Large Sulfur-Isotope Anomaly in Nonvolcanic Sulfate Aerosol and Its Implications for the Archean Atmosphere

Large sulfur-isotope anomaly in nonvolcanic sulfate aerosol and its implications for the Archean atmosphere

Robina Shaheena, Mariana M. Abaunzaa, Teresa L. Jacksona, Justin McCabea,b, Joël Savarinoc,d, and Mark H. Thiemensa,1

aDepartment of Chemistry and Biochemistry, University of California, San Diego, La Jolla, CA 92093; bPacific Ridge School, Carlsbad, CA 92009; cLaboratoire de Glaciologie et de Géophysique de l’Environnement, Centre National de la Recherche Scientifique, Unité Mixte de Recherche 5183, F-38041 Grenoble, France; and dLaboratoire de Glaciologie et de Géophysique de l’Environnement, Université Grenoble Alpes, Unité Mixte de Recherche 5183, F-38041 Grenoble, France

† Edited by Barbara J. Finlayson-Pitts, University of California, Irvine, CA, and approved July 18, 2014 (received for review April 8, 2014) Sulfur-isotopic anomalies have been used to trace the evolution of (1980–2002) have recently revealed how ENSO-driven changes oxygen in the Precambrian atmosphere and to document past affect the global transport and transformation of sulfate aero- volcanic eruptions. High-precision sulfur quadruple isotope sols from the troposphere to the stratosphere and across hemi- measurements of sulfate aerosols extracted from a snow pit at spheres (10). the South Pole (1984–2001) showed the highest S-isotopic anomalies (Δ33S =+1.66‰ and Δ36S =+2‰) in a nonvolcanic (1998–1999) Results and Discussion period, similar in magnitude to Pinatubo and Agung, the largest 2- The highest SO4 concentration in snow [154 parts per billion volcanic eruptions of the 20th century. The highest isotopic anom- (ppb)] is observed after volcanic activity (Pinatubo, June 1991; aly may be produced from a combination of different stratospheric Cerro Hudson, August 1991). The addition of volcanic sulfate to the sources (sulfur dioxide and carbonyl sulfide) via SOx photochemistry, including photoexcitation and photodissociation. The source of stratospheric baseline sulfate aerosol (or background sulfate aerosol SI Appendix anomaly is linked to super El Niño Southern Oscillation (ENSO) (1997– as defined in , Section 2) produced a significant de- 1998)-induced changes in troposphere–stratosphere chemistry and crease in heavy sulfur isotopes. The baseline sulfate aerosol value of 34 dynamics. The data possess recurring negative S-isotope anoma- δ SBG= 12‰ dropped to ∼3‰ (Fig. 1A) following the Pinatubo lies (Δ36S = −0.6 ± 0.2‰) in nonvolcanic and non-ENSO years, thus eruption, and δ33S, δ36S tracked this isotopic trend. The contribu- requiring a second source that may be tropospheric. The genera- tion of sea salt sulfate at the South Pole is <9%, indicating long- tion of nonvolcanic S-isotopic anomalies in an oxidizing atmo- range transported aerosol to be the main sulfate component (10). A sphere has implications for interpreting Archean sulfur deposits broad range in δ33S(1.61–11.32‰), δ34S(2–20‰), and δ36S(2.8– used to determine the redox state of the paleoatmosphere. 37‰)valuesforthesamplingtimeperiodindicatestheorigin of sulfate aerosols from various sulfur sources and chemical and UV photolysis | sulfur isotopes dynamical processes. A significant positive correlation of δ34S ulfur is a ubiquitous element on Earth. Its multiple valence − + Significance Sstates (S 2 to S 6) permit it to participate in a range of photochemical, geochemical, and biochemical processes, and its four stable isotope (32S, 33S, 34S, and 36S) allow tracing of chemical The highest S-isotope anomaly is observed in a nonvolcanic reactions at a molecular level. Multiple sulfur isotopes (δ33S, δ34S, period, and the magnitude of anomaly is similar to the largest ‡ and δ36S) and concomitant anomalies (Δ33S and Δ36S) in paleo- volcanic eruptions of the 20th century. S-quadruple isotope sediments [>2.5 giga-annum (Ga)] have been used to trace the data provided the first evidence of how super El Niño Southern Oscillation (ENSO) events (1997–1998) have affected the trans- EARTH, ATMOSPHERIC, origin and evolution of life and rise of oxygen in the Earth’s AND PLANETARY SCIENCES paleoclimatic history (1–3). In the present atmosphere, the port and transformation of aerosols to the stratosphere; thus, concentration of sulfate in ice cores and associated S-isotope record of paleo-ENSO events of this magnitude can be traced anomalies has served as a forensic tool to help understand the with the S-isotopic anomaly. High-resolution and high-precision dynamics of volcanic emissions, such as transport and trans- S-isotopic fingerprinting also revealed that the tropospheric formation of sulfur to the stratosphere and its impact on ozone sulfate produced during fossil-fuel and biomass burning con- 2- tributes to the stratospheric sulfate aerosol layer, a contribution chemistry (4–7). The low concentration of sulfate (SO4 )inice cores during volcanically quiescent periods and associated ana- previously unrecognized. The distribution of sulfur anomalies lytical challenges to analyze all four S-stable isotopes at high mimics the Archean isotope record, which is used to track the precision have restricted studies of the temporal distribution of origin and evolution of oxygen on earth. sulfur mass-independent signatures. Here, we present a high- – Author contributions: R.S., J.M., J.S., and M.H.T. designed research; R.S., M.M.A., and T.L.J. resolution seasonal record (1984 2001) of quadruple S-stable performed research; R.S., J.M., and M.H.T. contributed new reagents/analytic tools; R.S., isotopes and concomitant isotope anomalies of sulfate aerosols J.S., and M.H.T. analyzed data; and R.S. wrote the paper. extracted from a snow pit (1 × 1 m) at the South Pole (89.5° S, The authors declare no conflict of interest. † 17.3° W; 2,850 m) (8) to gain further insight into sources, pho- This Direct Submission article had a prearranged editor. tochemistry, and associated sulfur transformations of strato- 1To whom correspondence should be addressed. Email: [email protected]. spheric sulfate aerosols (SSAs). The time period encompasses This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. two major volcanic eruptions and three large El Niño Southern 1073/pnas.1406315111/-/DCSupplemental. ‡ Oscillation (ENSO) events. A recent study has attributed a global MIF. Here, delta denotes the ratio of the least abundant to the most abundant isotope warming hiatus (9) to a super ENSO event (1997–1998); therefore, {e.g., δ33S = [(33S)/(32S)sample/(33S)/(32S)std − 1) × 1,000]} relative to the same ratio in data from this period are timely for understanding changes in standard, which is Canyon Diablo Troilite (CDT) and expressed in parts per thousand (‰). Most natural processes fractionate S isotopes in proportion to mass differences and are stratospheric sulfate aerosol chemistry that play an important role in 33 34 36 34 described by δ S ≈ 0.515*δ S, and δ S ≈ 1.91*δ S, except UV photolysis of SO2. The mitigating global warming trends via scattering of incoming solar deviation from mass-dependent fractionation (MDF) is called anomalous or mass-inde- radiation. Oxygen triple isotope measurements of sulfate aerosols pendent fractionation (MIF) and quantified by Δ33S and Δ36S.

www.pnas.org/cgi/doi/10.1073/pnas.1406315111 PNAS | August 19, 2014 | vol. 111 | no. 33 | 11979–11983 Downloaded by guest on September 27, 2021 wavelength is available in the present-day atmosphere, as will be discussed. The S-isotopic anomalies observed in volcanic sulfate aerosols are accompanied by a significant increase in sulfate concentration (4, 5, 11). The Pinatubo and Cerro Hudson eruptions in 1991 produced a factor of 3.7 sulfate concentration increase in the snow record. The observed unprecedented S-isotope anomaly is not accompanied by a massive Pinatubo-size increase in sulfate concentration, thus requiring new, or highly perturbed, chemical, photochemical, and dynamical processes. Our high-resolution, seasonally resolved sulfate aerosol data reveal 36 the presence of a negative S-isotope anomaly (Δ S(avg) ∼−0.6‰) during nonvolcanic and non-ENSO baseline periods (SI Appendix), suggesting the presence of a second isotopically anomalous sulfate source. The S-isotopic anomaly during these time periods is within the range reported for tropospheric sulfate aerosols (15) of Δ36S= −0.3‰ to −2‰ (Fig. 2B). In the present atmosphere, short UV is blocked by the O3 layer; thus, the negative anomaly in tropospheric sulfate aerosol cannot be attributed to short-wavelength SOx photolysis. Romero and Thiemens (15) suggested possible transport of stratospheric S-isotope anomaly to the troposphere at low and mid latitudes. Considering the tropospheric S burden (16) (SO2 from fossil-fuel combustion ∼ − − 78 Tg S·yr 1, biomass burning ∼ 2TgS·yr 1, and natural sources − ∼ 25 Tg S·yr 1), it is unlikely that even a 10% transport of SSA − (0.01 Tg S·yr 1) can produce such a significant isotopic change in tropospheric sulfate aerosols. Alternatively, a 0.01% transport of anomalous sulfate from the troposphere to stratosphere can cause a significant change (Δ36S = −0.9‰) in the isotopic composition of SSA (SI Appendix), provided tropospheric sulfur is anomalous. Laboratory studies demonstrate that a negative S-isotope anomaly can be produced by nonphotochemical processes, such as primary sulfate produced during fossil-fuel combustion (Δ36S = −0.8‰ to −1.7‰) and biomass burning (Δ36S = −0.1‰ 33 34 36 to −2‰)(SI Appendix,TableS3) (17). The mechanism that gen- Fig. 1. (A) The concentration profile (1983–2000) of SO4 and δ S, δ S, δ S sulfate aerosol extracted from snow-pit samples at the South Pole. Note erates the negative anomaly in such processes is unknown (radical

anticorrelation between SO4 concentration and S-stable isotopes of SO4 aerosol driven or recombination reactions may be operative, for example), after Pinatubo and Cerro Hudson eruptions. (B) Sulfur-isotope anomaly (Δ33S but, clearly, high-temperature sulfur oxidation processes are a via- and Δ36S) of sulfate aerosols extracted from the snow-pit samples at the South ble source for the tropospheric negative anomaly. The S-isotopic – 34 33 36 Pole (1983 2000). For comparison, the ENSO-O3 Index is also shown. Note composition (δ S, Δ S, and Δ S) of baseline sulfate suggests that 1997–1998 biomass burning increased the O3 concentration in the upper transportofSO2 and SO4 to the stratosphere despite its normal troposphere/lower stratosphere (42) followed by a sharp decline in O3 and short tropospheric life time (∼2–5 d). concomitant increase in S-MIF. Purple bars indicate strong El Niño Southern – Oscillation events, and M stands for moderate ENSO. The observed positive sulfur anomaly during 1998 1999 requires a stratospheric photochemical process involving SO2.It is generally accepted that only explosive volcanic eruptions with δ33Sandδ36S(δ33S = 0.508*δ34S + 0.2, r = 0.97; δ36S = [volcanic explosivity index (VEI) > 4] have sufficient energy to 1.94*δ34S – 0.876, r = 0.99) is observed (SI Appendix, Fig. S2). transfer tropospheric boundary material into the stratosphere The maximum S-isotopic anomalies of Δ33S =+1.6‰ and that attain altitudes where short UV region at λ < 300 nm is Δ36S =+2.0 ‰ observed in 1998–1999 (Fig. 1B) occurs after the available. The Smithsonian database of global volcanic eruptions strongest El Niño event (1997–1998) of the decade. The peak is (www.volcano.si.edu) and Stratospheric Aerosol and Gas Experi- associated with a sharp increase in potassium (K) concentration ment II (SAGE II) archives (18), however, do not list any significant – (up to 42 ppb). This anomaly (Δ33S) is ∼2 times higher than the plinian volcanic activity in 1998 1999, ruling out volcanic SO2 input Pinatubo signal (Δ33S =+0.9‰) whereas the Δ36S is similar in to the stratosphere as a source for the observed positive S-isotopic magnitude to the Agung eruption (Δ36S =+2.5‰) (5), the anomaly. Increased SO4 concentrations from local (Antarctic) sources is ruled out because sea salt sulfate and sulfate produced largest volcanic eruption of the 20th century. from DMS oxidation is isotopically normal (5, 6). A potential new These S-isotopic anomalies are within the reported volcanic source of the increase in sulfate concentration and S-isotopic sulfate isotopic ranges between Δ33S = −1‰ to +0.9‰ and Δ36 = − ‰ + ‰ anomaly could be the higher altitude transport of SO2 and po- S 5 to 3 (5, 11), suggesting similar photochemical tentially from carbonyl sulfide (COS) by deep convection to the processes (Fig. 2A). Laboratory experiments have demonstrated tropical tropopause layer, followed by SO2 photochemistry upon that the S-MIF originates during SO2 photolysis at short wave- stratospheric COS oxidation (19). COS, the most abundant λ < Δ33 lengths ( 300 nm), producing sulfate with positive S and tropospheric S compound [∼500 parts per trillion (ppt)], is Δ36 Δ36 Δ33 negative S values, and a wide array of slopes ( S/ S), transported to the stratosphere (19) and removed by photolysis − − ranging from 1.1 to 4.3 depending on wavelength (12), (∼70%) to SO2 above 25 km. Increased COS (20–50%) in the pressure, and composition of the gaseous mixture (13) . The tropical tropopause layer (the main entry region to the strato- observations suggest that the S-isotope anomaly in sulfate sphere), along with a substantial increase in other tracers of – aerosol in 1998 1999 is a consequence of SOx (SO2,SO3) biomass burning (BB), including CO, HCN, CH3Cl, NOx,NOy in photochemistry (5, 6, 12, 14) in the short UV (<200 nm) region 1996 and 1999–2000, has been observed (20). Potassium (K), of the solar spectrum above the ozone layer (>25 km) where this a tracer of BB, can serve in certain circumstances as a tracer of

11980 | www.pnas.org/cgi/doi/10.1073/pnas.1406315111 Shaheen et al. Downloaded by guest on September 27, 2021 available from 1998 to 1999 BB events. Andreae and Merlet (24) estimated ∼1.8 Tg of S (∼15% contribution from COS), K (1.9 Tg), NOx (20.7 Tg), and CH3Cl (∼0.65 Tg) emissions from global BB events. El Niño Southern Oscillations are known to signifi- cantly affect chemistry, temperature, and dynamics of the troposphere and stratosphere (25) (SI Appendix). ENSO-induced variations in the upper troposphere/lower stratosphere (UT/LS) ozone levels (26) have been captured in the oxygen triple isotope data of sulfate aerosols (1981–2004) retrieved from the South Pole through its effect on the SO2 oxidation pathways (10). The 1998–1999 and 1984 peaks are the only deviants from the bulk S-isotope anomalies (Fig. 1B), and special processes are required for these two time periods. Assuming COS and SO2 from wildfires as the source of sulfur in this period with higher altitude transport to the stratosphere (18, 19) via pyrocumulus nimbus clouds (27), subsequent photochemistry of COS produced SO2 above 25 km could provide an extra S-isotope anomaly source. The altitude for both SO2 photo excitation and photolysis in this case likely differs from volcanic SO2 due to its production above the ozone layer. Additionally, S-MIF signatures recorded in ice cores after major volcanic eruptions are actually a mixture of anoma- lously produced sulfate via SOx photolysis and mass-dependently produced sulfate via SO2 + OH reaction (∼90% for Pinatubo sulfate), thus diluting the actual S-MIF signal (6, 7). The origin of the S-MIF is a function of the actinic light spectrum for both photoexcitation and photodissociation processes of SO2 (12, 28). Laboratory experiments have shown that COS photolysis at λ < 220 nm produces elemental sulfur S0 with no S-isotope anomaly (29). In an oxidizing environment, however, COS photolysis produces SO2 (19), and subsequent photochemical transformations at short wavelengths (<280 nm) can produce S-MIF in sulfate (12, 14). A recent model that considers SO2 photoexcitation rather than photolysis and volcanic plume chemistry (including heterogeneous stratospheric chemistry) suggests that UV photo excitation of SO2 is another route to the observed S-MIF in volcanic sulfate. This new mechanism may also provide information about the ozone-depletion chemistry in the plume (30) and is relevant for the present data. Laboratory experiments indicate that SO2 photodissociation is wavelength-dependent (Δ36S/Δ33S slopes vary from −1atλ = 193 nm to −4atλ = 248 nm as shown in Fig. 2), and the deviant sulfate circled points (1998– 1999) may result from photochemistry at shorter wavelength, likely

in the bands below 220 nm. There are no numeric simulations EARTH, ATMOSPHERIC, (including stratospheric heterogeneous chemistry and photo- AND PLANETARY SCIENCES chemical transformations) that are directly applicable to the present case where the dynamics and chemistry are perturbed as a result of changes in stratosphere–troposphere dynamics and intensive global BB following the super ENSO event (1997–1998), and it is difficult to quantify the excess sulfur reaching an altitude for the required wavelength (<220 nm). Based on the SO2 photolysis experiments at short wavelength (193 nm) by Farquhar 36 Fig. 2. (A) Comparison of the sulfur-isotopic anomaly in aerosols (1983– et al. (12), if Δ S(SO4)= 20‰ is assumed as an upper limit 36 2001) extracted from the snow pit (1 × 1 m) at the South Pole station (green {isotopic mass balance; Im [(Δ S(SO4) = 20‰) = (2‰_ENSO + – squares) with the volcanic sulfate aerosol retrieved from ice cores (4 6, 11). (B) 2.6‰_BG + ENSO)*40 ppb_background sulfate/excess SOx Comparison with tropospheric aerosol collected at various sites in the United from biomass burning]}, isotope mass balance suggests that the States (15) [La Jolla, CA (blue circles); Shenandoah National Park, VA (red incremental SOx required above background level in the strato- circles); and Bakersfield, CA (yellow circles] (SI Appendix, Table S4) and fossil- sphere (>25 km where λ < 200 nm light is available) to produce fuel and biomass burning signatures (magenta triangles) (SI Appendix, Table 36 36 33 Δ S = 2‰ is 5 ppb. S3). Δ S/Δ S slope of SO2 photolysis experiments using ArF excimer (193 nm) and KrF (284 nm) Xe-lamp (continuum from 220 nm to longer wavelength) The potential importance of different sulfur sources (e.g., are also shown for comparison(12). (C) S-isotope anomalies in sulfate and COS and differing SO2 photochemistry) and the second tropo- sulfide deposits from Precambrian rocks record (32) to the present-day sedi- spheric source may have further consequences in the Earth’s ments and comparison with aerosols in the present-day atmosphere. early atmosphere (31). Mass-independent isotopic compositions in S-bearing molecules have been observed in the Earth’s oldest rocks, which are interpreted as reflecting lowered oxygen and forest fires (21, 22), and ice-core data have revealed increased K ozone concentrations in the atmosphere allowing tropospheric concentration after intensive biomass burning events from 1750 SO2 photochemistry at short wavelengths (1). There is debate as to 1980 (23). There are no global measurements of trace gases to the oxidation state of Earth’s atmosphere–hydrosphere before

Shaheen et al. PNAS | August 19, 2014 | vol. 111 | no. 33 | 11981 Downloaded by guest on September 27, 2021 ∼2.4 Ga. Large mass-independent fractionation of S isotopes in environments, and the anomaly observed may be a consequence of pre-2.4 Ga sedimentary rocks and their absence in post-2.4 Ga both SO2 and COS photochemical transformations in early Earth. counterparts support the hypothesis for a reducing Archean This observation suggests that it is imperative to consider such atmosphere–hydrosphere (32, 33). Models assume that UV pho- reactions in the early Earth models to facilitate the optimal un- tolysis of volcanic gaseous SO2 in a low pO2 atmosphere (3, 14, derstanding of the Earth’s early atmosphere and its oxygen record. 34) is the source of S-isotope anomaly. The rate of volcanic The present work suggests that modeling efforts on the con- supply of SO2, COS, H2S, photochemical transformation, and sequences of COS emissions from past sources should be explored, + + further reactions to form sulfates (S 6), sulfite (S 4), elemental S especially isotopically, and should include aerial and subaerial − (S0), and sulfide (S 2) determines the overall signature preserved volcanoes, fumaroles, and oceans. It is apparent from the S-isotope in the rock (35, 36). Processes that can remove SO2 from the anomaly plots (Fig. 2) that short UV-processed aerosols produced atmosphere, such as homogeneous oxidation with OH radical by potentially amplified COS/SOx sources lie within the range and heterogeneous chemistry with O3 and H2O2 in the cloud, can reported for the early Archean sediments (sulfide and sulfates) and reduce the S-MIF in the atmospheric S pools (35). However, the for the two largest volcanic eruptions of the century, Pinatubo presence of large S-MIF in the Archean rock record (Δ33S = −2‰ and Agung; therefore, new models of the Archean folding in to +12‰ and Δ36S= −10‰ to +2 ‰) (36) is used to constrain COS and additional sulfur isotopic chemistry are needed to fur- − the upper limit of 10 5 present atmospheric level (PAL) of at- ther resolve the early Earth environment. A long-term record of mospheric oxygen levels in the Archean atmosphere (34). The S-quadruple isotopes along with other tracers of biomass burning, mixing of atmospherically derived S-MIF pool with the micro- such as soot and other tracers, is needed to quantify past ENSO bially derived S-MDF pools in the marine sediment lowers the and BB events and to assess future impact on stratospheric O3 overall magnitude of atmospherically derived S-MIF signal. chemistry during time periods of increased biomass burning in the Comparison of the S-isotope anomaly of marine paleo-sediments present-day atmosphere and stratospheric inputs. (sulfates and sulfide) with present-day sulfate aerosol, including tropospheric aerosols (Fig. 2C), reveals that the S-isotope Materials and Methods anomaly resides in a similar S-quadruple isotope space although Sulfate aerosol were extracted from a 1 × 1 m snow pit at the South Pole the magnitude of S-isotope anomaly in ice-core data is smaller (2,850 m high; snow accumulation rate 84 kg·m−2·a−1; mean annual tem- than the Precambrian record. The result suggests that both may perature −49.5 °C), Antarctica (8). The sample preparation for O-isotope analysis and SO2 collection for sulfur isotope analysis were described earlier be produced by the same SOx photochemical processes and that factors such as photolysis wavelength and pressure may be impor- (10). SO2 gas was converted to SF6 for S-quartet isotope analysis following the method developed earlier in our laboratory (3, 12, 43). The δ33S, δ34S, tant in accounting for some of the differences (12, 13). The pres- δ36S, Δ33S, and Δ36S showed standard deviations of 0.06‰,0.1‰,0.4‰, ence of S-MIF in the present-day atmosphere in nonvolcanic 0.05‰, and 0.2‰, respectively, over the course of one year (sample size = 1– aerosols (encircled points) after the super ENSO 1997–1998 sug- 2 μmole; SI Appendix, Table S2). gests that these two sources (SO2 and COS) could contribute to The mass independent signatures of S are measured as (3): sulfur-isotopic anomalies in the Archean. Thermodynamic gas 33 33 34 0.515 phase equilibrium shows COS to be a stable product of reactions Δ S = δ S − 1,000*[(1 + δ S/1,000) − 1] Δ36S = δ36S − 1,000*[(1 + δ36S/1,000)1.91 − 1]. such as CO2 + H2S → COS + H2OandCO+ H2S → COS+ H2 in reducing and oxidizing environments and has been detected in ACKNOWLEDGMENTS. We thank anonymous reviewers for critical evalua- terrestrial geothermal fluids and present-day volcanoes (37), the tion that greatly improved our manuscript. The National Science Foundation atmosphere of Venus (38), and dense molecular clouds and comets Atmospheric Chemistry Division and polar program are recognized for their (39, 40). Consequently, COS is another plausible S species in a re- support through Awards ATM0960594 and OPP0125761. J.S. thanks the Agence Nationale de la Recherche [ANR-NT09-431976-volcanic and solar ducing early earth environment. Numeric simulations of the CO2- radiative forcing (VOLSOL)] and the Centre National de la Recherche rich (1%) early Archean atmosphere suggest that COS (5 ppm) may Scientifique/Projet International de Coopération Scientifique exchange pro- have provided a greenhouse effect (41). If COS undergoes hydro- gram for their financial support for maintaining the collaboration with the lysis reactions, it could lead to SO2 in both oxidizing and reducing University of California, San Diego.

1. Thiemens MH, Shaheen R (2014) 5.6-Mass-independent isotopic composition of ter- 13. Masterson AL, Farquhar J, Wing BA (2011) Sulfur mass-independent fractionation restrial and extraterrestrial materials. Treatise on Geochemistry, eds Holland HD, patterns in the broadband UV photolysis of sulfur dioxide: Pressure and third body Turekian KK (Elsevier, Oxford), 2nd Ed, pp 151–177. effects. Earth Planet Sci Lett 306(3-4):253–260. 2. Thiemens MH (2006) History and applications of mass-independent isotope effects. 14. Pavlov AA, Kasting JF (2002) Mass-independent fractionation of sulfur isotopes in Annu Rev Earth Planet Sci 34:217–262. Archean sediments: Strong evidence for an anoxic Archean atmosphere. Astrobiology 3. Farquhar J, Bao H, Thiemens M (2000) Atmospheric influence of Earth’s earliest sulfur 2(1):27–41. cycle. Science 289(5480):756–759. 15. Romero AB, Thiemens MH (2003) Mass-independent sulfur isotopic compositions in 4. Baroni M, Savarino J, Cole-Dai J, Rai VK, Thiemens MH (2008) Anomalous sulfur iso- present-day sulfate aerosols. J Geophys Res 108(D16):4524. tope compositions of volcanic sulfate over the last millennium in Antarctic ice cores. J 16. Myhre G, Berglen TF, Myhre CEL, Isaksen ISA (2004) The radiative effect of the an- Geophys Res 113(D20):D20112. thropogenic influence on the stratospheric sulfate aerosol layer. Tellus B Chem Phys 5. Baroni M, Thiemens MH, Delmas RJ, Savarino J (2007) Mass-independent sulfur iso- Meterol 56(3):294–299. topic compositions in stratospheric volcanic eruptions. Science 315(5808):84–87. 17. Lee CCW, Savarino J, Cachier H, Thiemens MH (2002) Sulfur (S-32, S-33, S-34, S-36) and 6. Savarino J, Romero A, Cole-Dai J, Bekki S, Thiemens MH (2003) UV induced mass- oxygen (O-16, O-17, O-18) isotopic ratios of primary sulfate produced from combus- independent sulfur isotope fractionation in stratospheric volcanic sulfate. Geophys tion processes. Tellus B Chem Phys Meterol 54(3):193–200. Res Lett 30(21):2131. 18. Bauman JJ, Russell PB, Geller MA, Hamill P (2003) A stratospheric aerosol climatology 7. Savarino J, Slimane B, Cole-Dai J, Thiemens MH (2003) Evidence from sulfate mass from SAGE II and CLAES measurements. 2. Results and comparisons, 1984-1999. independent oxygen isotopic compositions of dramatic changes in atmospheric oxi- J Geophys Res 108(D13):16. dation following massive volcanic eruptions. J Geophys Res 108(D21):ACH7-1-6. 19. Bruehl C, Lelieveld J, Crutzen PJ, Tost H (2012) The role of carbonyl sulphide as 8. McCabe JR, Thiemens MH, Savarino J (2007) A record of ozone variability in South a source of stratospheric sulphate aerosol and its impact on climate. Atmos Chem Phys Pole Antarctic snow: Role of nitrate oxygen isotopes. J Geophys Res 112(D12):D12303. 12(3):1239–1253. 9. Tollefson J (2014) Climate change: The case of the missing heat. Nature 505:276–278. 20. Notholt J, et al. (2003) Enhanced upper tropical tropospheric COS: Impact on the 10. Shaheen R, et al. (2013) Tales of volcanoes and El-Niño southern oscillations with stratospheric aerosol layer. Science 300(5617):307–310. the oxygen isotope anomaly of sulfate aerosol. Proc Natl Acad Sci USA 110(44): 21. Pratt KA, et al. (2011) Flight-based chemical characterization of biomass burning aerosols 17662–17667. within two prescribed burn smoke plumes. Atmos Chem Phys 11(24):12549–12565. 11. Lanciki A, Cole-Dai J, Thiemens MH, Savarino J (2012) Sulfur isotope evidence of little 22. Vodacek A, et al. (2002) Remote optical detection of biomass burning using a potas- or no stratospheric impact by the 1783 Laki volcanic eruption. Geophys Res Lett 39(1). sium emission signature. Int J Remote Sens 23(13):2721–2726. 12. Farquhar J, Savarino J, Airieau S, Thiemens MH (2001) Observation of wavelength- 23. Whitlow S, Mayewski P, Dibb J, Holdsworth G, Twickler M (1994) An ice-core-based sensitive mass-independent sulfur isotope effects during SO2 photolysis: Implications record of biomass burning in the arctic and sub-arctic, 1750-1980. Tellus B Chem Phys for the early atmosphere. J Geophys Res Planets 106(E12):32829–32839. Meterol 46(3):234–242.

11982 | www.pnas.org/cgi/doi/10.1073/pnas.1406315111 Shaheen et al. Downloaded by guest on September 27, 2021 24. Andreae MO, Merlet P (2001) Emission of trace gases and aerosols from biomass 33. Domagal-Goldman SD, Kasting JF, Johnston DT, Farquhar J (2008) Organic haze, burning. Global Biogeochem Cycles 15(4):955–966. glaciations and multiple sulfur isotopes in the Mid-Archean Era. Earth Planet Sci Lett 25. Rieder HE, et al. (2013) On the relationship between total ozone and atmospheric 269(1-2):29–40. dynamics and chemistry at mid-latitudes. Part 2. The effects of the El Niño/Southern 34. Kasting JF, Pavlov AA (2001) Photochemical modeling of mass-independent sulfur Oscillation, volcanic eruptions and contributions of atmospheric dynamics and isotope fractionation in low-O2 atmospheres. Astrobiology 1(3):399–400. chemistry to long-term total ozone changes. Atmos Chem Phys 13(1):165–179. 35. Halevy I, Johnston DT, Schrag DP (2010) Explaining the structure of the Archean mass- – 26. Ziemke JR, Chandra S (2003) La Nina and El Nino-induced variabilities of ozone in the independent sulfur isotope record. Science 329(5988):204 207. tropical lower atmosphere during 1970-2001. Geophys Res Lett 30(3):1142. 36. Farquhar J, Wu NP, Canfield DE, Oduro H (2010) Connections between sulfur cycle 27. Fromm M, et al. (2005) Pyro-cumulonimbus injection of smoke to the stratosphere: evolution, sulfur isotopes, sediments, and base metal sulfide deposits. Econ Geol – Observations and impact of a super blowup in northwestern Canada on 3–4 August 105(3):509 533. 37. Chiodini G, Cioni R, Raco B, Scandiffio G (1991) Carbonyl sulphide (COS) in geothermal 1998. J Geophys Res 110(D8):D08205. fluids: An example from the Larderello field (Italy). Geothermics 20(5):319–328. 28. Danielache SO, et al. (2012) Photoabsorption cross-section measurements of S-32, 38. Krasnopolsky VA (2010) Spatially-resolved high-resolution spectroscopy of Venus 2. S-33, S-34, and S-36 sulfur dioxide for the (BB1)-B-1-X(1)A(1) absorption band. Variations of HDO, OCS, and SO2 at the cloud tops. Icarus 209(2):314–322. J Geophys Res 117(D24301). 39. Palumbo ME, Geballe TR, Tielens A (1997) Solid carbonyl sulfide (OCS) in dense mo- 29. Lin Y, Sim MS, Ono S (2011) Multiple-sulfur isotope effects during photolysis of car- lecular clouds. Astrophys J 479(2):839–844. bonyl sulfide. Atmos Chem Phys 11(19):10283–10292. 40. Mumma MJ, et al. (2003) Remote infrared observations of parent volatiles in comets: 30. Hattori S, et al. (2013) SO2 photoexcitation mechanism links mass-independent sulfur A window on the early solar system. Interpretation of the Remote and In-Situ Ob- isotopic fractionation in cryospheric sulfate to climate impacting volcanism. Proc Natl servations of Small Bodies, Advances in Space Research, eds Worms JC, Klinger J – Acad Sci USA 110(44):17656 17661. (Pergamon, Kidlington, UK), Vol 31, pp 2563–2575. 31. Hattori S, et al. (2012) Isotope effect in the carbonyl sulfide reaction with O(3P). J Phys 41. Ueno Y, Danielache SO, Johnson MS, Yoshida N (2009) Carbonyl sulfide (OCS) in the – Chem A 116(14):3521 3526. Archean atmosphere. Geochim Cosmochim Acta 73(13):A1358. 32. Farquhar J, Johnston DT (2008) The oxygen cycle of the terrestrial planets: Insights 42. Ziemke JR, Chandra S, Oman LD, Bhartia PK (2010) A new ENSO index derived from into the processing and history of oxygen in surface environments. Oxygen in the satellite measurements of column ozone. Atmos Chem Phys 10(8):3711–3721. Solar System, Reviews in Mineralogy & Geochemistry, eds MacPherson GJ, Mittlefehldt 43. Savarino J, Alexander B, Darmohusodo V, Thiemens MH (2001) Sulfur and oxygen DW, Jones JH, Simon SB (Mineralogical Society of America, Chantilly, VA), Vol 68, pp isotope analysis of sulfate at micromole levels using a pyrolysis technique in a con- 463–492. tinuous flow system. Analytical Chemistry 73(18):4457–4462. EARTH, ATMOSPHERIC, AND PLANETARY SCIENCES

Shaheen et al. PNAS | August 19, 2014 | vol. 111 | no. 33 | 11983 Downloaded by guest on September 27, 2021