The effects of land-surface heterogeneities on regionalclimate: A sensitivity study

Summary ?In order to further understand the land–atmosphere interactions and increase the predictability of climate models, it is very important to investigate the effects of land-surface heterogeneities. In this paper, we considered roughness-length and stomatal-resistance heterogeneities in the regional climate model RegCM2 (Giorgi et al., 1993) that employs BATS (Dickinson et al., 1993) as the land surface scheme. In representing the subgrid heterogeneities, a computationally efficient method, which is a combination of the mosaic approach and the analytical type of the statistical-dynamical approach, is applied. The method is also characterized by converting the probability distribution of fundamental variables to probability distributions of derived quantities. By using the 3-month observational data of 1991 Meiyu season over China, we conducted coupled-model experiments, and found that: (i) For the whole model domain, the consideration of the two heterogeneities, in which intrapatch variability plays a very important role, greatly affects the simulations for the surface flux, wind, temperature and precipitation fields. (ii) The temperature and heat fluxes are quite sensitive to the heterogeneities, which displays the following rule: for a sub-region, the mean sensible heat flux decreases, the mean latent heat flux increases, and the mean surface temperature decreases with the increase of the heterogeneities. Furthermore, the mean latent heat flux is more sensitive to the heterogeneities than the mean sensible heat flux. (iii) It seems that the influence of stomatal-resistance heterogeneity on the latent heat flux is greater than that of roughness-length heterogeneity. Therefore, it is necessary to appropriately represent subgrid land-surface heterogeneities so as to improve regional climate modeling. Received June 19, 2001; Revised January 21, 2002