Machine learning algorithms for mapping Prosopis glandulosa and land cover change using multi-temporal Landsat products: a case study of Prieska in the Northern Cape Province, South Africa

Invasive alien plants (IAPs) are responsible for loss in biodiversity and the depletion of water resources natural ecosystems. Prosopis species IAPs previously introduced by farmers to provide shade fodder livestock. In Northern Cape, spp. invasions associated with native resulting overgrazing degrading rangelands. Mapping glandulosa is essential management initiatives assist government minimising spread impact IAPs. This study aims evaluate performance two machine learning algorithms i.e., Support Vector Machine (SVM) Random Forest (RF) map spatial dynamics P. Prieska. The invasion extent was mapped using multitemporal Landsat data spanning period from 1990 2018. Validation results done through an estimated error matrix use proportion area estimates overall accuracy, user’s accuracy producer’s a 95% confidence interval. SVM RF classifiers showed similar Kappa statistics throughout years. These methods increase at least 3.3% invaded indicates importance imagery mapping historical current land cover change confirmed total invasion, which enables decision-makers improve monitoring eradication initiatives.