Dynamic and Thermodynamic Contributions to Late 21st Century Projected Rainfall Change in the Congo Basin: Impact of a Regional Climate Model’s Formulation

Abstract

Addressing the impacts of climate change requires, first of all, understanding the mechanisms driving changes, especially at the regional scale. In particular, policymakers and other stakeholders need physically robust climate change information to drive societal responses to a changing climate. This study analyses late 21st-century (2071–2100) precipitation projections for the Congo Basin under representative concentration pathway (RCP) 8.5, using the Rossby Centre Regional Climate Model (RCM) RCA4. Specifically, we examine the impact of the RCM formulation (reduction of turbulent mixing) on future change in seasonal mean precipitation by comparing the results of the modified model version (RCA4-v4) with those of the standard version (RCA4-v1) used in CORDEX (Coordinated Regional Climate Downscaling Experiment). The two RCM versions are driven by two global climate models participating in the Coupled Model Intercomparison Project phase 5 (CMIP5). The results show that seasonal precipitation is largely affected by modifications in the atmospheric column moisture convergence or divergence, and, in turn, associated with changes in the dynamic (ΔDY) and thermodynamic (ΔTH) components of the moisture-budget equation. Projected decreased precipitation in the dry seasons (December–January–February and June–July–August) is linked to increased moisture divergence driven by dynamic effects (changes in circulation), with most experiments showing ΔDY as the main contributor (>60%) to the total moisture budget. Overall, precipitation is projected to increase in the wet seasons (March–April–May and September–October–November), which can be attributed to both dynamic and thermodynamic effects, but with a larger thermodynamic contribution (changes in specific humidity, ΔTH > 45%), compared to the dynamic one (ΔDY > 40%). Through a comparison of the two model versions, we found that the formulation (reducing turbulent mixing) and boundary conditions (driving GCM) strongly influence precipitation projections. This result holds substantial value for ensuring the fitness of models for future projections intended for decision-makers.