Spatiotemporal Calibration of the Electromagnetic Induction-Salinity Relationship with Anocova Modeling.

Saline soils are often found in arid and semi-arid farmland. Soil salinity (EC_e) should be routinely mapped and monitored, to assure long-term sustainability of farming practices in these areas. Field-wide estimations of EC_e can be obtained through soil sampling directed by geospatial measurements of apparent soil electrical conductivity (EC_a). We present a modeling approach that allows calibrating the EC_a-EC_e relationship across the same geographical region. The calibration, which is essentially an analysis of covariance (ANOCOVA) between EC_a and EC_e, can be extrapolated spatially and temporally to other fields in the same region by employing as few as three additional soil cores per new field/timing. Two case studies for California are presented showing how the ANOCOVA calibration approach can be used in a spatiotemporal framework to map and monitor soil salinity. For the spatial framework, we show an ANOCOVA calibration and (spatially unbiased) cross validation over a regional-scale database from 22 fields where EC_a and EC_e were measured at 267 locations. For the temporal framework, we show a temporal ANOCOVA calibration and (temporally unbiased) cross validation using data from a long-term (1999-2012) monitoring study at a 32.4-ha saline field where EC_a and EC_e were monitored over time at the same 40 locations. Our results suggest that this approach is reliable at low salinity values (i.e., where common crops can grow). The ANOCOVA approach is a significant step towards enabling the routine use of EC_a mobile sensor technology for inexpensive soil salinity monitoring at high spatial and temporal resolutions.