Modeling Transient Soil Moisture Dichotomies Arising from Anthropogenic Land Cover Alterations.

The majority of the environmental monitoring networks around the world have been installed under grassland vegetation. However, the distinct vegetative dynamics of the different land covers can impose limitations to extrapolation of soil moisture products, and no methods that bridge the soil moisture regimes from one land cover to another are currently available. The objectives of our study were to: i) compare the soil moisture regimes between naturally occurring sod vegetation surrounding the Oklahoma Mesonet stations and soil moisture under wheat cropland, and ii) predict the soil moisture condition under wheat cropland from observed soil moisture under grassland vegetation. Observed soil moisture under grassland vegetation was obtained from the Oklahoma Mesonet. Soil moisture under the hypothetical scenario of having wheat cropland at each Mesonet station was simulated using the dual crop coefficient (dual Kc) soil water balance model. Root-zone (i.e. top 80 cm) fraction available water capacity (FAW) was estimated for both perennial warm-season grassland and annual winter wheat. Soil moisture dynamics were compared for 79 Mesonet stations that monitor soil moisture at 5, 25, and 55 cm depth. Results show that soil moisture regimes of winter wheat and grassland are intermixed at a spatial scale of <1 km, making them difficult to distinguish with existing soil moisture satellites, and limiting Mesonet soil moisture products as a direct surrogate of wheat soil moisture condition. Wheat FAW can be estimated from observed soil moisture under grassland, information that can be assimilated to improve: drought monitoring systems, calibration of remote sensing soil moisture products, and hydrological and crop models.