Remote sensing-based soil water balance to estimate mediterranean holm oak savanna (dehesa) evapotranspiration under water stress conditions

Abstract This paper aims to present the use of a remote sensing-based soil water balance to estimate holm oak woodland evapotranspiration (ET). The model is based on the assimilation of MODIS reflectance-based vegetation indices in the dual crop coefficient methodology. A daily water balance was performed on the root zone soil to estimate plant water stress. The methodology was evaluated with respect to the actual ET measured by eddy covariance in Mediterranean holm oak savanna (dehesa) for five consecutive years (2004 to 2008). The model adequately reproduced the absolute values and tendencies measured at daily and weekly periods. Root mean square error (RMSE) was 0.50 mm/day for daily values and 2.70 mm/week for weekly accumulated values. The analysis demonstrated the presence of a long period of water stress during the summer and at the beginning of fall. Measured ET dropped during these periods, and the model replicated this tendency accurately, reaching a stress coefficient value close to 0.2. To be operative, the proposed method required low ground data (reference evapotranspiration and precipitation) and the results indicated a simple, robust method that can be used to map ET and water stress in the dehesa ecosystem.