Abstract
Evapotranspiration (ET) estimation is important for understanding energy exchanges and water cycles. Remote sensing (RS) is the main method used to obtain ET data over large scales. However, owing to surface heterogeneities and different model algorithms, ET estimated from RS products with different spatial resolutions can cause significant uncertainties, whose causes need to be thoroughly analyzed. In this study, the Surface Energy Balance Algorithm for Land (SEBAL) model was selected to explore spatial resolution influences on ET simulations. Three satellite datasets (Landsat Thematic Mapper (TM), Moderate Resolution Imaging Spectroradiometer (MODIS), and Advanced Very High-Resolution Radiometer (AVHRR)) were selected to independently estimate ET in SEBAL model to identify the influence of the spatial scale on ET estimation, and analyze the effects and causes of scale aggregation. Results indicated that: (1) the spatial distributions of ET estimated from the three satellite datasets were similar, with the MODIS-based ET having the largest uncertainty; and (2) aggregating input parameters had limited changes in the net radiation and soil heat fluxes. However, errors in the sensible heat and latent heat fluxes were relatively larger, which were caused by changes in the selection of hot and cold pixels and the NDVI and surface albedo parameters during scale aggregation. The scale errors caused by the model mechanisms were larger than those caused by the land use/cover pattern in the SEBAL model. Overall, this study highlights the impact of spatial scale on ET and provides a better understanding of the scale aggregation effect on ET estimation by RS.
Highlights
It has been widely recognized that many hydrological processes and energy exchanges are scale-dependent [1,2,3,4]
In order to check the consistency of the Surface Energy Balance Algorithm for Land (SEBAL) performance for the three datasets, we compared the estimated ETdaily with the observed ETdaily from the eddy covariance data by extracting the pixel value corresponding to the geographic location of the flux tower site
The ET estimated using multiple sensors (ETLandsat, ETMODIS, and ETAVHRR) with different spatial resolutions were compared in this study
Summary
It has been widely recognized that many hydrological processes and energy exchanges are scale-dependent [1,2,3,4]. Scaling issues impact our ability to accurately model the exchanges of water and energy across the surface–atmosphere interface [5]. Several studies have illustrated that scale changes in hydrological models produce results that are considerably biased due to land surface spatial heterogeneities (e.g., variations in topography, land use/cover, and soil properties) [6,7,8,9]. Evapotranspiration (ET) is a key linchpin of the Earth’s hydrological cycle and energy balance system [10]. The accurate estimation of regional ET is crucial for water resources management, agricultural production, and ecosystem protection, in water-scarce regions [11,12,13].
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
