Industrial-era decline in Arctic methanesulfonic acid is offset by increased biogenic sulfate aerosol
Edited by Mark Thiemens, University of California, San Diego, CA; received May 5, 2023; accepted October 17, 2023
Significance
Decreasing methanesulfonic acid (MSA) concentrations across Greenland ice cores have suggested declining North Atlantic phytoplankton abundance, with implications for ocean ecosystems and climate. MSA is used to infer phytoplankton abundance because phytoplankton emits dimethyl sulfide (DMS), which is oxidized to MSA and sulfate. However, atmospheric chemistry controls how much DMS is converted to MSA vs. sulfate. Here, we measure DMS-derived sulfate compared to MSA in a Greenland ice core over the industrial era. We show that total biogenic sulfur, the sum of MSA and DMS-derived sulfate, is relatively steady over the industrial era, but MSA declines because anthropogenic pollution affects the ratio of MSA vs. DMS-derived sulfate. Therefore, interpreting MSA as a proxy for DMS emissions must consider atmospheric chemistry.
Abstract
Marine phytoplankton are primary producers in ocean ecosystems and emit dimethyl sulfide (DMS) into the atmosphere. DMS emissions are the largest biological source of atmospheric sulfur and are one of the largest uncertainties in global climate modeling. DMS is oxidized to methanesulfonic acid (MSA), sulfur dioxide, and hydroperoxymethyl thioformate, all of which can be oxidized to sulfate. Ice core records of MSA are used to investigate past DMS emissions but rely on the implicit assumption that the relative yield of oxidation products from DMS remains constant. However, this assumption is uncertain because there are no long-term records that compare MSA to other DMS oxidation products. Here, we share the first long-term record of both MSA and DMS-derived biogenic sulfate concentration in Greenland ice core samples from 1200 to 2006 CE. While MSA declines on average by 0.2 µg S kg–1 over the industrial era, biogenic sulfate from DMS increases by 0.8 µg S kg–1. This increasing biogenic sulfate contradicts previous assertions of declining North Atlantic primary productivity inferred from decreasing MSA concentrations in Greenland ice cores over the industrial era. The changing ratio of MSA to biogenic sulfate suggests that trends in MSA could be caused by time-varying atmospheric chemistry and that MSA concentrations alone should not be used to infer past primary productivity.
Sign up for PNAS alerts.
Get alerts for new articles, or get an alert when an article is cited.
Dimethyl sulfide (DMS) from marine phytoplankton is a dominant source of sulfate aerosol in the marine boundary layer (1). After emission to the atmosphere, DMS is oxidized through a complex system of gas-, aqueous-, and multi-phase reactions to form methanesulfonic acid (MSA), sulfur dioxide (SO2), and hydroperoxymethyl thioformate (HPMTF). These intermediate products can be deposited to the Earth’s surface or further oxidized to sulfate (SO42–) (2–4). Additionally, these products are an important source of cloud condensation nuclei, and DMS precursor emissions are one of the largest sources of uncertainty in estimating aerosol radiative forcing (5, 6).
To investigate whether and how much marine primary productivity has changed over time, ice core records of MSA concentration are used as a proxy for DMS emissions (7–12). Due to the coincident timing of peak DMS production with the onset of sea ice melt in spring, ice core records of MSA are also used as proxies for sea ice extent (13–17). These ice core MSA records are viewed as robust signals of DMS emissions because there is no other known natural or anthropogenic source of MSA (11). Importantly, MSA is well preserved in high accumulation ice core sites, although migration of the methanesulfonate ion in the ice can smooth seasonal to subdecadal signals (18).
Recently, declining industrial-era MSA concentrations since 1850 were observed in twelve Greenland ice cores and attributed to a decline in North Atlantic primary productivity due to slowing Atlantic Meridional Overturning Circulation (7) or a decline in Arctic sea ice extent (19). More recent Greenland ice core observations show an increasing trend in MSA concentrations from 2000 to 2015, and this trend has been attributed to increasing primary productivity due to earlier Arctic sea ice retreat (12). These somewhat contradictory studies, as well as preceding decades of research, assume that the relative yield of MSA from DMS oxidation is constant over time; however, different oxidation pathways of DMS have different production yields of MSA, and anthropogenic activity has altered emissions of precursors and thus the concentrations of several key oxidants since the preindustrial, including nitrogen oxides (NOx) (20–23), ozone (24, 25), and reactive halogens such as bromine monoxide and the chlorine radical (19, 26, 27). Thus, confidence in MSA’s role as a proxy for DMS emissions has been questioned due to potential changes in the partitioning of MSA, SO2, and sulfate during DMS oxidation (28).
Sulfate and its precursors (i.e., mainly SO2 and HPMTF) are the major products of DMS oxidation, but ice core sulfate is not used for investigating DMS emissions because DMS is not the only source of sulfate. Other sulfate sources to the Arctic and global atmosphere include volcanic degassing, sea salt, and anthropogenic sources, including fossil fuel combustion and metal smelting (29). Here, we leverage the distinct isotopic composition of DMS-derived biogenic sulfur compared to that of other sources (30–35) to quantify the contribution of DMS to total sulfate in a central Greenland ice core covering 1200 to 2006 CE. We compare this DMS-derived biogenic sulfate to measurements of MSA from the same ice core samples to assess MSA’s utility as a proxy for DMS emissions and to examine trends in DMS-derived biogenic sulfur aerosol over this time period.
Results
We measure concentrations of MSA and sulfate (SO42–) and sulfur isotopes (δ34S) of sulfate in ice core samples from 1200 to 2006 CE (Materials and Methods). Sulfur isotopes are used to estimate the relative contribution from each of the main sources of Arctic sulfate: DMS, volcanic, and anthropogenic emissions (30–35). The influence of sea salt is subtracted from the sulfate concentration and sulfur isotope measurements (Materials and Methods). In the preindustrial samples, when the anthropogenic contribution to sulfate concentrations is negligible, we assume that preindustrial non-sea salt sulfate (nssSO42–) is a sum of biogenic and volcanic sulfate (Eq. 1) and that δ34S(nssSO42–) is a weighted average of the isotopic signatures of sulfate from biogenic (δ34Sbio) and volcanic (δ34Svolc) sources (Eq. 2):where the isotopic signatures are estimated to be δ34Sbio = +18.8 ± 0.3‰ and δ34Svolc = +4.1 ± 0.5‰ [SI Appendix, Fig. S4 and Supplementary Text S3, (34)], and fbio is the biogenic sulfur fraction and fvolc is the volcanic sulfur fraction. The concentration of biogenic sulfate is calculated in each sample (bioSO4 = fbio ⋅ nssSO42–).
[1]
[2]
In the post-1850 samples, we assume that nssSO42– is a sum of anthropogenic, biogenic, and volcanic sulfate (Eq. 3) and that δ34S(nssSO42–) results from a weighted average of δ34Santhro, δ34Sbio, and δ34Svolc (Eq. 4):
[3]
[4]
where fanthro is the fraction of anthropogenic sulfate in the sample and δ34Santhro is the anthropogenic sulfur isotopic source signature, which is estimated to be +2.9 ± 0.3‰ [SI Appendix, Supplementary Text S3, (35)]. We assume that source signatures (δ34Sbio, δ34Svolc, and δ34Santhro) are relatively constant over time (SI Appendix, Supplementary Text S3).
The isotopic source signatures of volcanic, DMS-derived biogenic, and anthropogenic sulfur are determined via two methods described in detail in Jongebloed et al. (34, 35) and SI Appendix, Supplementary Text S3. The first method is a Keeling Plot, which is the regression of δ34S(nssSO42–) measurements against the reciprocal of the concentration measurements (1/[nssSO42–]) (36–38) and is used to determine δ34Svolc and δ34Santhro (SI Appendix, Fig. S4 B and D). The Keeling Plot is the preferred method for determining an isotopic source signature because it is only influenced by sulfate sources reaching Summit. The second method is a compilation of direct measurements of DMS-derived sulfur (DMS, DMSP, MSA, and nssSO42–) for δ34Sbio, volcanic sulfur (H2S, SO2, and bulk S) for δ34Svolc, and anthropogenic sulfur (coal and oil) for δ34Santhro. The global compilations in SI Appendix, Fig. S4 A, C, and E include sources too distant from Greenland to influence the ice core. For this reason, the Keeling Plot determination of sources is used for δ34Svolc and δ34Santhro. The global compilation of DMS-derived sulfur measurements is used for δ34Sbio because DMS-derived sulfur measurements do not vary significantly geographically (39). The comparisons and potential uncertainties in the Keeling Plot regression vs. direct measurement compilation are discussed in SI Appendix, Supplementary Text S3.
Possible changes in ice core δ34S(nssSO42–) due to changes in DMS and SO2 oxidation pathways could affect the interpretation of the record in Fig. 1 and thus quantification of bioSO4 and total bioS in Figs. 2 and 3 (40, 41). Previous studies in remote, polar environments suggest that fractionation due to transport and oxidation has a negligible effect on ice core δ34S(nssSO42–) (42, 43). Transects of sulfur isotopic composition of sulfate in Antarctica have shown consistent δ34S(nssSO42–) values across a geographic region where sources are expected to be nonvarying and thus the only factors influencing δ34S(nssSO42–) would be fractionation from transport or oxidation (42, 43). The consistent δ34S(nssSO42–) values indicate that fractionation during transport or from oxidation across these scales is a negligible factor influencing δ34S(nssSO42–). Additionally, the Keeling Plot determination of source signatures (δ34Svolc and δ34Santhro, SI Appendix, Supplementary Text S3 and Fig. S4) is applied to ice core samples with sulfate that have already been oxidized and transported to the ice sheet, and possible fractionation due to oxidation and transport is inherently included in this method of source signature determination.
Fig. 1.
Fig. 2.
Fig. 3.
Changes in isotopic composition due to changes in fractionation during DMS oxidation are more difficult to constrain. To our knowledge, there is no laboratory quantification of sulfate fractionation caused by different DMS oxidation pathways. However, the similarity in δ34S between DMSP, DMS, MSA, and DMS-derived nssSO42– suggests minimal fractionation (34). Future modeling and laboratory studies should quantify how changes in DMS and SO2 oxidation could cause fractionation and affect δ34S(nssSO42–) and consequently bioSO4 and total bioS.
Fig. 1 shows nssSO42– concentration and δ34S(nssSO42–) in ice core samples from 1200 to 2006 CE. The concentration of nssSO42– triples from the preindustrial mean of 9.6 ± 5.5 µg S kg–1 (mean + SD) to a peak concentration of 35.6 ± 7.1 µg S kg–1 between 1970 and 1980 (Fig. 1A), aligning with previous observations (45). Between 1970 and the top of the core, nssSO42– concentration declines to a mean of 12.9 ± 2.8 µg S kg–1 from 1996 through 2006. The pattern in nssSO42– concentration is mirrored by δ34S(nssSO42–), which decreases from +9.2 ± 1.4‰ in the preindustrial to a low of +4.9 ± 0.6‰ from the 1970s through the late 1990s and then increases to +6.6 ± 0.6‰ from 1996 to 2006 (Fig. 1B). The trends in δ34S(nssSO42–) reflect a shift from natural sulfur sources in the preindustrial (δ34Sbio = +18.8 ± 0.3‰ and δ34Svolc = +4.1 ± 0.5‰) to varying influence of anthropogenic sulfur (δ34Santhro = +2.9 ± 0.3‰), causing a decreasing and then increasing trend in ice core δ34S(nssSO42–) as anthropogenic emissions increase through the 1970s and then decrease following implementation of clean air policies.
Fig. 1C shows the fraction of nssSO42– concentrations attributable to anthropogenic, volcanic, and DMS-derived biogenic sulfur emissions, which are calculated using δ34S(nssSO42–) and Eqs. 1–4. The uncertainty range in the isotopic source signatures for δ34Svolc and δ34Santhro is calculated using the SE of Keeling Plot intercepts as well as the prediction interval of the Keeling Plot regression (SI Appendix, Supplementary Text S3 and Fig. S6). DMS-derived biogenic fraction of nssSO42– (fbio) is 34 ± 9% (mean + SD) of nssSO42– during the preindustrial and fbio decreased as anthropogenic sulfur emissions increased. During the years of maximum anthropogenic sulfur emissions (1972 to 1981) (44), the mean fbio is 12 ± 5%. Clean air policies in North America and Europe reduced anthropogenic sulfur emissions after 1975 (44). As a result, the relative importance of biogenic sulfate increases to fbio = 25 ± 4% of total nssSO42– by 2002 to 2006.
Fig. 2 shows the concentration of MSA and biogenic sulfate, which are both oxidation products of DMS, in ice core samples from 1200 to 2006 CE. MSA concentration (Fig. 2A) is on average 0.72 ± 0.22 µg S kg–1 during the preindustrial. Beginning around 1850, MSA concentration declines by 0.04 µg S kg–1 per decade until the late twentieth century. The negative trend in MSA concentration from 1850 to 1996 is statistically significant (p < 0.01) and consistent with declining MSA concentrations observed in other Greenland ice cores (7). The minimum MSA concentration is 0.32 ± 0.05 µg S kg–1 between 1969 and 1995 CE (SI Appendix, Fig. S1), a decrease of 55% from the preindustrial. Around the turn of the century, MSA concentration increases to an average of 0.50 ± 0.24 µg S kg–1 between 1996 and 2006 CE (Fig. 1C), a trend that is also consistent with a recent Greenland ice core (12).
Fig. 2B shows the concentration of bioSO4 and total bioS (MSA + bioSO4) from 1200 to 2006 CE. The bioSO4 concentration increases by 23% from the preindustrial mean of 3.2 ± 1.0 µg S kg–1 to a mean concentration of 4.0 ± 1.4 µg S kg–1 from 1850 to 2006 CE. The 0.8 ± 0.20 µg S kg–1 increase in bioSO4 concentration is greater than the 0.19 ± 0.36 µg S kg–1 decrease in MSA from the preindustrial to the industrial shown in Fig. 2A. As a result of an increase in bioSO4 concentration that is larger than the decrease in industrial-era MSA concentration, total bioS (MSA + bioSO4) concentration is 17% higher (p < 0.01) in the industrial era (4.6 ± 1.4 µg S kg–1) relative to the preindustrial (3.9 ± 1.1 µg S kg–1).
The uncertainty in isotopic source signatures contributes to uncertainty in bioSO4 and total bioS concentration. For example, if δ34Sbio, δ34Svolc, and δ34Santhro are determined from the compilation of direct measurements from only the back trajectory region (45˚ N to 90˚ N, 120˚ W to 30˚ E; back-trajectory mean shown as dashed lines in SI Appendix, Fig. S4 A, B, and D), the change in total bioS from the preindustrial to the industrial era is 3.6 to 4.0 µg S kg–1 (an increase of 13%), whereas the change in total bioS from the preindustrial to the industrial era from the global compilation of δ34Sbio and Keeling Plot determination of δ34Svolc and δ34Santhro is 3.9 to 4.6 µg S kg–1 (an increase of 17%). The increase of 13% vs. 17% is similar because δ34Sbio is very distinct from δ34Svolc and δ34Santhro. Additionally, we note that both mean estimates (3.6 vs. 3.9 µg S kg–1 in the preindustrial era and 4.0 vs. 4.6 µg S kg–1 in the industrial era) are within the uncertainty range (± 1.4 µg S kg–1) shown as error bars in Fig. 3B.
MSA is a minor component of total bioS throughout the record (Figs. 2D and 3A). During the preindustrial (1200 to 1850 CE), MSA is 19 ± 5% of total bioS (Figs. 2D and 3A), then declines to 8 ± 3% during the MSA minimum from 1969 to 1995 and increases to 13 ± 5% of total bioS in the last 10 y of the record (Figs. 2D and 3A). BioSO4 is the major oxidation product of DMS emissions and the main component of total bioS from 1200 to 1850 (81 ± 5%), 1969 to 1995 (92 ± 3%), and 1996 to 2006 (87 ± 5%). The higher total bioS concentrations during the MSA minimum (4.4 ± 1.8 µg S kg–1) and present day (4.0 ± 1.2 µg S kg–1) relative to the preindustrial (3.9 ± 1.1 µg S kg–1) (Fig. 2B) suggests that the DMS source of sulfur aerosol has not decreased since the preindustrial, but the relative yield of MSA vs. sulfate from DMS oxidation has changed (Figs. 2D and 3A).
Assuming a constant ratio of MSA to total bioS results in erroneous assumed trends in DMS-derived total sulfur aerosol in the Arctic and North Atlantic atmosphere over the industrial era. Fig. 3B demonstrates the difference between the observed (“true”) total bioS anomaly (relative to the observed preindustrial mean) and what the assumed (“false”) total bioS anomaly would be assuming [MSA]/[total bioS] is constant over the industrial era. The assumed normalized total bioS is calculated by multiplying the normalized preindustrial total bioS concentrations by the percent change in MSA concentrations between the preindustrial and shown time periods. The assumed total bioS during the MSA minimum (1969 to 1995), present day (1996 to 2006), and industrial era (1850 to 2006) is statistically significantly lower (60%, 31%, and 37%, respectively; p < 0.01) than the observed normalized total bioS during the same time periods. The difference between observed and assumed total bioS illustrates the flaw in assuming that MSA concentrations are a robust proxy for DMS emissions and total DMS-derived sulfur aerosol.
Discussion
The observed changes in ice core [MSA]/[total bioS] suggest changes in the relative yield of MSA vs. sulfate production from the oxidation of DMS. The ratio of DMS oxidation products is dependent on environmental conditions such as temperature, concentration of different oxidants, and the presence of clouds (3, 4, 46–48). The DMS oxidation mechanism involves both gas- and multi-phase chemistry, and different oxidants favor different DMS oxidation products. The most important oxidation pathways of DMS are thought to be gas-phase oxidation by the hydroxyl radical (OH) during the daytime (50 to 60%), which occurs by hydrogen abstraction and addition, and by the nitrate radical (NO3) during the nighttime (15 to 35%) (49, 50). Other DMS oxidants include bromine monoxide (BrO, 8 to 29%), the chlorine radical (Cl, 4 to 10%), and ozone (O3) in the gas and aqueous phases (2 to 8%) (49, 50).
Using MSA as a proxy for primary productivity assumes that the oxidation chemistry of DMS has remained relatively constant over long time periods; however, with the possible exception of OH, the abundance of all other DMS oxidants has changed due to anthropogenic emissions. In particular, emissions of NOx, which are precursors to the nitrate radical (NO3), increased from the preindustrial to 1970, remained elevated from 1970 to 1990, and then declined in the early 21st century [Fig. 2C; (20–22)] in regions affecting the Summit air-mass source region that includes eastern North America and western Europe (26).
DMS oxidation by NO3 is one of the main loss processes of DMS and favors the production of SO2 and HPMTF over MSA (50). All else being equal, elevated NO3 concentrations will increase the partitioning of DMS toward the SO2 and HPMTF oxidation pathway. The increase in NOx emissions from 1850 to the 1970s coincides with the decline in MSA concentration (Fig. 2 C and D and SI Appendix, Fig. S1). As NOx emissions decreased after 1990 (Fig. 2C), DMS oxidation shifted back toward the MSA oxidation pathway, as reflected in the increased [MSA]/[total bioS] at the end of the record (0.13 ± 0.05 from 1996 to 2006) relative to the [MSA]/[total bioS] during the MSA minimum (0.08 ± 0.03 from 1969 to 1995) (Fig. 2C) and increase in MSA concentration (SI Appendix, Fig. S1). This hypothesis is supported by atmospheric chemistry modeling by Fung et al. (50), which showed that the DMS + NO3 oxidation pathway is less important during the preindustrial due to lower NOx emissions, and even with modeling the same DMS emissions in the preindustrial and present-day, global MSA burden is 59% higher in the preindustrial compared to the present day.
DMS oxidation chemistry is complex, and it is possible that reactions other than the DMS + NO3 pathway could have contributed to or partially offset the industrial-era decline in MSA. For example, MSA destruction through aqueous and heterogeneous oxidation by OH could convert up to 75% of MSA to sulfate (50, 51) and could have increased in efficiency in the twentieth century due to increased cloud droplet acidity (52). Reactive halogens are DMS oxidants that have been perturbed by anthropogenic activities, including BrO (27) and Cl (26). Additionally, uncertainty remains in reaction intermediates, particularly HPMTF (48, 53), which has highly uncertain cloud uptake and loss process rates. Thus, changes in cloud properties, such as cloud droplet number concentration influenced by sulfate aerosol, could change oxidation and loss of HPMTF and other sulfur species. Additionally, varying relative humidity, organic peroxy radical concentration, and NO concentration change the relative yield of DMS oxidation products in ways that are not captured by models and require further study (54). Finally, understanding how the oxidation of DMS and other marine biogenic sulfur compounds such as methanethiol (55) have contributed to trends in MSA, SO2, and SO42– production requires further investigation.
The robustness of these conclusions depends on accurate determination of δ34Sbio, δ34Svolc, and δ34Santhro. The Keeling Plot regressions in SI Appendix, Fig. S4 B and D and compilations of direct measurements in SI Appendix, Fig. S4 B and E provide independent δ34Svolc and δ34Santhro that are statistically indistinguishable. Although δ34Sbio is determined only by a compilation of direct δ34S measurements of DMS-derived sulfur compounds (SI Appendix, Fig. S4A), Amrani et al. (39) find that δ34Sbio does not vary significantly by region. Furthermore, our analysis is qualitatively consistent even when the analysis is performed on a subset of the compilation from measurements only in the back-trajectory region: Total bioS does not decrease over the industrial era despite the decrease in industrial-era MSA concentrations (SI Appendix, Supplementary Text S3D).
We note that despite the increase in [MSA]/[total bioS] in 1996 to 2006 relative to 1969 to 1995 (Figs. 2 D and 3A), which reduces the relative yield of sulfate produced from DMS, fbio increases from ~1975 to 2006 due to decreasing anthropogenic sulfur emissions (Fig. 1C). As anthropogenic sulfur emissions decrease, the magnitude of anthropogenic sulfate aerosol cooling decreases (56), and the relative abundance and radiative importance of natural sulfate sources, including DMS, increases.
Conclusions
Our ice core measurements of biogenic sulfate (bioSO4) and MSA concentrations from 1200 to 2006 CE show that MSA concentrations declined from 1850 CE through the late twentieth century, while bioSO4 is higher in the industrial relative to the preindustrial. The cause for the decline in MSA is likely changing DMS oxidation chemistry from changes in atmospheric concentrations of key oxidants, including the nitrate radical, which increases the relative yield of SO2 and sulfate from oxidation of DMS (49). Although more investigation is required to quantify the effects of atmospheric chemistry on MSA and sulfate, our results show that relative yield of MSA from oxidation of DMS is variable over time and that MSA alone is not a reliable proxy for oceanic DMS emissions. Uncertainty in how changing sulfur isotope fractionation due to changing DMS and SO2 oxidation pathways over the industrial era affects interpretation of δ34S(nssSO42–) also requires future investigation. Overall, total bioS and bioSO4 concentrations throughout the ice core record suggest that industrial era DMS emissions are not significantly lower than, and possibly even slightly higher than, DMS emissions during the preindustrial. This finding contradicts previous studies inferring industrial-era primary productivity decline from observed ice core MSA concentrations (7, 12).
Since the preindustrial, anthropogenic sulfate has had a net cooling effect on the atmosphere (57), but as anthropogenic emissions of SO2 decrease, the cooling effect of anthropogenic sulfate is diminished (56). As a result, biogenic sulfate will become a relatively more important source of cloud concentration nuclei and potentially more important to Earth’s radiative balance. Assuming anthropogenic sulfur emissions will continue to decrease in the future (58), perturbations in biogenic sulfur could have a relatively larger role in Earth’s climate. Understanding how DMS emission have changed and will change in the future relies on robust proxies for primary productivity, which requires understanding of how changing atmospheric chemistry affects the interpretation of these proxies.
Materials and Methods
We measure MSA concentration, sulfate concentration, and sulfur isotopic composition of sulfate [δ34S(SO42–)] in samples from a Summit, Greenland ice core in years between 1200 and 2006 CE without influence from large volcanic eruptions (59, 60). Detailed methods for sample selection and analysis are described by Jongebloed et al. (34, 35) and in SI Appendix, Supplementary Text S1. We analyzed a total of 135 samples representing 184 sampled years, including one sample per decade from 1200 to 1750 CE at 2-y resolution, one sample every 4 y from 1750 to 1980 CE at 1-y resolution, and one sample each year at 1-y resolution from 1980 to 2006 CE. The aim of the sampling scheme is to quantify preindustrial (1200 to 1850 CE) biogenic sulfate (bioSO4) and MSA concentrations compared to recent decades. The uncertainty for sample δ34S(SO42–) measurement determined by replicate analysis of whole-process standards is ±1.20‰. The fraction of sea salt sulfate (fss) and non-sea salt sulfate (nssSO42–) in each sample is calculated using the mass fraction of bulk sea water SO42–/Na+ = 0.25 (61) and isotopic signature of sea salt sulfate (δ34Sss = +21‰; (62).
The similar isotopic signatures of volcanic and anthropogenic sulfur prevent clear separation of volcanic and anthropogenic sulfate in post-1850 samples. We therefore assume the industrial-era volcanic sulfate concentrations ([volcSO4]assumed) are equal to the preindustrial median and subtract this from the non-sea salt sulfate concentration and isotopic composition of non-sea salt sulfate (Eqs. 5 and 6):
[5]
[6]
We also subtract a range of volcanic sulfate concentrations ±2 SDs around the mean preindustrial volcanic sulfate concentration to investigate a range of possible solutions, such that [volcSO4]assumed, 97.5th percentile = and [volcSO4]assumed, 2.5th percentile = , where and σ are the mean and SD, respectively, of the preindustrial volcanic sulfate concentrations. We attribute the remaining sulfate to biogenic and anthropogenic sulfur and determine the concentration of biogenic sulfate in each sample. The range in [volcSO4]assumed is shown in SI Appendix, Fig. S2 and the resulting range in [bioSO4] is shown in SI Appendix, Fig. S3. The wide range in assumed volcanic sulfate concentrations contributes only 28% of the uncertainty (total uncertainty = ± 1.4 µg kg–1) in estimated biogenic sulfate concentrations because the isotopic signature of biogenic sulfur is distinct from the anthropogenic and volcanic signatures.
Sulfur isotopic source signatures for biogenic (δ34Sbio), volcanic (δ34Svolc), and anthropogenic (δ34Santhro) sulfur emissions are determined through Keeling Plot regressions and compilations of δ34S measurements from direct source gas emissions, described in detail in Jongebloed et al. (34, 35), SI Appendix, Supplementary Text S3 and Fig. S4. We assume that these source signatures are consistent over the preindustrial and industrial eras (SI Appendix, Supplementary Text S3). This assumption is supported by consistency between Keeling Plot determinations of source signatures with source signatures determined by compilations of measurements representing different regions and time periods (SI Appendix, Supplementary Text S3). However, the effects of possible changes in fractionation effects on δ34S(nssSO42–) due to changing oxidation pathways of DMS and SO2 must be investigated in future work to quantify how this might affect the quantification of bioSO4 from ice core δ34S(nssSO42–). A larger uncertainty range based on the prediction interval of the Keeling Plot regression is shown for the bioSO4 and total bioS concentrations in SI Appendix, Fig. S6.
Ice core samples for MSA measurement were prepared following established decontamination procedures (63). MSA concentrations were measured using a Sciex QTRAP 5500 tandem mass spectrometer. The mass spectrometric measurement method is similar to that used in previous studies of MSA in ice cores (64) and an internal standard (deuterated MSA) was used for quantitation. The detection limit for MSA is estimated with replicate measurement to be 0.03 µg S kg–1. To calculate the total biogenic sulfur (total bioS), we add ice core MSA concentration (µg S kg–1) and biogenic sulfate concentration (µg S kg–1).
Data, Materials, and Software Availability
All ice core data are available in the NSF Arctic Data Center at https://doi.org/10.18739/A2WW7717K, https://doi.org/10.18739/A2N873162, and https://doi.org/10.18739/A26T0GX7K (65–67). All study data are included in the article and/or SI Appendix.
Acknowledgments
U.A.J. and B.A. acknowledge NSF Division of Polar Programs (PLR) grants 1904128, 629363, and Atmospheric and Geospace Sciences (AGS) 1702266. J.C-D. acknowledges NSF grants PLR 0612461, PLR 1904142, and PLR 2230351. L.G. acknowledges the support from National Natural Science Foundation of China (41822605 and 41871051). W.C.P. and L.T. acknowledge NSF grant AGS 2155192.
Author contributions
A.J.S., J.C.-D., L.G., and B.A. designed research; U.A.J., A.J.S., J.C.-D., C.G.L., and B.A. performed research; U.A.J., A.J.S., C.G.L., W.C.P., L.T., S.Z., S.S., S.R.E., and B.A. contributed new reagents/analytic tools; U.A.J., A.J.S., J.C.-D., C.G.L., W.C.P., L.T., S.Z., S.S., S.R.E., and B.A. analyzed data; and U.A.J. and B.A. wrote the paper.
Competing interests
The authors declare no competing interest.
Supporting Information
Appendix 01 (PDF)
- Download
- 1.80 MB
References
1
M. O. Andreae, Ocean-atmosphere interactions in the global biogeochemical sulfur cycle. Mar. Chem. 30, 1–29 (1990).
2
A. R. Ravishankara, Heterogeneous and multiphase chemistry in the troposphere. Science 276, 1058–1065 (1997).
3
I. Barnes, J. Hjorth, N. Mihalopoulos, Dimethyl sulfide and dimethyl sulfoxide and their oxidation in the atmosphere. Chem. Rev. 106, 940–975 (2006).
4
E. H. Hoffmann et al., An advanced modeling study on the impacts and atmospheric implications of multiphase dimethyl sulfide chemistry. Proc. Natl. Acad. Sci. U.S.A. 113, 11776–11781 (2016).
5
K. S. Carslaw et al., Large contribution of natural aerosols to uncertainty in indirect forcing. Nature 503, 67–71 (2013).
6
A. Gettelman, Putting the clouds back in aerosol-cloud interactions. Atmos. Chem. Phys. 15, 12397–12411 (2015).
7
M. B. Osman et al., Industrial-era decline in subarctic Atlantic productivity. Nature 569, 551–555 (2019).
8
D. J. Polashenski et al., Denali ice core methanesulfonic acid records North Pacific marine primary production. J. Geophys. Res. Atmos. 123, 4642–4653 (2018).
9
E. S. Saltzman, I. Dioumaeva, B. D. Finley, Glacial/interglacial variations in methanesulfonate (MSA) in the Siple Dome ice core, West Antarctica. Geophys. Res. Lett. 33, 26649–26657 (2006).
10
E. S. Saltzman, P. Y. Whung, P. A. Mayewski, Methanesulfonate in the Greenland ice sheet project 2 ice core. J. Geophys. Res. Oceans 102, 26649–26657 (1997).
11
M. Legrand, P. Mayewski, Glaciochemistry of polar ice cores: A review. Rev. Geophys. 35, 219–243 (1997).
12
Y. Kurosaki, S. Matoba, Y. Iizuka, K. Fujita, R. Shimada, Increased oceanic dimethyl sulfide emissions in areas of sea ice retreat inferred from a Greenland ice core. Commun. Earth Environ. 3, 327 (2022).
13
E. C. Osterberg et al., Coastal ice-core record of recent northwest Greenland temperature and sea-ice concentration. J. Glaciol. 61, 1137–1146 (2015).
14
N. J. Abram, R. Mulvaney, C. Arrowsmith, Environmental signals in a highly resolved ice core from James Ross Island, Antarctica. J. Geophys. Res. Atmos. 116, D20116 (2011).
15
N. J. Abram, E. W. Wolff, M. A. J. Curran, A review of sea ice proxy information from polar ice cores. Quat. Sci. Rev. 79, 168–183 (2013).
16
M. A. J. Curran et al., Post-depositional movement of methanesulphonic acid at Law Dome, Antarctica, and the influence of accumulation rate. Ann. Glaciol. 35, 333–339 (2002).
17
M. A. J. Curran, T. D. van Ommen, V. I. Morgan, K. L. Phillips, A. S. Palmer, Ice core evidence for Antarctic Sea ice decline since the 1950s. Science 302, 1203–1206 (2003).
18
M. Osman, S. B. Das, O. Marchal, M. J. Evans, Methanesulfonic acid (MSA) migration in polar ice: Data synthesis and theory. Cryosphere 11, 2439–2462 (2017).
19
O. J. Maselli et al., Sea ice and pollution-modulated changes in Greenland ice core methanesulfonate and bromine. Climate of the Past 13, 39–59 (2017).
20
M. T. Lund et al., Concentrations and radiative forcing of anthropogenic aerosols from 1750 to 2014 simulated with the Oslo CTM3 and CEDS emission inventory. Geosci. Model. Dev. 11, 4909–4931 (2018).
21
D. A. Hauglustaine, Y. Balkanski, M. Schulz, A global model simulation of present and future nitrate aerosols and their direct radiative forcing of climate. Atmos. Chem. Phys. 14, 11031–11063 (2014).
22
L. Xu, J. E. Penner, Global simulations of nitrate and ammonium aerosols and their radiative effects. Atmos. Chem. Phys. 12, 9479–9504 (2012).
23
H. Wang et al., Increased night-time oxidation over China despite widespread decrease across the globe. Nat. Geosci. 16, 217–223 (2023), https://doi.org/10.1038/s41561-022-01122-x.
24
A. Gaudel et al., Tropospheric ozone assessment report: Present-day distribution and trends of tropospheric ozone relevant to climate and global atmospheric chemistry model evaluation. Elementa 6, 39 (2018).
25
L. Y. Yeung et al., Isotopic constraint on the twentieth-century increase in tropospheric ozone. Nature 570, 224–227 (2019).
26
S. Zhai et al., Anthropogenic impacts on tropospheric reactive chlorine since the preindustrial. Geophys. Res. Lett. 48, e2021GL093808 (2021).
27
T. Sherwen et al., Global impacts of tropospheric halogens (Cl, Br, I) on oxidants and composition in GEOS-Chem. Atmos. Chem. Phys. 16, 12239–12271 (2016).
28
M. Legrand et al., Sulfur-containing species (methanesulfonate and SO4) over the last climatic cycle in the Greenland Ice Core Project (central Greenland) ice core. J. Geophys. Res. Oceans 102, 26663–26679 (1997).
29
J. P. D. Abbatt et al., Overview paper: New insights into aerosol and climate in the Arctic. Atmos. Chem. Phys. 19, 2527–2560 (2019).
30
V. Wasiuta, A.-L. Norman, S. Marshall, Spatial patterns and seasonal variation of snowpack sulphate isotopes of the Prince of Wales Icefield, Ellesmere Island, Canada. Ann. Glaciol. 43, 390–396 (2006).
31
R. Ghahremaninezhad, A. L. Norman, J. P. D. Abbatt, M. Levasseur, J. L. Thomas, Biogenic, anthropogenic and sea salt sulfate size-segregated aerosols in the Arctic summer. Atmos. Chem. Phys. 16, 5191–5202 (2016).
32
N. Patris et al., First sulfur isotope measurements in central Greenland ice cores along the preindustrial and industrial periods. J. Geophys. Res. Atmos. 107, ACH 6-1–ACH 6-11 (2002).
33
A. L. Norman et al., Sources of aerosol sulphate at Alert: Apportionment using stable isotopes. J. Geophys. Res. 104, 11619–11631 (1999).
34
U. A. Jongebloed et al., Underestimated passive volcanic sulfur degassing implies overestimated anthropogenic aerosol forcing. Geophys. Res. Lett. 50, e2022GL102061 (2023).
35
U. A. Jongebloed et al., Sulfur isotopes quantify the impact of anthropogenic activities on industrial-era Arctic sulfate in a Greenland ice core. Environ. Res. Lett. 18, 074020 (2023). https://doi.org/10.1088/1748-9326/acdc3d.
36
D. E. Pataki et al., The application and interpretation of Keeling plots in terrestrial carbon cycle research. Global Biogeochem. Cycles 17, 1022 (2003).
37
C. D. Keeling, The concentration and isotopic abundances of atmospheric carbon dioxide in rural areas. Geochim. Cosmochim. Acta 13, 322 (1958).
38
C. D. Keeling et al., “A three-dimensional model of atmospheric CO2 transport based on observed winds: 1. Analysis of observational data” in Aspects of Climate Variability in the Pacific and the Western Americas, D. H. Peterson, Ed. (AGU, 1989), pp. 165–236.
39
A. Amrani, W. Said-Ahmad, Y. Shaked, R. P. Kiene, Sulfur isotope homogeneity of oceanic DMSP and DMS. Proc. Natl. Acad. Sci. U.S.A. 110, 18413–18418 (2013).
40
E. D. Sofen, B. Alexander, S. A. Kunasek, The impact of anthropogenic emissions on atmospheric sulfate production pathways, oxidants, and ice core Δ17O(SO42–). Atmos. Chem. Phys. 11, 3565–3578 (2011).
41
E. Harris et al., Sulfur isotope fractionation during oxidation of sulfur dioxide: Gas-phase oxidation by OH radicals and aqueous oxidation by H2O2, O3 and iron catalysis. Atmos. Chem. Phys. 12, 407–424 (2012).
42
R. Uemura et al., Sulfur isotopic composition of surface snow along a latitudinal transect in East Antarctica. Geophys. Res. Lett. 43, 5878–5885 (2016).
43
S. Ishino et al., Homogeneous sulfur isotope signature in East Antarctica and implication for sulfur source shifts through the last glacial-interglacial cycle. Sci. Rep. 9, 12378 (2019).
44
E. E. McDuffie et al., A global anthropogenic emission inventory of atmospheric pollutants from sector- and fuel-specific sources (1970–2017): An application of the Community Emissions Data System (CEDS). Earth Syst. Sci. Data 12, 3413–3442 (2020).
45
P. A. Mayewski et al., Sulfate and nitrate concentrations from a South Greenland Ice Core. Science 232, 975–977 (1986).
46
R. von Glasow, P. J. Crutzen, Model study of multiphase DMS oxidation with a focus on halogens. Atmos. Chem. Phys. 3, 6733–6777 (2004).
47
P. J. Hezel et al., Modeled methanesulfonic acid (MSA) deposition in Antarctica and its relationship to sea ice. J. Geophys. Res. Atmos. 116, D23214 (2011).
48
G. A. Novak et al., Rapid cloud removal of dimethyl sulfide oxidation products limits SO2 and cloud condensation nuclei production in the marine atmosphere. Proc. Natl. Acad. Sci. U.S.A. 118, e2110472118 (2021).
49
Q. Chen, T. Sherwen, M. Evans, B. Alexander, DMS oxidation and sulfur aerosol formation in the marine troposphere: A focus on reactive halogen and multiphase chemistry. Atmos Chem. Phys. 18, 13617–13637 (2018).
50
K. M. Fung et al., Exploring dimethyl sulfide (DMS) oxidation and implications for global aerosol radiative forcing. Atmos Chem. Phys. 22, 1549–1573 (2022).
51
E. L. Mungall, J. P. S. Wong, J. P. D. Abbatt, Heterogeneous oxidation of particulate methanesulfonic acid by the hydroxyl radical: Kinetics and atmospheric implications. ACS Earth Space Chem. 2, 48–55 (2018).
52
K. C. Kwong et al., Chemical transformation of methanesulfonic acid and sodium methanesulfonate through heterogeneous OH oxidation. ACS Earth Space Chem. 2, 895–903 (2018).
53
P. R. Veres et al., Global airborne sampling reveals a previously unobserved dimethyl sulfide oxidation mechanism in the marine atmosphere. Proc. Natl. Acad. Sci. U.S.A. 117, 4505–4510 (2020).
54
Q. Ye et al., Product distribution, kinetics, and aerosol formation from the OH oxidation of dimethyl sulfide under different RO 2 regimes. Atmos Chem. Phys. 22, 16003–16015 (2022).
55
G. A. Novak, D. B. Kilgour, C. M. Jernigan, M. P. Vermeuel, T. H. Bertram, Oceanic emissions of dimethyl sulfide and methanethiol and their contribution to sulfur dioxide production in the marine atmosphere. Atmos Chem. Phys. 22, 6309–6325 (2022).
56
D. Shindell, G. Faluvegi, Climate response to regional radiative forcing during the twentieth century. Nat. Geosci. 2, 294–300 (2009).
57
G. Myhre et al., Radiative forcing of the direct aerosol effect from AeroCom Phase II simulations. Atmos Chem. Phys. 13, 1853–1877 (2013).
58
S. Szopa et al., “Short-lived climate forcers” in Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change, V. Masson-Delmotte et al., Eds. (Cambridge University Press, Cambridge, United Kingdom and New York, NY, 2021), pp. 817–922.
59
E. Gautier et al., 2600-years of stratospheric volcanism through sulfate isotopes. Nat. Commun. 10, 466 (2019).
60
J. Cole-Dai et al., Two likely stratospheric volcanic eruptions in the 1450s C.E. found in a bipolar, subannually dated 800 year ice core record. J. Geophys. Res. Atmos. 118, 7459–7466 (2013).
61
H. D. Holland, The Chemistry of the Atmosphere and Oceans (Wiley, 1978).
62
C. E. Rees, W. J. Jenkins, J. Monster, The sulphur isotopic composition of ocean water sulphate*. Geochim. Cosmochim. Acta 42, 377–381 (1978).
63
K. Peterson, J. Cole-Dai, D. Brandis, T. Cox, S. Splett, Rapid measurement of perchlorate in polar ice cores down to sub-ngL−1 levels without pre-concentration. Anal. Bioanal. Chem. 407, 7965–7972 (2015).
64
E. S. Saltzman, I. Dioumaeva, B. D. Finley, Glacial/interglacial variations in methanesulfonate (MSA) in the Siple Dome ice core, West Antarctica. Geophys. Res. Lett. 33, 2005GL025629 (2006).
65
U. A. Jongebloed et al., Sulfur isotopes of sulfate measurements from a Greenland ice core (1850-2006). Arctic Data Center (2023), https://doi.org/10.18739/A26T0GX7K.
66
U. A. Jongebloed et al., Methanesulfonic acid measurements from a Greenland ice core (1850-2006). Arctic Data Center (2023), https://doi.org/10.18739/A2WW7717K.
67
U. A. Jongebloed et al., Sulfur isotopes of sulfate measurements from a Greenland ice core (1200-1850) and volcanic gas measurements from various volcanoes and hot springs in Iceland. Arctic Data Center (2023), https://doi.org/10.18739/A2N873162.
Information & Authors
Information
Published in
Cover image: Pictured is Construction of the Metropolitan Railway, an 1861 painting depicting the construction of the world’s first underground railway in London—an example of sustainability transition. A collection of articles in the Sustainability Transitions in Consumption–Production Systems Special Feature highlights recent research on the contemporary transition of electricity, mobility, and food systems toward sustainable development pathways. Together, the articles explore the multidimensional nature of sustainability transition processes shaped by social, political, economic, and cultural developments. See the Introduction to the Special Feature, e2206230120. Image credit: Percy William Justyne (artist), Wikimedia Commons/Materialscientist.
Classifications
Copyright
Copyright © 2023 the Author(s). Published by PNAS. This article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND).
Data, Materials, and Software Availability
All ice core data are available in the NSF Arctic Data Center at https://doi.org/10.18739/A2WW7717K, https://doi.org/10.18739/A2N873162, and https://doi.org/10.18739/A26T0GX7K (65–67). All study data are included in the article and/or SI Appendix.
Submission history
Received: May 5, 2023
Accepted: October 17, 2023
Published online: November 17, 2023
Published in issue: November 21, 2023
Keywords
Acknowledgments
U.A.J. and B.A. acknowledge NSF Division of Polar Programs (PLR) grants 1904128, 629363, and Atmospheric and Geospace Sciences (AGS) 1702266. J.C-D. acknowledges NSF grants PLR 0612461, PLR 1904142, and PLR 2230351. L.G. acknowledges the support from National Natural Science Foundation of China (41822605 and 41871051). W.C.P. and L.T. acknowledge NSF grant AGS 2155192.
Author contributions
A.J.S., J.C.-D., L.G., and B.A. designed research; U.A.J., A.J.S., J.C.-D., C.G.L., and B.A. performed research; U.A.J., A.J.S., C.G.L., W.C.P., L.T., S.Z., S.S., S.R.E., and B.A. contributed new reagents/analytic tools; U.A.J., A.J.S., J.C.-D., C.G.L., W.C.P., L.T., S.Z., S.S., S.R.E., and B.A. analyzed data; and U.A.J. and B.A. wrote the paper.
Competing interests
The authors declare no competing interest.
Notes
This article is a PNAS Direct Submission.
Authors
Metrics & Citations
Metrics
Citation statements
Altmetrics
Citations
Cite this article
120 (47) e2307587120,
Export the article citation data by selecting a format from the list below and clicking Export.
Cited by
Loading...
View Options
View options
PDF format
Download this article as a PDF file
DOWNLOAD PDFLogin options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginRecommend to a librarian
Recommend PNAS to a LibrarianPurchase options
Purchase this article to access the full text.