Abstract

Abstract Five decades of extensive research have highlighted the significant influence of ocean-emitted dimethylsulfide (DMS) on global climate, primarily through its role in marine aerosol and cloud condensation nuclei formation. However, uncertainties persist in seawater DMS concentrations, sea-air exchange rates, and atmospheric chemical pathways which determine its fate in the atmosphere. Current models of DMS emissions are predominantly informed by seawater observations, partly due to the lack of globally synthesised atmospheric DMS data. Here, I present the first synthesised global database of marine atmospheric DMS observations, comprising over 1.4 million measurements from 50 ship campaigns, 48 aircraft campaigns (>150 flights), and 11 stations, spanning a 37-year period (1987-2024) and covering all major ocean basins. Using the GFDL Atmospheric Model version 4 (AM4), I demonstrate that this database offers new constraints on seawater DMS , gas transfer parameterisations, and DMS lifetime. This database helps provide insights into the mechanisms governing DMS cycling and its impact on regional and global climate dynamics.

Bio George is a Postdoctoral Research Fellow in the Aerosols, Clouds, and Climate group at the University of Exeter, working in close collaboration with the UK Met Office. His research focuses on understanding the source regions of natural aerosols and their precursors, using tools such as Lagrangian source-receptor modelling and machine learning techniques. Previously, he led field-based research conducting dimethylsulfide (DMS) measurements in the Southern Ocean marginal ice zone, produced the first global analysis of controls on seawater DMS spatial variability, co-authored the latest global seawater DMS climatology, and worked extensively with Earth system models. Prior to his current role, he worked at Princeton University and NOAA ’s Geophysical Fluid Dynamics Laboratory, where he led an international collaborative effort to develop the first global atmospheric DMS observational database. In this talk, he will present key results from that effort and discuss implications for constraining DMS cycling in Earth system models.