Climate variability and change over Eastern South Africa

Abstract

South Africa is increasingly vulnerable to extreme weather events, including rising temperatures, droughts, and floods, particularly along the east coast. These phenomena have historically caused fatalities, infrastructure damage, and agricultural disruption. This study explores the historical and projected climate variability and changes over eastern South Africa, focusing on the near-future (2021-2050) and far-future (2070-2099) under high and low mitigation scenarios. The research was conducted in KwaZulu-Natal (KZN) and Mpumalanga (MP) provinces using observational and climate projection data from Copernicus Climate Change Service (C3S), managed by the European Centre for Medium-Range Weather Forecasts (ECMWF), and Coupled Model Intercomparison Project Phase 6 (CMIP6) models for future projections. Statistical and visual methods verified the CMIP6 models’ performance. Trend analyses of historical data from 1961-2020 were conducted using ERA5 reanalysis. Temperature indices (TX10P and TX90P) and rainfall indices (R10mm and R50mm) were calculated using RClimDex. Tools like Grid Analysis and Display System (GrADS), climate data operators (CDO), R and R Studio were used for data analysis. Results showed persistent cool days in 1968 for 29 days and more than 35 days in Kruger Mpumalanga International Airport and Durban respectively and an increasing trend in hot days and decreasing trend in heavy rainfall days at both locations. Furthermore, the frequency of rainfall exceeding 50mm/day trends varied by location. Statistical verification for Surface Air Temperature (SAT) showed a high correlation (≥0.9) and low Root Mean Square Error (RMSE) with ERA5 data, while rainfall model verification had a moderate correlation (0.5-0.7) and least RMSE, indicating higher uncertainty in precipitation modelling. The CMIP6 models performed well in projecting temperature and rainfall trends, with the highest temperatures and rainfall projected for KZN, while MP showed lower values. These findings align with previous studies on South African climate trends and they highlight the need for enhanced climate monitoring, continued high-resolution modelling, and improved climate projections at regional scales.