Multiscale Atmospheric Simulations over a Complex Terrain: Surface Variability and Land-Atmosphere Exchange

This study highlights the impacts of the surface variability on Land-Atmosphere interactions using multi-scale atmospheric simulations with the Weather Research and Forecasting model capacities, coupled with NOAH land surface model and Kusaka et al. (2001) urban canopy model. The 30-m resolution numerical datasets for land-use and topography are respectively given by National Land-Cover Dataset (NLCD 2001) and Shuttle Radar Topography Mission (SRTM). The finest simulation is performed with a grid size of 50 m, to study near-surface atmospheric dynamics and surface energy budget for a perfectly cloud-free day: atmospheric pressure changed within the day from 1015 hPa to 1019 hPa, recorded 2-m temperatures varied from 10°C to 27°C, specific humidity varied between 6 and 8 g/kg, while wind speed remained below a velocity of 9 m/s. Despite identical initial and boundary conditions (soil type, land use and soil moisture) as well as identical mesoscale meteorological forcing and radiative forcing, the real-case large eddy simulation indicates a large discrepancy of computed surface fluxes within the same land-use type. These discrepancies of surface fluxes, as large as 140W/m variation for water bodies, can be directly attributed to the surface variability around each grid cell imputed to the heterogeneit