Thesis (PhD (Civil Engineering))--University of Stellenbosch, 2011. / ENGLISH ABSTRACT: In South Africa, wind constitutes the most critical environmental loading affecting the design of the built environment. The wind climatic information, which is currently incorporated in structural design standards, is based on the analysis of records from a limited number of wind recording stations, mainly located in large cities, and was done several decades ago. In view of the size and the climatological diversity of South Africa, this information cannot be deemed to be adequate. Therefore, the incorporation of welldistributed and updated information on wind climate is essential. The present study endeavoured to address this issue. A strong wind climatology was developed with the use of observed climate data, with the most significant result that a mixed strong wind climate is prevalent in the greater part of South Africa. Statistical approaches to estimate extreme wind speeds were investigated with applicable wind data, with the optimum approach guided by the unique climatological environment and the statistical properties of the utilised data set: For the wind gust analysis the Peak-Over-Threshold method with the exponential distribution is recommended, while in a mixed strong wind climate the “mixed climate” approach is preferred. For the analysis of the hourly mean wind speeds the choice is between the Gumbel distribution and the mixed climate approach, depending on the strong wind climate. The estimation and incorporation of environmental correction factors to the measured wind speeds were necessary as the surroundings of most weather stations did not correspond to the reference Terrain Category. For some of the weather stations it was impossible to compensate for the inadequate exposure and surrounding complex topography, so that a reduced number of weather stations were available for the strong wind analyses. The values estimated for the design wind speeds, adjusted for the short lengths of data records, as well as techniques developed to guide the spatial interpolation of the quantiles, were utilised to develop updated maps of the regional design wind speeds. A comparative study between the results of this study, and that of the previous study on which the current loading code in South Africa is based, indicates that the present study should produce more reliable quantile estimations. / AFRIKAANSE OPSOMMING: Wind vorm die mees kritieke omgewingslading wat die ontwerp van die beboude omgewing in Suid-Afrika beïnvloed. Die windklimaat-inligting wat tans gebruik word in die ontwerp spesifikasies is gebaseer op die statistiese analiese van veskeie dekades gelede op ‘n beperkte aantal windmeting-stasies, hoofsaaklik gesentreer in groot stede. Indien die grootte sowel as die klimatologiese diversiteit van Suid-Afrika in ag geneem word kan hierdie inligting nie as voldoende gereken word nie. Die gebruik van heelwat beter verspreide en opgedateerde inligting oor die windklimaat is daarom noodsaaklik en die studie poog om hierdie leemte aan te spreek. ‘n Sterk-wind klimatologie van Suid-Afrika is ontwikkel deur die gebruik van waargenome klimaatdata, met die mees betekenisvolle bevinding dat ‘n gemengde sterk-wind klimaat in die grootste gedeelte van Suid-Afrika heers. Statistiese benaderings om ekstreme winde te beraam is ondersoek met die beskikbare winddata, met die optimale benadering wat sal afhang van die klimatologiese omgewing van die weerstasie en die statistiese eienskappe van die betrokke windrekord: Vir die wind-stoot analieses word die “Piek-Oor-Drumpel” metode met die eksponensïele verdeling aanbeveel, behalwe in ‘n gemengde sterkwind klimaat waar die “gemengde klimaat” benadering gebruik word. Vir die analiese van die uurlikse gemiddelde winde is die keuse tussen die Gumbel verdeling en die gemengde klimaat benadering, afhangende van die sterk-wind klimaat. Die skatting en toepassing van omgewingskorreksiefaktore vir die windspoed was nodig, aangesien die omgewings waarin die meeste weerstasies is nie ooreenkom met die verwysings Terrein Kategorie nie. Vir sommige weerstasies was dit onmoontlik om vir die onvoldoende blootstelling te vergoed, met die gevolg dat minder stasies beskikbaar was vir die sterk wind analiese. Die geskatte waardes vir die ontwerp-windsnelhede, asook tegnieke ontwikkel vir ruimtelike interpolasie, is gebruik vir die ontwikkeling van kaarte van die omgewings-ontwerpsnelhede, na verstellings van die waardes om te vergoed vir die kort data rekords wat gebruik is. ‘n Kritiese vergelykingstudie wat gedoen is tussen die resultate van die huidige studie, en die vorige waarop die huidige laskodes vir Suid-Afrika gebaseer is, dui aan dat die huidige studie betroubaarder skattings van die kwantiele behoort op te lewer.