Necessity of incorporating realistic land surface parameters for trend analyses of potential evapotranspiration and drought

Abstract

Potential evapotranspiration (PET) is a necessary input for various offline drought indices and, therefore, crucial for accurately evaluating drought conditions. Since representing land surface characteristics is needed for estimating PET, studying the impacts of land surface parameters on PET and offline drought indices is significant. Therefore, this study used the Food and Agriculture Organization of the United Nations’ report 56 (FAO-56) Penman-Monteith equation with the default and realistic land surface parameters to conduct two simulations worldwide from 1983 to 2020, based on the Standardized Precipitation-Evapotranspiration Index (SPEI) and various inputs (e.g., land use/cover, canopy height, and meteorological dataset). Then, the effects of land surface parameters on changes in PET and drought conditions were quantified by comparing these two simulations. Results suggested that in comparison to the default parameters, the realistic parameters-based results showed a more substantial and significant increase in annual PET, decrease in annual 3-month SPEI (SPEI3; i.e., drying), increase in drought duration (DD) and drought area (DA), and intensification of drought intensity (DI) globally. Despite the spatial similarity of the changes in annual PET and drought conditions, there were evident regional variations between the two simulations. The default parameters-based simulation showed weaker, stronger, and opposite trends in annual PET over 74.1 %, 16.7 %, and 9.2 % of the world, respectively, compared to the realistic parametrizations-based simulation. The default parameters could lead to weaker (more robust) drying and wetting over 50.9 % (36 %) of the world and opposite trends of annual SPEI3 over 13 % of the world. Moreover, the default parameters-based DD (DI) trends showed more than 46 %, 36 %, and 14 % of the world, respectively, with weaker, stronger, and opposite signs. Overall, our analyses highlight the importance of realistically representing the land surface characteristics for accurately capturing changes in PET and drought conditions.