Application of Data Mining and Machine Learning Techniques for Geohydrological Datasets in South Africa

Aug 4, 2026

Upload by Aurora

A desktop study was conducted to research data-driven modelling approaches for classifying relationships between borehole parameters and geological settings. Because borehole surveying and drilling are costly, applying data mining and machine learning to national groundwater databases and other spatial datasets can improve insight for managing groundwater resources. Five machine learning algorithms were evaluated on a consolidated dataset and benchmarked across three Vegter-region case studies. Results showed high-accuracy modelling for static water levels, while yield modelling was less reliable, leaving uncertainty about water-strike yield drivers.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages166

Reading Time6 h 58 min

AccessFree to download

Acknowledgements

Abstract

List of Tables

List of Figures

List of Equations

List of Abbreviations

Chapter 1: Introduction

Background

Problem statement

Aims and objectives

Basic hypothesis

Scope of research

Assumptions and limitations

Research contribution

Dissertation structure

Chapter 2: Literature Review

Introduction

Data mining

Datasets

Data-mining methods

Modelling and forecasting of geohydrological settings

Data-driven modelling techniques

Application of data mining and machine learning techniques for geohydrological datasets in South Africa C de Bruyn [orcid.org/0000-0003-3011-8563](orcid.org/0000-0003-3011-8563) Dissertation accepted in fulfilment of the requirements for the degree Master of Science in Environmental Sciences with Hydrology and Geohydrology at the North-West University Supervisor: Dr SR Dennis Graduation October 2023 24963623 ACKNOWLEDGEMENTS Firstly, I would like to thank YHVH, my Creator, and his Son, Yeshua, for giving me the strength to push onwards through this endeavour. I am thankful to Dr Rainier Dennis and the Centre of Water Sciences and Management who granted me this opportunity and who guided me through this process, equipping me with the proper tools and knowledge. To my parents, for all the love and support through this tough time. I would not have been able to pursue this degree, let alone finish it without them. Also, my brother for giving me advice on atopic which was completely new to me at the start of this study. And my fiancé, for encouraging me to finish what I started. Finally, I am grateful for having a friend in Lohan Bredenhann, who supported me and gave technical advice regarding this academic pursuit. ABSTRACT A desktop study was conducted to research data-driven modelling techniques to classify relationships between borehole parameters and the relevant geological setting. Borehole surveying and drilling is a costly endeavour and by applying data mining and machine learning techniques to national groundwater databases and other available national datasets such as spatial data, better insight and improvements on management of groundwater resources can result. Five machine learning algorithms were tested on a consolidated dataset and their performances compared in order to establish which algorithm yielded the most accurate results. It was established that Random Forest Regression and Classification could be used to model yield, and Support Vector Regression and Random Forest Classification could model static water levels. The algorithm was tested on three case study areas, based on Vegter regions. The results indicated that static water levels could be modelled with high rates of accuracy, but yield modelling was not as successful, and a lot of uncertainty still remains as to the drivers behind water strike yield. Keywords: data mining, machine learning, groundwater resource management, geohydrological datasets, data-driven modelling; water level modelling, yield modelling. TABLE OF CONTENTS ACKNOWLEDGEMENTS .......................................................................................................... I ABSTRACT .............................................................................................................................. II LIST OF TABLES .................................................................................................................... IX LIST OF FIGURES.................................................................................................................... X LIST OF EQUATIONS ...........................................................................................................XIV LIST OF ABBREVIATIONS ....................................................................................................XV CHAPTER 1: INTRODUCTION................................................................................................. 1 1.1 Background ...................................................................................................... 1 1.2 Problem statement ........................................................................................... 2 1.3 Aims and objectives ......................................................................................... 3 1.3.1 Aims ...........................................................................................................................

FAQs

What is the main goal of the study on South African geohydrological datasets?

To use data-driven modelling to classify relationships between borehole parameters and geological settings, supporting improved groundwater management.

Which machine learning algorithms performed best for static water levels and yield?

Random Forest Regression and Classification were used to model yield, while Support Vector Regression and Random Forest Classification were used for static water levels.

How did the results vary between static water level modelling and yield modelling?

Static water levels were modelled with high accuracy, but yield modelling was less successful, with significant uncertainty about drivers of water-strike yield.

Aug 4, 2026

Upload by Aurora