Large-scale origins of rainfall and temperature bias in high-resolution simulations over southern Africa

We use the Abdus Salam International Centre for Theoretical Physics (ICTP) regional climate model, RegCM3, nested in NCEP and ERA-Interim reanalyses (NC-RegCM and ERA-RegCM, respectively) to explore the effect of large-scale forcings on the model biases over a southern Africa domain at 25 km grid spacing. RegCM3 shows a generally good performance in simulating the location of the main rainfall features, temperature and synoptic scale circulation patterns, along with cloud cover and surface radiation fluxes. However, it shows a wet bias which varies substantially when fields from the 2 reanalysis products are used to drive the model, being greater in NC-RegCM than in ERA-RegCM. The wetter bias in NC-RegCM originates from larger moisture inflow, amplified cloud cover and upward motion caused by stronger low-level convergence. Similarly, negative temperature biases are present over most of the land areas, and the wetter NC-RegCM exhibits larger magnitudes of temperature bias. Surface flux analysis reveals that these lower temperature values in the NCEP-driven experiment are due primarily to the surface latent heat flux rather than cloud radiative forcing. As a result, the hydrologic cycle is more intense in the NC-RegCM than in the ERA-RegCM. Our results are relevant for a better understanding of the propagation of errors from the large-scale forcings to the regional model over the region of interest.