Impact of the land surface forcing on land-atmosphere feedback in the convection permitting modeling

The evaluation of the first of its kind multi-model convection permitting - regional climate model (CP-RCM) ensemble, produced within the framework of the international CORDEX - Flagship Pilot Study on Convective phenomena over the Mediterranean (CORDEX-FPS-CEM), showed the improved spatial representation, frequency and extremes of precipitation compared to coarser resolution counterparts (Ban et al., submitted 2019). It is important, though, to keep in mind that the quality of each of such high resolution simulations strongly depend on the quality of the forcing used to drive the model.

In this work we investigate the impact of the land-surface forcing on the model representation of land-atmosphere (LA) feedback. For that we used the 2 Weather Research and Forecasting (WRF) settings from the CORDEX-FPS-CEM ensemble to run 4 simulations for 2 seasons (summer and fall), combining 2 sets of data for the land cover (MODIS and CORINE) and the top soil texture (FAO and HWSD). In such a way we generated two ensembles with 8 simulations for each season. Additionally we run 4 simulations with perturbed starting date for the summer season to obtain an additional 5-member ensemble to investigate the impact of the internal variability on the final results.

The objective of this study is to investigate the model sensitivity to (1) land-surface static forcing, (2) season, and (3) the model configuration. To quantify the strength of LA coupling for each grid, we use coupling metrics appropriate for seasonal time scales. We applied mixing diagram approach to investigate the impact of the land-surface changes on the boundary layer evolution. We also investigate the impact of these changes on the potential for convection triggering, and we calculate correlation matrices to quantify the impact on the strength of the land-atmosphere coupling.

Preliminary results show evident effects of the land surface changes on surface variables, boundary layer evolution, atmospheric stability and humidity in the lower atmosphere. Strength of the sensitivity to a specific change in the land-surface forcing depend on the model configuration: WRF with less sophisticated parameterization schemes is more sensitive to land use changes, while more sophisticated configuration shows higher sensitivity to the soil texture changes in representing boundary layer evolution. Furthermore, the WRF shows stronger coupling and therefore stronger sensitivity to the land surface changes in the summer season.

Acknowledgement: This work is partially funded by the Spanish Government R+D programme through grant INSIGNIA (CGL2016-79210-R) co-funded by the ERDF/FEDER.

References: Ban N., and coauthors: The first multi-model ensemble of regional climate simulations at kilometer-scale resolution, Part I: Evaluation of precipitation, (submitted to Climate Dynamics, 2019)