Bromocarbons in the tropical coastal and open ocean atmosphere during the 2009 1 Prime Expedition Scientific Cruise (PESC-09)

31 Atmospheric concentrations of very short-lived species (VSLS) bromocarbons, including 32 CHBr 3 , CH 2 Br 2 , CHCl 2 Br, CHClBr 2 , CH 2 BrCl, were measured in the Strait of Malacca, the 33 South China and Sulu-Sulawesi Seas during a two month research cruise in June-July 2009. The highest bromocarbon concentrations were found in the Strait of Malacca, with smaller 35 enhancements in coastal regions of Northern Borneo. CHBr 3 was the most abundant 36 bromocarbon, ranging from 5.2 pmol mol -1 in the Strait of Malacca to 0.94 pmol mol -1 over 37 the open ocean. Other bromocarbons showed lower concentrations, in the range of 0.8-1.3 38 pmol mol -1 for CH 2 Br 2 , 0.1-0.5 pmol mol -1 forCHCl 2 Br and 0.1-0.4 pmol mol -1 for CHClBr 2 . 39 There was no significant correlation between bromocarbons and in situ chlorophyll- a but 40 positive correlations with both MODIS and SeaWiFS satellite’s chlorophyll- a . Together the 41 short-lived bromocarbons contribute an average of 8.9 pmol mol -1 (range 5.2-21.4 pmol mol - 42 1 ) to tropospheric bromine loading, which is similar to that found in previous studies from 43 global sampling networks (Montzka et al., 2011). Statistical tests showed strong Spearman 44 correlations amongst brominated compounds, suggesting a common source. Log-log plots of 45 CHBr 3 /CH 2 Br 2 versus CHBr 2 Cl/CH 2 Br 2 show that both chemical reactions and dilution into the background atmosphere contribute to the composition of these halocarbons at each 47 sampling point. We have used the correlation to make a crude estimate of the regional emissions of CHBr 3 and derive a value of 32 Gg yr -1 for the South East

Because of their short atmospheric lifetimes, the region where VSLS are emitted into 75 the atmosphere is significant and their O 3 depletion potentials (ODPs) vary accordingly (Ko 76 et al., 2003). Tropical regions are believed to be the most important location for rapid 77 transport of air from the surface to the upper troposphere and lower stratosphere. In the 78 tropics, deep convection provides a major pathway for rapid transport of insoluble gases from 79 the lower to the upper troposphere. Importantly, such convective transport appears to be 80 particularly strong over the western Pacific (Gettelman et al., 2002;Fueglistaler et al., 2004). 81 Furthermore, the warm, shallow waters of the tropical warm pool make them potentially 82 important source regions for biologically-produced halocarbons. Therefore, this region has 83 the potential to supply a proportion of the 'missing' ~6 pmol mol -1 of bromine, thought to be 84 related to VSLS, to the stratosphere. 85 Yokouchi et al. (1997) were the first to report atmospheric bromocarbon 86 measurements in the Strait of Malacca and the South China Sea. During a cruise between 87 Japan and the Bay of Bengal they measured mean concentrations of 0.77 pmol mol -1 (max 88 1.42 pmol mol -1 ) and 1.2 pmol mol -1 (max 7.1 pmol mol -1 ) for CH 2 Br 2 and CHBr 3 , 89 respectively. The highest levels were seen in harbour regions of Singapore and Penang and 90 the authors suggested a link between high CHBr 3 concentrations and high chlorophyll-a (chl- spanning a ten-year period, covering much of the world's oceans. Their tropical air mean 101 mixing ratios for CHBr 3 and CH 2 Br 2 were 1 pmol mol -1 (0.4-2.1 pmol mol -1 ) and 0.9 pmol 102 mol -1 (0.6-1.3 pmol mol -1 ) respectively.

103
Despite these measurements collected over the last two decades, the seas and about 1 pmol mol -1 . The data were used to make an estimate of the regional emission strength 113 of CHBr 3 which, depending on assumptions, ranged between 21 and 50 Gg/yr in SE Asia

218
In the following sections we will, first, present an overview of the halocarbon data collected 219 during PESC-09 (section 3.1). We then discuss possible causes of any observed variability, 220 focussing on an analysis of air parcel trajectories and in situ measurements of chl-a, which 221 are a possible proxy for biological activity (section 3.2). In section 3.3 we explore 222 correlations between the measured species and provide a rough estimate of regional  Table 1 along with some comparison with data from the literature. Figure

388
Since CHBr 3 has the shorter lifetime of the two species, an increase in the ratio would be 389 consistent with more aged air masses, in which CHBr 3 has been removed at a faster rate. suggest that it is not reasonable to derive global emission estimates from regional data for a 418 short-lived species such as bromoform. They showed that data from Borneo could instead 419 only be used to constrain regional emissions.

420
Using data from the cruise, CHBr 3 /CH 2 Br 2 is plotted against CHBr 2 Cl/CH 2 Br 2 in on data collected in a recent western Pacific cruise. If we assume that this emission ratio of 5 429 is appropriate to the SE Asian Region, (10°N to 20°S, 90°E to 160°E, as used by Pyle et al.

430
(2011)) and recalling that the air mass histories in the two periods of these two studies are 431 quite similar, then we could derive an emission estimate for CHBr 3 from the regional CH 2 Br 2 432 emission. That value is not observationally constrained; instead we use the SEA regional region are beginning to be better characterised.

493
The bromocarbons were well correlated, suggesting common sources. Previously, 494 such relationships have been used to obtain global emission estimates. For a short-lived gas 495 like bromoform, with an atmospheric lifetime of ~26 days, we do not think this approach is 496 suitable; the footprint sampled by the cruise is regional at most. Instead, we used an approach 497 as applied by Yokouchi et al. (2005) to provide a very rough estimate of the regional (S.E. on CH 2 Br 2 emissions, would be required to confirm this estimate, but the use of species 503 correlations for regional emission estimates looks to be a promising approach.

504
Chl-a was also measured on the cruise. There was little obvious correlation between