Two-decade dynamics of MODIS-derived Secchi depth in Peninsula Malaysia waters

Abstract

Secchi disk depth (Z sd ) is an essential environmental factor for studying ecosystem dynamics and biogeochemical processes in aquatic environments. Monitoring the long-term changes in water transparency is critical to predict the cascading impacts of climate change on marine ecosystems. We investigated the seasonal and interannual dynamics of Z sd in the eastern coast of Peninsula Malaysia (ECPM) and the Straits of Malacca (SoM) using a 21-year time series of MODIS ocean color data. To enable the reliable assessment of Z sd and its long-term variability, the performance of existing and regional algorithms was investigated using in-situ optical measurements collected during different monsoon seasons and in various environmental conditions. Our validation results showed that the existing Z sd algorithms performed adequately, but exhibited large errors, especially at relatively high Z sd values. On the other hand, the regional empirical algorithm based on a direct relationship between remote sensing reflectance and Z sd showed significant improvements by reducing the overall bias observed in existing Z sd schemes. The results indicated that the monthly climatological Z sd over the period 2000–2020 showed distinct patterns in different seasons. The ECPM waters had deeper Z sd than SoM waters. Maximum transparency usually occurred during the southwest and spring inter-monsoon and minimum transparency occurred during the northeast monsoon. Long-term seasonal evolution of Z sd reveals that widespread and persistent anomalies dominated the ECPM and SoM regions. Interannual trends indicate notable and complex changes in Z sd that were probably driven by variability in the ocean-atmosphere dynamics of Niño-Southern Oscillation (ENSO) and local environmental conditions. This study highlights the extensive analysis of Z sd status and its spatio-temporal pattern from space, which can significantly benefit long-term ocean monitoring efforts in the ECPM and SoM regions. • A new regional empirical algorithm was proposed to derive Z sd from satellite ocean color data. • The new algorithm showed the best performance compared to other existing models. • Strong seasonal and interannual variations of Z sd were observed in the ECPM and SoM. • Long-term variations in Z sd were associated with the ENSO cycle and environmental conditions.