Sulfur cycle

Atmospheric sulfur plays an important role in the radiative balance of the atmosphere [169-174]. Anthropogenic sources are dominant in highly-industrialized regions, such as those in the mid-latitudes of the Northern Hemisphere, and are well defined. Natural sources and sinks of sulfur gases are much less well defined, but have received greater scrutiny in recent years due to their potential involvement in the regulation of climate in remote parts of the ocean. The major source of natural sulfur gases is the sea [170,173]. Of particular interest are DMS and COS. Both of these compounds are formed predominantly in aerobic marine environments, i.e. the upper layers of the ocean, and their sources and sinks are affected by solar UVR.

5.4.1 Dimethyl sulfide

DMS reactions in the troposphere are believed to lead to enhanced reflectivity of marine clouds [171] and thus DMS emissions may have a cooling influence on the atmosphere. One of the best demonstrations of the link between the natural atmospheric sulfur cycle and the physical climate system are the observations that link the satellite derived stratus cloud optical depth and observed DMS derived cloud condensation nuclei (CCN) concentrations at Cape Grim, Australia [175]. Statistical evidence indicates that the optical depth of the clouds is correlated with the number of CCN in the atmosphere. Thus, any UV-related changes at the surface of the ocean that result in the alteration in DMS flux to the atmosphere and the subsequent formation of CCN would also alter the atmospheric radiation budget for the affected region.

DMS is the predominant volatile sulfur compound in the open ocean. It is derived from the transformation of dimethylsulfonium propionate (DMSP), an organosulfur compound that is synthesized by marine algae. Most marine phytoplankton synthesize DMSP, but their DMSP content is highly variable [176]. Concentrations of DMSP are highest in dinoflagellates and cocco-lithophores, algal classes that represent only a small fraction of the algae in the sea. Within major blooms of these algae there are readily discernable relationships between DMSP concentrations and algal biomass. High concentrations of DMS and DMSP occur in the Southern Ocean during early to mid spring as the sea ice melts [177], a time in which that region also experiences intense UV exposure due to ozone depletion. To first order, one might expect that DMS concentrations will decrease as surface UV increases, because DMS production is closely related to phytoplankton photosynthesis, which is inhibited by increased UV. Such decreases were observed under ozone-depleted regions of the Southern Ocean during 1993 [178]. However, this oversimplified view must be tempered by the fact that DMS concentrations are affected by a variety of biological and chemical processes that could be affected by UV exposure. For example, other research has shown that DMS emissions increase when the phytoplankton are stressed by zooplankton grazing [179]. Therefore, the effects of increased UV on zooplankton discussed earlier could have a positive effect on DMS release. Moreover, recent studies indicate that changes in mixing depth also can influence DMS emissions by controlling the conversion efficiency of DMSP into DMS [180,181]. Decreases in mixing depth are proposed to reduce assimilatory metabolism of DMSP via UV-induced photoinhibition (or perhaps nutrient limitation) thus enhancing the conversion of DMSP into DMS.

In addition to possible effects of solar UVR on the biological production of DMS, DMS emissions to the atmosphere also are strongly affected by biological and photochemical sinks of DMS in the upper ocean. DMS photoreactions accounted for 7% to 40% of the total turnover of DMS in the surface mixed layer of the equatorial Pacific Ocean [182]. The photoreaction involved conversion into dimethyl sulfoxide, but the yield of the conversion was only 14%, much less than was expected. DMS absorbs little or no light at wavelengths >295 nm, so its photolysis under solar radiation in pure water is very slow. Apparently, natural photosensitizers in the seawater initiated this indirect photooxidation (see Section Although earlier results indicated that this reaction may be induced by visible light [182], more recent research using Sargasso Sea water indicates that solar UV is mainly responsible [183].

5.4.2 Carbonyl sulfide

COS is the most concentrated sulfur gas in the troposphere [170,173] and it is believed to play a role in the maintenance of the stratospheric sulfate layer [184,185], although this role may be more limited than was originally believed [185],

COS is produced in surface seawater by the photochemical degradation of DOM and it is degraded mainly by hydrolysis [94,141,186-192]. Comparison of the quantum yield spectra for COS from various regions of the sea again show that, as in the case of CO (Figure 4), COS is produced most efficiently in the UV region and that there appears to be a major difference between the quantum yield spectra in coastal regions and in the open ocean (Figure 8). Quantum yields are generally higher in coastal areas [189,193], especially in the UV-A region, and so are observed COS concentrations and fluxes [186]. Potential annual photoproduction of COS in the open ocean and coastal regions at various 10 degree latitude bands is computed in Figure 9 by cross-multiplying the fluxes, computed as described in section and the Appendix, and the areas of the open ocean and coastal ocean. Pos et al. [191] provided evidence that the photoproduction of COS and CO in the sea may be linked by competitive reactions involving free radical species, possibly helping to explain the similarity in the fractional change in quantum yields with increasing wavelength for CO and COS (see Figures 4 and 8).

In an impressive effort to model global air-sea fluxes of COS based on known information about its sources and sinks, Preiswerk and Najjar [141] estimated that the open ocean was a net source of COS (2.1 Gmol yr_1), taking into account the possibility that dark production occurs. This is close to another global estimate of Ulshôfer and Andreae [192]. The estimates of Preiswerk and Najjar [141] did not include coastal contributions. Early observations and global estimates of Andreae and Ferek [186] suggested that COS emissions predominately come from such regions. Although the computed production comparisons indicate that this likely is not the case (Figure 9), it appears that substantial amounts of COS are produced in coastal regions, a large fraction of

300 320 340 360 380 400 Wavelength, nrrt

Figure 8. Quantum yield spectra for photoproduction of carbonyl sulfide in the ocean. Circles: Coastal regions in North Sea (filled) and Gulf of Mexico (open) [193]; triangles: Average in predominantly open Pacific Ocean during 1993-1994 [189]. Not shown are additional data of Weiss et al. [189] which indicate that coastal quantum yields are significantly higher.

Figure 8. Quantum yield spectra for photoproduction of carbonyl sulfide in the ocean. Circles: Coastal regions in North Sea (filled) and Gulf of Mexico (open) [193]; triangles: Average in predominantly open Pacific Ocean during 1993-1994 [189]. Not shown are additional data of Weiss et al. [189] which indicate that coastal quantum yields are significantly higher.


Figure 9. Latitudinal variations in potential production of carbonyl sulfide in the ocean: filled circles: open ocean estimated using apparent quantum yields for Pacific Ocean provided by Weiss et al. [189] and open ocean areas [8, 173]; open circles: coastal photoproduction estimated using average coastal AQYs determined by Zepp and An-dreae [193] and coastal ocean areas [67,173]. Integrated production was 58.3 Gmol yr_1 for the open ocean and 12.8 Gmol yr-1 for coastal regions, uncorrected for cloud cover. Light attenuation by clouds reduces UVR by approximately 30-40% on average [103], A substantial fraction of the COS is hydrolyzed before it can escape to the atmosphere [141,

which occurs in cold Northern latitudes where hydrolysis rates are much slower.

Factors that result in the higher quantum yields and concentrations of COS in coastal areas are not well understood, although it seems likely that the different chemical composition of CDOM and the higher concentrations of organosulfur compounds near the continents are mainly responsible. The observed regional differences in absorption spectrum and photoreactivity of CDOM may be due to UV-induced photobleaching (see section Moreover, differences in the nature and concentration of dissolved organic sulfur (DOS) may account for the higher photoproduction quantum efficiencies in coastal regions. Zepp and An-dreae [193] presented evidence that COS formation can involve photosensitized reactions whose rates depend upon both CDOM as well as DOS concentrations (see section Thus, higher COS quantum yields in coastal areas potentially could be attributable to higher reactive DOS concentrations. Research on the distributions of organosulfur compounds in coastal areas [194] and open ocean [Cutter, peronal communication] is just starting to appear. Thiol concentrations in coastal North Sea waters were found to correlate with chlorophyll and appeared not to have significant riverine inputs [194]. On the other hand, Uher and Andreae [188] found that the photoproduction rates of COS in the North Sea correlated with absorption coefficients in the near UV (i.e., with CDOM concentrations), indicating that reactive DOS and CDOM must co-vary in this region. Whether this is the case in other parts of the sea is unknown.

COS may also influence metal cycling in the ocean. The hydrolysis of COS in the upper ocean results in the production of sulfide [195], a ligand that can reduce the biological lability of metals by chelation or formation of insoluble metal sulfides. Certain metal sulfide complexes photoreact efficiently when exposed to solar UVR. Other interactions of UV with metals cycling are considered in the next section.

0 0

Post a comment