Improving Water Quality Time-Series Prediction in Hong Kong using Sentinel-2 MSI Data and Google Earth Engine Cloud Computing

TL;DR

This study tackles coastal water-quality forecasting in Hong Kong by fusing Sentinel-2 MSI time-series with Google Earth Engine to predict $Chl-a$, $SS$, and turbidity using an LSTM model. It builds robust predictors from band reflectances, including band ratios and a line-height metric, and validates the approach with cross-validated temporal sequences (300 training points, 52 validation) across 2015–2021. The results show improved prediction for $SS$ and turbidity and competitive performance for $Chl-a$ relative to prior works, highlighting the value of remote sensing for continuous, large-scale water-quality monitoring in dynamic, case-2 coastal waters. The work demonstrates the practicality of cloud-based, satellite-driven time-series modeling for proactive water resource management in densely urban coastal systems, while outlining avenues for enhancement with additional architectures and ensemble methods.

Abstract

Effective water quality monitoring in coastal regions is crucial due to the progressive deterioration caused by pollution and human activities. To address this, this study develops time-series models to predict chlorophyll-a (Chl-a), suspended solids (SS), and turbidity using Sentinel-2 satellite data and Google Earth Engine (GEE) in the coastal regions of Hong Kong. Leveraging Long Short-Term Memory (LSTM) Recurrent Neural Networks, the study incorporates extensive temporal datasets to enhance prediction accuracy. The models utilize spectral data from Sentinel-2, focusing on optically active components, and demonstrate that selected variables closely align with the spectral characteristics of Chl-a and SS. The results indicate improved predictive performance over previous methods, highlighting the potential for remote sensing technology in continuous and comprehensive water quality assessment.

Figures (1)

• Figure 1: Time-Series Visualizations of the Models