Idavada Volcanics, Faulting, and the Formation of Lake Idaho

The oldest volcanic rocks present in the study area (ca. 12–10 million years ago [Ma] (Wood and Clemens, 2002)) are the rhyolites and basalts of the Idavada Volcanics (Malde and Powers, 1962) that outcrop in the eastern Danskin Mountains and the Mount Bennett Hills and are associated with the southwest migration of the continental plate over the Yellowstone hot spot (Pierce and Morgan, 1992). The western Snake River Plain formed as an intercontinental rift basin with normal faulting in the Mount Bennett Hills representing 2.8 kilometers (km) of offset and occurring from 11 to 9.5 Ma (Wood and Clemens, 2002). Geophysical and deep borehole data suggest that mafic intrusive rocks, consistent with rifting, underlie the central axis of the plain near Mountain Home (Mabey, 1982; Wood and Clemens, 2002) although basaltic tuff, interpreted as Idavada Volcanics, was intercepted below 7,150 ft at the far eastern end of the plain (Arney and others, 1984; Clemens and Wood, 1993; Shervais and others, 2002). Rhyolites altered by interaction with water (Jenks and Bonnichsen, 1989) suggest that the formation of paleo Lake Idaho, a large, deep lake in the western Snake River Plain, preceded the rhyolitic volcanism and faulting (Wood and Clemens, 2002), potentially forming as early as 12 Ma (Bonnichsen and others, 2016). Lake Idaho is indicated by the occurrence of thick fluvial and lacustrine sediments, interbedded with spatially discrete basalt units in varying proportions, that make up the conventional Idaho Group and extend over 100 miles along the Snake River (Malde and Powers, 1962). The Chalk Hills Formation of the Idaho Group represents the first phase of deep Lake Idaho, dated between approximately 10–6 Ma (Wood and Clemens, 2002); maximum lake level is thought to be from 1,100 to 1,160 m (3,610 to 3,800 ft) based on preserved shoreline features (Jenks and Bonnichsen, 1989; Bonnichsen and others, 2016).

Miocene Basalts and Lake Idaho

Basalts interbedded with sediments represent a period of volcanism occurring from 9 to 7 Ma (White and others, 2002); this volcanism occurred under varied subaerial and sublacustrine conditions, with basalt flow pathways likely influenced by Lake Idaho and in turn influencing deposition of sediment associated with Lake Idaho (Jenks and Bonnichsen, 1989). In the Mountain Home area, these basalts alternately have been described as pre-Lake Idaho (Shervais and others, 2002) or as subaqueously deposited water-affected basalts (Godchaux and Bonnichsen, 2002) and have been intercepted below sediments in multiple deep boreholes (Shervais and others, 2002). Conventionally, this unit was defined as the Banbury Basalt of Idaho Group (Malde and Powers, 1962) but multiple later researchers have ceased to use this term because its widespread assignment and correlation outside the type section rendered it unuseful (Bonnichsen and Jenks, 1990; Wood and Clemens, 2002); researchers now favor naming only contiguous basalt units (Jenks and others, 1998; Shervais and others, 2002; Wood and Clemens, 2002). During a break in volcanic activity from approximately 7 to 2.2 Ma (White and others, 2002), the level of Lake Idaho likely fluctuated, including a substantial drawdown around 6 Ma (Wood and Clemens, 2002). The second phase of deep Lake Idaho is associated with the Glenns Ferry Formation, dated from approximately 4.3 to 2.5 Ma (Wood and Clemens, 2002). The Glenns Ferry Formation includes deep lacustrine facies as well as mixed fluvial and floodplain facies at the margin of the lake (Malde, 1972) and prograding delta deposits representing lake drawdown (Wood and Clemens, 2002). The Glenns Ferry Formation crops out at the surface at the eastern end of the study area and has been interpreted as occurring at depth in deep boreholes in the central part of the study area (Arney and others, 1984). The repeated fill and drawdown of Lake Idaho may be related to the reorganization of the Snake River drainage, which was driven by eastward migration of the continental divide due to thermal doming associated with continental passage over the Yellowstone hot spot (Pierce and Morgan, 1992; Wood and Clemens, 2002; Staisch and others, 2021).

Pleistocene Basalts and Surficial Sediments

Basaltic volcanism reinitiated at about 2 Ma (White and others, 2002), about the same time as Lake Idaho began to drawdown for the final time (Wood and Clemens, 2002). These eruptions formed relatively thin, laterally extensive, coalescing flows on top of Lake Idaho sediments or older basalts and created the topographic plateau of the Mountain Home area, flowing into Lake Idaho in some places (Shervais and others, 2002; Wood and Clemens, 2002). These plateau-forming basalts are analogous to the Bruneau Formation (Malde and Powers, 1962). Many styles of basaltic volcanism occurred during the Pleistocene, including maar and cinder cone styles that indicate subaerial eruption in swampy, recently drained, or groundwater-intercepting conditions (Godchaux and Bonnichsen, 2002). Depending on their age, these basalts may have erupted along the shores of Lake Idaho or through streams or groundwater after the final draining of Lake Idaho (Godchaux and Bonnichsen, 2002). Finally, unconsolidated surficial sediments, including alluvial fans, melon gravels associated with the Bonneville flood, landslide deposits, and alluvium were deposited on the surface of the study area (Lewis and others, 2012).

Granite

Cretaceous granitic rocks, primarily granodiorite and two-mica granite (unit K g) of Lewis and others (2012), crop out at the surface in the western Danskin Mountains (fig. 2). Also grouped in this hydrogeologic unit (fig. 3) are very minor amounts of Cambrian to Neoproterozoic Windermere Supergroup metamorphic rocks (unit _ Late_< s of Lewis and others, 2012) and Neoproterozoic granitic gneiss (unit Late_< i of Lewis and others, 2012), both exposed in the House Mountain metamorphic complex at the far northern study boundary. Similarly, a minor amount of silicic Challis intrusive rocks (unit T e i of Lewis and others [2012]) are exposed at the far northeastern study boundary. Functionally, all of these crystalline rocks have low storativity and hydraulic conductivity, with secondary permeability associated with jointing and fracturing (tables 2, 3) and are thus reasonably combined into the granite hydrogeologic unit.

Rhyolite

Rhyolite crops out extensively at the surface in the eastern Danskin Mountains and in the Mount Bennett Hills (fig. 2); it was defined at the surface from the Miocene rhyolite unit ( T m r) of Lewis and others (2012; fig. 3). This unit occurs in the subsurface occasionally in boreholes that were drilled within or just south of the rhyolite units that were mapped at the surface. Permeability in this unit is secondary, associated with jointing and fracturing (Ralston and Chapman, 1968; Whitehead, 1992) and the unit has low storativity and hydraulic conductivity (tables 2, 3).

Basalt

As discussed in the “Geologic History” section, there are likely basalts of at least two ages in the study area—Miocene and Pliocene basalt (unit T p m b of Lewis and others, 2012) and the younger Pliocene and Pleistocene basalt (unit Q T b of Lewis and others, 2012; fig. 3). These units occur extensively at the surface and in the subsurface through the central and southern parts of the study area and to a lesser extent in the eastern study area (fig. 2). The basalt hydrogeologic unit includes the following subunits: basalt (undifferentiated), cinders, and voids as interpreted from the well-driller reports. This detailed lithology is retained in part because there is likely to be significant heterogeneity within and between basalt units. In the eastern Snake River Plain, basalts are recognized to have low permeability in massive basalts as well as high permeability zones in the porous, broken interflow zone between lava flows, through networks of interconnected open fissures (voids), and in scoria (cinders) near vents (Anderson and others, 1999; Welhan and others, 2002; Ackerman and others, 2006). Therefore, the hydraulic properties of a given basalt unit are highly dependent on its characteristics and likely to be highly spatially heterogeneous, representing a range of storativity values and hydraulic conductivities from low to high (tables 2, 3). Transmissivity of this unit was estimated to be 48,200 feet squared per day (ft²/d) in the study area (Young, 1977), but could be as low as 20 ft²/d (Mann, 1986; table 2).

Fine-Grained Sediments

Fine-grained sediments were defined at the surface of the study area (fig. 2) based on map units of Lewis and others (2012): Q s (fluvial and lake sediment) and Q T p m s (sedimentary rocks, including consolidated to weakly consolidated sandstone and siltstone deposits of Lake Idaho; fig. 3). The fine-grained sediments unit occurs at the surface in the southeastern part of the study area, along the Snake River in the south, and near the foothills in the northwest (fig. 2); it also occurs at depths greater than 500 ft in deep boreholes on the Mountain Home Air Force Base (Deep-MH-1, Lewis and Stone [1988] and Deep-MH-2, Shervais and others [2012]). The northern extent of this unit is unclear since much of the study area is covered in basalt. However, assuming a high lake level from 1,100 to 1,160 m (3,610 to 3,800 ft) and minimal subsequent tectonic alteration (Jenks and Bonnichsen, 1989), Lake Idaho may have extended nearly to the foothills of the Danskin Mountains and Mount Bennett Hills, an altitude of approximately 1,200 m or 3,940 ft (Phillips and others, 2012).

The Lake Idaho deposits posed a substantial challenge for defining hydrogeologic units; these deposits are described in part as containing clastic beds that “are consolidated, although usually not lithified” (Malde and Powers, 1962, p. 1,202). In well-driller reports, presumed Lake Idaho deposits are variably described. For example, in an area with mapped Lake Idaho deposits and within a short horizontal distance, one well-driller report described shale and sandstone (for example, 352841; Zinsser and others, 2024), whereas a nearby well-driller report described clay and sand (for example, 307071; Zinsser and others, 2024). Because of this discrepancy in driller interpretation of rock versus sediment, the fine-grained unit includes the following primary lithologies as interpreted from the well-driller reports: silt, clay, shale, siltstone, sedimentary rock, and hardpan/caliche. Although lumping these units substantially improves lithologic continuity in the hydrogeologic framework model and is similar to the approach taken in the nearby Treasure Valley (Bartolino, 2019), it is likely that the hydraulic properties of these materials differ substantially. In general, hydraulic properties of the fine-grained sediments unit are intermediate (tables 2, 3). Transmissivity of this unit ranges from 970 to 3,325 ft²/d (table 2).

Coarse-Grained Sediments

Coarse-grained sediments were defined at the surface of the study area (fig. 2) based on the map units of Lewis and others (2012): Q a (alluvial deposits), Q l s (landslide deposits), Q a f (alluvial fan deposits), Q b s (Bonneville flood deposits), and Q T s (sediments and sedimentary rocks, including old terrace gravels; fig. 3). This unit occurs at the surface primarily along the range front, in tongues tracing the location of streams across the “Mountain Home plateau”, and along the “Snake River canyon” (fig. 2). Coarse-grained sediments also occur extensively at all depths interbedded with basalts and fine-grained sediments. Although grain size is poorly sorted in many of these deposits, coarser grain sizes (sand, gravel) are the defining feature of this unit. Similar to the fine-grained unit, the coarse-grained unit includes multiple primary lithologies interpreted from the well-driller reports: sand, gravel, sandstone, conglomerate, and boulders. The hydraulic properties of these materials likely differ spatially but are generally high (tables 2, 3). Transmissivity of the unit may range from 500 to 230,000 ft²/d (table 2).

Model and Database

The surface geology of the three-dimensional hydrogeologic framework model reasonably represents the surface location of the hydrogeologic units (figs. 2, 4) as defined by simplified geology from (Lewis and others, 2012). The model contains more variation at the surface than the simplified geology because the model was constrained by borehole lithology which describe units at a finer scale than the simplified geology and thus provides the model with more detail in areas with high borehole density (for example, in the central study area). The surficial expression of the hydrogeologic units in the model was improved in areas with low borehole density (for example, in the Danskin Mountains, Mount Bennett Hills, and southeastern “Mountain Home plateau”) by incorporating the surficial hydrogeologic unit locations as additional data points.

At depth, the three-dimensional hydrogeologic framework model reasonably represents expected geology throughout the study area (fig. 5) except below the mountain front, where the model incorrectly extrapolates fine-grained sediments extending under granite bedrock. In general, fine-grained sediments interlayered with minor coarse-grained sediments underlie most of the study area, thickening to the southeast (fig. 6). In the central and eastern parts of the study area, this unit also thickens toward the south (figs. 7, 8). Basalt, with minor and generally spatially discrete cinder units and minor interlayered fine- and coarse-grained sediments, predominates in the central study area (underlying the City of Mountain Home; fig. 8) and in the southwestern study area (although there is limited borehole data in this area; fig. 5). The coarse-grained sediments unit occurs in substantiative amounts on the Danskin Mountain range front (fig. 9) and in the Hammett area (fig. 7). Granite and rhyolite occur in the Danskin Mountains, from west to east, respectively, although the model inaccurately depicts these units as underlain by fine-grained sediments and basalt because of a lack of borehole data in the mountains (figs. 5, 9). Rhyolite also occurs in the model at depth below the Mount Bennett Hills range front (fig. 7), which is generally accurate, but the model-depicted interlayering of fine-grained sediments and rhyolite is unlikely.

Volumetrically (table 4), the largest hydrogeologic unit in the three-dimensional hydrogeologic framework model is fine-grained sediments (37 percent), followed by basalt (about 32 percent), coarse-grained sediments (about 19 percent), rhyolite (9.5 percent), and granite (2.2 percent). The volumetric percentages of rhyolite and granite are almost certainly an underestimate because the model does a poor job of representing these units at depth in the Danskin Mountains and Mount Bennett Hills because of a lack of deep boreholes in the mountains. Overall, basalt (about 49 percent) was the most common lithology by length in the borehole database (table 4), followed by fine-grained sediments (24 percent) and coarse-grained sediments (about 13 percent).

Limitations

There are inherent limitations in the borehole database and three-dimensional hydrogeologic framework model. The quality of borehole location data varies; boreholes with more certainty in their locations (for example, locations were determined by a global positioning system) were preferentially selected whenever possible, but some borehole locations are only known to the center of the nearest quarter-quarter section (as in, within 285 m [Lifton and others, 2021], assuming the quarter-quarter section was correctly identified). The quality of lithology and construction information also varies between well-driller reports, and the hydrogeologic units and subunits were drastically simplified from lithology to generate the model. Longer boreholes were preferentially selected over shorter boreholes, when possible, to get better lithology coverage, and database groundwater levels represent static water levels as reported when the wells were drilled, which encompasses more than 50 years. There is evidence that faulting both impedes and enhances groundwater flow (see section, “Faulting and Geothermal”) within the study area, but faults are not explicitly represented within the hydrogeologic framework model.

The choice of modeling algorithm and parameters also fundamentally limits the model; in this case, somewhat generous search distances allowed lithology to be interpreted for the full model, thereby preventing geologic gaps at the expense of poor interpretations in other areas (for example, the projection of sediment units under granite in the mountains). Unit transitions are also modeled abruptly, which allows the model to be repeatable, but unit transitions are therefore not geologically realistic in many situations. Model interpretations should be considered with caution in areas with sparse data (for example, under the mountains, the southwestern study area, and depths greater than 1,000 ft). The three-dimensional hydrogeologic framework model has the most complexity and is likely most representative of actual conditions within 1,000 ft of the surface and in areas with good borehole coverage.

Groundwater Occurrence

The borehole database is a helpful but imperfect tool to describe groundwater occurrence. For example, the quality of information varies in the well-driller reports, construction information is often incomplete, and boreholes are frequently open to multiple formations. Completion depth and lithology may not coincide with the water-bearing unit, and not all boreholes intercepted water or were intended for water production. For example, several very deep boreholes were drilled for oil and gas or deep geothermal exploration, and several others were drilled for gas pipeline cathode protection. Deep boreholes were also preferentially selected over shallow boreholes to get more complete lithology information, which may bias summary borehole statistics. Nonetheless, the 540 boreholes described in the database represent a rich data source to describe general patterns in water occurrence and are used here as such.

Based on reported water-bearing intervals and well completion lithologies in the borehole database, each hydrogeologic unit is capable of producing water. Wells are most commonly completed, in order, in basalt (undifferentiated subunit), fine-grained sediments (silt), unknown, coarse-grained sediments (sand subunit), and in the cinders subunit of the basalt hydrogeologic unit (fig. 10). Proportionally, water-bearing units occur most often in the cinders subunit (fig. 10). The median borehole depth was 491 ft, and over 95 percent of the boreholes are less than 1,000 ft. The total depth of wells tends to be shallowest (less than 250 ft) proximal to the range front and in lowlands adjacent to the Snake River (fig. 11). The deepest wells (greater than 1,000 ft) mostly occur in the Cinder Cone Butte Critical Groundwater Area and in the eastern study area (fig. 11).

Historically, both artesian and geothermal groundwater conditions have been observed in the study area (Ralston and Chapman, 1968). Although not all well-driller reports include water temperature at completion, water temperature was indicated in 235 of 540 reports; of these, 13 were recorded as greater than 85 degrees Fahrenheit. Most of the geothermal wells are located along the range front in a strikingly subparallel arrangement to northwest-trending faults, although several geothermal wells also occur along the central axis of the study area (fig. 12). Non-geothermal artesian conditions also occur in at least one well in the City of Glenns Ferry (308657; Zinsser and others, 2024) and were historically reported in the vicinity of the geothermal artesian wells in the Mount Bennett Hills range front (Ralston and Chapman, 1968).

Groundwater Altitude and Flow Direction

Two water-table contour maps were developed for the study area: one for the regional aquifer from the foothills in the north to the Snake River in the south (fig. 13A), and the other for the perched groundwater zone near the City of Mountain Home (fig. 13B). The regional water-table contour map was developed using groundwater altitudes from 139 wells. Depth to water in wells in the regional aquifer ranges from 150 to 765 ft. Previous studies have identified multiple perched groundwater zones near the City of Mountain Home (Young and others, 1992). The perched groundwater zone water-table contour map for this report represents a generalized water table from groundwater altitudes measured in 24 wells. Depth to water in wells in the perched groundwater zone ranges from 30 to 100 ft. However, 7 additional wells measured in the perched zone area had groundwater altitudes inconsistent (higher and lower) with nearby wells and may represent either multiple perched zones or are open to both the perched groundwater zone and the regional aquifer with groundwater levels influenced by both (fig. 13A, B; wells in unknown zone). The water table presented therefore represents one out of many possible interpretations of variable groundwater levels. Four wells were determined to represent perched groundwater conditions in the Mayfield area and were not included in contouring. Additional perched groundwater zones occur below the Mountain Home Air Force Base and in the Cinder Cone Butte Critical Groundwater Area but are not a significant source of supply and were not included in the water-table maps (SPF Water Engineering LLC, 2017; Air Force Civil Engineer Center, 2022). In four wells outside of the perched groundwater zone, groundwater levels indicated that conditions in the wells were not representative of the regional aquifer; thus, the four wells were not included in the regional contouring (fig. 13B; wells in unknown zone). Two wells were not included in contouring because the wells were being pumped or had been recently pumped at the time of measurement.

Regional groundwater generally flows from the foothills in the north of the study area toward the Snake River in the south-southwest (fig. 13A). Depth to water in wells in the regional aquifer ranges from 150 to 765 ft. The hydraulic gradient is steep in the north (100 feet per mile [ft/mi]), flattens south of the City of Mountain Home and Cinder Cone Butte Critical Groundwater Area (10 ft/mi), then steepens again at the “Snake River canyon” (50 ft/mi). Groundwater discharge is thought to be to the Snake River based on concurrence between the projected water table and Snake River surface-water altitudes (fig. 13A). This interpretation is consistent with previous researchers (Nace and others, 1957; Ralston and Chapman, 1968), a numerical groundwater flow model (Newton, 1991), and estimated groundwater discharge to the Snake River in this reach (Wood and others, 2014).

Generally, the spring and autumn water-table contours are consistent. The water-table contours moved upgradient in areas that experienced groundwater-level declines between the spring and autumn 2023 synoptics, but the direction of flow was consistent (fig. 13A). Water-table contours are uncertain in the southwestern study area and between the Cinder Cone Butte Critical Groundwater Area and the City of Mountain Home because of a lack of wells in those areas (fig. 13A). Low hydraulic gradients south of the City of Mountain Home and the Cinder Cone Butte Critical Groundwater Area could be related to long-term groundwater-level declines (Tesch, 2013) or could be due to higher transmissivity of the aquifer in these areas (Lindholm and others, 1988). The regional water-table contours from 2023 are generally consistent with groundwater flow described in previous reports although groundwater altitudes are lower in the Cinder Cone Butte Critical Groundwater Area and south of the City of Mountain Home than previously reported (Young, 1977; Young and others, 1992).

The general groundwater flow direction of the perched groundwater zone north of I−84 is to the southwest, whereas south of I−84 the flow direction is more westerly (fig. 13B). The groundwater level in the perched groundwater zone north of I−84 is approximately 50 ft above the regional aquifer. The perched groundwater zone and regional aquifer become closer in depth south of I−84 on the western extent of the perched groundwater zone which may be an indication of downward infiltration of perched groundwater into the regional aquifer. Because of the closeness in depth of the perched groundwater zone and regional aquifers on the western extent, it is possible to make multiple valid interpretations of the water-table contours. In the southern extent of the perched groundwater zone, the regional aquifer deepens and the perched groundwater zone shallows, resulting in greater depth between the two. As groundwater flows southwesterly toward the edge of the perched groundwater zone, the groundwater percolates downward to the regional aquifer through preferential pathways (Young, 1977; Bendixsen, 1994). The perched groundwater zone may also discharge at Rattlesnake Spring (Norton and others, 1982; Bendixsen, 1994). However, the water table shows flow to the west, implying only localized perched groundwater discharges at Rattlesnake Spring and most groundwater percolates through the unsaturated zone to the regional aquifer on the western boundary of the perched groundwater zone (fig. 13B).

Information about vertical hydraulic gradients in the study area is limited as there are limited monitoring wells open to different depths (rather than open to all depths). A consulting company measured groundwater levels in a dedicated groundwater-level monitoring well screened from 418 to 540 ft below land surface and an adjacent, deep 1,120-ft production well screened from 890 to 1,090 ft below land surface. Groundwater level is higher in the deep production well than the shallower monitoring well, suggesting there is an upward hydraulic gradient (HDR, 2024). However, it is unclear if there is truly upward flow or if this reflects the occurrence of a locally confining layer at depth. A limited, detailed study of three-dimensional flow within two monitoring wells on the Mountain Home Air Force Base indicated downward intraborehole flow within both wells (Newhouse and Hanson, 2002). Finally, the disappearance of the perched groundwater zones to the south have long been interpreted to imply that shallow perched water is percolating down to the regional aquifer (Young, 1977; Bendixsen, 1994). Vertical groundwater movement associated with faults is discussed further in the section “Faulting and Geothermal.”

Groundwater Change

Change in groundwater altitude from spring 2023 to autumn 2023 varied across the study area (fig. 14A, B). In the perched groundwater zone, groundwater levels rose in most wells with greater rises closer to Canyon Creek and rises ranging overall from 0.6 to 15.44 ft (fig. 14B). Groundwater levels declined in the perched groundwater zone in the well near Rattlesnake Creek (6.01 ft) and the two wells closest to the outlet of Mountain Home reservoir (1.66 to 4.07 ft). The most southern perched groundwater zone well only had a slight decline in groundwater level (0.17 ft). Differences in groundwater-level changes in the perched groundwater zone may be related to heterogeneity within the zone, groundwater use, and (or) proximity to surface-water recharge (fig. 14B).

In general, groundwater level in most wells in the regional aquifer declined (0.03 to 22.01 ft) from spring 2023 to autumn 2023. In the Cinder Cone Butte Critical Groundwater Area, groundwater-level declines ranged from 0.08 ft in the most upgradient well to 20.04 ft in the most down gradient well (fig. 14A). Groundwater levels declined in all wells near the Mountain Home Air Force Base with declines ranging from 3.33 ft to 10.63 ft. There was less change in groundwater levels east of the City of Mountain Home and in the far western study area where substantial groundwater withdrawals have not been developed. Groundwater levels rose from 5.70 ft to 11.09 ft in wells near the City of Grand View, likely due to recharge from applied irrigation water from the Snake River. Groundwater levels in the communities of King Hill, Hammett, and the City of Glenns Ferry rose in wells near the Snake River and declined distally; rises may occur in areas where surface water is used to irrigate, and declines may occur in areas with less surface-water irrigation and (or) more intensive groundwater withdrawals.

Groundwater storage change over the 2023 irrigation season was estimated using groundwater levels measured in the spring and autumn 2023 synoptics. Storage change estimates are provided for five regions: (1) the Cinder Cone Butte Critical Groundwater Area, (2) the Mountain Home Groundwater Management Area (which legally excludes [1]), (3) the combined groundwater management areas, (4) the study area excluding the two groundwater management areas, and (5) the perched groundwater zone. The autumn 2023 regional and perched groundwater zone water tables were intersected with the three-dimensional hydrogeologic framework model; the percentage of each hydrogeologic framework unit within this intersected surface is provided in table 5. The estimated storage change in acre-ft is provided in table 6 along with a range of estimates using the lower and upper values of storativity (table 2). Groundwater storage in the combined groundwater management areas is estimated to have declined 37,310 acre-ft (8,860 to 64,930 acre-ft) from groundwater withdrawals during the 2023 irrigation season (table 6). This estimate is broadly consistent with a previous annual withdrawal estimate of 43,000 acre-ft (SPF Water Engineering LLC, 2017) and an annual water-budget deficit of 30,900 acre-ft (Harrington, 2004). Groundwater storage in the study area excluding the combined groundwater management areas is estimated to have increased 4,810 acre-ft (1,230 to 10,670 acre-ft; table 6). The groundwater storage in the perched zone is estimated to have increased 6,390 acre-ft (1,410 to 11,690 acre-ft; table 6).

The IDWR maintains a network of groundwater-level monitoring wells in the “Mountain Home plateau” which have long-term groundwater-level data since 1960 (fig. 15). Hydrographs of groundwater altitudes for selected wells show different interannual and long-term changes depending on the location and depth of the well. In general, regional aquifer wells in the Cinder Cone Butte Critical Groundwater Area, City of Mountain Home, and Mountain Home Air Force Base have declined since 1980 (figs. 16, 17). In contrast, groundwater levels along the “Snake River canyon” in the community of Hammett and the City of Glenns Ferry (fig. 18) and in the perched groundwater zone near the City of Mountain Home (fig. 19) do not show long-term declines.

Wells in the Cinder Cone Butte Critical Groundwater Area have the greatest groundwater-level declines in the study area, with wells in the southwestern part of the critical groundwater area (03S 05E 07BDD1, 02S 04E 22CCC2; fig. 15) showing declines of over 100 ft since about 1980 (fig. 16). Wells in the eastern part of Cinder Cone Butte Critical Groundwater Area (02S 05E 26BDB1, 02S 05E 36BBB1) do not show the same magnitude of long-term change as those within the south and west but have nonetheless declined since 1980. In a well located northwest of the Cinder Cone Butte Critical Groundwater Area (01S 04E 30AAC1; fig. 15), groundwater level rose until 2000 (about 7 ft from 1967 to 2000) and then has declined since (about 7.5 ft from 2000 to 2023; fig. 16). These recent declines could be related to the groundwater in these wells being captured in the cone of depression within the Cinder Cone Butte Critical Groundwater Area since 2000 (Tesch, 2013). Wells north of I−84 but within the critical groundwater area boundary show no decline (02S 05E 03BAB1) or an overall decline (02S 05E 11AAB1) which is a third of the magnitude of the declines observed in wells in the southwest part of the Cinder Cone Butte Critical Groundwater Area (03S 05E 07BDD1, 02S 04E 22CCC2) over the period of record (figs. 15, 16).

Groundwater levels have also declined over 50 ft since 1980 in wells south of the City of Mountain Home and in the vicinity of Mountain Home Air Force Base (fig. 17). In the eastern part of the study area, well 04S 10E 30BBA1 declined about 15 ft between about 1968 and 1988 but has been stable overall since with periods of small declines and recovery (fig. 20). Similarly, monitoring wells near the community of Hammett and the City of Glenns Ferry (05S 08E 36CCC1, 05S 09E 36DAB1) showed relatively stable groundwater levels, although groundwater levels in well 05S 09E 36DAB1 rose from 2006 to 2012 (fig. 18). Since other monitoring wells did not experience the same pattern, this rise in water-level may be related to local changes in irrigation or pumping practices.

Since the 1970s, long-term groundwater monitoring wells in the perched groundwater zone showed short periods of decline (over several years) followed by recovery within a few years that resulted in overall stable groundwater levels (fig. 19). Periods of decline and recovery seem to be related to periods of abundance or lack of surface water, as indicated by the streamflow in Canyon Creek (fig. 19). For example, groundwater-level declines in the early-1990s are associated with very low streamflow, followed by the return of high streamflow and groundwater-level recovery in the mid-1990s (fig. 19). Starting in 2016, during spring runoff in years of high flow, streamflow has been intentionally diverted from Canyon Creek for groundwater recharge. Continuous groundwater-level measurements show a sharp increase in groundwater level during times of diversion, and recovery of groundwater levels during the spring runoff of Canyon Creek in general even during years without diversions (fig. 21). Taken together, these patterns suggest the importance of surface-water supply to recharge the perched groundwater zone.

Perched Groundwater Zone

Perched groundwater has been described around the City of Mountain Home since at least the 1950s, when Nace and others (1957) described perched water occurring in shallow gravels that was recharged by underflow and losses from irrigation and local streams. Nace and others (1957) also recognized that impermeable basalts, caliche, and fine sediments could create perching conditions, whereas permeable basalts and sediments would allow percolation to the regional aquifer. Ralston and Chapman (1968) described an area of shallow (less than 300 ft) well development in the Mountain Home area with perching conditions caused by a series of clay lenses and a clay layer on top of basalt. Young (1977) recognized perched conditions in multiple locations along the range front within the study area, with recharge derived from surface water and irrigation, suggesting that such leakage was responsible for the formation of the perched zones, and explicitly delineated the approximate area of the perched water zones. Later researchers largely adopted this (Norton and others, 1982; Bendixsen, 1994) or a similar (Lindholm and others, 1988) delineation and described the perched groundwater zone around the City of Mountain Home as occurring primarily in unconsolidated sediments (Norton and others, 1982; Bendixsen, 1994; SPF Water Engineering LLC, 2017).

The three-dimensional hydrogeologic model demonstrates that shallow fine-grained and coarse-grained sediments units underlying the City of Mountain Home are largely discontinuous (figs. 5, 8). A query of the borehole database to identify the lithology of shallow (less than 150 ft) units identified in the well-driller reports as water-bearing around the City of Mountain Home lends insight into the perched zone (table 7). Most of the perched units, particularly north of I−84 and west of U.S. Route 20, occur in basalt or cinders subunits approximately 30–100 ft below the surface, whereas perched water is noted in shallower (0–25 ft below the surface) sand and gravel subunits south of I−84 (figs. 22, 23). Basalt (undifferentiated) is the most common subunit described by drillers as not bearing water below a shallow water-bearing unit (table 7; fig. 23). This suggests that there is not an extensive fine-grained sediment unit creating perching conditions in many locations, although it is possible that minor fine-grained sediments (such as a thin soil over basalt) could contribute to perching conditions but potentially be missed in well-driller reports. Thus, based on these observations, perched groundwater near the City of Mountain Home occurs at multiple altitudes, primarily in basalt and some shallow unconsolidated sediments, and is most commonly underlain by basalt.

The lack of an extensive fine-grained sediments unit makes the lateral extents of perched groundwater conditions near the City of Mountain Home uncertain. It is unclear what method Young (1977) used to derive the approximate outline of the perched unit as the boundary extends beyond the shallow groundwater-level data in that study’s map. Figure 24 shows an updated perched groundwater zone outline based on: (1) 2023 synoptic groundwater-level data, (2) USGS groundwater-level data since 1955, and (3) wells with water-bearing units less than 150 ft deep identified in the borehole database. There is also an interesting concurrence of the original perched groundwater zone delineation with the surficial mapping of a coarse-grained sediments unit (Lewis and others, 2012) and the location of major irrigation infrastructure (fig. 24). The importance of local surface-water recharge to the occurrence of perched groundwater near the City of Mountain Home has long been recognized (Nace and others, 1957; Young, 1977; Norton and others, 1982). Additional areas of shallow groundwater levels that fit neither the perched groundwater zone near the City of Mountain Home nor the regional aquifer were identified in the groundwater synoptics (fig. 13A, B); these areas generally occur along the range front and are consistent with outlying perched groundwater zone locations described by Young (1977) and Lindholm and others (1988). While all of these methods are imperfect to delineate the zone within which perched groundwater conditions may occur, they do generally agree.

Conceptually, these findings suggest a system in which perched conditions can occur at multiple altitudes, locations, and units depending on the precise hydrogeologic conditions encountered in the subsurface and contingent on proximal surface recharge (fig. 25). Precipitation falls on the mountains and foothills, feeding streams and infiltrating into the fractured rhyolite, basalt, and sediments via both deeper flow paths to the regional aquifer and shallow flow paths where local conditions allow. High hydraulic conductivity can occur within the basalt hydrogeologic unit along interflow zones (highly porous, broken zones between discrete flows), through vertical fissure networks (Anderson and others, 1999; Welhan and others, 2002), in scoria (Anderson and others, 1999), and through improperly constructed wells (wells open to multiple aquifer zones). The basalt hydrogeologic unit near the City of Mountain Home is highly complex, representing multiple overlapping shield volcanoes and cinder cone volcanic events (Shervais and others, 2002). In the subsurface, this complexity translates to intermittent zones with high horizontal hydraulic conductivity (for example, in interflow zones and horizontally distributed cinders), punctuated by zones of enhanced vertical hydraulic conductivity (for example, fissure networks and near-vent scoria) that allow perched groundwater to percolate downward toward the regional aquifer (fig. 25). Reported and measured water-levels at different altitudes supports the idea of multiple horizontal water-bearing units at different altitudes. The disappearance of perched groundwater conditions toward the south indicates the downward migration of perched groundwater to the regional aquifer, and vivid observations from Mountain Home Reservoir demonstrate the impact that vertical fractures can have on water loss (Harrington-Thornton, 1995). Aiming to eliminate “uncontrolled loss of reservoir water” (p. 3), the reservoir was drained and excavated to a depth of approximately 10 ft in 1995, revealing basalt bedrock with a large vertical fracture with a visible depth of at least 2–3 ft, which was then partially filled with concrete with two standpipes emplaced for managed recharge (Harrington-Thornton, 1995). The horizontal and vertical complexity in the basalt hydrogeologic unit, combined with abundant available recharge via streams, canals, and irrigation near the City of Mountain Home and near the range front, create complex zones of perched groundwater.

Faulting and Geothermal

Faulting likely impacts water occurrence and movement within the study area in variable ways, including enhancing local water production, altering hydraulic conductivity, or offsetting water-bearing zones, depending on the location. While the locations of some faults are known through geologic mapping (Lewis and others, 2012; Phillips and others, 2012) and subsurface geophysical surveys (Liberty, 2012; fig. 2), the location and impact on hydrology of other faults is inferred from various lines of evidence. A likely fault was identified during the development of the borehole database, where two newly drilled wells (473257 and 473699; Zinsser and others, 2024) in the same housing development south of Mayfield had a difference in static groundwater levels of over 110 ft despite being located only about 1,700 ft apart. Similarly, these wells had offset groundwater levels (112 ft) when measured as part of the autumn 2023 synoptic. It is possible that the fault offsets a water-bearing unit, resulting in the lower groundwater level on the down-dropped sediments south of the fault, and (or) that the fault created a zone of decreased hydraulic conductivity, causing lower groundwater levels on the downgradient side of the fault (fig. 26). Down-dropping to the southwest is generally consistent with regional faulting although the offset is not clearly resolvable in the unconsolidated sediment units described in the well-driller reports. A similar aquifer-offsetting fault was inferred by Norton and others (1982) along I−84 between Orchard and the City of Mountain Home. Faults interpreted by Liberty (2012) near Mayfield offset the granitic bedrock surface by 100–300 ft and dip steeply to the southwest, and offset water-table reflectors in a seismic profile south of Indian Creek suggest a buried fault offsetting the water table (Liberty, 2012). Thus, there exists evidence in geophysical data, groundwater-level data, and borehole lithology to suggest that unmapped faults can inhibit groundwater movement, lowering the water surface altitude on the down-dropped and down-gradient side of the faults.

There is also evidence to suggest that faults may enhance groundwater movement within the study area. Welhan and others (2002) interpreted steepened groundwater-surface gradients along the range front as an indication of deep groundwater recharge along faults; this hypothesis is supported by some geochemical data indicating deep geothermal water may contribute to aquifer recharge in the Mayfield area (Hopkins, 2013). In the eastern study area, groundwater levels in a series of deep, artesian geothermal wells are likely controlled by a series of faults. These wells have been subject to extensive litigation, study, and decree since the 1960s. Lithology descriptions in the well-driller reports are ambiguous, with rocks described by color (variously, grey, black, red, green, and cream) and competency (hard, broken), but generally the wells are interpreted to be completed within faulted silicic-volcanic rocks (Castelin, 1993) which is consistent with area geology (Lewis and others, 2012). Three wells (424832, 424830, and 352841) are hydraulically connected, while a fourth well (423043) is not (fig. 12; Castelin and MacAllister, 1978; Castelin, 1993) and is completed in a sand unit. Confining conditions are likely caused by overlying fine-grained sediments, which are variously described as clay or shale, with water supplied and heated via deep flow paths and transmitted by faults (fig. 27). The occurrence of multiple, unconnected geothermal artesian zones suggests multiple faults may be present. However, artesian flow at the connected wells is seasonal, and in practice, flow occurs at lower rates than the decreed volumes (Castelin, 1993). Extensive exploration of deep geothermal water resources has been undertaken on the Mountain Home Air Force Base (Shervais and others, 2012) but this resource is not being used at present. Thus, while there is evidence of deep geothermal recharge to the aquifer in some areas, there is not compelling evidence that this represents a substantiative source of water to the regional aquifer. Whether enhancing or inhibiting groundwater flow, it is likely that additional unmapped faults occur within the study area and may affect groundwater occurrence and movement.

Data Gaps

Data gaps and limitations in our knowledge of the study area remain. In the three-dimensional hydrogeologic framework model, units below the mountains are poorly represented, and units in the southwestern study area and greater than 1,000 ft below land surface are largely unconstrained because of a lack of borehole data in these areas. Faults are not represented within the model, and the full scope of fault influence on groundwater movement is unknown. Similarly, the importance of deep geothermal recharge to the regional aquifer is not well understood. Generally, the three-dimensional hydrogeologic framework model represents the hydrogeologic conditions best where there is abundant borehole data (for example, at depths less than 1,000 ft and where there are many boreholes) but poorly represents large-scale geologic relations (for example, range front contacts). In contrast, the conceptual diagrams represent large-scale geologic relations (for example, faults, contacts between bedrock and sediments) more accurately but generalize near-surface conditions to demonstrate conceptual understanding of water movement.

In the water-table contour map, there is limited primary groundwater data in the eastern and southwestern study area and in the mountains and foothills because of a lack of wells and (or) the inability to access and measure existing wells. Many wells in the perched groundwater zone are open to multiple units, which obscures understanding of perched and regional groundwater levels and may contribute to the observation of groundwater levels that do not clearly belong to either zone. Similarly, the relative importance of precipitation, mountain front recharge, and incidental surface recharge to the persistence of the perched groundwater zone is not well understood. Installation of wells which intentionally isolate the perched groundwater zone and regional aquifer would be helpful for understanding the dynamics of each, and in resolving vertical groundwater gradients. A study evaluating sites for a water supply well for the City of Mountain Home estimated transmissivity of 58,580 ft²/d (6,760–481,450 ft²/d) from 18 well tests (Brockway Engineering, 2019). Nonetheless, estimates of transmissivity within the study area are limited and additional aquifer tests could be helpful for potential future numerical groundwater modeling, particularly if such estimates were clearly tied to distinct hydrogeologic units.