Soil Moisture Content in 3D + Time: Visualization and Modeling at the Field-Scale.

Comprehension of soil moisture dynamics at the field scale requires measurements with high spatial and temporal resolution. We installed a soil moisture sensor network on a dryland Long-Term Agro-Ecosystem Research farm stationed on complex terrain in the Palouse region, which hosts variable soils, microclimates, and landscape positions. At 42 geo-referenced locations distributed across the 37 ha farm, sensors were installed at five depths (30, 60, 90, 120, and 150 cm). Volumetric soil moisture content was recorded each hour by buried data loggers and downloaded via radio transmission. We visualized soil moisture dynamics in Google Earth using the plotKML package in R. With regression-kriging, we produced predictions of soil moisture across space and through time. These predictions can assist in exploring relationships between soil moisture and soil physical properties, cropping systems, and landscape positions. Furthermore, soil moisture predictions can integrate with biophysical models and decision support tools for informing precision management practices.