minute quadrangle, a strand of the Live Oak Canyon fault zone dips about 35 degrees to the northwest (fig. 5; the fault was observed in a geotechnical trench). The fault clearly is a thrust fault that appears to be an extension of the Live Oak Canyon fault zone, and not a strand of the normal-slip Crafton Hills fault zone as mapped previously by Matti and others (2003a). The newly recognized change in structural style probably applies to fault strands directly northeast of the 35-degree dip depicted in figure 5. The structural character of the boundary zone between extensional and contraction fault complexes has not been resolved and is beyond the scope of this report.

Normal-Slip Faults Associated with the Yucaipa Groundwater Subbasin

A series of northeast-trending, normal-slip faults traverses and bounds the Yucaipa subbasin and is responsible for much of its topographic and structural relief (Burnham and Dutcher, 1960, p. 112, fig. 3). The normal-slip structures include (from west to east) the Redlands fault, the Reservoir Canyon fault, the Crafton Hills fault zone, the Yucaipa graben complex (including the previously named Oak Glen fault), the Chicken Hill fault, and probably the Casa Blanca fault (fig. 5). Matti and others (1985, 1992a)57 and Matti and Morton (1993) interpreted the normal faults as a zone of extensional stress created by a late Quaternary right step between the San Andreas and San Jacinto fault zones. Whether the faults were active prior to the late Quaternary is not known.

South Mesa Barrier

The South Mesa barrier is a westerly-northwesterly trending structure that was identified as a barrier to groundwater flow by Moreland (1970; fig. 5), who suggested that the structure was probably associated with the San Andreas fault zone and Banning fault and had down-to-the-south displacement. The mapped distribution of the structure was based on observed disparities in groundwater levels between several wells, geophysical data, and alignment with a mapped fault trace east of the area of investigation (Moreland, 1970). The structure is not included in Quaternary Fault and Fold Database of the United States (U.S. Geological Survey and California Geological Survey, 2016), and surface expression of the structure is not identifiable (Moreland, 1970; Matti and others, 2003a), indicating that movement on the structure had ended prior to the deposition of Quaternary materials.

San Gorgonio Pass Fault Zone and Cherry Valley Thrust Fault

The San Gorgonio Pass fault zone is a series of Quaternary reverse, thrust, and tear faults that extends from the eastern portal of San Gorgonio Pass (east of Cabazon, fig. 2) westward to the town of Calimesa (figs. 2, 5). Regionally, the fault zone has a distinctive zigzag character in which the elongate, northwestward segments appear to be high-angle tear faults having oblique right-lateral displacements and in which the shorter eastward-northeastward segments are thrust and reverse faults (fig. 2). The northwest-oriented segments have approximately the same orientation as right-lateral fault strands of the San Andreas fault zone (Matti and others, 1992a).

In the Yucaipa subbasin, the San Gorgonio Pass fault zone is represented by the Cherry Valley thrust fault of Bloyd (1971), a north-dipping thrust that appears to be the westernmost part of the zone of crustal convergence that is the San Gorgonio Pass fault zone (Matti and others, 2015). The age and duration of faulting of the Cherry Valley thrust fault are not documented, but contraction probably developed in the Pleistocene and continued intermittently for much of Quaternary time (Matti and others, 2015). East of the town of Calimesa, the Cherry Valley thrust fault carries folded strata of the San Timoteo Formation in the hangingwall south over younger sedimentary materials of the footwall. Traced westward across Interstate 10, the fault forms a scarp within the sedimentary deposits of Live Oak Canyon and locally places these rocks over middle Pleistocene alluvial deposits (unit Qof2 of Matti and others, 2015). Trench exposures show that the fault dips north about 15 degrees along the eastern segment (CHJ, Inc, 2004; Matti and others, 2015), and between 11 and 45 degrees along the western segment (Rasmussen and Associates, 1988a, b; Petra Consultants, 2004; Matti and others, 2015). The geometry of the Cherry Valley thrust fault at depth is uncertain; the fault could steepen northward and root into the Banning fault, or the fault could retain its shallow north dip beyond, and truncate, the Banning fault (Matti and others, 2015).

The Cherry Valley fault scarp is directly along trend with caliche zones (calcrete) that Burnham and Dutcher (1960, p. 100) originally associated with the Banning fault. Matti and others (2015) did not recognize the scarp west of where they depict it on their geologic map, but fracturing and pulverization of sedimentary material in the sedimentary deposits of Live Oak Canyon might have continued for an unknown distance beyond the westernmost scarp segment. If so, then displacement on the Cherry Valley thrust fault could be responsible for the carbonate accumulation observed by Burnham and Dutcher (1960) that they attributed to the Banning fault. Moreover, as a youthful fault that breaks not only sedimentary deposits of Live Oak Canyon but also middle and upper Pleistocene alluvial deposits, the Cherry Valley thrust fault south of the town of Calimesa may well account for local differences in groundwater levels that originally led Burnham and Dutcher (1960) and Bloyd (1971) to position the Banning fault and cite that structure as a barrier to groundwater flow.

Crystalline Basement

Crystalline basement forms the bottom boundary of the subbasin in which basin-fill aquifer materials are deposited (table 1; figs. 7, 9). The crystalline basement does not hold large volumes of groundwater; however, small-scale fractures, joints, and faults that are open and permeable provide conduits for transmitting precipitation and runoff into the crystalline subsurface and then laterally into sedimentary materials of the Yucaipa subbasin. Outcrops of crystalline basement in the Crafton Hills, San Bernardino Mountains, and Yucaipa hills (fig. 9) are cohesive blocks in many places but also include substantial regions of cataclastic, sheared, fractured, and broken rock (Matti and others, 2003a, 201561) that have potential to act as conduits for groundwater recharge and pathways for groundwater flow in the subsurface.

A zone of weathered saprolitic basement material is evident above the crystalline basement surface at depth and in outcrop (figs. 7, 8A, 8B; described in the “Interpretation of Crystalline Basement” section below). This 60–100 ft thick weathered saprolite zone is consistent with modern saprolite zones in temperate climates (Sequeira Braga and others, 2002; Dixon and others, 2009) and is observed in select USGS multiple-depth monitoring-well sites in the Yucaipa subbasin; however, the full geographic extent of this zone is unknown. A fairly continuous zone of weathered saprolitic basement material may be present above crystalline basement throughout the study area. In addition, this weathered saprolite zone may act as a conduit to flow or be a local source of groundwater storage. Uncertainty in the location, thickness, and spatial distribution of the weathered saprolite zone led to its exclusion from the HFM.

Consolidated Sedimentary Materials

The consolidated sedimentary materials unit comprises the deep subsurface sedimentary rocks south of the Banning fault and, if present, the Western Heights subarea, and the San Timoteo Formation (figs. 7, 9; table 1). The consolidated sedimentary materials unit crops out in the southwest part of the Yucaipa subbasin, mainly in the The Badlands where the San Timoteo Formation is present at land surface (figs. 5, 9; Morton and Matti, 2001; Matti and others, 2003b; Matti and others, 2015). The only other surface exposure of the consolidated sedimentary materials is on the hangingwall of the Cherry Valley thrust fault (fig. 9).

The parts of the consolidated sedimentary materials that represent the middle member of the San Timoteo Formation are in the southern half of the exposed San Timoteo Formation in the Yucaipa subbasin and generally consists of light-gray, sheetlike layers of pebble-cobble conglomerate, with medium to thick intervals of gray-brown fine- to coarse-grained sandstone and minor amounts of siltstone and mudstone intervals (figs. 5, 9; Frick, 1921; Matti and others, 2015). The parts of the consolidated sedimentary materials that represent the upper member of the San Timoteo Formation are in the northern half of the exposed San Timoteo Formation in the Yucaipa subbasin to the south and west of San Timoteo Canyon and the Live Oak Canyon fault zone and have a relatively consistent lithology of “gray, coarse grained, moderately indurated sandstone and conglomerate” (figs. 5, 9; Frick, 1921; Morton and Matti, 2001).

Consolidated sedimentary materials are more compacted, consolidated, cemented, and have a greater abundance of clay and silt relative to the overlying unconsolidated sediment and surficial materials. The consolidated sedimentary materials are likely the least transmissive basin-fill hydrogeologic unit in the study area; Dutcher and Fenzel (1972) calculated a permeability of about 5 gallons per day per square foot (gpd/ft²) in what they identify as the lower member of the San Timoteo Formation, which corresponds to the consolidated sedimentary materials hydrogeologic unit classified in this study. Similarly, Rewis and others (2006) designated the San Timoteo Formation as impermeable and inactive in their groundwater-flow model of the Beaumont plain area.

Unconsolidated Sediment

The unconsolidated sediment unit comprises the sedimentary deposits of Live Oak Canyon unit and likely parts of the middle and upper Pleistocene alluvial deposits (figs. 7, 8, 9; table 1). The unconsolidated sediment unit crops out north of the The Badlands where the sedimentary deposits of Live Oak Canyon are present at land surface (fig. 9). This unit is exposed along the walls of San Timoteo Canyon. In the subsurface, the unconsolidated sedimentary unit may be distinguished, in part, from the overlying unconsolidated surficial materials by reddish paleosol beds and local unconformities that are are found at the contact between the sedimentary deposits of Live Oak Canyon and the middle Pleistocene alluvial deposits (Matti and others, 2003a, 201561).

The unconsolidated sediment unit is the primary aquifer unit in the Yucaipa subbasin. The unit consists of unconsolidated materials (although the unit includes parts of the sedimentary deposits of Live Oak Canyon that are consolidated) and is dominated by outcrops of sand- and gravel-bearing deposits, with fewer outcrops of mud-bearing deposits. The unit is the most extensive and voluminous sedimentary unit in the subbasin, and the water table sits almost exclusively within it (about 200 to more than –300 ft below land surface; figs. 7B, 7D; Cromwell and others, 2022a). The unconsolidated sediment unit is substantially more coarse grained (sand, gravelly sand, and gravel) than fine grained (silt, mud, and clay), indicating likelihood of high permeability and capacity for groundwater flow and storage. Dutcher and Fenzel (1972) calculated a permeability of about 220 gpd/ft², in what they identified as the upper San Timoteo Formation, now defined as sedimentary deposits of Live Oak Canyon, which are part of the unconsolidated sediment unit.

Surficial Materials

The surficial materials unit consists of latest Quaternary alluvial deposits and likely parts of the middle and upper Pleistocene alluvial deposits (figs. 5, 7, 8 and 9; table 1). The unit appears as continuous outcrops along the broad valley floor and in eroded stream channels incised into the crystalline bedrock and older hydrogeologic units (figs. 5, 9). The surficial materials unit is comprised mostly of alluvial-fan or alluvial axial-valley deposits, with local outcrops of landslide, wash, and colluvial materials. Alluvial-fan sediments are coarser grained, gravel rich, and more poorly sorted than the axial-valley sediments; axial-valley sediments typically have more fine-grained layers of silt and clay interbedded with sand and gravel beds (Matti and others, 2003a). This unit is relatively thin and is about 30–160 ft thick in most of the Yucaipa subbasin.

The surficial materials unit is often entirely above the water table (about 200–300 ft below land surface; figs. 7B, 7D; Cromwell and others, 2022a); therefore, this unit likely does not play a substantial role in groundwater storage, but the coarse-grained materials likely to allow water added to the hydrologic system to promote rapid recharge groundwater in the basin. Dutcher and Fenzel (1972) calculated a permeability of about 40 gpd/ft² in materials that correspond, at least in part, with the surficial materials unit.

Data Sources

Surface and subsurface hydrogeologic data were used to quantify the subsurface extent of each hydrogeologic unit and evaluate the within-unit distribution of different lithology classes. Surface hydrogeology (fig. 9) was compiled from digital versions of geologic maps (Dibblee and Minch, 2003a, 200820; Morton and Miller, 2006; Rewis and others, 2006; fig. 5) and supported by additional geologic mapping investigations (Morton and Matti, 2001; Matti and others, 2003a, b, 20156061). The interpolated surface of the top of crystalline basement is derived from gravity-derived depth-to-basement studies of the Yucaipa subbasin area (Anderson and others 2004; Langenheim and others, 2005; Mendez and others, 2016; fig. 6).

Drillers’ lithology log descriptions and borehole geophysical measurements were the primary data used to evaluate subsurface geology; these data were acquired from a variety of sources, including DWR (https://water.ca.gov/Programs/Groundwater-Management/Wells/Well-Completion-Reports), the USGS National Water Information System (U.S. Geological Survey, 2018), a USGS report (Mendez and others, 2018), and a consultant report (Geoscience Support Services, Inc., 2014a). In total, lithology and geophysical logs from 272 boreholes were compiled within the study area (fig. 9). Four multiple-depth monitoring-well sites were previously drilled and constructed by the USGS (YVWC, YVDA, YV6E, and YVEP; figs. 8, 9; table 2). These well sites provide detailed lithology descriptions, continuous geophysical measurement data, drill cuttings and cores, and groundwater information at discrete depths from nested piezometers, and are some of the deepest boreholes in the study area (Mendez and others, 2018). Drillers’ lithology descriptions for all boreholes used in this study were standardized to a common set of textural classes and are available as part of a USGS ScienceBase data release (Cromwell and others, 2022b).

The HFM incorporates faults that have structural and (or) hydrologic importance (table 3; fig. 9). Spatial data for the model faults were compiled from three sources: (1) the surface (two-dimensional) distribution of faults from the Quaternary Fault and Fold Database of the United States (U.S. Geological Survey and California Geological Survey, 2016); (2) the surface and subsurface (three-dimensional) geometry of faults from the Southern California Earthquake Center’s Community Fault Model version 4.0 (v4.0; Nicholson and others, 2013) and version 5.0 (v5.0; Nicholson and others, 2014; see also https://www.scec.org/research/cfm); and (3) surface (two-dimensional) spatial data from the geologic maps of Matti and others (2003a, b, 20156061). Although the geologic map databases for the Yucaipa, Redlands, and El Casco 7.5-minute quadrangles (Matti and others, 2003a, b, 20156061) contain attributed geospatial data for many faults in the Yucaipa subbasin, these sources do not cover the entire extent of the subbasin nor do they consistently have fault-dip information. For these reasons, fault information for the HFM was compiled primarily from the Quaternary Fault and Fold Database of the United States and the Community Fault Model v4.0 and v5.0. These databases provide seamless geospatial coverage for faults in the Yucaipa subbasin, and the Community Fault Model especially models the subsurface geometry of faults in the subbasin.

In order to simplify structural features for the HFM, model faults were selected and (or) modified from the source datasets to best represent the predominant orientation and character of each mapped fault; fault zones having more than one individual trace (for example, discontinuous en echelen fault strands) are portrayed as continuous line features. The HFM represents some mapped faults and faults zones using a single model fault but others using two or three model faults. Single model faults include the Casa Blanca, Cherry Valley thrust, and Chicken Hill faults, the Crafton Hills and Live Oak Canyon fault zones, and the South Mesa barrier (table 3; figs. 5, 9). These model faults are derived from the digital datasets listed in table 3. Southeast of the Yucaipa subbasin the HFM incorporates another single-model fault, which is labeled as “Wildwood Canyon fault” (after Matti and others, 1992a) in table 3 and figure 9; the Quaternary Fault and Fold Database of the United States (U.S. Geological Survey and California Geological Survey, 2016) identifies this fault as part of the San Gorgonio Pass fault zone.

The Banning fault occurs in the southern and southeastern parts of the Yucaipa subbasin, and in the HFM is represented by “active” and “inactive” segments of a singular model-fault feature (table 3; fig. 9). The “inactive” segment applies to the Miocene Banning fault as mapped by Matti and others (1992a, 2003a, b, 2015)596061. The “active” segment applies to a segment southeast of the town of Calimesa that Matti and others (2015) referred to as the Miocene Banning fault that possibly was reactivated in Quaternary time (discussed above in the “Faults” section). The Quaternary Fault and Fold Database of the United States (U.S. Geological Survey and California Geological Survey, 2016) refers to this fault as the Banning fault while the Community Fault Model v4.0 (Nicholson and others, 2013) does not recognize this structure.

In the HFM the Yucaipa graben complex is narrower in geographic and geologic scope than indicated in the geology section of this report. Geologically, the graben complex is defined on the east by the Chicken Hill fault and by multiple north-trending subparallel fault scarps in the east of the Crafton Hills (Matti and others, 2003a; fig. 5). The HFM emphasizes two of the latter as singular model faults that are labelled as “western” and “eastern” faults of the Yucaipa graben complex (table 3; fig. 9).

In the HFM the San Andreas fault zone is represented by three model faults, labeled as “northern”, “central”, and “southern” San Andreas model faults (table 3; fig. 9). Fault usage between available sources is variable: (1) Matti and others (1992a, 2003a)59 applied individual fault-strand names to various San Andreas traces, including the Mission Creek and Mill Creek strands and the San Bernardino strand for the modern neotectonic trace (fig. 2); (2) the Quaternary Fault and Fold Database of the United States assigns these strands to their “San Andreas fault zone, San Bernardino Mountains section”; and (3) the Community Fault Model v4.0 applies the names “San Andreas fault”, “Mission Creek fault”, and “Mill Creek fault” to faults in this area, but somewhat differently than Matti and others (1992a, 2003a)59.

In the HFM, the northern model fault of the San Andreas zone coincides with the San Bernardino strand as mapped by Matti and others (1992a, 2003a59; figs. 2, 9), although its surface distribution and subsurface geometry are from the Community Fault Model v4.0 (Nicholson and others, 2013). The Community Fault Model v4.0 also is the source for the surface and subsurface extent of the central San Andreas model fault (fig. 9). The southern model fault is derived from geospatial information from the Quaternary Fault and Fold database of the United States (U.S. Geological Survey and California Geological Survey, 2016). The central model fault cuts southeast across the Yucaipa subbasin in the Triple Falls Creek subarea (fig. 4C) where it then trends southeast along the trace of the southern San Andreas model fault (fig. 9).

The central and southern San Andreas model faults differ from how Matti and others (1992a, 2003b)60 interpret structural relations in this part of the Yucaipa subbasin. This difference centers on the Oak Glen fault as originally named and mapped by Burnham and Dutcher (1960; also see Dutcher and Burnham, 1959). These studies recognized an east-trending fault scarp at the north end of Yucaipa Valley that they interpreted as a western segment of their Oak Glen fault (see “Faults” section), although they did not interpret the subsurface geometry or slip style for the Oak Glen fault. Matti and others (2003a) mapped curving en echelen faults at the north end of Yucaipa Valley, and interpreted these structures as bounding the north margin of their extensional Yucaipa graben complex (see above “Faults” section; note, Matti and others [1992a] originally included these faults within their “Crafton Hills horst-and-graben complex”). Subsequently, the Quaternary Fault and Fold Database of the United States (U.S. Geological Survey and California Geological Survey, 2016) and the Community Fault Model v4.0 (Nicholson and others, 2013) reinterpreted fault-distribution patterns and fault nomenclature at the northern head of Yucaipa Valley: the Quaternary Fault and Fold Database of the United States assigned the name “San Andreas fault zone, San Bernardino Mountains section” to the scarps mapped by Matti and others (1992a, 2003a59; see Harden and Matti, 1989, fig. 2); the Community Fault Model v4.0 does not incorporate faults mapped by Matti and others (2003a; Harden and Matti, 1989, fig. 2), but assigns the name “San Andreas fault” to a fault in this area having a broadly similar orientation and distribution. Owing to these interpretive differences—and especially owing to the need for a seamless tectonic framework—the central and southern San Andreas model fault in the HFM uses the geospatial distribution of the “San Andreas fault” reported by the Community Fault Model v4.0, and a trace of the “San Andreas fault zone, San Bernardino Mountains section” as reported by the Quaternary Fault and Fold Database of the United States.

Interpretation of Crystalline Basement

Crystalline basement was interpolated from gravity-derived depth-to-basement studies of the region surrounding the Yucaipa subbasin (Anderson and others 2004; Rewis and others, 2006; Mendez and others, 2016). Figure 6 shows the depth to the crystalline basement unit from the HFM and the extent of the input data incorporated from each of the three gravity-derived depth-to-basement studies. Figure 7 shows two sections, A-A’ and B-B’, through the Yucaipa subbasin, with figures 7A and 7C showing the interpolated crystalline basement surface underlying a conceptualization of the geology, and figure 7B and 7D showing the interpolated crystalline basement surface underlying the interpolated hydrogeologic units from the HFM. The previously published depth-to-basement studies estimated depth-to-basement over different, but overlapping, geographic extents using gravity, aeromagnetic, and seismic geophysical methods. The study area of Mendez and others (2016) covers most of the Yucaipa subbasin (fig. 6). The study areas of Anderson and others (2004) and Langenheim and others (2005) focused on the San Bernardino valley to the west, and the San Gorgonio Pass area to the east of the subbasin, respectively, and each study included parts of the Yucaipa subbasin and the surrounding YVW. Digital depth-to-basement point data were compiled from Mendez and others (2016) and Langenheim and others (2005) at resolutions of about 164 and 1,640 ft, respectively. Depth-to-basement contours were digitized from Anderson and others (2004) and converted to point data along the resulting contours. The combined depth-to-basement point data and surface data from geologic maps were interpolated using EarthVision, a 3D geologic-modeling software package (Dynamic Graphics, Inc., 2015), that uses a biharmonic cubic-spline algorithm that utilizes a minimum-tension (minimum curvature) gridding technique designed to generate horizon grids from 3D point data (Dynamic Graphics, Inc., 2015). The result of this interpolation is a continuous estimate of the top of crystalline basement at a resolution of about 492 ft (fig. 6).

Crystalline basement is exposed at land surface in the Crafton Hills, the Yucaipa hills, and the San Bernardino Mountains and varies in elevation from more than 3,000 ft above NAVD 88 to more than 8,700 ft above NAVD 88. Within the Yucaipa subbasin, crystalline basement forms a shallow depression, the orientation of which generally parallels that of the northeast-trending faults that transect the subbasin. In the Yucaipa subbasin, the crystalline basement unit is deepest in the Western Heights subarea between the Chicken Hill fault, the Live Oak Canyon fault zone, and Crafton Hills fault zone at about 4,000 ft below land surface (figs. 6, 7). At the southern margin of the Yucaipa subbasin and south of the Banning fault, crystalline basement forms a deep trough greater than 8,000 ft below land surface (figs. 6, 7). West of the Yucaipa subbasin and at the western end of the YVW, crystalline basement reaches depths between about 3,000 and 6,000 ft below land surface and underlies the adjoining San Bernardino groundwater subbasin.

Gravity-derived estimates of the depth-to-crystalline basement are complex and demonstrate the density contrast between the dense basement rocks and the less dense basin-fill materials. In the Yucaipa subbasin, the gravity-derived depth-to-crystalline basement included constraints at points where the depth-to-crystalline basement was estimated from direct observations in 15 boreholes, including USGS multiple-depth monitoring-well sites YV6E and YVEP (Mendez and others, 2016). The estimated depth-to-basement in these two well sites (900 and 800 ft below land surface, respectively; figs. 8A, 8B, “Geologic unit” columns) were derived from drillers’ lithology descriptions and geophysical logs and verified by cores from the bottom of each site (Mendez and others, 2016).

The depth of the gravity-derived crystalline basement surface at well sites YV6E and YVEP was 13 and 68 ft lower, respectively, than what was estimated from borehole data; in fact, the gravity-derived surface at YVEP was modeled to be 15 ft below the bottom of the borehole. The difference between the modeled surface and estimates from the borehole data could be due to the presence of weathered basement material that can be difficult to distinguish from overlying semi-consolidated alluvium (Mendez and others, 2016). The interpolated elevation of the crystalline basement surface in the HFM is lower than the gravity-derived surface of Mendez and others (2016) at well sites YV6E and YVEP (fig. 7) despite using the gravity-derived surface points from that study as an input dataset. The reason for this discrepancy is that the HFM surface was interpolated over a grid spacing that was three times coarser than the gravity-derived surface of Mendez and others (2016); therefore, the crystalline basement surface in the HFM is an approximation of the gravity-derived surface of Mendez and others (2016).

The borehole data for USGS multiple-depth monitoring-well sites YV6E and YVEP were analyzed for the presence of weathered basement material, or saprolite, as part of this study. The presence of weathered basement material in YV6E and YVEP can be inferred from the offset of the gravity-derived crystalline basement surface and the estimated depth-to-basement from borehole data by Mendez and others (2016) and from the lithologic descriptions and geophysical log data (figs. 8A, 8B). For example, Mendez and others (2018) noted the presence of mafic-rich material near the bottom of YVEP (840–850 ft below land surface), that was thought to be weathered mafic bedrock. The interpreted top of weathered basement material is depicted in the “Geologic unit” column for well sites YV6E and YVEP in figures 8A and 8B, and as dashed gray lines in figure 7, as is borehole-estimated top of crystalline basement from Mendez and others (2016). The gravity-derived crystalline basement surface depths at YV6E and YVEP (913 and 868 ft below land surface, respectively; Mendez and others, 2016) are not shown in figure 8.

Lithologic descriptions, geophysical logs, and drill cores (figs. 8A, 8B; Mendez and others, 2016, 2018) for YV6E and YVEP showed regular, alternating coarse gravel (including large cobbles and boulders) with clay- and sand-rich intervals that likely indicate cohesive boulders of crystalline rock surrounded by weathered, saprolitic basement material (similar to observed outcrops in the Yucaipa hills). The zones of weathered basement for YV6E and YVEP are estimated to be about 140 and 130 ft thick, respectively, above the gravity-derived, crystalline basement of Mendez and others (2016). This thickness is consistent with other weathered saprolite zones in temperate climates (Sequeira Braga and others, 2002; Dixon and others, 2009). Water-chemistry data from USGS multiple-depth monitoring-well site piezometers installed in and near this zone are consistent with groundwater stored in and recharged through regional basement material (Rewis and others, 2006; Cromwell and others, 2022a). This zone of weathered saprolitic basement material may be present above crystalline basement throughout the study area. Subsurface observations of this zone were limited to USGS multiple-depth monitoring-well sites YV6E and YVEP, prohibiting confident extrapolation across the study area; therefore, this zone is not included as a separate hydrogeologic unit in the HFM.

Interpretation of Basin-Fill Hydrogeologic Units

Geologic maps, drillers’ lithology log descriptions, and geophysical logs were the primary data sources used to interpret the subsurface basin-fill hydrogeologic units. Data were interpreted in two steps. First, basin-fill hydrogeologic units were identified and characterized in the four USGS multiple-depth monitoring-well sites (fig. 8), where the availability of high-quality lithology descriptions, geophysical logs, and physical sample material increased confidence in interpretations and descriptions of each unit. Next, hydrogeologic units identified in the multiple-depth monitoring-well sites were correlated to other boreholes by iteratively analyzing lithologic and geophysical logs relative to the multiple-depth monitoring-well sites and geologic maps. Figure 10 shows the subsurface input point locations for each basin-fill hydrogeologic unit.

The consolidated sedimentary materials unit was not identified in any USGS multiple-depth monitoring-well sites (figs. 8, 10A); however, the unit was identified in boreholes located in The Badlands and in wells in the hangingwall of the Cherry Valley thrust fault (fig. 10A). This unit was identified in drillers’ lithology logs based on the presence of consistent fine-grained materials, often in contrast to the substantially coarser-grained overlying material. The exact subsurface extent of this unit is unknown, but geologic information about the San Timoteo Formation, on which the consolidated sedimentary materials unit is primarily based, was used to inform the subsurface modeling of the unit. Geologic maps (Morton and Matti, 2001; Matti and others, 2003b, 201561; Morton and Miller, 2006) show that the San Timoteo Formation generally dips about 10–35 degrees to the northeast in the study area due to the regional influence of the San Timoteo anticline. The geologic cross section in figure 7A shows the formation dipping to the northeast and onlapping onto crystalline basement east of the Banning fault. Furthermore, the San Timoteo Formation was not identified in deep boreholes north and east of San Timoteo Canyon. Control points based on mapped dip angles were projected into the subsurface to provide structural controls of the consolidated sedimentary materials unit where it plunges beneath land surface (fig. 10A). The resulting modeled grid horizon of the unit was later manually adjusted to accommodate the inferred onlap of the unit across the Banning fault (fig. 7B). In the southeastern part of the study area, the modeled top of the San Timoteo Formation of Rewis and others (2006) was included in the HFM as the top of the consolidated sedimentary materials unit (fig. 10A).

The unconsolidated sediment unit was assumed to fill much of the basin to within several tens of feet of land surface, overlying the crystalline basement and consolidated sedimentary materials units (figs. 7B, 7D). Unconsolidated sediment was identified in all four USGS multiple-depth monitoring-well sites and in boreholes throughout the study area (figs. 8, 10B). The sedimentary deposits of Live Oak Canyon, middle and upper Pleistocene, and latest Quaternary alluvial deposits are all comprised of unconsolidated materials, but differentiating between the geologic units from lithologic and geophysical records in subsurface boreholes is difficult (see “Geologic Units” section). The top of the unconsolidated sediment unit was therefore interpreted to be the basal contact of the first (most shallow) prominent coarse-grained interval recorded in drillers’ lithologic logs or geophysical logs (fig. 8). Using this interpretation, the unconsolidated sediments hydrogeologic unit “crosses” stratigraphic boundaries. In drillers’ logs, this contact was generally interpreted to be where lithologic descriptions changed from primarily gravel to a mix of sand and gravel. In geophysical logs, the contact was generally interpreted to be at a decrease in the amplitude and magnitude of geophysical resistivity curves (fig. 8; Mendez and others, 2018). In several boreholes, red, indurated lithology intervals were identified and classified as the top of the unconsolidated sediment unit; these intervals were interpreted to represent pedogenic-soil horizons at the contact between the sedimentary deposits of Live Oak Canyon and the overlying middle and upper Pleistocene alluvial deposits (Matti and others (2003a, 2015)61. Control points based on mapped dip angles in the western part of the study area were projected into the subsurface to provide subsurface structural controls of the unit (fig. 10B).

The surficial materials unit was identified in the upper 50–150 ft of each USGS multiple-depth monitoring-well site (fig. 8) and in boreholes throughout the study area (primarily in the Yucaipa Valley area of the subbasin and along San Timoteo Canyon and Live Oak Canyon; fig 10C). The unit was interpreted to represent the first (most shallow) prominent coarse-grained interval recorded in drillers’ lithologic logs or geophysical logs (fig. 8). In drillers’ logs, the unit was generally classified as predominantly gravel rich with sands and minor clays, and was distinguished in geophysical logs from the underlying hydrogeologic units by the presence of high-amplitude and -magnitude resistivity curves (fig. 8; Mendez and others, 2018). The interpreted thicknesses of these deposits in USGS multiple-depth monitoring-well sites were consistent with field observations of Matti and others (2003a). Similar thicknesses of the unit were identified in the remainder of the borehole dataset (fig. 10C); therefore, an equivalent thickness is expected throughout the study area wherever the unit is present.

Model Construction

Digital datasets representing the top of each hydrogeologic unit were compiled as input data for the HFM. Each hydrogeologic unit dataset contains outcrop information from geologic maps. With the exception of the crystalline basement unit, each hydrogeologic dataset also contains subsurface picks from drillers’ lithology logs and geophysical logs.

Digital fault data for the HFM were created by generating a set of horizontal (x, y) and vertical (z) data points along the surface-model fault trace (table 3; fig. 9) and determining x, y, and z at depth from the fault-dip angle and -dip azimuth. The geologic modeling software utilizes fault-dip azimuth and fault-dip angle values to generate digital 3D fault planes for use in the model interpolation. Fault-dip azimuth is the geographic orientation at which the fault plane is projected into the subsurface, and fault-dip angle is the angle below the horizontal at which the fault plane descends into the subsurface. Fault-dip azimuth and fault dip angle values estimated for this report are listed in table 3. Fault dip-azimuths were estimated from the digital data sources (table 3). Fault-dip angles were assumed to be vertical for all faults, except for the Cherry Valley thrust fault and the northern strand of the San Andreas fault zone (table 3; fig. 9). Because of limitations of the geologic modeling software, most “vertical faults” were incorporated into the HFM with 85-degree-dip angles. The inactive strand of the Banning fault and the South Mesa barrier were assumed to only offset crystalline basement and were therefore incor­porated in the HFM as proper vertical faults (90-degree dip angle) within that unit (table 3; fig. 7B; the geologic modeling software permits faults within a unit to be vertical). The Cherry Valley thrust fault is a low-angle structure, and the assigned dip angle of 20 degrees (table 3) is consistent with the range of measured dips from the geologic map of the El Casco 7.5-minute quadrangle (Matti and others, 2015); subsurface. Subsurface points along the northern San Andreas fault were taken directly from the Community Fault Model v4.0 (the Mission Creek strand of the San Andreas fault zone of Nicholson and others, 2013) and are therefore variable.

The geologic modeling software used to build the HFM, EarthVision, generates horizon grids from 3D point data (Dynamic Graphics, Inc., 2015)—in this case, the hydrogeologic unit datasets described above. Grid spacing of all hydrogeologic unit horizons was 492.1 ft in the x and y horizontal directions. Control points were added where subsurface borehole data were limited to enforce conceptual geologic structures and improve constraint of horizon interpolations (fig. 10). Each hydrogeologic unit horizon grid was constructed in stratigraphic order, oldest (deepest) to youngest (most shallow), with subsequently younger horizons built on top of previously constructed horizon grids. The total thickness of each hydrogeologic unit was defined as the difference between a given unit horizon and the upper surface of the unit horizon below. Finally, all horizon grids were clipped at land surface using a discretized digital elevation model (DEM) grid based on the National Elevation Dataset 10-meter DEM. The HFM is formed from a set of independent blocks (called fault blocks) that represent hydrogeologic materials within a particular volume defined by faults. Horizon tops are interpolated independently for each fault block, and offsets across each fault are maintained during the gridding process by utilizing data point elevations on either side of the fault surface for each hydrogeologic unit. In this manner, each of the fault blocks are modeled independently and then combined to form a single model. The resulting EarthVision horizon grids were exported to Esri ArcGIS version 10.7.1 GIS software and were manually adjusted to ensure that the grids were consistent with geologic principles, the geologic understanding of the Yucaipa subbasin area, and to enforce a minimum thickness of 9.8 ft for each hydrogeologic unit.

Modeled Hydrogeologic Units

The interpolated crystalline basement unit is exposed at land surface in the Crafton Hills, the Yucaipa hills, and the San Bernardino Mountains and underlies the three basin-fill hydrogeologic units (figs. 7, 11). The high elevations where crystalline basement is at land surface contrast with the top of crystalline basement in the relatively narrow, deep structural valley south of the inactive Banning fault (figs. 6, 7, 11). The unit varies at depth from tens of feet to more than 8,000 ft below land surface. In the HFM, all faults offset the crystalline basement unit, and areas of greatest structural relief are proximal to faults, especially the Live Oak Canyon fault zone, the Chicken Hill fault, and the inactive strand of the Banning fault (figs. 6, 7A, 7C, 10).

The consolidated sedimentary materials unit is limited in extent to the southern part of the HFM. The unit outcrops at land surface to the south and west along the The Badlands, and dips to the northeast with an average dip of about 20 degrees below the horizontal (figs. 7, 10, 11). The subsurface projection of the consolidated sedimentary materials unit carries it down to the northeast into the structural low south of the inactive Banning fault, but then it levels out and drapes onto the crystalline basement, pinching out northeast of the inactive Banning fault. (figs. 7B, 10, 11). In the HFM, all faults, with the exception of the inactive Banning fault and the South Mesa barrier, are permitted to displace the consolidated sedimentary materials unit; however, displacement is limited to those faults that intersect that unit, which are the Live Oak Canyon fault zone and the Cherry Valley thrust fault.

The consolidated sedimentary materials depicted in the Western Heights subarea structural low in cross section B-B’ (fig. 7D) were not modeled as part of the HFM (fig. 10A). The presence of these materials were not verified in available borehole data and were assumed to be located beneath the depth of hydrogeologic investigation (Cromwell and others, 2022a); therefore, for numerical simplicity, the unit was not included in this area in the HFM. Instead, the unconsolidated sediment unit was modeled to fill the entirety of the Western Heights structural low (fig. 10B; see below). Below depths of about 1,500 ft below land surface, this unit subarea should be considered to have the hydrogeologic properties of the consolidated sedimentary materials.

The unconsolidated sediment unit overlies the consolidated sedimentary materials unit in the southern part of the HFM and sits unconformably on top of crystalline basement in the northern half of the HFM where the consolidated sedimentary materials unit is absent (figs. 7B, 7D, 10, 11). The unconsolidated sediment unit fills in areas of structural relief above the consolidated sedimentary materials unit and crystalline basement, resulting in varying thickness (fig. 10B). The unconsolidated sediment unit is thinnest at the northern and southern margins of the HFM and thickest in the vicinity of the inactive Banning fault (fig. 10B). All faults in the HFM except the inactive strand of the Banning fault and the South Mesa barrier are active through this unit.

As indicated above, the unconsolidated sediment unit was modeled to fill the entirety of the Western Heights structural low (fig. 10B), even though consolidated sedimentary materials are likely present at depth (fig. 7D). Although the unconsolidated sediment unit comprises most of the basin-fill of the HFM in the Western Heights subarea, the hydrogeologic properties of the modeled unit are expected to vary with depth. At shallow depths where the unit is identified in available boreholes (from land surface down to about 1,500 ft below land surface; fig. 7D), the unconsolidated sediment is expected to have hydrogeologic properties consistent with the borehole observations. At depths below the available borehole data (deeper than about 1,500 ft below land surface; fig. 7D), at which the consolidated sedimentary materials are likely present, the unconsolidated sediment is expected to have hydrogeologic properties consistent with the consolidated sedimentary unit. See the “Textural Analysis of Basin-Fill Hydrogeologic Units” section below for more information.

The surficial materials unit is laterally extensive across the study area and overlies each of the older units at various locations (figs. 7B, 7D, 10C, 11). The surficial materials unit is continuous in the broad plains of the Yucaipa Valley and the Beaumont plain; elsewhere, the unit is present where stream channels incised the underlying hydrogeologic units (figs. 10C, 11). The surficial materials unit is generally less than about 100 ft thick, although the unit may be as thick as about 350 ft in parts of the Yucaipa Valley and Beaumont plain (fig. 10C; note that the total depth of some wells is less than the modeled thickness of the unit). All model faults except the inactive Banning fault and the South Mesa barrier are active through this unit.

Model Uncertainty

The HFM was constructed at a resolution appropriate for basin-scale hydrogeologic studies and for inclusion in an integrated hydrologic model. Emphasis was placed on quantifying the hydrogeology within the Yucaipa subbasin, which is the area of focus for this study. Areas outside the Yucaipa subbasin, and areas within the subbasin with little subsurface data (fig. 10), could be targets for future hydrogeologic modeling efforts.

The modeled hydrogeologic unit surfaces are interpolations of available input data; therefore, the modeled surfaces contain uncertainties related to the distribution and quality of the input data and uncertainties inherent to the interpolation algorithm used by the geologic modeling software. Input points derived from previously published geologic maps, gravity-derived depth-to-basement data, and fault catalogs are distributed across the study area and provide consistent and reliable interpretations of the Yucaipa subbasin geology and hydrogeology, within the documented uncertainties of those original publications. Input points derived from borehole data (or from structural control points) are discontinuous across the study area, and interpretations from these data are based on variable descriptions of downhole lithology and variable geophysical methods and measurements. These input points are consistent in their representation of the subsurface hydrogeology, but modeled surfaces in areas with more and higher quality borehole data are likely to be more representative of the subsurface hydrogeology than areas with fewer, or lower quality data. The EarthVision geologic modeling software uses a biharmonic cubic-spline algorithm that utilizes a minimum-tension (minimum curvature) gridding technique, meaning that the software does not “connect the dots” between input points but instead finds the easiest path through the available data; because of this algorithm, the resulting hydrogeologic unit surfaces are not expected to precisely match the input data.

Figure 10 shows calculated differences between the subsurface elevations or thicknesses of the basin-fill hydrogeologic units from input data (borehole data, structural control points, and other subsurface data) and from the interpolated hydrogeologic units from the HFM. In general, the differences between hydrogeologic unit elevations and thicknesses from borehole data and those of the interpolated model units were less than about 50 ft, and the differences between structural control points and the interpreted model units were about 100–150 ft (fig. 10). For the consolidated sedimentary materials, input points derived from Rewis and others (2006) were within about 100 ft of the interpreted model elevation. The largest differences between input point elevations and interpolated model elevations were at depths of several hundred to thousands of feet below land surface, such as for the structural control points and some compiled from Rewis and others (2006). Larger differences between input data and interpolated model elevations occurred in areas where manual adjustments were made to the model unit elevations to conform with geologic concepts. Manual adjustments were made, for example, where the consolidated sedimentary materials are believed to drape across the inactive Banning fault, and near hydrogeologic contacts where minimum thicknesses were enforced to ensure proper pinching-out thinning units.

Textural Analysis of Basin-Fill Hydrogeologic Units

Variability of lithology and grain size was analyzed for the consolidated sedimentary materials, unconsolidated sediment, and surficial materials unit. The primary variable used was sediment grain size, tabulated as a percentage of coarse-, medium-, and fine-grained textures. For each well, each downhole lithologic interval was generalized as either “coarse-grained,” “medium-grained,” or “fine-grained” based on the main texture from the drillers’ log descriptions. In this study, coarse-grained intervals were those with main textures classified as boulders, cobbles, conglomerate, gravel, pebbles, and rock. Medium-grained intervals were those classified as sand, sandstone, or hardpan, and fine-grained intervals were those classified as clay, shale, loam, or topsoil. Intervals with main textures classified as silt were generalized as fine-grained if the primary texture modifier was fine-grained (such as “clayey”) or defined as medium-grained if the primary texture modifier was medium- or coarse-grained (such as “sandy” or “gravelly”).

Textural data were generated by digitally intersecting the downhole lithologic interval data in each borehole with the elevation of the hydrogeologic unit tops from the HFM, thereby assigning specific lithologic intervals to a hydrogeologic unit at each borehole. The percentage of coarse-, medium-, and fine-grained material for a borehole within a given hydrogeologic unit was calculated as the aggregate thickness of coarse-, medium-, or fine-grained intervals penetrated by the borehole divided by the total thickness of the unit in that borehole. The percentage of the generalized grain-size groupings at each borehole are shown in figure 12, and the total percentages of each generalized grain-size grouping for each unit are shown in figure 13.

Textural variability in the basin-fill sediment ultimately depends on the sedimentary facies, environment of deposition, and depositional history of the area. Boreholes in the consolidated sedimentary materials unit are predominantly fine grained across the modeled extent of the unit, which is consistent with previous geologic observations, and the total percentage of fine-grained materials in the unit is nearly 50 percent (figs. 12A and 13). Boreholes in the unconsolidated sediment have varying percentages of each of the three grain-size groupings across the modeled extent of that unit, although medium-grained materials comprise more than 40 percent of the unit (figs. 12B, 13). In the Yucaipa Valley, the unit is predominantly composed of coarse- and medium-grained materials, while elsewhere, the unit is predominantly composed of medium- and fine-grained materials. The predominance of fine-grained materials in the Western Heights subarea may be related to springs and peaty land that were historically present in this area (Moreland, 1970).

Boreholes in the surficial materials unit penetrated predominantly medium-grained material with nearly equal percentages of coarse- and fine-grained materials; the total percentage of medium-grained materials in the unit is about 40 percent, and the total percentage of both coarse- and fine-grained materials is about 30 percent (figs. 12C, 13). The spatial distribution of grain size in the modeled surficial materials unit is similar to the unconsolidated sediment unit with more coarse- and medium-grained materials in Yucaipa Valley and generally more medium- and fine-grained materials across the remainder of the modeled extent. An exception to this comparison is in the northern part of the study area between the San Bernardino Mountains and the Yucaipa hills where boreholes penetrating the surficial materials unit encountered mostly coarse-grained materials. In this area, fractured and weathered crystalline basement rocks are likely present in what is modeled as surficial materials; drillers’ descriptions often classify the main texture as “rock” which is generalized as coarse grained in this analysis.

Potential Barriers to Groundwater Flow

The geometry of the Yucaipa subbasin and the thick sedimentary sequences that overlie crystalline bedrock demonstrate potential for groundwater storage and extraction. Structural features in the HFM, especially the northeast dipping consolidated sedimentary materials unit and northeast-trending normal faults, control local stratigraphy and may affect groundwater flow. The northeasterly dip of the consolidated sedimentary materials in the HFM produces a non-horizontal, low-permeability barrier to groundwater flow (Rewis and others, 2006) at the contact with the unconsolidated sediment (figs. 7B, 11). The subsurface quantification of this contact and the extent and structure of the crystalline basement are important to understanding the total storage volume of the Yucaipa subbasin.

The HFM provides a quantitative framework with which to evaluate the potential effect of faults on groundwater flow in the Yucaipa subbasin. Faults can inhibit groundwater flow in permeable unconsolidated materials (such as the unconsolidated sediment and surficial materials units) through the presence of fine-grained gouge material, chemical cementation of proximal sediments, and the juxtaposition of layers across faults caused by sharp folds or vertical or horizontal displacement of beds. Faults in the HFM within the Yucaipa subbasin that were previously interpreted to inhibit groundwater flow include strands of San Andreas fault zone, strands of the Yucaipa graben complex, the Chicken Hill fault, the Casa Blanca fault, the South Mesa barrier, the Cherry Valley thrust fault, and to a lesser extent, the inactive Banning fault (Burnham and Dutcher, 1960; Bloyd, 1971; Dutcher and Fenzel, 1972; Moreland, 1970). Further evaluation of the effects of these faults and the Live Oak Canyon and Crafton Hills fault zones on groundwater flow can be found in Cromwell and others (2022a) and Alzraiee and others (2022).

The inactive Banning fault and the South Mesa barrier are modeled in the HFM as only present in crystalline basement and are not interpreted to directly offset or juxtapose layers within the basin-fill hydrogeologic units. These structures were previously identified as partial barriers to groundwater flow, but because the structures are not interpreted in the HFM to directly offset basin-fill materials, a different mechanism must be inferred. The inactive Banning fault and the South Mesa barrier may indirectly cause thinning or pinching out of hydrostratigraphic layers, or both, within the basin-fill hydrogeologic units that “drape” across structural crests in crystalline basement (fig. 7); the thinning of hydrostratigraphic layers may restrict the movement of groundwater and form a partial barrier to groundwater flow, as observed in groundwater basins near Los Angeles (Reichard and others, 2003).