Modelling sea-ice and oceanic dimethylsulfide production and emissions in the Arctic
Recent field observations suggest that the radiative forcing of aerosol and clouds in the Arctic may be seasonally regulated by the oceanic emissions of the climatically-important biogenic trace gas dimethylsulfide (DMS). However, the validity of the proposed argument is challenged by the limited spatio-temporal coverage of these earlier studies in this difficult-to-access region. In particular, little is known about the pan-Arctic distribution of the oceanic DMS emissions, its temporal variability, and the impacts of sea-ice biogeochemistry on these emissions. In this dissertation, I investigated these unexplored subjects through numerical modelling. Using a one-dimensional (1-D) column modelling framework, I developed a coupled sea ice-ocean biogeochemical model and assessed the impacts of bottom-ice algae ecosystems on the underlying pelagic ecosystems and the associated production and emissions of DMS. The model was calibrated by time-series measurements of snow and melt-pond depth, ice thickness, bottom-ice and under-ice concentrations of chlorophyll-a and dimethylsulfoniopropionate (DMSP), and under-ice irradiance obtained on the first-year landfast sea ice in Resolute Passage during May-June of 2010. Many of the model parameters for the DMSP and DMS production and removal processes were derived from recent field measurements in the Arctic, which is advantageous over the previous Arctic-focused DMS model studies as their model parameters were based on the measurements in extra-polar regions. The impacts of sea-ice biogeochemistry on the DMS production in the underlying water column and its potential emissions into the overlying atmosphere were quantified through sensitivity experiments. To extend the study domain to the pan-Arctic, I implemented the sea-ice ecosystem and the coupled sea ice-pelagic DMS cycling components of the 1-D column model into a three-dimensional (3-D) regional modelling framework. A multi-decadal model simulation was performed over the period 1969-2015 using realistic atmospheric forcing and lateral boundary conditions. The results of the simulation were evaluated by direct comparisons with available data products and reported values based on field and satellite measurements and other model simulations. The decline of Arctic sea ice was successfully simulated by the model. The magnitude of the pan-Arctic sea-ice and pelagic annual primary production and their general spatial patterns were comparable to other model studies. The mean seasonal cycle and the spatial distribution of the model-based surface seawater DMS climatology within the pan-Arctic showed some similarities with in situ measurement- and satellite-based climatologies. However, at the same time, the comparison of the DMS climatologies was challenged by the bias in the measurement-based climatology, emphasizing the need to update this data product, which was created almost a decade ago, by incorporating data acquired during the recent field campaigns. The analysis of the modelled fluxes of DMS at the ice-sea and sea-air interfaces revealed different responses to the accelerated decline of sea ice over the recent decades (1996-2015). There was no trend in the pan-Arctic ice-to-sea DMS flux due to the counteracting effect of vertical thinning and horizontal shrinking of sea ice that drove ice algal production. In contrast, the pan-Arctic sea-to-air DMS flux showed a consistent increase (about 40 % over the last two decades) driven by the reduction of sea ice cover that promoted outgassing and biological productivity. This finding suggests that the climate warming in the Arctic causes an increase in DMS emissions, and encourages further exploration of the biological climate regulation in the Arctic.
oceanography, sea ice, arctic, biogeochemistry, dimethylsulfide, numerical modelling, DMS, ecosystem