Philosophiae Doctor - PhD (Earth Science)

Permanent URI for this collection

https://hdl.handle.net/10566/17483

Browse

Now showing 1 - 20 of 73

Using machine learning algorithms to develop a remotely-sensed framework for drought monitoring in different climate regions in South Africa
(University of the Western Cape, 2025) Bhaga, Trisha Deevia
Droughts pose a significant threat to rainfed smallholder farming systems, particularly in regions with varying climatic conditions. This study aimed to develop a spatial modelling framework for assessing the occurrence and frequency of droughts across different climatic zones, with a focus on rainfed smallholder farms. Specifically, this study sought to develop a spatial modelling framework for assessing the occurrence and frequency of droughts on rainfed smallholder farms across different climatic zones in South Africa i.e. the Limpopo, North West, Mpumalanga and Gauteng Provinces. By integrating satellite-derived vegetation indices, specifically the Normalised Difference Vegetation Index (NDVI) and the Modified Soil-Adjusted Vegetation Index 2 (MSAVI2), with state-of-the-art machine learning algorithms, the accuracy of drought mapping in rainfed smallholder farms was significantly enhanced. The study employed the Gradient Boosting Decision Tree (GBDT), Support Vector Machine (SVM), Convolutional Neural Network (CNN), Random Forest (RF) and Transformer models to capture complex spatial and temporal patterns of drought dynamics. The results showed that the Transformer model was effective in detecting rainfed smallholder farms (with an Overall Accuracy of 0.85 and a mean Intersection over Union (IoU) of 0.86). The study further evaluated the agricultural and meteorological drought conditions from 2004 to 2023. The Vegetation Condition Index (VCI) from SPOT VEGETATION 1 and PROBA-V data and the Standardised Precipitation Index (SPI) from CHIRPS data were computed.
The use of cover crops for sustainable citrus orchard performance
(University of the Western Cape, 2025) Silwana, Sibongiseni
Cover crops play a pivotal role in maintaining and improving agricultural sustainability through enriching organic matter content. They are commonly planted between tree rows to enhance soil quality, suppress weeds and improve production and fruit quality in citrus orchards. However, there is limited information regarding plantation of cover crops in citrus orchards in South Africa. Current conventional agricultural production systems in citrus orchards rely heavily on a high level of agrochemicals. This study is an attempt to alleviate some of the direct input costs and high application of agrochemicals. A cover crop field study was conducted at Lamara farm at Franschhoek, South Africa to assess the effect of cover crop species on soil quality, weed suppression, soil enzymes, citrus fruit yield and soil water content under two termination methods (slashed and non-slashed). Cover crop species tested were vetch, medics, oats and control (no cover crop). The experiment was set up in a randomised block design with a factorial treatment structure (Factor 1 - cover crop species; Factor 2 – cover crop termination methods) replicated six times. In the first study, effect of cover crop species on soil organic carbon, weed control and citrus fruit weights using two termination methods (slashed and non-slashed) were assessed. Soil samples were taken before planting cover crops and one year after planting cover crops to determine soil fertility. Dry weight samples of cover crop and weed were collected during the termination stage and fruit weight measurements were taken at harvest.
Earth observation technologies for improved agricultural decision support systems in South Africa
(University of the Western Cape, 2025) Mpakairi, Kudzai Shaun
This work investigates the application of remotely sensed data and advanced machine learning techniques in enhancing sustainable agricultural practices in South Africa, focusing on crop monitoring, water use efficiency, and land management. Firstly, a systematic review of remote sensing applications in Southern African agriculture, evaluating key advancements, challenges, and opportunities was conducted to document the key scientific knowledge gaps that then informed the focus of this study. The findings of the review revealed that the adoption of remotely sensed data and machine learning algorithms in agriculture remains in its infancy. Building on these insights, this study proposed a methodological framework for delineating irrigated and rainfed croplands in South Africa. By leveraging high-resolution Sentinel-2 satellite imagery and advanced machine learning techniques—including Deep Learning Neural Networks (DNN) and Random Forest (RF)—the study demonstrated the effectiveness of these models in generating accurate, large-scale agricultural land-use maps, achieving an overall classification accuracy of 0.71. Further, a novel approach for crop classification was also introduced by integrating unsupervised learning techniques and spectral matching algorithms, enabling accurate identification (OA = 0.84, p-value = 0.01) of major crop species across South Africa’s diverse agricultural landscapes. Additionally, the study employed multi-temporal MODIS satellite imagery to quantify annual crop water use (CWU) and crop water productivity (CWP), revealing substantial spatiotemporal variations between irrigated and rainfed croplands. Irrigated croplands generally had higher annual CWP (>0.002 kg/mm3/yr), while rainfed croplands consistently showed low CWP in forestry (0.001 kg/mm3/yr) and sugar (0.0012 kg/mm3/yr) agricultural regions.
Remotely sensed delineation of groundwater dependent ecosystems in Khakhea-Bray Transboundary Aquifer
(University of the Western Cape, 2025) Rampheri, Mangana Berel
Remote sensing of groundwater-dependent ecosystems (GDEs) has increased substantially in recent years. Of significant prominence, is the delineation and mapping of groundwater-dependent vegetation (GDV), species diversity, and water quality in these ecosystems. Groundwater-dependent ecosystems provide several ecological services such as habitat for wildlife fauna, carbon sequestration and water purification. The recent technological advancements and readily accessibility of new satellite sensors with improved sensing characteristics have resulted in numerous state-of-the-art applications for GDEs assessment and monitoring. These studies were done at varying scales, essentially in light of global climate change and variability. In this study, progress on the remote sensing of GDEs in semi-arid environments is assessed. The study presents the key trends in GDEs remote sensing that underpin many of the recent scientific research milestones and application developments. In addition, it observed a considerable shift towards the use of advanced spatial modelling techniques, using high-resolution remotely sensed data to further improve the characterisation and understanding of GDEs. Thus, literature shows the successful use of freely available remotely sensed data in mapping GDEs.
Multispectral remote sensing of groundwater dependent ecosystems in the Kruger National Park, South Africa
(University of the Western Cape, 2025) Chiloane, Chantel Nthabiseng
Groundwater dependent vegetation (GDV) provides crucial ecosystem services for ecological and socio-economic development. The significance of GDV is amplified in semi-arid rangelands where they provide habitat, and forage for wildlife during the dry periods when surface water resources are scarce. Continuous assessment and conservation of groundwater-GDV in semi-arid rangelands, such as Kruger National Park (KNP) is crucial for maintaining ecosystem resilience and biodiversity amid the challenges posed predominantly by climate change. This study investigates the integration of remote sensing, machine learning, and climate modeling to delineate, map, and monitor the spatial distribution, diversity, and climate vulnerability of GDV in the Kruger National Park, South Africa. By utilizing cloud computing platforms like Google Earth Engine (GEE) in conjunction with open-source datasets and crowd-sourced field observations, this research sought to demonstrate the potential for conducting large-scale and fine-grained ecological assessments. To address the aim of this study the following specific objectives were drawn (i) to develop remote sensing-based techniques for delineating and mapping the spatial distribution of GDV in the Kruger National Park, (ii) to assess the potential of spatially explicit techniques to determine species diversity within groundwater dependent ecosystems (GDEs), (iii) to explore the spatio-temporal variations of vegetation diversity in GDEs and the driving forces for the observed changes and (iv)to determine climate change effects on the habitat suitability for GDV under the moderateclimate change scenario.
Assessment for sustainable water utilization from the Huixian karst wetland in Guilin of China
(University of the Western Cape, 2025) Qi, Jihong
Karst wetland as a special type of the inland wetland has significant ecological functions and associates socio-economic values. Huixian karst wetland located in Guangxi Zhuang Autonomous Region in the southwest of China providesthe irreplaceable goods and services for local development, as the focus of the study in this thesis. With the urbanization in terms of population growth and housing construction, however, Huixian wetland was kept in change in land use practices and increase in local use of water resources over time, especially in the past 40 years. The wetland suffers from shrinkage in size at an alarm rate and serious degradation of ecological functioning. It takes much negative effects on local residential livelihood and economic development. This thesis focuses on the core area of Huixian wetland as the study area aiming to assess its water utilization while maintaining the currently ecological health. Considering the complex hydrogeologic regimes of the wetland itself and the lack of historical monitoring data on hydrology locally, a methodology consisting of multimethods and techniques were reviewed and adapted at a macro scale of the problem, including field surveys, water balance, remote sensing (RS), geographic information system (GIS), ecological water requirement and numerical simulation. At the beginning of the study, the dedicated field geomorphological investigations for many times were carried out, which assisted for understanding of the relationship of each component and the water circulation paths of the study area and paved the way for the follow-up methods. The application of the RS and GIS were used to extract the spatial and temporal https://uwcscholar.uwc.ac.za/home 2 change of the hydrological regimes, land use and growth condition of aquatic vegetations over the study area during long-term period, which were used to patch up the absence of long-term monitoring data. Based on the analysis results, ecological water requirement of wetland was determined in terms of selected the indicator of the aquatic vegetation for limiting factor to control the water utilization, in order to avoid continued loss of the wetland function further more. And then, a hydrogeological conceptual model was built to simulate the groundwater flow regime. A zone of mobile groundwater divides within the wetland catchment between Lake Mudong and Fenshui Pond was conceptualized and verified as opposed to a postulated single fixed line of a watershed. The zone or belt of drifting divides occurred, whereas a crucial role of Lake Mudong it the belt formulation was analyzed. Based on this study, it was concluded that the impact of the supply water scenario of less than 20% withdrawal of groundwater recharge for meeting sustainable water utilization of local community was the acceptable margin of the fluctuation in order to maintain the current wetland integrity and health without getting further degradation.
Impact of bush encroachment on African savannah rangelands in the face of climate change and variability
(University of the Western Cape, 2024) Maphanga, Thabang; Dube, Timothy
In the African continent, especially in the southern part of the continent, climate change has been associated with significant changes in temperature and precipitation patterns. Some of the signs of climate variability noted in Southern Africa are low and erratic precipitation that leads to drought, along with increases in temperature. Due to the climate variability, terrestrial ecosystems such as savannah rangelands have been adversely affected which has led to the disruption of ecosystem services and spatial configuration. For example, studies have shown that rainfall variability in the arid savannah regions is the primary determinant of the rate of bush encroachment. In rangelands, climate change has a significant impact on the transition from grass-dominated species to woody species, which is influenced by the specific weather patterns in the area. In this regard, it is pivotal to understand the rate of bush encroachment within the savannah rangelands of South Africa. The impact of bush encroachment on savannah rangelands amid climate variability in (un)protected areas remains under explored, despite extensive works in bush encroachment in general. In order to gain a thorough understanding of how environmental changes affect the growth of bushes, both in untouched landscapes and in human-manipulated environments. The main aim of this study was to assess the utility of remotely sensed data in detecting and mapping the extent of bush encroachment in the African savannah rangelands in the face of climate change and variability. To address the overarching aim of this study, four specific objectives were set. Firstly, a review of literature on advancements in satellite sensing of the impacts of climate and variability on bush encroachment in savannah rangelands was conducted.
Estimating groundwater recharge in Jinci karst groundwater system, North China
(Universty of the Western Cape, 2024) Shen, Haoyong; Xu, Yongxin
More than 25% of the world’s population relies on karst groundwater as the primary source for drinking and irrigation. However, due to the global climate change as well as population expansion, the contradictory between supply and demand of water resource becomes more and more pronounced. Karst groundwater supply drinking water and irrigation water to more than thirty large cities in northern China due to its high concentration, stable groundwater flow and good water quality. Meanwhile, due to the excessive development of karst groundwater, a series of aquatic ecological problems have arisen, such as over 30% karst large springs are in process of drying up, the regional karst groundwater level declining and the degradation of karst groundwater quality in the past forty years.
Reservoir characterization and sedimentary architecture of the siliciclastic sediments in the southern Pletmos basin, South Africa
(Universty of the Western Cape, 2021) Oghenekome, Monica Enifome; Chatterjee, Tapas
The present research characterizes the siliciclastic reservoirs sediments of the southern Pletmos Basin and provide insights into the reservoir mineralogical variation, rock type, depositional environment for accurate identification and quantification of the hydrocarbon
Application of geochemical characterization, multivariate statistics and geological modelling in assessment and prediction studies on selected coal and gold mine waste in South Africa
(University of the Western Cape, 2023) Abegunde, Oluseyi Blessed; Opuwari, Mimonitu
Over the years, South Africa has generated vast amounts of coal fly ash and gold slime tailings, constituting over 70% of the country's waste materials. These byproducts contain elevated levels of trace metals, posing a potential threat to the environment upon release. Addressing this issue requires a comparative study of the environmental impact of coal fly ash and selected mine tailings on water resources and land pollution. This research aims to investigate and compare leachability, metal release, oxidation effects, and environmental pollution between coal fly ash and gold tailings. By contrasting these aspects, the study seeks to enhance understanding of the potential risks associated with these materials, aiding informed decision-making for their management and regulation. Additionally, the research explores the correlation between gold tailings' acid potential generation and coal fly ash's alkaline potential generation in terms of leachability, metal release, oxidation effects, and environmental pollution. The research employed comprehensive laboratory experiments and analytical investigations, including leaching tests under simulated weathering conditions. A total of 51 gold tailings samples and 66 coal fly ash samples were analysed through SEM and XRD for mineralogical insights and ICP-MS and XRF for geochemical analyses. Statistical analysis revealed the significant roles of pH, Fe ions, Ca2+, and Mg2+ in metal extraction from both materials. Notably, the study identified key factors contributing to the environmental impact of coal fly ash and gold tailings. SEM imagery highlighted heterogeneous characteristics in gold tailings, while factor analysis indicated the potential release of ferrous ionic species, contributing to acidity. Trace elements like Ni, Zn, Pb, and Cu were predominantly associated with Fe/Mn oxides during leaching experiments, facilitating their mobilization with acid-generating ions.
Rainfall induced transient pressure wave mechanisms and pore water pressure dynamics in tailings
(University of the Western Cape, 2023) Theron, Charné; Yongxin, Xu
Tailings Storage Facilities are some of the largest engineered structures responsible for the containment of mining waste, yet some physical/hydrological processes causing slope instability are still poorly understood. Previous studies conducted on the causes of failure of TSFs indicate rainfall-induced slope instability as the main trigger. However, the generation of certain physical/hydrological processes and the behaviour of soil hydraulic properties in the vadose zone when exposed to high intensity rainfall events, have only been considered on natural hillslopes. The purpose of this study is to investigate one such process resulting from high rainfall intensities on partially saturated conditions. In particular, it is hypothesised in this work that the wetting front advance from high rainfall intensities, anticipated under the force of climate change, may cause transient air pressure waves to change pore water suction and elevate the phreatic surface, subsequently impacting on the effective stress. The mechanisms that lead to the mobilisation of pre-event water through transient air pressure waves are known as Groundwater Ridging and the Lisse Effect. In this study, it was hypothesized that these phenomena contribute significantly to slope instability.
An assessment of water use by riparian and non-riparian invasive alien plants and the effects on water resources at the catchment scale
(University of the Western Cape, 2023) Mkunyana, Yonela Princess; Mazvimavi, Dominic
The diversity and increasing extent of invasive alien plants (IAPs) are currently a global ecological and economic problem, especially in semi-arid regions. The introduction of IAPs worsens the strain in South Africa’s water resources by reducing the mean annual runoff by approximately 7%. Further work needs to be done to ensure that those IAPs that have the greatest impact receive priority. Effective management of IAPs promotes the conservation and sustainable use of the natural environment, the economies it supports, and the high quality of life in society. The monitoring of the distribution of IAPs is therefore an important consideration in catchment management. Remote sensing and cloud computing platform, Google Earth Engine (GEE) were used to map land use/landcover changes between the 1973 – 2021 period using Landsat imagery.
Spatial and temporal dynamics of flows and pools along non-perennial rivers in arid and semi-arid areas, South Africa
(University of the Western Cape, 2023) Maswanganye, Sagwati Eugene; Dube, Timothy
This study aimed to assess the use of multi-source remotely sensed data for monitoring the spatial distribution and dynamics of pools in two distinct sites (Touws and Molototsi) in South Africa. Various water extraction indices, including NDWI, Modified NDWI, and NDVI, were employed, along with a random forest classifier and Sentinel-1 SAR data, to map pools and their dynamics in both locations. The remote sensing methods effectively detected and mapped pools with satisfactory accuracy, except for small pools below 400 square meters. The study identified flow occurrences and rainfall as significant factors influencing changes in pool sizes. However, the interaction between pools and groundwater required further investigation and showed no conclusive evidence in this study.
Hydro geochemical characteristics of the basement aquifers in namaqualand
(University of the Western Cape, 2003) Titus, R.A; Xu, Y
At the onset of this research programme it became apparent that there is a dearth of research studies focusing on the groundwater resources of the region. As a result, a conceptual representation of the hydrogeological system (i.e. specific aquifer systems, groundwater flow regime, etc.) for the perceived problematic aquifer systems in the Namaqualand region did not exist. The research project contributes significantly to conceptual representation of the aquifer system, in particular a quantitative understanding of the regions' groundwater resources in terms of its hydrochemical development and the construction of a groundwater flow model at regional scale as well as at a local scale. The groundwater resources for the Namaqualand region are developed predominantly in the basement rocks. The infiltration and flow of water is controlled by the prevailing complex fracture network and can vary in space and time. Such observations relate to structurally controlled flow systems and varying water chemistry amongst closely spaced fracture systems.
A decision-making framework for groundwater management in arid zones (with a case study in Namaqualand).
(University of the Western Cape, 2004) Pietersen, Kevin Claude; Xu,Yangwin and Green, George Clifford
The main aim of the work on which this thesis is based was to develop a framework for sustainable management of groundwater resources in arid zones with emphasis on Namaqualand in the north western region of South Africa. The first part of the thesis focuses on describing the groundwater resource base and legislative framework for groundwater management. Most aquifers in South Africa occur in fractured rock ranging in age from earliest Pre-Cambrian to Jurassic. Primary aquifers are mostly restricted to the coastal plains and river deposits. Characterisation of the fractured rock aquifers has been limited. Thus, an inadequate knowledge base exists in a number of hydrogeological domains to understand the attributes and dynamics of fractured rock aquifers. A serious shortcoming in our knowledge base is to understand the institutional arrangements necessary for proper resource management. This is particularly important in view of the National Water Act of 1998, where groundwater is subject to the same protection measures as surface water. These measures are sophisticated and require tools and technologies to be developed to support sustainable groundwater management and utilisation. The technical, economic, social, legal, political, and environmental issues affecting groundwater management of arid zones of South Africa were analysed. The objectives and appropriate measures to overcome the barriers for sustainable development were also presented. This was necessary to structure the decision problem and to generate and identify the decision alternatives. Thus decision alternatives have been formulated for the various elements of the objectives in order to achieve preferred scenarios. The main part of the thesis was the development of a decision-making framework for groundwater management in the arid zones of Namaqualand. A multiple criteria decision analysis (MCDA) approach was adopted to assist in the formulation of the decision-making framework. ln order to select the most appropriate MCDA technique, some background was required on the theoretical aspects. A value function method was selected, which provides decision support by interval SMART/SWING. This method incorporates informational uncertainty through interval judgements. The decision problem supported the selection of this method because of the discrete alternatives and uncertainty associated with groundwater management. Further, the method provided an interactive technique to interrogate various decision alternatives based on prior knowledge of the decision-maker. The software WINPRE was utilised in the value tree construction and analyses, A number of value trees and attributes were defined. These attributes were evaluated against the identified alternatives. This provided a systematic framework for the analytical understanding of the problem. As a result a number of preferred alternatives were elicited. The analyses resulted in a model for groundwater management in arid zones. The ideal state for groundwater management was presented, but in practice this is difficult to implement, mainly because of human and financial resources. As a result, a critical path was established based on the analyses done in this thesis. This was applied to the Namaqualand example. ln this example the activities to support a Catchment Management Strategy was identified. This means that the decision-maker is able to focus on issues that are deemed important. More importantly, the method allows the decision-maker to develop the various consequences of the alternatives with stakeholders. As a result of the decision model a number of strategies were proposed for sustainable groundwater management in arid zones with a case study in Namaqualand. A consequences table were developed which could be used for M&E purposes.
Applicability of big data analytics to support groundwater management in Southern Africa
(University of the Western Cape, 2022) Gaffoor, Zaheed; Kanyerere, Thokozani
Groundwater is a vital resources for member states in the Southern African Development Community (SADC). However, sustainable managing groundwater resources in SADC is a challenge. Amongst the many issues facing groundwater managers in SADC, limited access to high fidelity data, as well as a poor understanding of the techniques needed to transform the data into information, has hampered the decision making process. Big data relates to large, voluminous and heterogenous datasets which are being generated through numerous activities, such as computer simulations, remote sensing, commercial transactions, internet activity, monitoring networks, IoT sensors, historical documents, social media and many others.
Developing an integrated remotely sensed framework for the detection and monitoring of seasonally-flooded wetlands in semi-arid environments of southern Africa
(University of the Western Cape, 2022) Gxokwe, Siyamthanda; Dube, Timothy
Wetlands are among the most important ecosystems on earth; they cover approximately 4-6% of the earth’s surface and offer critical eco-hydrological services. However, these ecosystems are under threat from anthropogenic activities, droughts and climate variability, as well as from global environmental change. It is estimated that over 60% of the world’s wetlands have been lost due to climate change and variability, as well as other anthropogenic influences. There is, therefore, a need for their routine monitoring and assessment to ensure the sustainable use and management of these systems on a national, regional and local scale, and prevent their further degradation and loss. This study aimed at developing an integrated cloud-computing-based, remotely-sensed framework for the detection and monitoring of small and seasonally-flooded wetlands along the semi-arid Limpopo Transboundary River Basin of southern Africa, which was previously a challenging task when using the traditional assessment and monitoring methods.
Streamflow generating mechanisms in a mountainous Jonkershoek catchment, South Africa
(University of the Western Cape, 2022) Mokua, Retang Anna Lapalemabu; Mazvimavi, Dominic
The Jonkershoek valley catchment in a mountain region that forms the headwaters of the Eerste River, which contributes water to the Cape Winelands District Municipality and parts of the City of Cape Town in the Western Cape Province, South Africa. These areas include highly urbanized and agriculturally advanced regions, which are vulnerable to occasional water shortages. Understanding the processes and factors accounting for the spatial and temporal variations of streamflow in this area is critical. This information can be used to build appropriate conceptual models of this catchment’s functioning to support the management of water resources.
Geological and hydrgeological characteristics of acid mine drainage from an abandoned coal mine: A case study of Shandi coal mine in Niangziguan spring catchment, Shanxi
(University of Western Cape, 2022) Wang, Zhaoliang; Xu, Yongxin
As Shanxi Province is rich in coal resources, the output of raw coal approximately accounts for one quarter of China's production. It is one of the most important energy and heavy chemical bases and plays a significant role in the sustainable development of the national economy and energy security. With the continuing exploitation of coal resources in Shanxi, water environmental problems such as the destruction of water resources and deterioration of water quality have become increasingly prominent. Especially with the closure of many depleted coal mines, water pollution caused by acid mine drainage (AMD) has become more and more serious, which aggravated the shortage of water resources and threatens the safety of local drinking water supply. Since 2008, more than 100 coal mines have been abandoned in the Yangquan coalfield of Shanxi, and the AMD has polluted the surface water and groundwater in the Niangziguan spring catchment.
GIS and remote sensing-based integrated modelling of climate and land use change impacts on groundwater quality: Cape Flats Aquifer, South Africa
(University of Western Cape, 2021) Gintamo, Tesfaye Tessema; Kanyerere, Thokozani
The need to ensure groundwater security is vital, particularly in urban areas. Assessing the impact of land use and climate variables on groundwater quality can help improve sustainable management. The vulnerability mapping of groundwater contamination identifies high-risk areas. Using models and technologies that forecast the distribution of contamination risk over time and place can help prioritize groundwater monitoring. Based on such needs, the Cape Flats aquifer in Cape Town, South Africa, was chosen as the case study for assessing the potential for groundwater contamination risk in urban and coastal hydrogeological settings. The Cape Flats aquifer has been highlighted as an alternate water supply source to augment current supply sources in Cape Town. However, the shallow aquifer is under pressure from agricultural and industrial activities and long-term climate variables, among other factors.

Browse

Recent Submissions