Advancing Water Quality Monitoring and Prediction Capability of USGS NGWOS Program with Satellite and Drone Remote Sensing Technologies

Rivers and lakes provide valuable freshwater resources for human consumption, agriculture, industry, fishery, and other economic activities. Due to global climate changes and destructive land use practices, freshwater resources in the United States and many other countries are increasingly experiencing widespread deterioration, manifested as poor water clarity, high turbidity, eutrophication, and Harmful Algal Blooms (HABs). This project will develop comprehensive testbed datasets, computational algorithms and methods, and software tools for satellite and drone remote sensing of river and lake water quality. A set of innovative remote sensing models and computational algorithms will be developed to assess and monitor key water quality parameters, including Chlorophyll-a/algal blooms, turbidity/clarity/suspended sediment/Secchi depth, and Colored Dissolved Organic Matters (CDOM).

The primary remote sensing observations used in this project will include satellite multispectral images from Sentinel-2A/B, Landsat-8/9, Planet SuperDove, as well as drone hyperspectral images. Extensive fieldwork will be conducted in multiple sites in the Southeastern and Northeastern US across hydrological and ecological regimes, and in situ sensing with multi-parameter sondes and laboratory-based analysis of grab water samples will generate testbed datasets for the calibration, validation, accuracy, and transferability assessment of water quality remote sensing algorithms and models. In collaboration with USGS scientists, this project will compare and evaluate different atmospheric and adjacency effect correction strategies and methods to create aquatic (water-leaving) reflectance. A robust method will be developed to assess and quantify adjacency effect in terms of the size and geometry of rivers and lakes and their surrounding topography and vegetation canopy. By exploiting state-of-the-art machine learning and Geospatial Artificial Intelligence (Geo-AI) techniques, this project will create remote sensing models and software tools that are not only able to produce highly accurate water quality data but also spatially transferable and temporally repeatable, hence representing a major technical breakthrough in water quality remote sensing applications.

This project will greatly expand the spatial-temporal coverage of water quality measurements, hence significantly enhancing USGS’s operational capability in routine monitoring and predicting the water quality, as well as enabling water resource planners and managers to detect and track HABs, to issue public alerts and advisories, and to assess the effectiveness of remedial measures. This project will foster the next generation of water scientists and remote sensing specialists through mentoring postdoctoral researchers and graduate students and make substantial contribution to USGS’ Next Generation Water Observing System (NGWOS) Program and National Water Information System Modernization Program.