Evaluating and Optimizing VIIRS Retrievals of Chlorophyll-a and Suspended Particulate Matter in Turbid Lakes Using a Machine Learning Approach

The Visible Infrared Imaging Radiometer Suite (VIIRS) instrument was launched to continue the legacy of the Moderate Resolution Imaging Spectroradiometer (MODIS). Despite recent studies demonstrating the use of VIIRS observations over inland waters, VIIRS has not been widely used to generate water quality products (e.g., chlorophyll-a (Chia) and suspended particulate matter (SPM)] in relatively large turbid lakes. This study examines the quality of VIIRS-derived remote sensing reflectance (R-rs) from four different atmospheric-correction processors with matchups from 13 lakes sized between 107 and 2573 km(2) across the eastern plain of China. The operational R-rs products from the National Oceanic and Atmospheric Administration (NOAA) outperformed R-rs retrieved from other state-of-the-art algorithms with mean uncertainties of 57%, 33%, 20%, and 28% for R-rs (486), R-rs (551), R-rs (671), and R-rs (745), respectively. These uncertainties induced similar to 55% uncertainty in satellite-retrieved SPM and Chl-a from recently developed algorithms in the studied lakes. To improve accuracy in Chl-a and SPM retrievals, a deep neural network was developed for simultaneous retrievals of Chl-a and SPM from VIIRS Rayleigh-corrected reflectance. The model with satisfactory accuracy (mean uncertainties of 28% for Chl-a and 20% for SPM) outperformed other machine learning approaches and nearly halved uncertainties compared to those obtained from satellite-derived R-rs products. Within the 2012-2020 period, high-quality VIIRS-derived Chl-a and SPM across 61 lakes in eastern China had evident interannual variability in SPM hut insignificant temporal variations in Chl-a. This study provides validated, high-quality, basin-scale VIIRS-derived Chl-a and SPM products in eastern China over the past decade. Our results offer a strategy for improving regional water quality products from VIIRS observations.

Automated summary

This study examines the quality of water quality products generated from VIIRS observations and develops a deep neural network for improving the retrieval accuracy of chlorophyll-a and suspended particulate matter. The results provide high-quality VIIRS-derived water quality products in eastern China over the past decade.