Thesis (MScEng)--Stellenbosch University, 2012. / ENGLISH ABSTRACT: The Western Cape has been subject to extreme rainfall storms in the last decade, which has been responsible for millions of Rands worth of damage to public infrastructure and human settlements. Current climate model projections under climate change suggest that the Western Cape can expect greater rainfall intensities, with magnitude of rainfall events increasing, but the frequency of occurrence of events decreasing. The aim of this research is to assess if historical short duration rainfall data provides any evidence to support increasing intensities in rainfall in the Western Cape. Short duration rainfall data (< 24 hours), consisting of digitised autographic- and automatic weather station data (AWS), was selected from the South African Weather Services’ database for the Western Cape and wider South African region, in order to combine the data for an extension of the effective record length. Numerous difficulties were encountered with the data that required the application of editing and quality control procedures. The digitised autographic data contained many errors and generally compared poorly to standard gauge daily rainfall totals of the same period. After the application of editing and quality control procedures, seven stations were selected and their autographic and AWS data was combined for further detailed analysis. Analysis was divided into two sections, one for the analysis of the magnitude of the rainfall, the other for the frequency of occurrence of rainfall events. For the magnitude analysis, nonstationary extreme value theory was applied by implementing a parametric and non-parametric non-stationary approach to both the generalised extreme value distribution (GEV) with an annual maximum series, and a generalised Pareto distribution with a peaks over threshold series. The parametric approach entailed fitting linear models to the parameters of the extreme value distributions, and the non-parametric approach tested if return levels of the distributions remained constant over a moving window period. The frequency analysis entailed the recording of the annual number of rainfall events exceeding predetermined threshold values. For the magnitude analysis, the parametric approach only produced two stations with significant non-stationarity and only for the GEV, whilst the non-stationary approach produced results that can indicate non-parametric behaviour, but this was most likely because of the combination of autographic and AWS data. The frequency analysis produced results that indicated no observable results for some stations, while others produced trends that are consistent with a change from the autographic to AWS data. This lead to the conclusion that with the available data, no clear or significant evidence supporting increasing intensities or any other change in short duration rainfall was found regarding the magnitude and frequency of occurrence of rainfall events. / AFRIKAANSE OPSOMMING: The Wes-Kaap het die afgelope dekade deur uiterse storms deurgeloop wat miljone Rande se skade aan publieke infrastruktuur en nedersettings veroorsaak het. Huidige klimaatsmodel resultate wat klimaatsverandering inkorporeer, beraam dat die Wes-Kaap hoër reënvalintensiteite kan verwag, waar die grootte van die en reënvalgebeurtenisse toeneem, maar die frekwensie van die reënvalgebeurtenisse afneem. Die doel van hierdie navorsing is om te bepaal of historiese kortduurtereënvaldata in die Wes-Kaap enige toenemende reënvalintensiteite aandui. Kortduurtereënvaldata (< 24 uur) wat uit versyferde autografiese- en automatiese weerstasie (AWS) data bestaan, was uit die Suid Afrikaanse Weerdienste se databasis gekies vir stasies in dieWes-Kaap en die wyer Suid Afrikaanse gebied, sodoende om die effektiewe rekordlengte te verleng. Redigering- en kwaliteitsbeheerprosudures moes toegepas word na ’n aantal probleme met die data ontdek was. Die versyferde autografiese data het baie foute bevat en oor die algemeen swak vergelyk met die standaard daaglikse totaalmetings van dieselfde periode. Na die redigering- en kwaliteitsbeheerprosudures toegepas was, was sewe stasies gekies en hul autografiese en AWS data was saamgestel vir verdere analise. Die analise het uit twee afdelings bestaan: die analise van die grootte van die reënvalgebeurtenisse, en die frekwensie van die reënvalgebeurtenisse. Nie-stasionêre ekstreemwaardeteorie was toegepas op die grootte van die reënvalgebeurtenisse deur gebruik te maak van parametriese en nie-parametriese nie-stasionêre metodiek. Die veralgemeende ekstreemwaardeverdeling (AEW) met ’n jaarlikse maksimum reeks, en die veralgemeende Paretoverdeling met ’n pieke bo drempelwaarde reeks was gebruik. Die parametriese metode het bestaan uit die passing van lineêre modelle aan die parameters van die estreemwaardeverdelings. Die nie-parametriese metode het getoets of die terugkeervlakke van die verdelings konstant gebly het oor ’n bewegende vensterperiode. The frekwensie analise het bestaan uit die aantekening van die jaarlikse reënvalgebeurtenisse wat bepaalde drempelwaardes oorskry. Die parametriese metode het net twee beduidende nie-stasionêre staties opgelewer, uitsluitlik vir die AEW. Die nie-parametriese metode het resultate gegenereer wat moontlike niestasionarieit aandui, maar dit is heel moontlik as gevolg van die samestelling van die autografiesen AWS data. Die frekwensie analise het óf geen sigbare neiging vir die resultate vir ván stasies gegenereer nie, óf het resultate gegenereer wat aandui op die samestelling van die autografiesen AWS data. Dit het tot die gevolgtrekking gelei dat daar geen aanduiding in die kortduurtereënvaldata bestaan wat toenemende reënvalintensiteite ondersteun nie vir beide die grootte en frekwensie van reënvalgebeurtenisse.

Thesis (MSc)--Stellenbosch University, 2012. / Severe floods in the Western Cape have caused significant damage to hydraulic structures, roads and other infrastructure over the past decade. The current design criteria for these structures and flood return level calculations are based on the concept of stationarity, which assumes that natural systems vary within an envelope of variability that does not change with time. In the context of regional climate change and projected changes in rainfall intensity, the basis for these calculations may become unrealistic with the passage of time. Hydraulic structures and other infrastructure may become more vulnerable to damaging floods because of changing hydroclimatic conditions. This project assesses the changes in extreme rainfall values over time across the Western Cape, South Africa. Using a Generalised Pareto Distribution, this study examines the changes in return levels across the Western Cape region for the periods 1900-1954 and 1955-2010. Of the 137 rainfall stations used in this research, 85 (62%) showed an increase in 50-year return level, 30 (22%) a decrease in 50-year return level and 22 (16%) stations displayed little change in rainfall intensity over time. While there were no clear spatial patterns to the results, they clearly indicate an increase in frequency of intense rainfalls in the latter half of the 20th and early 21st century. The changes in return level are also accompanied by a change in the frequency of high intensity 2-3 day long storms. 115 (84%) of the 137 rainfall stations showed an increase in the frequency of long duration, high intensity storms over the data record. This change generates a shifting risk profile of extreme rainfalls, which, in turn, creates challenges for the design of hydraulic structures and any infrastructure exposed to the resulting damaging floods. It can therefore be argued that it is inappropriate to design structures or manage water resources assuming stationarity of climate and that these principles should be assessed in order to reduce the risk of flood damage owing to increasing storm intensity. KEY WORDS Flood Risk, Stationarity, Disaster Risk, Hazard, Extreme Rainfall, Generalized Pareto Distribution, Climate