Statistical analysis of rainfall seasonality across South Africa
Date
2019
Authors
Roffe, Sarah Jane
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Straddling the subtropics and mid-latitudes, the climate of South Africa is influenced by the seasonal migration of the Intertropical Convergence Zone, the Antarctic sea ice extent and associated displacement of the Southern Hemisphere Westerlies. As a consequence of this synoptic mechanism, South Africa has a unique rainfall climatology of summer, winter and year-round rainfall zones (SRZ, WRZ and YRZ). Since the first known publication in 1938, much research has attempted to classify the spatial distribution of these rainfall zones. Different methods with different definitions of rainfall seasonality have been applied, and as such the classification of rainfall seasonality is disputed throughout this literature. Unlike research attempting to classify rainfall seasonality, there has been little research investigating how the seasonal characteristics of rainfall have and may possibly change under anthropogenically induced climate change. Across these studies different seasonality metrics have been applied making spatial and cross-study comparisons difficult. This is problematic for South Africa which has a heavy reliance on the seasonal characteristics of rainfall for activities such as agriculture, water resource management and tourism. To address the aforementioned gaps and issues in literature, this research sought to develop a standard approach to quantify, classify and understand rainfall seasonality and the changes thereof. A consistent dataset of daily rainfall and temperature spanning the period of 1987-2016 from 46 weather stations spread across South Africa was applied to objectively test four seasonality metrics which statistically discriminate between SRZ and WRZ conditions. Linear correlation analysis was applied to the annual seasonality quantities to investigate changes in rainfall seasonality. The results of the classification of rainfall seasonality across the four approaches were not fully consistent, further highlighting the necessity of a standard approach. Due to the different applications of seasonality metrics, two standard approaches were proposed. The first metric quantifies climatological seasonality based on the relationship between temperature and rainfall and returns a seasonality score. This method was suggested as it produced the best classification of rainfall seasonality, with the YRZ extending northwest from the southern coast to the Namibian border, spanning the transition between the SRZ and WRZ. The second metric is based on the timing of accumulation of 10% and 90% of daily rainfall representing boundaries for the wet season and was suggested due the societal relevance of this metric for agricultural seasonality. These standard metrics are important for future climatological research across South Africa and are valuable to facilitate development of more accurate climate model forecasts and projections. Few statistically significant (8%) changes were quantified. Although this was the case, the changes quantified demonstrated much agreement across the seasonality metrics and spatially, and were broadly consistent with literature. For the majority of SRZ locations, an increase in SRZ seasonality with a reduction in the wet season duration due to later commencement was quantified, which is consistent with an expansion of the tropics. The results for the WRZ locations predominantly suggest a decrease in WRZ seasonality and a longer wet season duration which coincided with a reduction in rainfall during many winter months; this has been linked to a poleward migration of the Westerlies. For the YRZ locations along the southern coast, an increase in WRZ seasonality consistent with a reduction in the wet season duration was quantified. These quantified changes in the seasonal characteristics of rainfall have implications for crop yields and water supplies across South Africa and are particularly important for developing adaptation strategies to reduce potentially detrimental consequences of climate change.
Description
A Thesis submitted in fulfilment of the academic requirements for the degree of Doctor of Philosophy
School of Geography, Archaeology and Environmental Studies
University of the Witwatersrand, Johannesburg, October 2019
Keywords
Citation
Roffe, Sarah Jane, (2019) A statistical analysis of rainfall seasonality across South Africa, University of the Witwatersrand, Johannesburg, https://hdl.handle.net/10539/29571