The sensitivity of dimethyl sulfide production to simulated climate change in the Eastern Antarctic Southern Ocean

Abstract

Dimethyl sulfide (DMS) is a radiatively active trace gas produced by enzymatic cleavage of its precursor compound, dimethyl sulfoniopropionate (DMSP), which is released by marine phytoplankton in the upper ocean. Once ventilated to the atmosphere, DMS is oxidised to form non-sea-salt sulfate and methane sulfonate (MSA) aerosols, which are a major source of cloud condensation nuclei (CCN) in remote marine air and may thus play a role in climate regulation. Here we simulate the change in DMS flux in the Eastern Antarctic ocean from 1960–2086, corresponding to equivalent CO2 tripling relative to pre-industrial levels. Calibration to contemporary climate conditions was carried out using a genetic algorithm to fit the model to surface chlorophyll from the 4-yr SeaWiFs satellite archive and surface DMS from an existing global database. Following the methodology used previously in the Subantarctic Southern Ocean, we then simulated DMS emissions under enhanced greenhouse conditions by forcing the DMS model with output from a coupled atmospheric–ocean general circulation model (GCM). The GCM was run in transient mode under the IPCC/IS92a radiative forcing scenario. By 2086, the change simulated in annual integrated DMS flux is around 20% in ice-free waters, with a greater increase of 45% in the seasonal ice zone (SIZ). Interestingly, the large increase in flux in the SIZ is not due to higher in situ production but mainly because of a loss of ice cover during summer–autumn and an increase in sea-to-air ventilation of DMS. These proportional changes in areal mean flux (25%) are much higher than previously estimated for the Subantarctic Southern Ocean (5%), and point to the possibility of a significant DMS–climate feedback at high Southern latitudes. Due to the nexus between ice cover and food-web structure, the potential for ecological community shifts under enhanced greenhouse conditions is high, and the implications for DMS production are discussed.