Buthelezi, Nomcebo Siphesihle2020-09-092020-09-092019Buthelezi, Nomcebo Siphesihle (2019) Delineating current and potential distributions of prosopis glandulosa in the SKA SA Karoo Site, University of the Witwatersrand, Johannesburg, <http://hdl.handle.net/10539/29576>https://hdl.handle.net/10539/29576A research report submitted to the Faculty of Engineering and the Built Environment, University of the Witwatersrand, Johannesburg, in partial fulfilment of the requirements for the degree of Master of Science in Geographical Information Systems and Remote Sensing, 2019Prosopis species (also known as Mesquite), in particular P. glandulosa (Honey Mesquite) have a negative impact on indigenous biodiversity and the livelihood of communities in the semi-arid and arid parts of South Africa. The spread of these species is a threat to the environments in which they have been introduced as they spread at high rates, increase the mortality of indigenous trees and disrupt important ecosystem processes such as hydrological and nutrient cycles. Due to the negative impacts of Prosopis on important ecosystem services and South Africa’s native biodiversity, it is essential for the distribution of these species to be identified, controlled and monitored in order to mitigate their spread and restore damaged ecosystems. The objectives of this study were to use Remote Sensing and Geographic Information Systems (GIS) tools to: (i) delineate the distribution of Prosopis using high resolution satellite imagery, (ii) determine the changes in spatial distribution of these species in the period 2003-2017, and (iii) use moderate spatial resolution satellite imagery and ancillary environmental data to predict areas susceptible to future invasion.. The study area used in this investigation is the Square Kilometre Array (SKA SA) site, situated in Northern Cape Province, South Africa. Satellite images were classified using Multi-layer Perceptron (MLP) Neural Network classification algorithm to improve the land use land cover classification accuracy. A WordView-3 image with 1.24 m spatial resolution was used to delineate the distribution of Prosopis in the study area for the year 2016. Landsat images from the years 2003, 2008, 2013 and 2017 were used to conduct a change detection analysis. The prediction model developed in the study was able to predict Prosopis cover for the years 2017 and 2022 cover using ancillary environmental data and land use land cover maps. The study was also able to quantify the area covered by Prosopis species for the years 2017 and 2022.Online resource (83 pages)enVegetation mapping--South AfricaMesquiteCrop improvementThesis