Keywords

1 Introduction

Anthropogenic sources emit sulfur dioxide (SO2) into the atmosphere which is oxidized by gas- and aqueous-phase chemical reactions to form sulfate \(\left( {{\text{SO}}_{4}^{2 - } } \right)\). Dimethylsulfide (DMS), emitted from oceans, can undergo atmospheric reactions to produce SO2. Here, we study the annual, seasonal, and spatial impacts of DMS chemistry on \({\text{SO}}_{4}^{2 - }\) using the coupled hemispheric Community Multiscale Air Quality (CMAQ) model (Mathur et al. 2012).

2 Method

The model uses a horizontal resolution of 108 km and 44 vertical layers. Emissions and meteorological fields previously described by Sarwar et al. (2015) are also used here. Annual anthropogenic SO2 emissions of ~54 Tg are used in the model (http://edgar.jrc.ec.europa.eu/index.php). Two model simulations are performed for 2006. One simulation uses the Carbon Bond chemical mechanism (CB05) without any DMS chemistry while the other simulation uses the CB05 with DMS. Differences in the results are attributed to DMS chemistry. In CMAQ without DMS chemistry, \({\text{SO}}_{4}^{2 - }\) can be formed via oxidation of SO2 by one gas-phase reaction and five aqueous-phase reactions. DMS chemistry contains seven additional reactions (Sander 2011) (two reactions with hydroxyl radical and five reactions with nitrate, chlorine radical, chlorine monoxide, iodine monoxide, and bromine monoxide) for oxidation of DMS to SO2. DMS emissions are calculated using the monthly oceanic climatological DMS concentrations and the total resistance to gas-transfer at the air/sea interface (Lana et al. 2011). We calculate an annual DMS emissions total of 16.1 Tg for the Northern Hemisphere compared to the global annual DMS emissions total of ~28 Tg reported by Kloster et al. (2006) and Lana et al. (2011).

3 Results and Discussion

Predicted annual-mean SO2 and \({\text{SO}}_{4}^{2 - }\) over the seawater in the Northern Hemisphere without and with DMS chemistry are shown in Fig. 55.1. Predicted SO2 and \({\text{SO}}_{4}^{2 - }\) concentrations generally decrease with altitude, with DMS chemistry effectively increasing SO2 and \({\text{SO}}_{4}^{2 - }\) levels particularly near surface.

Fig. 55.1
figure 1

Annual mean SO2 and \({\text{SO}}_{4}^{2 - }\) over seawater without and with DMS chemistry

Full size image

DMS chemistry enhances annual mean \({\text{SO}}_{4}^{2 - }\) over seawater by 48% compared to the simulation without DMS chemistry. It increases \({\text{SO}}_{4}^{2 - }\) by 42% in winter, 42% in spring, 70% in summer, and 43% in fall. The largest impact occurs in summer due to the combination of high DMS emission rates and high oxidant concentrations.

The spatial distribution of the predicted annual-mean surface SO2 and \({\text{SO}}_{4}^{2 - }\) concentrations are shown in Fig. 55.2. The model without DMS chemistry predicts higher concentrations of SO2 and \({\text{SO}}_{4}^{2 - }\) over land and small concentrations over seawater reflecting mainly the impact of anthropogenic SO2 sources. DMS chemistry effectively enhances SO2 and \({\text{SO}}_{4}^{2 - }\) concentrations over seawater and many coastal areas. Higher enhancements are predicted over seawater than over coastal areas. DMS chemistry enhances SO2 concentrations by >40 pptv in many oceanic areas and \({\text{SO}}_{4}^{2 - }\) levels by >0.8 μg/m3 in some areas of the Pacific Ocean, Atlantic Ocean, Arabian Sea, and Caribbean Sea.

Fig. 55.2
figure 2

(Top row) predicted annual mean SO2 without DMS chemistry and enhancement due to DMS chemistry (bottom row) predicted annual mean \({\text{S}}{\mathbf{O}}_{4}^{2 - }\) without DMS chemistry and enhancement due to DMS chemistry

Full size image

Predicted \({\text{SO}}_{4}^{2 - }\) levels are compared to observed data from the Chemical Speciation Network (CSN) and the Interagency Monitoring of Protected Visual Environments (IMPROVE) sites in the U.S. (Fig. 55.3a). The model without DMS chemistry underestimates the summertime observed data. DMS chemistry only marginally improves the model performance when compared to observed data from all sites. However, it affects the model performance by larger margins when compared to the observed data from coastal sites (Fig. 55.3b). It improves the model performance from March to September but deteriorates the performance in other months.

Fig. 55.3
figure 3

A comparison of model predicted \({\text{S}}{\mathbf{O}}_{4}^{2 - }\) with observed data from CSN and IMPROVE sites for a the entire U.S. and b sites only in the Gulf Coast areas of the U.S

Full size image

4 Summary

We performed annual hemispheric CMAQ model simulations without and with DMS chemistry. The model without DMS chemistry predicts only small levels of SO2 and \({\text{S}}{\mathbf{O}}_{4}^{2 - }\) over seawater. However, DMS chemistry substantially enhances SO2 and \({\text{S}}{\mathbf{O}}_{4}^{2 - }\) over seawater and coastal areas. It enhances the annual mean surface \({\text{S}}{\mathbf{O}}_{4}^{2 - }\) concentrations by 48% compared to that obtained without DMS chemistry. It enhances summertime surface \({\text{S}}{\mathbf{O}}_{4}^{2 - }\) by the largest margin among all seasons. The model without DMS chemistry tends to under-predict \({\text{S}}{\mathbf{O}}_{4}^{2 - }\) compared to observed data in the US. DMS chemistry only marginally affects the model performance when compared to the observed data for the entire U.S. However, DMS chemistry affects model performance for \({\text{S}}{\mathbf{O}}_{4}^{2 - } \varvec{ }\) by larger margins in U.S. coastal areas.