Projecting Regional Climate Change in the Tropics

With much of the world’s population that is most vulnerable to current and near-future climate change residing in the topics, confident projections on regional space scales are greatly needed. Computational resources limit the horizontal resolution of GCMs for these projections, especially when ensemble simulations are needed to distinguish between climate trends and natural variability, to evaluate the sensitivity to assumptions, and to build confidence through the understanding of physical processes. Regional climate models use the same set of governing equations as GCMs, but their application over a limited area allows for higher resolution and the optimization of parameterization choices for local climate. However, these advantages are counterbalanced by a certain loss of global connectivity. Methodologies for designing regional model simulations for tropical applications are evaluated, including the specification of resolution, domain, parameterizations, initial conditions, lateral and surface boundary conditions, and various choices for generating ensembles.

Several examples of the use of regional modeling to analyze observed trends and project greenhouse gas-induced climate change over Africa on space and time scales useful for planning and adaptation are presented. These include an evaluation of the observed and projected precipitation increases in the Sahel, changes in growing seasons in Malawi, and the potential for changes in extreme events across the African continent.