Assessing the utility of drone technology in estimating surface water temperature, total suspended solids (TSS) and Chromophoric dissolved organic matter (CDOM) in reservoirs: a case study in the uMngeni Catchment.

Abstract

Over the past few decades, South Africa has faced severe water shortages, primarily due to the declining quality of its natural water supplies. This decline has further strained irrigation standards, directly impacting crop yields, livestock health and soil fertility. This has emphasised the need for advanced, near-real-time approaches to assess and monitor key water quality parameters affecting irrigation water quality, such as water temperature, total suspended solids (TSS), and Chromophoric dissolved organic matter (CDOM). This research explores the utility of UAV-based remote sensing for monitoring water quality parameters in small reservoirs to address the limitations of traditional remote sensing and ground-based methods, which are often labour-intensive, costly and lack sufficient spatial and temporal coverage. Chapter 1 introduces the research problem, highlighting the increasing pressure on water resources in southern Africa’s resources due to climate change, population growth and land-use changes. It outlines the study's objectives, which include developing a robust methodology for using UAV-derived data to monitor key water quality parameters and improving decision-making in water resource management at the farm scale. Chapter 2 presents a global systematic review of the literature for UAV-based remote sensing for water quality monitoring. It critically evaluates advancements in sensor technology, machine learning algorithms and statistical approaches, identifying key research gaps. The chapter emphasizes the potential of UAVs to provide high- resolution, real-time data but notes challenges such as cost, regulatory constraints and the lack of standardized validation protocols. Chapter 3 provides a case study of the High Flight Farm dam in the uMngeni catchment, illustrating the application approach of UAV-derived data in monitoring water temperature, TSS, and CDOM. The study demonstrates the integration of UAV-based observations with machine learning techniques and model development to produce high-accuracy predictive spatial maps that inform sustainable agricultural practices. Finally, Chapter 4 synthesises the findings, addressing limitations such as weather and operational constraints while offering recommendations for future research. These include expanding research on underrepresented water bodies and promoting interdisciplinary collaborations to enhance the accessibility and scalability of UAV technology in water quality monitoring.