Impact of Volcanic Aerosols on Sahel Rainfall

Abstract

Large volcanic eruptions have been one of the dominant sources of natural climate change throughout the course of the Earth's history. Among other effects, these explosive eruptions inject a cloud of sulfate aerosols into the stratosphere that scatter solar radiation back to space, cooling the Earth's surface and causing various climatic effects. Because of its climate, the semi-arid African Sahel region is particularly sensitive to volcanic radiative perturbations. The region is highly vulnerable to changes in annual precipitation, as most of the region's workforce is employed in the agricultural industry. The Sahel experienced extreme famine and economic collapse after droughts in the 1970s and 1980s. There is evidence to suggest that spatially asymmetric volcanic aerosol clouds produce different hydroclimatic responses based on their hemispheric symmetry, both globally and in the Sahel. In this thesis, volcanic eruptions are grouped into three categories: symmetric forcing, Northern Hemispheric forcing, and Southern Hemispheric forcing, exemplified in the respective hemispheric forcing structure of three specific volcanic eruptions: Mount Pinatubo in 1991, Santa Maria in 1902, and Mount Agung in 1963. We use data from a 30-ensemble experiment run on the Forecast-Oriented Low Ocean Resolution (FLOR) version of the Geophysical Fluid Dynamics Laboratory (GFDL) climate model CM2.5 to explore the regional responses to symmetric and asymmetric volcanic eruptions in the Sahel. Our results are consistent with the hypothesis that the meridional structure of volcanic forcing is order-one important to the West African rainfall response. We find that there is a decrease in annual Sahelian rainfall when forcing is stronger in the Northern Hemisphere, and an increase in Sahelian rainfall when forcing is stronger in the Southern Hemisphere. We then characterize the mechanisms behind these changes using an analysis of the moist static energy budget in the atmospheric column above the Sahel. Finally, we take advantage of the millennia of simulations and forcing data provided by the Community Earth System Model Last Millennium Ensemble (CESM-LME) to classify 46 large eruptions dating back to 870~CE into symmetry categories based on their forcing structure in order to assess the robustness and model-sensitivity of these results. Using the 13-member Last Millennium Ensemble to examine the response of Sahelian rainfall to these eruptions, grouped by their symmetry type, we find again that Northern Hemisphere eruptions tend to cause a drying of the Sahel, while Southern Hemisphere eruptions tend to cause a wetting, although the annual cycle response in the LME differs from that found in the FLOR experiments. The inter-model differences in our results likely arise from the ways the two models simulate the West African monsoon, illustrated in the differing control Sahel rainfall climatologies. In addition, FLOR and LME use different historical aerosol data. Our results highlight the need for accurate meridional structures in historic volcanic forcing data used for climate models as well as the need for further study on regional effects of hemispherically asymmetric radiative forcing, especially as they might pertain to aerosol geoengineering.