Quantifying the impacts of urbanisation on urban agriculture over three decades, the case of Ekurhuleni, South Africa
Human beings affect the functioning of the earth system through the key process of Land use and Land cover change (LULCC). Among the many drivers of LULCC today are the rapid growth in population, the spread of urban centres, an increase in demand for food production and the scarcity of agricultural land. One of the main threats to urban agricultural system is urbanisation and urban extension. In South Africa, approximately 13% of the country is agricultural land and at least 22% is highly arable. However, this small agricultural land has continuously been lost due to intensive and rapid growths and extension of the industrial and residential land uses. This study aims is to investigate the impacts of the rapid urban growth in the Ekurhuleni on urban agriculture from 1985 to 2017 using Landsat multi-temporal data. Eight Landsat images were used to cover the period 1985 to 2017 with a 5-year interval to quantify the impacts of urbanisation on urban agriculture. The images were geometrically corrected and georeferenced to ensure that they conform to one another. 2017 image was used as the reference image. SVM was used to classify the images and the holdout test dataset was used for validation and accuracy assessment. Change detection techniques were then used to quantify the temporal extent of the urban area on the agricultural land. The overall accuracies from SVM for the images varied between 70 and 95%. The Kappa coefficient ranged from 0.7 to 0.9. Change detection statistics were produced from the classified maps in ENVI classic and the overall results showed that there was a decrease in the agriculture LULC class and an increase in the built-up area over the period of 30 years. The simulated LULC for 2049 was done using the cellular automata coupled with the artificial neural network model in QGIS MOLUSCE plugin. The prediction results for the year 2049 showed that built up area will continue to increase as a result of population growth and economic growth while urban agricultural land will continue decreasing. It is recommended in this study that higher resolution imagery should be used in further studies. Ancillary data such as socio-economic relationships will aid in better understanding LULCC in Ekurhuleni. This study recommends that more effort is put in place to protect land that is zoned out for agricultural purposes.
A research report submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in partial fulfilment of the requirements for the degree of Master of Science (Geographical Information Systems and Remote Sensing) at the School of Geography, Archaeology & Environmental Studies May 2019