Toward confident and useful climate change projections: West Africa and the Sahel [INVITED]

Analysis of AOGCMs run for the most recent IPCC report (AR4) introduced uncertainty about how West Africa and the Sahel will change as climate warms. Using annual or seasonal precipitation to characterize climate change, most of the AOGCMs predicted relatively little change for the region, but disagreed on the sign of the change. Simulations with one outlier model produced extreme drying, and another produced extremely wet conditions.

Projections for 2081-2100 using a 9-member ensemble of regional model simulations constrained on their lateral boundaries and at the surface by anomalies from 9 different AR4 AOGCMs produce a much higher level of agreement despite, for example, differences in future projections of SSTAs. A process study indicates that overall ocean warming produces the first-order response across the Sahel and West Africa, except near the west coast. Agreement among the ensemble members, validation of present day climate features and dynamics, and process studies designed to illuminate the physics of climate change contribute to confidence in the projections. Seasonal- and annual-mean changes in precipitation are similar to those of most of the AR4 AOGCM projections in the sense that they do not reveal an exceptionally strong climate change signal, but an examination on sub-seasonal time scales reveals stark changes in precipitation patterns that are important for agriculture and increases in heat indices across the Sahel that would threaten health. For example, rainfall events in the 21st c. over West Africa have rainfall intensities that are roughly double those of the 21st c. over West Africa in late summer and fall, with numerous individual events that are unprecedented in the 20th c. climate.

While climate change projections for the end of the century can produce clear anomalies that are more satisfying for scientific analysis, projections for the next 20-50 years are more actionable for policymakers. Continent-wide, 90-km simulations for 2030-2060 with a regional model are constrained with anomalies from an average of AOGCM simulations, and used to identify regions of potentially significant climate change by mid-century. Additional integrations place high-resolution (up to 10 km) nested grids over the active regions to produce projections and study climate change processes on the mesoscale. Analysis of changes in growing season length, heat wave incidence, and drought for West Africa and the Sahel are presented.