A thesis submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg,
in fulfilment of the requirements for the degree of Doctor of Philosophy.
November 2017, Johannesburg. / The main aim of this PhD research is to use accurate, relatively long-temperature data sets (spanning 1851-2016) available in South Africa, to establish a more precise understanding of temporal temperature variability over this region. This is achieved through the use of robust methodology (i.e. data quality control and homogenization procedures, using a reference series) to produce high quality data sets for temperature trend analysis. Data quality and homogenization of data sets are conducted using RClimDex, ProClim and AnClim software. Long-term temperature trends are determined using a nonparametric statistic (i.e., Mann Kendall test). The core objectives are to provide a homogeneous dataset for the Western Cape region in South Africa and establish long-term (97 years) annual and seasonal instrumental temperature trends for this region. The second objective is to extend past work by examining temperature trends for KwaZulu-Natal (KZN) over a considerably longer period, with a specific focus on contrasting coastal and adjacent higher lying interior regions. The last objective is to assess the extent and timing of El Niño/La Niña impacts on maximum temperature (Tmax) and minimum temperature (Tmin) over selected South African stations for the period 1884-2016, and to establish any possible direct change in such temperature responses to ENSO over time.
Long-term temperature trends for the Western Cape Province indicate statistically significant increasing annual temperature trends for adjusted (homogenized) Tmax and Tmin data over the period 1916-2013, with the exception of an insignificant decreasing Tmax trend at Cape St Blaize. For the common period (1937-2001), a statistically significant increasing trend (0.13°C/decade) is recorded for all stations in Western Cape Province, with a 0.04°C/decade stronger warming signal than before homogenization was applied. The seasonal trends also support significant increasing trends, with the exception of Tmax trends for summer (-0.03°C/decade) and autumn (0°C/decade) at Kirstenbosch.
For KZN, cooler conditions are noted for the second half of the 19th century (i.e. 1851, 1858, 1860-1864) compared to more recent times. For the period 1930-2015, annual Tmin records significant warming at all stations, with the rates of warming (°C/decade) over the interior being double that for the coast. Tmin warming trends are 1.31 (summer), 2.75 (autumn), 3.33 (winter) and 1.25 (spring) times faster over the interior compared to the coast. Thus, higher
lying interior regions are warming considerably faster than adjacent coastal areas mainly due to Tmin warming trends.
The impact of El Niño and La Niña events on Tmax and Tmin over South Africa is investigated using composite and Fourier series analysis (1884-2016). For 17 stations spread across South Africa, it is concluded that strong coherent relationships exist between El Niño/La Niña events and Tmax/Tmin. Contingency tables were constructed for each station during the austral summer season (December, January and February) and analysed using the Chi-squared statistic. The most notable finding is that El Niño events have had a stronger warming effect over many regions in South Africa after the late 1970s, the most prominent regions being the northern interior and southeast coast, where associated Tmax record an average of 0.41°C and 0.45°C higher values over the period 1979-2016 compared to the earlier period (1940-1978) respectively. The Chi-squared statistic results indicate that ENSO phases exert a stronger influence over the interior of South Africa compared to that over the coast.
Keywords: homogeneity, temperature, trend analysis, Mann Kendall, historical climate, Western Cape, KwaZulu-Natal, South Africa, ENSO. / LG2018
Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:wits/oai:wiredspace.wits.ac.za:10539/25018 |
Date | January 2017 |
Creators | Lakhraj-Govender, Rakhee |
Source Sets | South African National ETD Portal |
Language | English |
Detected Language | English |
Type | Thesis |
Format | Online resource (xvii, 200 leaves), application/pdf |
Page generated in 0.009 seconds