Ocean color satellites provide essential insights into water quality and ecosystem dynamics by estimating chlorophyll, suspended matter, and dissolved organic material. Atmospheric correction, the process of removing scattering and absorption from satellite signals, is central to these analyses. Traditional algorithms assume that near-infrared signals from seawater are negligible, a simplification that often fails in turbid coastal regions. Compounding this issue, most models treat seawater absorption as constant, overlooking natural variability driven by temperature shifts. Such oversights can distort data products used for fisheries, pollution tracking, and climate studies. Due to these problems, there is a pressing need for algorithms that explicitly account for environmental variability.
Researchers from the Ocean University of China and collaborators report (DOI: 10.34133/remotesensing.0731) a major step forward in remote sensing accuracy in the Journal of Remote Sensing (September 5, 2025). Their study introduces ACiter-T, an upgraded atmospheric correction algorithm that incorporates seawater temperature effects into satellite processing. Using simulated data and more than 500 satellite–in situ matchups, the team demonstrated that the temperature-adjusted method significantly improves retrievals in turbid and cold waters. This innovation strengthens the reliability of global ocean color monitoring systems.
The team focused on a common correction scheme, the Near-Infrared Iterative ACiter. While robust for many conditions, ACiter assumes a fixed absorption coefficient for pure seawater at 22 °C. This study revealed that deviations from this baseline, especially greater than ±10 °C, can skew reflectance measurements, particularly in the blue spectral bands.
To resolve this, the researchers created ACiter-T, which integrates temperature-adjusted absorption coefficients derived from laboratory and field data. Tests using synthetic datasets across a wide range of chlorophyll, suspended particulate matter, and water temperatures showed that errors in reflectance decreased systematically when temperature was considered. For highly turbid and cold waters, mean absolute percentage differences dropped by more than 50% compared to the baseline algorithm.
Validation using 528 matchups from AERONET-OC coastal monitoring sites confirmed these improvements. At Belgium's MOW1 site, for example, errors in blue-band reflectance decreased from over 23% to less than 10%. In less turbid waters, however, temperature corrections made little difference—indicating that ACiter-T is most impactful under dynamic, complex conditions. Together, these results highlight the critical role of thermal variability in shaping satellite ocean color retrievals.
“Our findings demonstrate that seawater temperature is not just a background factor but a key driver of accuracy in satellite ocean color products,” said lead author Junwei Wang. “By refining the atmospheric correction process with temperature-sensitive parameters, we can significantly improve the reliability of data used for monitoring coastal ecosystems. This ensures that policy makers, fisheries managers, and climate scientists are basing decisions on more precise information, particularly in regions where environmental changes are most rapid and impactful.”
The ACiter-T algorithm represents a practical enhancement for operational ocean color missions, including those run by NASA and other agencies. By correcting for temperature-driven optical changes, satellites can deliver more reliable assessments of chlorophyll concentrations, sediment transport, and water quality in coastal and estuarine systems. This has direct applications for aquaculture, pollution monitoring, and climate adaptation strategies. Beyond immediate benefits, the study underscores the importance of incorporating dynamic environmental variables into Earth observation models. Such advances will be critical as global warming increases seawater temperature variability, challenging current remote sensing frameworks.
###
References
DOI
10.34133/remotesensing.0731
Original Source URL
https://doi.org/10.34133/remotesensing.0731
Funding information
This work was supported by the National Natural Science Foundation of China (grant numbers T2222010 and 42306197), the Civilian Aerospace Technology Pre-research Program (grant number D010202), and the National Key Research and Development Program of China (grant numbers 2022YFC3104900 and 2022YFC3104903).
About Journal of Remote Sensing
The Journal of Remote Sensing, an online-only Open Access journal published in association with AIR-CAS, promotes the theory, science, and technology of remote sensing, as well as interdisciplinary research within earth and information science.