Reproduce “Improving Water Quality Time-Series Prediction in Hong Kong using Sentinel-2 MSI Data and Google Earth Engine Cloud Computing” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Rohin Sood, Kevin Zhu
Researchers from Del Norte High School and Algoverse AI Research developed Long Short-Term Memory (LSTM) models for predicting chlorophyll-a (Chl-a), suspended solids (SS), and turbidity in Hong Kong's coastal waters. Utilizing Sentinel-2 satellite data and Google Earth Engine, these models achieved R² values up to 0.91 for Chl-a and significantly improved accuracy for SS and turbidity compared to previous methods.
Key insights
- LSTM networks, when trained on extensive multi-year Sentinel-2 datasets, demonstrate superior capability in capturing temporal variations of water quality parameters compared to traditional machine learning models.
- The integration of Sentinel-2 data with Google Earth Engine provides a scalable and efficient framework for remote sensing-based water quality monitoring in complex coastal environments.
- Careful selection of spectrally relevant band ratios and derived variables from Sentinel-2 significantly enhances the sensitivity and accuracy of predictive models for Chl-a, SS, and turbidity.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.