Multispectral Insights into Turbidity Variations Over Time in The Derawan Island

Abstract This investigation delves into the temporal and spatial dynamics of turbidity in the Derawan Islands, Indonesia, utilizing a comprehensive approach that combines multispectral satellite imagery from Sentinel-2, in-situ measurements via AAQ Licor sensors across 165 observation stations, and continuous data collection from a Turbidity Logger. Situated within the Coral Triangle, understanding the turbidity variations in the Derawan Islands is crucial for the conservation and sustainable management of its marine ecosystems. Our research developed and validated empirically derived algorithms to accurately estimate turbidity, utilizing a strategic partition of in-situ data—70% for model development and 30% for validation. This approach resulted in robust models, demonstrating their efficacy with Root Mean Square Error (RMSE) values as low as 0.85 and R-Squared (R 2 ) values up to 0.56, indicating a high degree of accuracy in satellite-derived turbidity measurements. The study unveiled significant spatial and temporal turbidity heterogeneity, linking these variations to both natural and anthropogenic factors. The high-resolution data from the Turbidity Logger revealed critical diurnal fluctuations and short-term turbidity events, providing insights into the dynamic marine environment of the Derawan Islands.