Surface ocean CO2 concentration and air-sea flux estimate by machine learning with modelled variable trends
Jiye Zeng, Yosuke Iida, Tsuneo Matsunaga, Tomoko Shirai
Abstract
The global ocean is a major sink of anthropogenic carbon dioxide (CO 2 ) emitted into the atmosphere. Machine learning has been actively used in the past decades to estimate the oceanic sink, but it is still a challenge to obtain an accurate estimate due to scarcely available CO 2 measurements. One of the methods to deal with data scarcity was normalizing multiple years’ CO 2 values to a reference year to increase the spatial coverage. The practice assumed a constant CO 2 trend for the normalization. Here, we used three machine learning models to extract variable ocean CO 2 trends on a decadal scale and proposed a method to use the extracted ocean CO 2 trends to correct the decadal atmospheric CO 2 trends for data normalization. The method minimizes assumptions of using the extracted ocean CO 2 trends directly. Comparisons of our CO 2 flux estimate with machine learning products included in Global Carbon Budget 2021 indicates that using the variable trends improved the bias resulted from using a constant trend and that the trends are a critical factor for machine learning methods. Our dataset includes monthly distributions of surface ocean CO 2 concentration and air-sea flux in 1980-2020 with a spatial resolution of 1×1 degree.