Testing a Dry Soil Layer scheme in SURFEX and Harmonie-Arome

Abstract

Total evapotranspiration (ET) is a key variable in Land surface/Earth System models, which needs to be accurately simulated in order to avoid excessive biases in simulated energy, water, and carbon cycles. Models, which usually simulate ET by formulating resistance-based parameterizations for plant transpiration and direct evaporation from the soil and the plant canopy, often overestimate ET on semi-arid environments. A new soil evaporation resistance parameterization in which soil evaporation is approximated as water vapour diffusion through a dry soil layer (DSL) following [1] is available in SURFEX, and has shown good performance in offline tests for two semi-arid sites from the LIAISE field campaign [2]. Recent Harmonie-Arome (H-A) tests activating DIF and MEB ISBA settings have shown a degradation in near surface humidity variables, leading to an overall dry bias, specially in areas with contrasting seasons where the soil gets quite dry in late spring and summer. In this study we test the DSL parameterization in H-A experiments over Spain.