Evaluation of Surface Energy Balance System (SEBS) Model for Estimation of Evapotranspiration in Eastern Ethiopia

Abstract

Quantification of evapotranspiration (ET) is needed for a wide array of applications especially for water resources management. ET estimation by using either ground based direct measurement or modeling by using in-situ based meteorological variables is found cost inefficient or for some geographic areas it might be impractical. This issue is specially challenging in data scarce regions. This study evaluates remote sensing based surface energy balance system (SEBS) in arid and semi arid eastern Ethiopia. After undertaking a series of pre-processing and processing phases, surface parameters has been generated from MODIS level 1B products for dry season and dekadal composites from spot vegetation, down welling surface short web flux from LSA SAF and near surface meteorological variables from GLDAS has been utilized. By ingesting these parameters at prepackaged SEBS model, ET has been estimated. For comparative analysis, MSG ET with FAO-PM has been utilized. Our findings have shown that estimated ET ranges from 0 to 10 mm day ^-1 with large spatial variability. In areas that are well vegetated, SEBS provides larger values both at dry and wet seasons but at arid and semi-arid vast expanses of the study site there is found very lower ET value at dry season and increases with vegetation regeneration in the summer season. It was observed that at both dry and wet seasons SEBS ET values were over estimated than MSG ET product and underestimated than FAO-PM values. These under and over estimations were very large at arid and semi-arid areas which is attested from point based analysis undertaken using meteorological stations located at different climatic characteristics. Our study confirmed that, though SEBS model provides reasonable results it is sensitive for vegetation cover of a site. Taking these into consideration there should be comparative analysis of the model with direct field measurements for more certainty.

Keywords: Evapotranspiration, Eastern Ethiopia, SEBS, Remote sensing