To reflect the mixing of soil horizons in the calculation of the deep percolation factor, the depth-weighted average of Ksat for the entire soil profile was used in place of the lowest Ksat for each profile. We reduced the deep percolation factor rating for soils with claypans by 20% to reflect the risk that modified claypans will reform, which can occur in as little as four years in soils with weak structure . Cemented layers were assumed to have been removed by deep tillage and were not included in the weighted average. Data below the restrictive horizon was included in the depth-weighted average if populated in the database. The depth-weighted average of Ksat was used in place of the harmonic mean to estimate hydraulic conductivity for the root zone residence time factor. Our study area included over 17.5 million acres of agricultural land as identified by the state Farmland Mapping and Monitoring Program. Based on our initial modeling, which did not initially consider the effects of deep tillage, soils in the Excellent, Good and Moderately Good suitability groups comprised over 5 million acres, or 28% of the study area . These highly rated soils were most abundant on broad alluvial fans on the east side of the Central Valley stemming from the Mokelumne, Stanislaus, Merced, Kern and Kings rivers . Excellent, Good and Moderately Good ratings are also found throughout much of Napa, Salinas and Santa Maria valleys and in patchesalong the Russian River in Mendocino and Sonoma counties and the northern parts of the Coachella Valley. The best soils — the Excellent and Good groups — occupied about 3.2 million acres, representing 18% of the study area .

Some areas of Good and Excellent ratings were found on sandy floodplains of rivers and streams,large plastic pots especially along the Sacramento and Feather rivers. Floodplains may not be ideal locations for groundwater banking because of the potential for applied water to flow, by subsurface transport, into rivers and streams. Thus, these systems should not be prioritized for groundwater banking unless it is known that the surface water bodies are losing streams — that is, surface water bodies that discharge to groundwater. Most major streams that traverse the San Joaquin Valley, for instance, are known to be losing streams. Extensive Moderately Good areas were mapped on the western margins of the San Joaquin and Sacramento valleys where soils tend to be finer textured and sometimes salt-affected . Moderately Good groups were also mapped in basin alluvium where low energy flood events have deposited fine sediments. Moderately Good groups occupied 1,786,972 acres or 10% of the study area. These areas may require careful consideration for groundwater banking. The majority of land in the study area was classified as Moderately Poor, Poor or Very Poor SAGBI groups . Soils with low SAGBI scores were abundant throughout the basin margins of the Sacramento and San Joaquin valleys as well as across land interstratified between recent alluvial fan deposits of the Mokelumne, Tuolumne, Stanislaus, San Joaquin, Kings and Kern rivers. Very Poor and Poor ratings are also found on the northern portions of the Salinas and Santa Maria Valleys and throughout most agricultural regions in Sonoma County and southern parts of the Coachella Valley. Of the SAGBI components, the deep percolation factor was limiting over the greatest area . These limiting conditions arise from different characteristics of soils. For example, old, highly developed soils found along the margins of the Central Valley contain water-restrictive horizons . The center of the valley contains young soils with fine texture throughout the soil profile. Both of these soil landscapes contain at least one soil horizon with low permeability. In contrast, high deep percolation scores were found on coarse-textured soils derived from recent alluvial fans with drainages sourced in granitic terrain of the Sierra Nevada and the Salinian block within the Coast Range.

Areas limited by the root zone residence factor typically had soils with uniformly fine texture throughout the soil profile and poor drainage. Poorly and very poorly drained soils have properties or conditions that promote saturation in the upper parts of the soil profile, such as high clay content, water restrictive layers or regionally shallow water tables. The least suitable soils in this factor were those with poor drainage or high shrinks well properties. Low scores for root zone residence factor were widespread along the west side of the San Joaquin and Sacramento valleys in soils weathering from Coast Range alluvium . Poor drainage and fine textured soils were also found in the basin alluvium towards the center of valleys. Low scores for this factor were also found on alluvial fans that have drainages confined to the metamorphic portions of the Sierra Foothills such as the Calaveras River fan, which tend to have fine textured sediments compared to fans sourced in granitic terrain in the high Sierra Nevada. Chemical limitations had a localized influence on the distribution of SAGBI ratings. Most of the salt-affected soils are present along the west side of the San Joaquin Valley and to a lesser extent along the western margin of the Sacramento Valley . The distribution of salt affected soils results from a combination of the salt-rich nature of the marine sediments within the Coast Range and poor drainage conditions on the west side that prevent salts from leaching out of soil. There are other chemical limitations of soils we could not evaluate that would influence groundwater banking, most notably the concentration of residual nitrate in soil. Crops with high nitrogen demand or high residual nitrate in soil in the fall after harvest may not be suitable for groundwater banking . The surface condition factor was weighted lowest among all other factors because compaction from standing water can be fixed with tillage and amendments. Low surface condition factor ratings were abundant in soils with loamy surface textures or high SAR and were located throughout the study area but tended to be concentrated on the west side of the Central Valley where sodium affected soils are common . Soil landscapes with low slope factor ratings were limited to the margins of the valleys . This sloping terrain is a result of uplift by the Coast Range and Sierra Nevada over geologic time scales, which increased slope gradients and accelerated erosion.

The natural erosion of the valley margins has created gentle to steeply undulating land forms .When deep tillage on orchard and vineyard croplands was incorporated into the model, the Excellent, Good and Moderately Good SAGBI suitability groups increased from 28% to 31% of the land area, adding 550,494 acres of suitable agricultural land for groundwater banking . A majority of improved SAGBI scores were located in the eastern San Joaquin Valley and Tulare Basin,black plastic plant pots bulk where soils with restrictive horizons are common . It is possible that over time, more suitable land for groundwater banking will become available as marginal soils continue to be developed and modified for agricultural purposes . The final SABGI that accounts for deep tillage represents the best estimate of soil suitability for groundwater banking. Over 12 soil survey areas are classified as out-of-date in agricultural regions of California and do not accurately document the extent of soil modification by deep tillage. These modified SAGBI ratings provide an updated assessment of the current state of soils in the study area. There are approximately 5.6 million acres of land with soils in Excellent, Good and Moderately Good SAGBI suitability groups, a significant amount of agricultural land capable of accommodating deep percolation with low risk of crop damage or contamination of groundwater by salts. Most suitable soils for agricultural groundwater banking occur on or near alluvial fans created by rivers draining the Sierra Nevada. Perhaps not coincidentally, these are also the areas that have California’s most successful groundwater banking programs . Our preliminary survey of UCCE perennial crop experts suggests that pears, wine grapes and some root stocks of various Prunus species are best suited for groundwater banking if planted on suitable soils and managed appropriately, especially after bud break. While extensive in acreage, almonds may be less ideal because of the trees’ sensitivity to saturated conditions and high nitrogen demand . Walnuts may be an option given that bud break typically occurs in late April. Wine grapes may be the best option because of the extensive acreage planted, low nitrogen demand and tolerance to standing water . Almonds with plum root stocks may also be suitable; however, currently almonds with water tolerant root stocks are generally planted in soils that are poorly drained and thus less likely to be suitable for groundwater banking. A preliminary calculation based only on soil properties and crop type shows that landscapes rated Excellent or Good could be used to bank as much as 1.2 million acre-feet of water per day. This estimate assumes 1 foot per day of water infiltration on lands in the Excellent and Good categories that are planted with grapes or alfalfa , or fallowed . There are significant limitations to this estimate. Most importantly, California lacks the infrastructure to accommodate and route such large volumes of water to the fields in such a short time . Plus, the heterogeneity in precipitation across the state makes this estimate improbable . Offsetting these limitations to some degree are other crop types that would be suitable for recharge but were not included in this estimate.

The financial risk associated with crop loss may exceed the potential benefits of water savings. Perennial crops carry particular risks and uncertainties. For instance, while trees and vines are generally more tolerant of saturation before bud break than after , determining a reliable cutoff date for this increased tolerance is difficult. Tree and vine roots generally start to grow several weeks before bud break, so damage from water logging can occur well before bud break. Moreover, bud break for a given species varies by location across the state. In addition, standing water on trunks can lead to aerial Phytophthora or other diseases. Investigating this opportunity in less valuable cropping systems, such as alfalfa, irrigated pasture and annual crops may be more promising until further research on tree crop sensitivity to standing water has been conducted. If groundwater banking on agricultural lands becomes a priority, coordination at the policy, market and planning levels would be needed to provide an adequate land base ready to opportunistically capture floodwaters. Adoption of this practice would likely require some form of support to mitigate or protect growers from the risks of crop failure. For example, growers who make their land available for floodwater capture and groundwater banking could receive credits from municipalities or irrigation districts. They could also receive credits from irrigation districts for enrolling in a long term program. Long-term commitments from growers likely would be needed for basin-scale planning purposes. Although not included among the crops listed in table 1, alfalfa may be an ideal crop for groundwater banking because it requires little or no nitrogen fertilizer, reducing the risk that groundwater recharge would transport nitrates into aquifers. Alfalfa is sensitive to flooding and saturated conditions; thus the timing of flooding should coincide with older fields slated for replanting. Because the financial risk associated with crop damage is lower in alfalfa than in tree and vine crops, the financial incentive needed to drive grower participation in groundwater banking programs likely would be lower as well. Most annual cropping systems would be suitable for groundwater banking if water is applied when land is fallow. The major risk in annual crops is leaching of residual pesticides or fertilizer in the soil. Appropriate management practices for groundwater banking with specific annual crops would need to be developed. If agricultural groundwater banking becomes an important water security practice, the SAGBI may provide valuable information to guide future changes in cropping systems. SAGBI can be a powerful aid to decision makers and stakeholders when considering the trade offs associated with the implementation of groundwater banks utilizing agricultural land for direct recharge. It was also developed with the intention of informing growers of the potential hazards associated with this practice.