Application of SEBAL model for mapping evapotranspiration and estimating surface energy fluxes in south-central Nebraska

Knowledge of spatiotemporal distribution of evapotranspiration (ET) on large scales, as quantified by satellite remote sensing techniques, can provide important information on a variety of water resources issues such as evaluating water distributions, water use by different land surfaces, water allocations, water rights, consumptive water use and planning, and better management of ground and surface water resources. The objective of this study was to assess the operational characteristics and performance of the surface energy balance algorithm for land (SEBAL) model for estimating crop ET (ETc) and other energy balance components, and mapping spatial distribution and seasonal variation of ETc on a large scale in south-central Nebraska climatic conditions. A total of seven cloud free Landsat Thematic Mapper (TM)/Enhanced Thematic Mapper (ETM) satellite images (May 19, June 20, July 22, August 7, September 8, September 16, and October 18, 2005) were processed to generate ETc maps and estimate surface energy fluxes. Predictions from the SEBAL model were compared with the Bowen ratio energy balance system (BREBS)-measured fluxes on an instantaneous and daily basis. The ETc maps generated by the model for seven Landsat overpass days showed a very good progression of ETc with time during the growing season in 2005 as the surface conditions continuously changed. The predictions for some surface energy fluxes were very good. Overall, a very good correlation was found between the BREBS-measured and SEBAL-estimated ETc with a good r(2) of 0.73 and a root-mean-square difference (RMSD) of 1.04 mm day(-1). The estimated ETc was within 5% of the measured ETc. The model was able to predict growing season (from emergence to physiological maturity) cumulative daily corn ET reasonable well within 5% of the BREBS-measured values. The model overestimated the surface albedo by about 26% with a RMSD of 0.05. The difference between the measured and predicted albedo was the greatest on May 19, early in the growing season before the full canopy cover. The second largest difference between the two albedo values was on October 18, a day after harvest. The model significantly under predicted soil heat flux with a large RMSD of 80 W m(-2) and most of the underestimation occurred in the late growing season. Local calibration of soil heat flux significantly improved the agreement between the measured and predicted values. Furthermore, the sensible heat flux was underestimated between September 20 (after physiological maturity) and October 18 (a day after harvest). While our results showed that SEBAL can be a viable tool for generating ETc maps to assess and quantify spatiotemporal distribution of ET on large scales as well as estimating surface energy fluxes, its operational assessment for estimating sensible heat flux and ETc, especially during the drier periods for different surfaces, needs further development.