Dataset from: Ngamile, S, Kganyago, M, Madonsela, S, & Mvandaba, V (2025) Determining the Sensitivity of Sentinel-2 Bands to Optically and Non-Optically Active Water Quality Parameters Under High- and Low-Flow Conditions in the Cradle of Humankind World Heritage Site South Africa

This dataset includes in-situ water quality measurements, laboratory results, and Sentinel-2 MSI reflectance values collected in the Cradle of Humankind World Heritage Site, South Africa. The data support the study on the sensitivity of Sentinel-2 bands to optically and non-optically active water quality parameters under high- and low-flow conditions. Data Collected In-situ measurements: Dissolved Oxygen (DO), Electrical Conductivity (EC), pH, and Temperature. Laboratory-derived measurements: Chlorophyll-a and Suspended Solids. Sentinel-2 MSI surface reflectance values extracted for each sampling point. Sampling Details Data were collected during: High-flow (wet season): 14 & 16 March 2024 Low-flow (dry season): 28 & 31 August 2024Measurements were taken between 09:00 and 15:00 SAST. Water samples were collected at 6 purposively selected sites using a Hach HQ40d meter and analysed at the CSIR Hydrology and Water Resources Laboratory within 72 hours. Contents of the ZIP File The ZIP archive contains CSV files for each sampling date and season. Each CSV includes: Field measurements Laboratory results Sentinel-2 reflectance values Sampling coordinates Processing Data were checked for consistency, merged by sample ID, and linked to Sentinel-2 Level-2A reflectance values extracted using cloud-masked, atmospherically corrected scenes. Intended Use The dataset can be used for: Remote sensing-based water quality modelling Machine learning sensitivity analysis Hydrological and environmental studies Comparing high-flow vs low-flow water quality conditions Ethical Considerations Sampling followed standard environmental protocols, and no human data or restricted biological materials were collected. Keywords: Sentinel-2, Water Quality, Inland Waterbodies, Remote Sensing, Machine Learning