Assessment of Turbidity and Chlorophyll-a Variations in Lagos Lagoon Using Multispectral Remote Sensing

Abstract

Turbidity and chlorophyll-a are important indicators of aquatic environmental conditions because they provide information about suspended particles, phytoplankton abundance, and changes in water quality. Lagos Lagoon, located in southwestern Nigeria, is exposed to multiple environmental pressures associated with urban development, wastewater discharge, surface runoff, sediment transport, and changing hydrological conditions. Conventional water quality monitoring provides essential measurements but may not adequately capture the spatial and temporal variability of these parameters across extensive aquatic environments. Multispectral remote sensing offers an effective complementary approach by enabling repeated observations of water surface reflectance and the identification of spatial patterns associated with turbidity and chlorophyll-a concentrations. This research examines the application of Sentinel-2A and Landsat-8 imagery, combined with artificial intelligence and machine learning for advanced data analytics, to assess variations in these water quality indicators across Lagos Lagoon. The proposed methodology incorporates satellite image preprocessing, atmospheric correction, water masking, spectral index extraction, machine learning regression, temporal analysis, and validation using in situ measurements. Random Forest, Support Vector Regression, and Gradient Boosting are considered for estimating turbidity and chlorophyll-a from multispectral observations. An experimental framework is developed to compare sensor-specific and integrated models using statistical performance metrics, including the coefficient of determination, root mean square error, and mean absolute error. Since processed satellite imagery and field measurements are not provided for this study, the numerical experimental results are presented as hypothetical examples rather than verified observations. The research provides a reproducible framework for investigating seasonal variability, identifying potential areas of elevated turbidity or chlorophyll-a, and improving the integration of satellite remote sensing with environmental monitoring. The proposed approach can support future research, aquatic ecosystem assessment, and evidence-based water resource management in Lagos Lagoon.