AUTHOR=Li Wei , Liu Chunli , Zhai Weidong , Liu Huizeng , Ma Wenjuan 

TITLE=Remote sensing and machine learning method to support sea surface pCO2 estimation in the Yellow Sea

JOURNAL=Frontiers in Marine Science

VOLUME=Volume 10 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2023.1181095

DOI=10.3389/fmars.2023.1181095

ISSN=2296-7745

ABSTRACT=Due to the limited availability of in-situ spatial and temporal distribution data, the current status of the sea surface carbon dioxide partial pressure (pCO2) in the Yellow Sea is unclear. In addition, the physical and biological environment has changed in the coastal seas, and these changes will affect the sea surface pCO2. Therefore, this study aimed to establish an inversion model for pCO2 in the Yellow Sea, before exploring the trend in the interannual variation in pCO2. Models were trained and tested using 14 cruise data sets from 2011 to 2019, and the Moderate Resolution Imaging Spectroradiometer (MODIS) derived sea surface temperature, chlorophyll concentration, diffuse attenuation of downwelling irradiance, and in-situ salinity were used as the input variables. The model was applied to satellite remote sensing data from between January 2003 and December 2021 to determine the spatial, seasonal, and interannual variations in pCO2. The results showed that the model developed for this study performed well, with a root mean square difference of 43 atm and coefficient of determination (R2) of 0.67. Moreover, pCO2 increased at a rate of 0.36 atm year−1 (R2 = 0.27, p < 0.05) in the Yellow Sea, which is much slower than the rate of atmospheric carbon dioxide (CO2) rise. The reason behind it needs further investigation.