The Inversion of HY-1C-COCTS Ocean Color Remote Sensing Products from High-Latitude Seas

China's first operational ocean color satellite sensor, named the Chinese Ocean Color and Temperature Scanner (HY-1C-COCTS), was launched in September 2018 and began to provide operational data in June 2019. However, as a polar orbiting ocean color satellite sensor, HY-1C-COCTS would inevitably encounter regions impacted by large solar zenith angles when observing the high-latitude seas, especially during the winter. The current atmospheric correction algorithm used by ocean color satellite data processing software cannot effectively process observation data with solar zenith angles greater than 70 degrees. This results in a serious lack of effective ocean color product data from high-latitude seas in winter. To solve this problem, this study developed an atmospheric correction algorithm based on a neural network model for use with HY-1C-COCTS data. The new algorithm used HY-1C-COCTS satellite data collected from latitudes greater than 50 degrees N and between April 2020 and April 2021 to establish a direct relationship between the total radiance received by the satellite and the remote sensing reflectance products. The evaluation using the test dataset shows that the inversion accuracy of the new algorithm is relatively high under different solar zenith angles and different HY-1C-COCTS bands (the relative deviation is 3.37%, 7.05%, 5.10%, 5.29%, and 10.06% at 412 nm, 443 nm, 490 nm, 520 nm, and 565 nm, respectively; the relative deviation is 1.07% when the solar zenith angle is large (70 similar to 90 degrees)). Cross comparison with MODIS Aqua satellite products shows that the inversion results are consistent. After verifying the accuracy and stability of the algorithm, we reconstructed the remote sensing reflectance dataset from the Arctic Ocean and surrounding high-latitude seas (latitude greater than 50 degrees N) and successfully retrieved chlorophyll-a data and information on other marine ecological parameters.

Automated summary

China's first operational ocean color satellite sensor, HY-1C-COCTS, encountered challenges with large solar zenith angles in high-latitude seas during winter, leading to a lack of effective ocean color product data. To address this issue, a new atmospheric correction algorithm based on neural network model was developed for use with HY-1C-COCTS data. The algorithm successfully established a direct relationship between total radiance received by the satellite and remote sensing reflectance products, improving inversion accuracy and enabling retrieval of marine ecological parameters.