Quantifying and understanding dimethylsulfide variability and its influence on the climate system

Abstract

The significance of the role dimethylsulfide (DMS) plays in regulating the Earth’s climate has been the subject of extensive study. In recent years, advancements in measurements, analytical techniques, and Earth system modelling have led to the rapid expansion of DMS research studies from which an increasingly nuanced understanding of DMS within the climate has evolved. This thesis contributes to the community research effort on all three fronts and provides new insights from which recommendations for future work are made.

I present new measurements of DMS concentrations from the southeast Atlantic sector of the Southern Ocean marginal ice zone during springtime. The relationship between observations of extremely high DMS concentrations in sea ice (>400 nM) and elevated concentrations in seawater (~40 nM) can be explained by a physical dilution and mixing mechanism. Observed DMS concentrations and estimated fluxes are presented, and their importance is discussed in the context of previous regional observations and large-scale climatological estimates.

An objective global analysis of DMS mesoscale and submesoscale spatial variability is presented. DMS is found to vary on the order of tens of kilometres in all ocean basins, and at different times of the year. DMS variability length scales are uncorrelated with DMS concentrations which enables this analysis to help identify mechanisms underpinning DMS variability. Almost 80 % of DMS variability can be explained using the variability of sea surface height anomalies, density, and chlorophyll-a. These results imply that existing large-scale parameterisations are using appropriate parameters but that regional contrasts in DMS variability point to unresolved drivers.

The sensitivity of aerosol–cloud interactions to the distribution and variability of seawater DMS concentration is tested in UKESM1.1. I test the three latest independent climatological DMS estimates, a previous benchmark climatology, a climatology derived from the model’s native DMS scheme, and a control experiment without marine DMS. The latest observationally driven estimates of DMS concentration produce a similar impact in driving the seasonal cycle of Southern Ocean cloud droplet number concentration. The large negative offset in cloud droplet number relative to satellite data indicates that non-DMS emissions and atmospheric processes remain elusive in the model. The regional contrasts in DMS concentration between different estimates lead to a 12 % (0.11 W m¯²) uncertainty in aerosol effective radiative forcing.