Using land cover changes to mitigate the impacts of hot temperatures in South African cities

Chambers, Timothy

Using land cover changes to mitigate the impacts of hot temperatures in South African cities

Thesis / Dissertation

2025

Publisher

University of Cape town

Department

Department of Environmental and Geographical Science

Faculty

Faculty of Science

Abstract

South African cities are negatively impacted by hot temperatures, which have severe impacts on human health. Climate change and population growth are expected to exacerbate these impacts in the future, given that Southern Africa has been projected as a hotspot for future warming and growth. The added socio-economic factors leave a large proportion of South Africa's population vulnerable to these hot temperatures. Although several studies have identified strategies for adapting to, or mitigating, the impacts of these hot temperatures, these strategies require extensive resources. The effective implementation of the strategies in a local context would need to identify and prioritise areas with highest risk, and with greatest reward. This thesis addresses these issues by examining the spatial distribution of temperatures in ten hot South African cities and investigating the impact of land cover changes on both the current distribution of temperatures, and in reducing future temperatures. The study analyses observation and simulation datasets, using Landsat 8 satellite observations to produce high-resolution land surface temperature (LST) maps over each city. The 2020 South African land cover (SANLC2020) dataset is then incorporated into the Urban Canopy model (UCM) and the Weather Research and Forecasting model (WRF) to simulate conditions over the cities.To investigate the effectiveness of using land cover changes as potential cooling strategies, two idealised but realistic scenarios were implemented into the model: (a) increasing tree coverage to 60% across the urban area and (b) increasing the albedo of all building roofs to 40%. The results of the study reveal substantial temperature differences (up to 10°C) between neighbourhoods within the cities. These differences show strong correlations to observed variations in land cover, which itself shows links to socioeconomic status and historical policies. In all but one of the study areas, the cities are found to be cooler than the surrounding, less vegetated rural areas, illustrating an Urban Cool Island (UCI) effect. This is caused by the importation of trees, moisture, and reflective surfaces by people. The addition of the converted SANLC land cover dataset significantly enhances the performance of the WRF model. Using this, the model showed good skill in reproducing observed spatial patterns and values of temperature. The model-based sensitivity tests showed that both scenarios could reduce the surface temperature in the cities by up to 3°C. The information gathered in the study highlights the significant potential there is for targeted land cover changes to mitigate the impacts of hot temperatures, reduce thermal inequality, and adapt South African cities into thermal refuges

Keywords

Environmental and Geographical Science

Reference:

Collections

Masters

Full item page