Machine-Learning-Based Estimation of Marine CHBr3 Emissions around Asia and the Implication on Bromine Radicals and Ozone Depletion

Abstract

Ocean-produced, very short-lived bromine substances, such as bromoform (CHBr₃), can significantly affect the tropospheric and stratospheric bromine loading and ozone budget. However, large uncertainties exist for marine CHBr₃ emission estimations, especially around Asia. In this study, we establish a new marine emission inventory of CHBr₃, considering the physical and biogeochemical forces in the interaction of the ocean and atmosphere. The surface seawater concentrations are predicted based on expanded in situ CHBr₃ measurements and the machine-learning approach. The marine CHBr₃ emission is estimated to be 55.9 Gg Br yr^–1 around Asia, accounting for 20.4% of the global total. Our estimate is 18.7 Gg Br yr^–1 higher than the previously widely used bottom-up emission inventory, primarily due to discrepancies in the Western Pacific and the South China Sea. Higher estimated marine CHBr₃ emissions significantly affect the stratospheric bromine radicals and ozone over Asia, leading to a maximum increase of 13.6% in Br and 12.2% in BrO concentrations and thus up to a 2.5 ppbv reduction in stratospheric ozone. This study highlights the need for additional marine CHBr₃ observations in data-sparse regions to better quantify marine CHBr₃ emissions and assess their potential atmospheric impacts. This new approach also provides a valuable model framework to calculate sea–air fluxes for other compounds of interest.