Understanding land use and land cover dynamics and their effects on the surface water resources at Letaba Catchment in Limpopo, South Africa

Abstract

This work investigates the intricate dynamics of land use and land cover (LULC) due to pressures from population growth and environmental changes. The implications of these changes are particularly significant for water security, especially in semi-arid and water-scarce regions. Leveraging on the capabilities of Google Earth Engine (GEE), this comprehensive study seeks to unravel the spatial, seasonal, and long-term impacts of LULC changes on surface water bodies and water quality within these vulnerable regions. By exploring the interconnectedness of land dynamics and water resources, this research contributes valuable insights essential for sustainable water management in the face of evolving environmental conditions. This research encompasses a multi-faceted exploration of the Letaba catchment, offering a comprehensive understanding of its evolving LULC dynamics over a 31-year period (1990, 2000, 2010, 2021). LULC changes were classified using the random forest (RF) computed in GEE. The method of using remote sensing, especially GEE effectively mapped LULC dynamics with satisfactory accuracies ranging from 93 to 99% within four timeframes. The analysis unveils a noteworthy decline in natural forests (-3%), grasslands (-3%), bare surfaces (-2%), and water bodies (-4%), when compared with increases in shrublands (24%), plantations (41%), and built-up (47%) areas. This transformation underscores the imperative for adaptive strategies, with a particular focus on the pivotal role of climate-smart agriculture in harmonizing ecosystem services and human well-being. Further, the research assesses the effectiveness of Sentinel-2-derived spectral indices, such as Normalized difference water index (NDWI), Modified normalized difference water index (MNDWI), and Sentinel 2 water index (SWI), in conjunction with landscape metrics, to delineate the spatial and seasonal variations of surface water between 2017 and 2022 in GEE. Notably, remarkable overall accuracies ranging from 96% to 100% were achieved across both dry and wet seasons. This unveils surprising seasonal fluctuations, challenging conventional assumptions about water resource availability and advocating for adaptive water resource management strategies that account for such variabilities. Furthermore, the study delves into the impact of LULC alterations on water quality parameters. By integrating in-situ measurements with Sentinel 2 remote sensing data, strong correlations between LULC, spectral bands, water indices, and key water quality variables are revealed. These findings confirm the pollution status of the catchment, accentuating the urgent need for effective land planning and management. This not only challenges conventional notions about water resource availability but also underscores the necessity of adopting adaptive management approaches and sustainable strategies to address changing climatic conditions and human impacts on water quantity and quality in semi-arid tropical environments.