Towards a High Resolution Soil Moisture Map of Oklahoma.

Large scale (> 1 km²) estimates of soil moisture have been validated by using data from single points of long-term networks with sparse soil moisture measurements and/or by using data from short-term experiments with dense soil moisture measurements. Long-term soil moisture data that match or can be scaled to match the spatial resolutions of large scale (e.g. satellite and model) estimates are needed. Our goal is to produce daily soil moisture maps of Oklahoma at quarter-section (~800 m) resolution that are suitable for not only scientific purposes — validation of modeled and remotely sensed soil moisture, for example — but also for operational purposes — weather forecasting, land management, etc. Initial results towards this goal have been developed by combining Oklahoma Mesonet data with simple physical and spatiostatistical models. The Oklahoma Mesonet is a network of 120 automated weather and soil monitoring stations covering the entire state of Oklahoma at an average spacing of one station every 30 km. Mesonet soil moisture measurements are only taken under grassland conditions, so we must also account for other land cover conditions. We have developed a soil moisture map that estimates plant available water (to a depth of 80 cm) for both grass and wheat land covers. The mapping method involves estimating soil moisture under (hypothetical) wheat at each Mesonet site, then combining maps of "normal," grassland Mesonet soil moisture measurements and maps of wheat soil moisture estimates. Maps will be developed further by adding other relevant data sources (soil property maps, radar precipitation estimates) and by using more advanced spatiostatistical methods. We will also incorporate findings from a roving cosmic-ray soil moisture sensor ("COSMOS rover") in the development and validation of our soil moisture maps.