Evaluation of diverse approaches for estimating sea-surface DMS concentration and air–sea exchange at global scale

Abstract

Environmental context As climate models increasingly include detailed, process-based models of aerosol formation, they need to represent dimethylsulfide emissions from the ocean. Options for this include data-based climatologies and empirical or process-based models; there are diverse examples of each in the literature. This paper presents the first global-scale comparison of all available approaches and evaluation of their skill relative to observations and their possible roles in future climate models. Abstract Ocean emission and subsequent oxidation of dimethylsulfide (DMS) provides a source of sulfate in the atmosphere, potentially affecting the amount of solar radiation reaching the Earth’s surface through both direct and indirect radiative effects of sulfate aerosols. DMS concentration in the ocean is quite variable with season and location, which in turn leads to high spatial and temporal variability of ocean DMS emissions. This study tested currently available climatologies and empirical and prognostic models of DMS concentration in the surface ocean against each other and against observations. This analysis mainly reveals the limitations of estimating DMS with an empirical model based on variables such as chlorophyll and mixed-layer depth. The various empirical models show very different spatial patterns, and none correlate strongly with observations. There is considerable uncertainty in the spatial and temporal distribution of DMS concentration and flux, and in the global total DMS flux. Global total air–sea flux depends primarily on global mean surface ocean DMS concentration, and the spatial distribution of DMS concentration and the magnitude of the gas exchange coefficient are of secondary importance. Global total flux estimates range from 9 to 34TgSyear–1, with a best estimate of 18–24 Tg. Both empirical and prognostic models generally underestimate the total compared with the best available data-based estimates.