Mapping and monitoring the distribution and invasion of Mesquite (Prosopies Glandulosa) along the AL Gash River Kassala, Sudan using remote sensing techniques
No Thumbnail Available
Date
2020
Authors
Mokgehle, Dineo Revinwa
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The accelerated invasion of mesquite (Prosopis) has brought negative socio-economic and ecological impacts, both at local and global scale. According to the International Union for conservation of Nature (2004), mesquite has been identified as one of the top 100 worst invasive species in the world. The lack of information regarding the spatial and temporal variability of mesquite invasion has compromised the implementation of monitoring and control efforts. Hence, the mapping and monitoring of mesquite is vital in order to obtain precise and up to date spatial and temporal data about its invasion dynamics. This study focused on investigating the ability of Sentinel-2 data in mapping the current spatial distribution of mesquite invasion along the Al Gash River in Kassala, Sudan using Support Vector Machine and Random Forest classifiers. Sentinel-2 image, which covered the study area was obtained during the dry period (March). The utilisation of Random Forest classifier achieved an overall accuracy of 93.44%, whereas the Support vector machine classifier achieved an overall accuracy of 87.57%. Additionally, multitemporal Landsat earth observation data were used to monitor the spatio-temporal dynamics of the mesquite invasion over a period of 30 years from 1989 to 2019, with five years intervals. The change detection statistics depicted that the invasion of mesquite has increased over the years. The lowest mesquite areal coverage was found in 1989, with 13.7% (38967.78ha) of the study area. However, mesquite rapidly increased to 34.46% (124365ha) in 2014. The year 2019, witnessed a decline in mesquite coverage, covering 101214.6ha (26.84% of the study area). Overall, the study demonstrates the ability of Sentinel-2 to detect and discriminate mesquite from other land use and land cover types. The spatial and temporal variability of the new generation multispectral sensor enables continuous monitoring of the invasive species, through the provision of up-to-date and readily available data, at reasonable resolutions
Description
A research report submitted to the Faculty of Science, University of the Witwatersrand, Johannesburg, in partial fulfilment of the requirement for the degree of Master of Science in GIS and Remote Sensing at the School of Geography, Archaeology & Environmental Studies, 2020