Can shallow-layer measurements at a single location be used to predict deep soil water storage at the slope scale?

- Variations in soil water storage were related to soil water status.
- Strong temporal stability was identified along an agricultural slope.
- Deep-layer soil water storage was predicted with shallow soil water measurements.

SummaryKnowing the amount of soil water storage (SWS) in agricultural soil profiles is important for understanding physical, chemical, and biological soil processes. However, measuring the SWS in deep soil layers is more expensive and time consuming than in shallower layers. Whether deep SWS can be predicted from shallow-layer measurements through temporal stability analysis (TSA) remains unclear. To address this issue, the soil water content was measured at depths of 0-1.6 m (0.2-m depth intervals) at 79 locations along an agricultural slope on 28 occasions between July 2013 and October 2014. SWSs values were then calculated for the 0-0.4, 0.4-0.8, 0.8-1.2, 1.2-1.6, and 0-1.6 m soil layers. The SWS exhibited strong temporal stability, with mean Spearman's ranking coefficients (rs) of 0.83, 0.92, 0.83, and 0.79 in the 0-0.4, 0.4-0.8, 0.8-1.2, and 1.2-1.6 m soil layers, respectively. As expected, the most temporally stable location (MTSL1) accurately predicted the average SWS of the corresponding soil layer, and the values of absolute bias relative to mean (ARB) were lower than 3% for all of the investigated soil layers. Using TSA, deep-layer SWS information could be predicted using a single-location measurement in the 0-0.4 m soil layer. The mean ARB values between the observed and predicted mean SWS values were 2.9%, 4.3%, 3.9%, and 2.7% in the 0.4-0.8, 0.8-1.2, 1.2-1.6, and 0-1.6 m soil layers, respectively. The prediction accuracy of the spatial distribution generally decreased with increasing depth, with linear determination coefficients (R2) of 0.93, 0.79, 0.72, and 0.84 for the four soil layers, respectively. The proposed method could further expand the application of the temporal stability technique in the estimation of SWS.