Estimating spatially distributed cotton water use from thermal infrared aerial imagery

High resolution thermal infrared (TIR) observations of irrigated lands have the potential to retrieve spatially distributed estimates of evapotranspiration (ET) and thereby assess crop stress and refine water scheduling. To assess this possibility we use a remote sensing data set from a 2003 central Arizona cotton experiment in combination with ground-based observations to estimate instantaneous ET. The spatially distributed estimates are compared with those derived from soil neutron probe water observations. We retrieve ET from TIR observations using a two source energy balance (TSEB) approach, which models surface energy fluxes from distinct soil and vegetation sources. This distinction of energy sources is important for heterogeneous and sparsely vegetated surfaces since soil and plant ET processes are significantly different. TSEB requires radiometric surface temperatures, vegetation densities from Normalized Difference Vegetation Index (NDVI) and near surface meteorological observations. Incorporation of TIR observations into other remote sensing surveys could help realize near-real-time water use monitoring at local and regional scales. Copyright ASCE 2005.