The paper presents an investigation on the influence of land cover changes on rainfall variability, and in particular daily rainfall extremes in southern Africa. It examines the effect of increasing the amount of vegetation over the majority of southern Africa, using the Regional Climate Model (RCM) and General Circulation Model (GCM). The report presents the results for the desert run and savanna run cases. For the desert run case, results indicate that due to complete lack of vegetation, higher surface temperatures and a reduction in moisture are experienced. This results in a reduction in evaporation, which produces less vertical uplift. In response, an increase in near-surface pressure and associated anticyclonic winds is shown, and therefore a decrease in rainfall. The report further notes that the decreases are consistent regardless of whether the model is run in regional or global mode suggests that this result is not simply due to biases within the RCM lateral boundary data, and that precipitation changes resulting from localised vegetation change are not being overridden by larger scale atmospheric circulation features as simulated by the GCM. In the savanna run (using the RCM only), an increase in rainfall is suggested over regions where an increase in vegetation was imposed. The report notes that the results suggest that the relative increase in vegetation produces more available moisture at the surface. Due to the resulting increase in evaporation, more latent heat is released and therefore higher surface temperatures are again shown. This warmer surface, and increase in available moisture, may produce an increase in vertical motion, giving lower near-surface pressure and therefore an increase in rainfall. The report notes that going by the results, and theoretical considerations, it is proposed that the model’s ability to simulate rainfall in response to vegetation changes can be used as an indicator of the model’s ability to simulate the atmosphere-land surface feedback. This is because both rainfall extremes and the atmosphere-land surface feedback are dependent on latent heat fluxes, as opposed to global mean rainfall which is thought to be constrained by the energy budget of the troposphere. It recommends that:
future work is needed to continue to assess the atmosphere-land surface feedback, in terms of how different models are able to simulate changes in rainfall
more research is needed to improve upon these experiments, applying the vegetation anomaly at an appropriate time of the year
an improved understanding of extreme rainfall and its controls is crucial for increasing resilience and the adaptive capacity of society to these events.