Land surface effects on heavy precipitation over subtropical South America

ERNESTO H. BERBERY; OMAR V. MÜLLER

Melbourne

Congreso; IUGG 2011 General Assembly. Earth on the Edge: Science for a Sustainable Planet; 2011

International Union of Geodesy and Geophysics

Subtropical South America is a region where typical climate models have difficulties in reproducing the pattern or magnitude of precipitation, and the reasons are not fully understood. One possible explanation (although not the only one) may be due to changes in surface conditions not represented in models. Vast areas of South America are suffering from human-induced changes in land cover that may affect ecosystem-climate feedbacks, with deforestation and land-clearing for agriculture and cattle ranging being the most important ones. Land-clearing produces an increase in albedo, a reduction of transpiration, and a net release of CO2. On the other hand, other extensive land-use changes in South America, such as grassland afforestation, produce a decrease in albedo, a rise of evapotranspiration, and greater surface roughness.These kinds of ecosystem-climate feedbacks are variable in time, but many climate models use fixed land cover types that are insensitive to interannual variability of the surface states. In this research, we use Ecosystem Functional Types (EFTs) to develop a consistent set of the land surface physical properties to represent the interannual variability of the land states in southern South America and propose a method to replace the traditional land-cover types in regional climate models by the time-varying EFTs. Model simulations with conventional land cover types were compared against simulations using EFTs as lower boundary conditions. The results show that the use of EFTs seems to better represent the changes in land-atmosphere interactions and thus to reduce the precipitation biases in regions of heavy rainfall.