Submitted to: Irrigation Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: May 29, 2008
Publication Date: November 1, 2008
Citation: Chavez Eguez, J.L., Neale, C.M., Prueger, J.H., Kustas, W.P. 2008. Daily evapotranspiration estimates from extrapolating instantaneous airborne remote sensing ET values. Irrigation Science. 27:67-81. Interpretive Summary: Remote sensing of land surface energy balance provides essentially instantaneous estimates of latent heat flux or evapotranspiration. These estimates are used in the prediction and monitoring of spatially distributed daily crop water use. However, instantaneous evapotranspiration values are of relative importance unless they can be employed to predict daily crop water consumption. Therefore, the determination of an accurate method for extrapolating remote sensing instantaneous evapotranspiration estimates to daily values is imperative. In this study, several extrapolation methods proposed in the literature were evaluated using multispectral airborne remote sensing imagery, weather data, and eddy covariance energy balance systems. Results indicated that daily crop water use estimates based on the evaporative fraction method better compared to measured values. Errors were below 10% on corn and soybean fields grown under rainfed agricultural conditions.
Technical Abstract: In this study, six extrapolation methods have been compared for their ability to estimate daily crop evapotranspiration (ETd) from instantaneous latent heat flux estimates derived from digital airborne multispectral remote sensing imagery. Data used in this study were collected during an experiment on corn and soybean fields, covering an area of approximately 12 km x 22 km, near Ames, Iowa. ETd estimation errors for all six methods and both crops varied from -5.7 ± 4.8 % (MBE ± RMSE) to 26.0 ± 15.8 %. Extrapolated ETd values based on the evaporative fraction (EF) method gave best estimates of measured values. This method reported an average ETd estimate error for corn fields of -0.3 mm d-1, with a corresponding error standard deviation of 0.2 mm d-1, i.e., about 5.7 ± 4.8 % average under prediction when compared to average ETd values derived from eddy covariance (EC) energy balance systems. A solar radiation-based ET extrapolation method performed relatively well with ETd estimation error of 2.2 ± 10.1 % for both crops. An Alfalfa reference ET-based extrapolation fraction method (ETrF) worked better for soybean fields, yielding an overall ETd overestimation of about 4.0 ± 10.0 % for both crops. It is recommended that the average daily soil heat flux not be neglected in the calculation of ETd when utilizing the EF method EF. These results validate the use of the airborne multispectral RS-based ET methodology for the estimation of instantaneous ET and its extrapolation to ETd. In addition, all methods need to be further tested under a variety of environmental and climatological conditions, fundamentally method ETrF that showed larger biases under rainfed agriculture.