Retrieval of Water Quality Parameters in Tasik Kenyir, Malaysia Using Sentinel-2 MSI Satellite Idris M.S. 1*, Khalil I.1 and Mat Amin R.1
1Faculty of Science and Marine Environment, Universiti Malaysia Terengganu, 21030 Kuala Nerus, Terengganu, Malaysia
Abstract
Tasik Kenyir, the largest man-made lake in Malaysia, is a vital freshwater ecosystem that supports biodiversity, fisheries and ecotourism. However, its water quality (WQ) is increasingly threatened by human activities, underscoring the need for long-term monitoring. This study aimed to evaluate the S2-MSI capability to derive key WQ parameters- chlorophyll-a (Chl-a), colored dissolved organic matter (CDOM), suspended particulate matter (SPM) and Secchi disk depth (SDD) in Tasik Kenyir. Using an in-situ dataset of WQ parameters and optical measurements collected during multiple field campaigns, we developed WQ algorithms for the S2-MSI sensor. To ensure reliable evaluation, several atmospheric correction (AC) models were assessed for their accuracy in retrieving water reflectance (Rrs) from S2-MSI data. Among the AC models tested, C2RCC yielded the most accurate Rrs, demonstrating strong agreement with field measurements. Validation results revealed that Chl-a can be accurately estimated using the NIR-blue band ratio, while the blue-green band ratio was more responsive to changes in CDOM and SDD. The results also showed that a single-band Rrs based on the red wavelength was the most reliable predictor of SPM in the study area. The best-performing algorithms for each parameter were then applied to cloud-free S2-MSI images acquired over a 5-year period to assess seasonal variations in WQ. The analysis revealed significant seasonality in all parameters, with maximum values during the northeast monsoon and minimum values during the spring inter-monsoon. Despite these seasonal fluctuations, the spatial distribution of WQ parameters within each season remained relatively uniform, with similar patterns of water characteristics, indicating minimal spatial variation across the lake. This study demonstrates the applicability of S2-MSI for retrieving WQ parameters, offering valuable support for long-term monitoring programs, which are currently insufficient in Tasik Kenyir.
Keywords: water quality, Tasik Kenyir, algorithms, Sentinel-2
