6–10 Jul 2026
University of the Western Cape
Africa/Johannesburg timezone
**Tours now open!** Registration is now closed - All registration payments are due before 23:39 SAST on 26 June.

Towards Physics-Informed Machine Learning for Water Quality Prediction: A Data Assessment Study in South Africa

7 Jul 2026, 11:20
20m
Lecture Hall DL3 (University of the Western Cape)

Lecture Hall DL3

University of the Western Cape

Oral Presentation Track F - Applied Physics Applied Physics

Speaker

Precious Mabidi (University of Venda)

Description

Water quality monitoring is critical for sustainable water resource management, particularly in water-stressed regions such as South Africa. However, the availability, accessibility, and consistency of water quality remain a major challenge, which may be contaminated by waste from different sources like mines, industries and agricultural activities. Some areas in South Africa face serious water scarcity, where consumers are compelled to buy water, and those who cannot afford it use water from rivers or dams, which may be contaminated by waste from various sources, such as mines, industries. This study employs machine learning models including Linear Regression, Decision Trees, and Random Forests to predict water quality. The dataset contains water quality parameters such as pH, turbidity, dissolved oxygen, electrical conductivity, and nutrient concentrations. For electrical conductivity Linear Regression shown the highest predictive performance (R² = 0.625), indicating predominantly linear relationship. Random Forest and Decision Tree models showed moderate performance (R² ≈ 0.56), suggesting limited nonlinear interactions within the dataset. The findings will support the integration of conventional machine learning approaches to enhance water quality prediction, water monitoring, and resource management.

Apply for student award at which level: PhD
Consent on use of personal information: Abstract Submission Yes, I ACCEPT

Authors

Precious Mabidi (University of Venda) David Tshwane (NiThecs) Lufuno Takalani (University of Venda) Tshifhiwa Ranwaha (University Of Venda) Reginah Maphanga (Sol Plaatje University)

Presentation materials