Abstract
Extreme storms in South Africa and specifically in the Western Cape have been responsible for widespread destruction to property and infrastructure, even leading to displacement and death. The occurrences of these storms have been increasingly linked to human-induced climate change that is expected to cause more variable weather. Studies on climate circulation models for future climate conditions project that rainfall in the Western Cape and wider South African region is to become more intense and extreme. Sub-daily rainfall for 3 stations in the Western Cape and 4 stations in the rest of South Africa were analysed in order to determine if any trends towards more intense and extreme rainfall are observed and whether the trend is unique to the Western Cape or indicates a wider trend. This study explores this expectation by using historical short-duration rainfall (less than 24 h) for 7 stations in the Western Cape and South African region. Digitised autographic and automatic weather station 5-min rainfall data were combined to extend the effective record length. Both the magnitude and frequency of occurrence of rainfall events were analysed to assess if rainfall intensities are showing any evidence of increasing over time. For the magnitude of rainfall events, extreme value theory was applied to non-stationary sequences, using both a parametric and non-parametric approach for both event maxima and peaks over threshold modelling. The frequency analysis entailed measuring the frequency of exceedance of rainfall events over a certain threshold value. Both the magnitude and frequency analysis indicated that the combination of the two record types influenced the results of some of the stations, while the others showed no consistent evidence of changing rainfall intensities. This led to the conclusion that, from the available observed short-duration record, no evidence was found of trends or indications of changes in rainfall intensities. Keywords : climate change, short duration rainfall, extreme value theory, non-stationary
Highlights
Developments in climate sciences over the past 20 years are attributing the increasing occurrence of extreme weather events like droughts and floods to changing climate conditions as a result of increases in atmospheric greenhouse gas concentrations caused by rapid human development since the industrial age
To improve decision making in this regard, an understanding of the mechanisms at work is required and a number of critical questions can be asked: What role does climate change play with the frequency and intensity of these storms in South Africa? Are design practices used to anticipate and design for storms, based on the stationarity of the climate, still adequate? If not, is this a phenomenon unique to the Western Cape or is it observed in other parts of South Africa?
Van Wageningen and Du Plessis (2007), analysing 5-min rainfall data for the Molteno reservoir rainfall station in Cape Town in the Western Cape over the period 1961–2003, found that the occurrence of rainfall events decreased from the 1990s, while the average magnitude of the 5-min events increased over the same period, indicating increasing intensities
Summary
Developments in climate sciences over the past 20 years are attributing the increasing occurrence of extreme weather events like droughts and floods to changing climate conditions as a result of increases in atmospheric greenhouse gas concentrations caused by rapid human development since the industrial age. IDF curves are typically derived by fitting statistical distributions to different storm duration rainfall data and obtaining the specific intensity design levels for the relevant return periods. The magnitude analysis focussed on the amount of rainfall calculated over the aforementioned storm durations using non-stationary sequences on extreme value distributions in order to determine if rainfall events are becoming more extreme over time. The third approach was based on a methodology where the shape parameter for the data set, consisting of the number of exceedances above a selected threshold value, was calculated. For all combinations of return and window periods, the number of significant slopes and corresponding sign of the slopes would be recorded This procedure is applied to all storm durations for both the GEV distribution and GPD. Rainfall above a threshold value was used instead of using all the data, as in Van Wageningen and Du Plessis (2007), since the interpolation technique for the autographic data includes low rainfall values (< 0.2 mm) interpolated over long durations, resulting in extremely low and arguably unrealistic rainfall values that will be counted as an event
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
