Climate Change Adaptation Data
Description
This dataset reflects the two methodological approaches used in the study. Quantitative rainfall and temperature data were analysed to illustrate climatic variability and extreme events in the study area, as well as their impacts on local communities. The qualitative data, drawn from interviews, focus group discussions, and questionnaires, explored socio-economic and environmental conditions, perceptions of climate risks, adaptation strategies, and the roles of local government. Together, the quantitative and qualitative findings informed the development of a climate change adaptation framework for the study area.
Files
Steps to reproduce
Quantitative climate data consisted of daily rainfall and temperature records (1970–2023) obtained from the South African Weather Service (SAWS) and the South African Sugar Research Institute (SASRI). Data were sourced from multiple stations across northern and southern Durban to capture spatial climate variation in eThekwini. All records were checked for accuracy, consistency, and completeness, with missing values replaced using neighbouring stations with comparable long-term records. To reduce short-term fluctuations, daily data were converted to monthly averages and then to mean annual temperature and rainfall. During reformatting, dates, units, and variable names were standardised so that all stations followed a uniform structure. This cleaning and organisation process, completed in Microsoft Excel, produced a coherent dataset suitable for long-term climate trend analysis. A key limitation is that while the dataset reflects broader municipal-scale patterns, it cannot fully capture microclimatic variability, such as local rainfall intensities or heatwaves, which strongly influence community-level climate impacts. The cleaned datasets were analysed in Python using a customised script to assess temporal changes in rainfall and temperature. Python’s efficiency and reproducibility make it widely used in climate research. Key libraries included Pandas for data manipulation, NumPy for numerical computation, Matplotlib and Seaborn for generating climate trend visualisations, and Graphviz for mapping relationships between variables. This workflow supported the identification of long-term climate variability, including fluctuations in maximum and minimum temperatures and rainfall extremes, with drought and flood years highlighted for disaster-risk assessment. Qualitative data were collected through semi-structured interviews and a focus group using mixed-method questionnaires that explored socio-economic and environmental conditions, climate risk perceptions, adaptation strategies, and local government roles. Forty interviews were conducted in four purposively selected communities—Nazareth, Springfield, Sydenham, and Westridge—chosen for their high exposure to climate hazards and socio-economic diversity. Additional interviews were held with Gift of the Givers and the South Durban Community Environmental Alliance, along with a youth focus group at Centenary Secondary School to understand adolescent perspectives on climate change and resilience. With consent, all discussions were recorded and transcribed. The qualitative data were analysed in NVivo, which facilitated systematic coding, theme identification, and comparison across stakeholders. This enabled insights into institutional responses, policy gaps, and community-based adaptation, supporting a more comprehensive understanding of climate vulnerability and resilience within the municipality.
Institutions
- University of KwaZulu-Natal - Westville CampusKwaZulu-Natal, Durban