skip to main content

Estimation of Evapotranspiration of Rangeland Cover Using SEBAL Algorithm in Robat Mahidasht Region, Kermanshah, Iran



Rangelands are the most important plant ecosystems in Iran that have multiple and vital roles in economic stability and food security in the country. In recent decades, with increasing population, increasing forage consumption, climate change and rainfall fluctuations, most of the country’s pastures have been destroyed or faced with a degradation trend. Therefore, in order to strengthen this vital ecosystem in the country, rangeland management components have a special priority and importance. One of the main components of rangeland management is the study on the water requirement of rangeland plants. The water required by the plant is equivalent to its evapotranspiration. Estimation of evapotranspiration using meteorological station data can be done at different time intervals; But determining its spatial distribution on a large scale is not possible. Remote sensing techniques and evapotranspiration estimation algorithms based on the surface energy balance of the earth are among the methods that are able to produce maps with appropriate temporal and spatial coverage. In this study, 4 images of LANDSAT 5 processed (from June to July of 2010), and evapotranspiration of the Mahidasht region, Kermansh province, Iran were estimated. Then, Sentinel 2 images were used to identify the existing rangelands in the Robat Mahidasht region by Maximum Likelihood classification method. The evapotranspiration of the rangelands was obtained uing Surface Energy Balance Algorithm for Land (SEBAL) maps. Based on the results, it was observed that the difference of estimating the actual evapotranspiration between the SEBAL algorithm and lysimetric measurements was a maximum value of 9.7%, which is acceptable. The coefficient of determination between SEBAL and lysimetric data was (R2=0.99) and mean absolute difference was 0.53 mm/day. The estimated evapotranspiration rates of rangeland cover at the four Landsat imaging dates were 2.1, 3.46, 3.4, and 3.44 mm/day. Also, the results showed less rangeland evapotranspiration than other coverings like forest and agriculture, which is due to the dryness of the topsoil profile and shallow depth of most of the rangeland plants, especially annuals species. It was concluded that SEBAL algorithm is a suitable method for estimating the evapotranspiration of rangeland cover with acceptable accuracy.

Graphical Abstract