Thermal and visible remote sensing for estimation of evapotranspiration of rainfed agrosystems and its impact on groundwater in SE Australia

Rainfed agrosystems are important components of the world's food production system and account for 65 - 95% of total agriculture. In contrast to irrigated production systems, relatively little attention has been paid to understanding the hydrological interactions between the components of rainfed agrosystems and their impact on water resources, especially groundwater. A new model, the Surface Energy Balance Algorithm for Rainfed Agriculture (SEBARA), has been developed to estimate the spatial pattern of evapotranspiration in these agrosystems using satellite images (thermal, infrared and visible spectra). The model was calibrated for two competing land uses (Eucalyptus globules tree plantations and pastures) in adjacent catchments in western Victoria, southeastern Australia. Using measurements from a flux tower in the pasture catchment and adjusted sapflow measurements in the plantation catchment, an estimation accuracy of 95% was achieved. The tree plantations had higher available net radiation, lower soil heat flux and higher latent heat flux, resulting in 15-20% higher evapotranspirative demand than the pasture, depending upon the age and canopy of plantations. The evapotranspiration rate of plantations declines where groundwater depth is >12m or where shallow groundwater is saline. The shallow root system of the pasture means that it relies solely on soil moisture to meet its water requirements and thus has lower evapotranspiration, which varies according to the pasture species.