Remotely sensed retrieval of air-sea carbon flux and acidification risk in Chinese Bohai Sea based on a semi-analytical mechanism model with hour-level GOCI image and ERA5 reanalysis data

Marine carbon sinks act as a buffer against global warming, but raise the risk of acidification, especially in the marginal shelf seas which are rich in terrigenous carbon input. The Chinese Bohai Sea performs generally as a weak carbon source while carbon fluxes (fCO2), pH, and aragonite saturation states (omega arag) vary in time and space under intensive land -sea interaction. However, there are still 1) insufficient spatiotemporal resolution in existing remotely sensed retrieval of carbon flux, 2) inadequate analytic mechanism for existing empirical models, which are not suitable for case II waters in Bohai Sea, and 3) limited research on remotely sensed retrieval of acidification risk expressed as pH and omega arag. Thus, we proposed a semi-analytical mechanism algorithm (MeSAA) to gain seawater partial pressure of CO2 (pCO2sw) with hour-level GOCI imagery and seawater carbonate equilibrium equation (CO2SYS). With the assistance of ERA5 reanalysis data, the gained pCO2sw was then used to obtain fCO2 by using the sea -air CO2 partial pressure difference (Delta pCO2) method. Similarly, with the assistance of remotely sensed retrieval, two indices, pH and omega arag were also gained from CO2SYS to identify the acidification risk. The results showed that Bohai Sea was a weak carbon source with the positive values of fCO2 (0.5-6 mmol C & sdot;m 2 & sdot;day 1) and total emission (0.956 Tg C & sdot;yr 1). It suffered from a high acidification risk, especially in three bays with low values of omega arag (1.15-1.3) from May to September 2011. Although photosynthetic carbon sequestration was intensive near shore, it could not consume the large amount of the rich carbon input, and resulting a monthly increase for pCO2sw and fCO2 and a monthly decrease for pH and omega arag from May to September. The distribution of pCO2sw was in accord with a former study, but the values were not. The local parameter adjustment of MeSAA in the Bohai Sea was analyzed for this issue, so did the effect of uncertainty analysis of mapping at different time scales (hourly and daily). Moreover, the contrast of absolute deviation and relative deviation on different time scales verified the advantage of hour-level GOCI imagery in improving assessment accuracies. This study gained a more precise change trend ofpCO2sw, fCO2, pH, and omega arag in the Bohai Sea from May to September 2011, which would be beneficial to the study of the carbon cycle in the marginal sea of the shelf under the conditions of climate change.