Skip to content

Changes in Extreme Rainfall Events in South Africa

Literature › Works

Simon J. Mason; Peter R. Waylen; Gillian M. Mimmack; Balakanapathy Rajaratnam; J. Michael Harrison (1999). Changes in Extreme Rainfall Events in South Africa. Climatic Change 41. Kluwer Academic Publishers.

Type journal article
Scope Comparative and methods
Method descriptive quantitative
Period studied 1931–1990
Regions South Africa, east coast / KwaZulu-Natal, winter rainfall region (south-west), north-east (Lowveld)
Themes methods data sources, agriculture land
Citation key mason1999

Argument

A climatology note testing whether the intensity of extreme daily rainfall in South Africa changed between 1931–1960 and 1961–1990. Using annual daily-rainfall maxima from stations with at least 50 complete years and no site relocations, the authors fit beta-κ (right tail) and beta-P (left tail) distributions to each 30-year half and assess the significance of differences in 5- to 50-year return-period intensities with a distribution-free resampling test (1000 random splits per station). They find significant increases in high-rainfall intensity over about 70% of the country, with the largest percentage increases for the most extreme events and increases above 50% on the east coast, partly but not only driven by cyclone Demoina (1984). Decreases occur in the north-east, north-west and the south-western winter rainfall region. They argue the changes are at least partly real despite untested instrumentation inhomogeneities, and are consistent with enhanced-greenhouse expectations, though attribution to a single synoptic system or to warming is not possible.

Evidence and method

Daily rainfall records 1931–1990 for c.314–316 South African stations supplied by the Computer Centre for Water Research; annual maxima extracted; GCM simulation literature and prior regional climatology for interpretation. Research funded by Eskom.

Findings

  • Significant increases in extreme rainfall intensity between the two periods cover about 70% of South Africa. (p. 1)

    Significant increases in the intensity of extreme rainfall events between 1931–1960 and 1961–1990 are identified over about 70% of the country.

  • 10-year event intensity rose by more than 10% over large areas, excluding parts of the north-east, north-west and the winter rainfall region. (p. 1)

    The intensity of the 10-year high rainfall events has increased by over 10% over large areas of the country, except in parts of the north-east, north-west and in the winter rainfall region of the south-west.

  • Station selection required at least 50 years of complete daily records and excluded relocated sites, but instrumentation changes were not tested for. (p. 2)

    Stations that involved site relocations were not included, but in-homogeneities resulting from changes in instrumentation have not been tested for because of a lack of metadata.

  • East-coast increases exceed 50%, partly reflecting cyclone Demoina in January 1984. (p. 4)

    Along the east coast, increases of over 50% in the intensity of 10-year high rainfall events have been experienced. The large increases here are partly a reflection of the impact of cyclone Demoina in January 1984

  • The north-east's decline in extremes matches falling annual rainfall since the late 1970s. (p. 4)

    The north-eastern part of South Africa has experienced a decrease in annual rainfall totals since the late-1970s (Ma-son, 1996), which is reflected in the decrease in extreme rainfall events in the area.

  • 65 of 314 stations show significant increases at the 90% level, exceeding the 43 needed to claim real change under independence; field significance not computed due to spatial dependence. (p. 5)

    At the 90% confidence level, 65 stations have experienced increases in 10-year high rainfall intensities in the right tail between 1931–1960 and 1961–1990.

  • Low annual maxima became significantly less intense at 46 stations, but changes are small. (p. 5)

    there is a total of 46 stations countrywide where the low annual maxima have become significantly less intense. The decreases are, however, small and mostly less than 20%.

  • There is little evidence of long-term trends in annual rainfall totals nationally, except possibly the far east. (p. 6)

    Over South Africa itself, there is little evidence of significant long-term trends in annual rainfall totals in the country, except possibly for a decrease in the far eastern part of the country

  • Heavy rainfall impacts on agriculture, dams, water management and insurance motivate the study. (p. 7)

    The impacts of heavy rainfall events on agriculture, engineering structures such as dams, water resource management, the insurance industry and human liveli-hood are considerable