An evaluation of CORDEX regional climate models in simulating precipitation over Southern Africa

• Published in: Atmospheric science letters Vol. 16; no. 3; pp. 199 - 207
• Main Authors: Shongwe, Mxolisi E., Lennard, Chris, Liebmann, Brant, Kalognomou, Evangelia‐Anna, Ntsangwane, Lucky, Pinto, Izidine
• Format: Journal Article
• Language: English
• Published: Chichester, UK John Wiley & Sons, Ltd 01.07.2015; John Wiley & Sons, Inc
• Subjects: Agreements; Annual variations; Atmospheric precipitations; Atmospheric sciences; Bias; Climate; Climate change; Climate models; Computer simulation; Convergence zones; CORDEX – Africa; Hydrologic data; Interannual variability; Monthly rainfall; Physics; Precipitation; Precipitation patterns; Precipitation probability; Probability; Probability density functions; Probability theory; Rain; Rainfall; rainfall characteristics; Rainfall frequency; Rainfall simulators; Rainy season; Regional analysis; regional climate model evaluation; Regional climate models; Regional climates; Seasons; Theoretical physics; Tropical climate; Tropical Convergence Zone; Wet season; South Africa; Mozambique; Africa
• ISSN: 1530-261X; 1530-261X
• DOI: 10.1002/asl2.538

This article evaluates the ability of the Coordinated Regional Downscaling Experiment (CORDEX) regional climate models (RCMs) in simulating monthly rainfall variation during the austral summer half year (October to March) over southern Africa, the timing of the rainy season and the relative frequencies of rainfall events of varying intensities. The phasing and amplitude of monthly rainfall evolution and the spatial progression of the wet season onset are well simulated by the models. Notwithstanding some systematic biases in a few models, the simulated onset and end of the rainy season and their interannual variability are highly correlated with those computed from the reference data. The strongest agreements between the reference and modelled precipitation patterns are found north of about 20°S in the vicinity of the Inter Tropical Convergence Zone. A majority of the RCMs adequately capture the reference precipitation probability density functions, with a few showing a bias towards excessive light rainfall events.