Using satellite imagery and rainfall data to track climate variability in South Africa

The ARC-Institute for Soil, Climate and Water hosts an archive of data from regional weather stations in South Africa dating from 1900 and a database of National Oceanic and Atmospheric Administration-Advanced Very High Resolution Radiometer (NOAA-AVHRR) satellite data dating from1985. These two data sets provide an excellent opportunity to monitor climate and vegetation variation over time. Two case studies, combining these two data sets, were recently conducted on a regional scale in South Africa. The one, concerned with the semiarid Limpopo River valley, combined point measurements of rainfall with local Normalized Difference Vegetation Index (NDVI) values to investigate the correlation between point rainfall and the NDVI response in the data sets. The results show a strong log-linear correlation between rainfall data and NDVI due to saturation of the NDVI at high rainfall and cloud contamination. The NDVI seems to be more sensitive to drought conditions than high rainfall. Putting this into perspective, it was deemed useful as a potential tool to monitor the movement of the 500 mm isohyte over the central parts of South Africa. The 500-mm rainfall isohyte is an important meteorological feature, which is significant for dryland agriculture and crop production in South Africa. Areas receiving more than this amount of rain annually are regarded as being suitable for dryland agriculture. However, rainfall over South Africa shows a high degree of interseasonal variability. This subsequent case study intended to track the movement of the isohyte as well as its impact on vegetation. The NDVI was used to monitor shifts in vegetation growth over 17 years, while rainfall surfaces were created to give an idea of the historical interseasonal migration of the 500 mm isohyte since 1930 including the 17 years of vegetation data. Results demonstrate that a threshold value for NDVI during a season can be used to approximate the area receiving sufficient rainfall for dryland agriculture. Available archived satellite data is insufficient to observe vegetation change in the context of natural climate variability. Historical rainfall data is therefore used as a guiding reference for using satellite data in monitoring regional vegetation conditions. Results show a potential for using satellite imagery for operational monitoring of global change with rainfall as ancillary data. This approach can be used on a regional scale to assess the suitability, for different types of agriculture, of specific areas, thereby aiding agricultural zoning.