Foot-print Scale Real-time Modeling Tool of Surface Energy Fluxes, Evapotranspiration, Soil Moisture, and Soil Temperature: Application in the the Southern Great Plains

Abstract:

A benefit of training a process-based model is the capacity to use it as a complement to standard weather stations to estimate evapotranspiration (ET), energy fluxes, soil temperature, and soil moisture. Additionally a real-time modeling tool becomes an accessible option in response to the high cost and complexity of surface energy budget (SEB) field measurements that limit the availability and spatial distribution these data. This study evaluates the capability and transferability of a physically-based hydrological model, that uses dynamically changing vegetation parameters derived from remotely sensed imagery to simulate the temporal dynamics (1-hour resolution) over different locations in Oklahoma with crops and grasses as dominant land cover type. The tool is aimed to serve as a "virtual tower" for real-time energy, evapotranspiration soil moisture and temperature estimations. The calibration and validation of the model was performed using the Eddy Covariance Tower (ECT) observations available from the FluxNet network at the US-Arm Lamont (crops) and the MOISST-Marena (grassland) eddy covariance sites in north-central Oklahoma. The data to parametrize the model in regards to vegetation and its properties (e.g. canopy capacity, stomatal resistance and albedo) were obtained from remote sensing sources (i.e. MODIS MCD15A3H, MCD43A). The transferability of static soil parameters satisfactorily resulted in high modeling scores at the two test (non-calibrated) sites (US-A32 and US-A74). Uncalibrated simulations provided confidence on the model as a real-time virtual tool that is able to capture diurnal and annual cycles of the simulated variables with minimal, standard input forcing. Future work includes testing the model over different types of vegetation and ecosystems, including those with complex terrain.