The role of land-surface interactions for surface climate in the EC-Earth3 earth system model

Abstract. Land-surface conditions have prominent effects on local and regional climate through the exchanges of energy and moisture with the atmosphere and on global climate by the exchanges of carbon dioxide. Therefore, it is important that land-surface components of earth system models (ESMs) like EC-Earth3 can simulate the processes governing the energy and water cycles and the carbon cycle realistically. The aim of this study is twofold, first to evaluate the quality of simulation of surface climate by the land-surface component of the EC-Earth3 ESM, combining the HTESSEL land-surface model and the LPJ-GUESS dynamical vegetation model, and second to assess the role of the coupling of the land surface with the atmosphere for the simulation of the surface climate in EC-Earth3. To this end, two simulations with different configurations of the EC-Earth3 ESM are considered: an offline simulation with HTESSEL+LPJ-GUESS forced with meteorological data from the ERA5 re-analyses and a simulation with the atmospheric component of EC-Earth3, where the land-surface conditions (soil moisture and vegetation characteristics) are prescribed from the offline simulation. The land-surface component of EC-Earth3 is characterized by marked regional biases in various aspects of surface climate. These are, for instance, too warm land-surface temperatures in the tropics and in the mid- and high latitudes of the Northern Hemisphere, resulting in a warm overall bias. Surface soil moisture, on the other hand, is characterized by a dry bias in the subtropics and parts of the extra-tropics and a wet bias in the tropics and the eastern part of Asia, resulting in a slightly negative overall bias. The incoming net radiation is underestimated by the model over much of the global land area, causing a negative overall bias. For the fluxes of sensible heat, the model also shows a negative overall bias with a clear tendency to underestimate the sensible heat fluxes in regions, where they are relatively strong, and underestimate them in regions where they are rather weak. The biases in the fluxes of latent heat generally correspond to the biases in the sensible heat fluxes (with opposite sign) with an underestimation of the fluxes of latent heat in regions where the sensible heat fluxes are too strong and an overestimation in the regions where the sensible heat fluxes are too weak. The coupling with the atmosphere leads to somewhat stronger biases in the aspects of surface climate considered in the study. The most pronounced effect of the coupling is found for land-surface temperature, including a change in the sign of the overall bias from a warm overall bias in the simulation with the land-surface component to a considerable cold bias in the atmospheric component of EC-Earth3. For surface soil moisture, the coupling with the atmosphere changes a dry overall bias of the land-surface component to a wet bias in the atmospheric model. Analysing the correspondence between the global patterns for the simulations and the reference data reveals relatively large effects of the atmospheric coupling on land-surface temperature as well as on net radiation and sensible heat flux but small effects on surface soil moisture and latent heat flux.