Geophysical Quantification of Water Percolation Quotient in an Alluvial Agricultural Soil

Efficient water use planning is crucial for the sustainability of irrigated agriculture in California, where alluvial geological materials with indigenous salts impinge on crop growth. To facilitate irrigation scheduling and cultivation planning, it is necessary to determine water percolation quotients (WPQ) required for removal of excess salts from the rhizosphere. In order to estimate real-time WPQ, we conducted electromagnetic geophysical surveys at a saline farmland followed by stochastic computations. Results showed a wide variability in salinity that reached 16 dS m in some locations. About 95% of the surveyed samples surpassed 2 dS m. Despite spatially dependent asymmetric variability and skewness (-0.13 to 1.90), the WPQ distribution patterns were consistently quantified with low errors (< 0.06). The sensor responses in the fields reached 100% cumulative frequency at a threshold of 13.6 dS m. Up to 49% of WPQ data ranged from 0.1 to 0.2. The WPQ decreased with increasing salinity and the zones with low quotient values represented areas where plant growth could be impaired. High WPQ levels demarcated zones with potential solute dissolution and dispersion. Overall, evaluation of WPQ can benefit irrigation planning and crop management practices while enhancing water use efficiency for agricultural production in farms that have been affected by drought and water shortage, and crop growth can be sustained at WPQ level that maintains salts below the crop tolerance threshold.


Introduction
Geological parent materials in many agricultural lands of California are predominated by alluvial geomorphic structures that are primarily composed of shale and sandstone deposits containing elevated levels of indigenous salts (McNeal & Balisteri, 1989). High water tables, shallow clay layers and inadequate drainage in these areas result in reduced water percolation and salinity buildup that ultimately impair soil structure and crop growth (USGS, 2015). Agricultural productivity of these farmlands is heavily dependent on irrigation. However, extensive application of poor-quality water often exacerbates the salinity problem. Nearly 45% of irrigated agricultural lands in California are impacted by soil or water-induced salinity (Letey, 2000).
For mitigating these adverse conditions, it is necessary to develop a precise water percolation quotient (WPQ) in order to remove salts from the rhizosphere and maintain a tolerable salinity for plants. Application of irrigation water in excess of plant requirement can achieve this removal (Hanson et al., 2006). The percent of excess water flowing past the rhizosphere depends on several factors including soil texture and salinity, electrical conductivity of the applied irrigation water and salt tolerance of the planted crops. Subsequently, estimation of precise water quantity in field conditions is a challenging task for the growers as conventional measurement methods of using electrode probes and soil sampling are laborious, slow and costly (Davis et al., 1999). Under variable environmental conditions, the electromagnetic geophysical sensing (EGS) technique can be utilized for rapid and reliable quantification of WPQ over large areas. The sensing approach allows for real-time above-ground measurements and provides a better, rapid and economical option as compared to the invasive traditional methods (Hendrickx et al., 1992;Diaz & Herrero, 1992;McKenzie et al., 1997;Sudduth et al., 2003).
Versatility of the EGS technique has been reported for diverse environments. The sensing approach was found to be effective for assessing groundwater recharge (Cook et al., 1992), wetland (Paine et al., 2004), soil moisture in agroforestry (Huth & Poulton, 2007), and coastal agriculture (Yao et al., 2016). Cassel et al. (2015) already detailed the concept of applying electromagnetism for rapid and high-resolution assessment of salinity. The EGS techniques have been applied for salinity and water management across the globe including China (Yao et al., 2016), New Zealand (El-Naggar et al., 2017, Utah (Abdu et al., 2017), the Nile delta (Aboelsoud & Abdel-Rahman, 2017) and Spain (Pedrera-Parrilla et al., 2017). Lia et al. (2015) used EGS measurements for studying the impact of saline water irrigation on desert ecology. Huang et al. (2017a) predicted soil water dynamics using EGS and assimilating artificial neural network and physical model data. Recently, Cassel (2017) demonstrated the application of electromagnetism for site-specific yield optimization in central California. In another study, Cassel and Sharma (2017) described the application of electromagnetism for spatial analysis of salt heterogeneity in a grape field.
Considering the multipurpose benefits of EGS, the current research concentrated on water quantity analyses with specific focus on water percolation to application ratio. Thus, the objective of this study was to estimate WPQ in some salt-affected alluvial agricultural soils in California. Based on the electromagnetic survey and stochastic computations, we quantified soil salinity distribution and WPQ in real-time despite high parametric variability. This approach can help growers and agricultural decision makers develop selective soil reclamation and irrigation management practices depending on specific crop tolerance levels.

Geophysical Survey
We performed an EGS survey in a Californian farmland with alluvial silty clay soils that had slowly degraded due to salt accumulation (USDA, 2006). The geomorphic feature included fan remnants with alluvial parent materials derived from sedimentary and igneous rocks. The land was characterized by flat topography with < 1% slope, alkaline soil pH (> 7.9), silty loam surface horizon and silty clay subsurface soils. In this area, occurrence of clayey and saline parent materials and inadequate drainage led to salt buildup that was exacerbated by intensive irrigation practices (Cal EPA, 2006;USDA, 2014). Typical vegetation comprised irrigated crops including tomato as well as salt-tolerant shrubs and grasses.
Our study concentrated on two 800 m × 800 m contiguous fields within the farm, designated hereafter as fields A 1 and A 2 . In these fields, soils within a 1.2 m rhizosphere were investigated using dual-dipole electromagnetic sensors (38DD) aligned in perpendicular positions. The sensors were set to measure soil electrical conductivity at 14.6 kHz frequency across 0.8 and 1.5 m lateral and vertical depths, respectively. Potential signal interferences from terrestrial metals were eliminated by securing the sensors in a protective carrier-sled located about 3 m behind a tow-vehicle. Precise survey points were obtained using a global positioning system with differential correction capability at the sub-meter accuracy level. The real-time data were recorded using an on-board computer connected through digital interfaces.

Optimal Sampling and Laboratory Analyses
Following the EGS data recording, an optimal soil sampling plan with spatial characteristics of the whole survey area was devised using an ESAP analysis (Lesch et al., 2000). For subsequent calibration, the response sample design methodology was applied to select the survey locations that yielded minimal spatial auto-correlations and best described the spatial variability in apparent electrical conductivity. Within 48 hours of the survey, physical soil samples were collected at specific plan-defined points across the rhizosphere. Afterward, these soil samples were processed and analyzed for electrical conductivity (EC), volumetric water content (θ) and water saturation percentage (SP) using procedures described by Dane and Topp (2002) and Gavlak et al. (2003).

Stochastic Analyses and Mapping
Soil salinity and WPQ were determined based on electrical conductivity characterization by geophysical surveys, laboratory assay, as well as statistical and stochastic modeling analyses described by Lesch et al. (2000). The EGS response data were also analyzed for dispersion, variance, symmetry and distributions. For each field, the model that estimated WPQ was calibrated using the EGS response values as well as their mean and range across the rhizosphere. The principal component de-correlation analysis and validation routines were applied to expel the outliers, and center, scale and de-correlate the conductivity data. Results from the field surveys and the physical soil data were utilized for stochastic analyses, and EC, θ and SP data were integrated to compute WPQ. The signal and trend surface parameters were estimated using multi-linear regression analyses. The WPQ was then determined based on the estimated rhizosphere conductivity values by using a conversion routine available in Lesch et al. (2000). Next, raster maps of soil salinity and WPQ were plotted using ArcGIS (ESRI, 2012).

Results and Discussion
Relevant soil physicochemical properties of the plan-defined locations within the rhizosphere of fields A 1 and A 2 are summarized in Table 1. The electrical conductivity (EC) values in the surveyed fields ranged from 1 to 26 dS m -1 , and were characterized by wide variations in salinity with 45 to 80% coefficient of variation (CV). The mean salinity levels were 9 and 6 dS m -1 in fields A 1 and A 2 , respectively. In this farmland, most vegetable crops would be adversely affected by these levels of salinity. For example, tomato was reported to tolerate salinity up to a threshold of 2.5 dS m -1 (Hanson et al., 2006). Likewise, onions yields start decreasing when salinity reaches 1.2 dS m -1 . Even cotton, considered a salt-tolerant crop with a salinity threshold of 7.7 dS m -1 , would exhibit yield declines across most of field A 1 and in the north-west corner of field A 2 . Average SP values ranged from 27 to 60% with the higher values suggesting more clayey soil texture. Greater average and maximum SP levels were observed in field A 1 , which also exhibited the higher mean EC level. The soil water contents were near the field capacity conditions in both fields with levels ranging from 18 to 40%. Abdu et al. (2017) employed electromagnetic mapping at varied water contents of a fallow field with mostly homogeneous silt-loam alluvial soils under xeric moisture regime in Utah. By utilizing the geophysical method, the authors were able to determine the textural patterns and observed that the lowest electrical conductivity zone had high correlations with coarse geological materials.
Considering multi-layer contributions from zero to infinity (), the cumulative response (R(d)) and the corresponding electrical conductivity (EC(d)) were integrated to determine the total EGS response (R egs ):  Vol. 9, No. 12;2017 the areas exposed to deep drainage risks. The geophysical assessment can help growers in understanding the soil water dynamics for farm management planning. The spatial distribution of WPQ can assist evaluating soil water storage and hydraulic transport processes and the overall approach can be applied to a wide range of agricultural operations including irrigation scheduling, water budgeting and site-specific crop selection. Efficient use of water resources is crucial for the sustainability of irrigated agriculture in California and many parts of the world, where continuous droughts have accentuated the need to conserve water and improve on-farm water application.
The study pertains to critical challenges in California including limited water supplies for agricultural industry and declining water and soil quality. Evaluation of WPQ can facilitate development of crop water requirements and optimization of water use efficiency under integrated irrigation scheduling. The WPQ evaluation can be integrated with agricultural monitoring methods and precise data acquisition processes to provide real-time plant water utilization data that can subsequently enhance efficient irrigation planning decisions to sustain crop productivity and implement soil reclamation strategies. The overall approach of this study can benefit decision support tools for planning water-related farming practices.