Parametrization of extremely heterogeneous land-surface processes

The land surface plays a crucial role in the climate system, significantly influencing the exchanges of energy, mass, and momentum among the atmosphere, biosphere, and lithosphere. While land-surface processes in homogeneous terrains are well understood and effectively integrated into the parameterization schemes of existing weather models, our understanding of these processes in extremely heterogeneous regions remains insufficient. This gap in knowledge limits our capacity to accurately parameterize land-surface interactions in such areas. Extremely heterogeneous surfaces are characterized by a variety of soil types and pronounced orographic features, such as mountains or steep slopes.

State-of-the-art weather models commonly utilize the Monin-Obukhov similarity theory (MOST) for parameterizing surface momentum, heat, and moisture fluxes. However, these similarity functions are based on empirical data obtained from field campaigns conducted in homogeneous environments. When these functions are applied to extremely heterogeneous regions, they can produce large biases between modeled and observed surface sensible or latent heat fluxes. Furthermore, in large-eddy simulations (LES), the underlying assumptions of MOST - such as horizontal homogeneity and stationarity - are often violated. Additionally, inconsistencies arise between the fluxes calculated using subgrid closure schemes and those derived from MOST in the surface layer.

To tackle these challenges, we propose an alternative approach that circumvents the use of MOST for parameterizing surface fluxes. In land-surface-parameterization schemes, surface fluxes are often determined using resistance networks. Instead of estimating these resistances using MOST, our aerodynamic resistance approach (ARA) uses the eddy viscosity/diffusivity calculated by the subgrid closure schemes.

First tests in idealized large-eddy simulations (LES) using the Weather Research and Forecasting model (WRF) show that the ARA-calculated surface fluxes are more consistent with the subgrid closure calculations than the MOST-derived fluxes. Next, the ARA will be tested in real-case simulations of the Tengchong site (China) on the Tibetan plateau which is known for its heterogeneous landscape. Moreover, the simulation results will be compared to observational data which has been available at the site for more than 12 years.