Statistical downscaling for precipitation projections in West Africa

Andrew Polasky, Jenni L. Evans, Jose D. Fuentes

Research output: Contribution to journalArticlepeer-review

Abstract

The West Africa region (5 to 20 N and 10 E to 20 W) is particularly vulnerable to climate change due to a combination of unique geographic features, meteorological conditions, and socio-economic factors. Drastic changes in precipitation (e.g., droughts or floods) in the region can have dramatic impacts on rain-fed agriculture, water availability, and disease risks for the region’s population. Quantifying these risks requires localized climate projections at a higher resolution than is generally available from general circulation models. Using self-organizing maps, we produce station-based downscaled precipitation projections for medium and high-emission climate scenarios for this region. Compared to historical observations, the downscaled values are able to match the historical range of the distribution, and recreate the seasonal variability for the inland portions of the region, but struggle with the seasonal cycle along the coast. We find a decrease in the interior Sahel region by an average of 10% by 2100 under the high greenhouse gas-emission scenario of Shared Socioeonomic Pathway 5 - 8.5. Precipitation decreases in the Sahel are primarily driven by reductions in the number of rainy days during the wet season, rather than by consistent decreases in the magnitude of the precipitation amounts or decreases in the average length of the wet season.

Original languageEnglish (US)
Pages (from-to)327-347
Number of pages21
JournalTheoretical and Applied Climatology
Volume155
Issue number1
DOIs
StatePublished - Jan 2024

All Science Journal Classification (ASJC) codes

  • Atmospheric Science

Cite this