Atmosphere (Jun 2022)

Performance of CMIP6 HighResMIP on the Representation of Onset and Cessation of Seasonal Rainfall in Southern West Africa

  • Francis Nkrumah,
  • Kwesi Akumenyi Quagraine,
  • Kwesi Twentwewa Quagraine,
  • Caroline Wainwright,
  • Gandomè Mayeul Leger Davy Quenum,
  • Abraham Amankwah,
  • Nana Ama Browne Klutse

DOI
https://doi.org/10.3390/atmos13070999
Journal volume & issue
Vol. 13, no. 7
p. 999

Abstract

Read online

Changes in rainfall onset and cessation dates are critical for improving decision making and adaptation strategies in numerous socio-economic sectors. An objective method of determining onset and cessation date is employed over Southern West Africa (SWA) in this study. The method is applied over 34 years of the quasi-global rainfall dataset from the Climate Hazards Group InfraRed Precipitation with Stations (CHIRPS) and five High Resolution Model Intercomparison Project (HighResMIP) model datasets under the Coupled Model Intercomparison Project Phase 6 (CMIP6) experiment. Generally, a strong agreement exists between CHIRPS and the HighResMIP models in capturing the behaviour of seasonal rainfall over SWA, with models able to capture the bimodal rainfall season. The ability of models in capturing onset and cessation dates as observed in CHIRPS shows the strength of these models in representing the short break between the two wet seasons that is otherwise known as the ‘Little Dry Season’. Patterns observed in the onset and cessation dates over the SWA region are consistent with the northward and southward displacement of the Intertropical Convergence Zone (ITCZ). The seasonal timing of the models shows good agreement with observations such that most mean onset/cessation dates agree within 26 days. While IPSL-CM6A-ATM-HR, a model among the five HighResMIPs used in the study, best agrees with CHIRPS in representing onset and cessation dates during the unimodal rainfall season, no one model best agrees with CHIRPS during the bimodal season, with models outperforming each other in representing onset/cessation dates with little variation.

Keywords