Yuzhen Yan

The ability of ice cores to trap ancient atmosphere naturally raises the question whether one could understand past atmospheric chemistry from the trapped bubbles. Yet, understanding the history of atmosphere chemistry, particularly the oxidizing power of the atmosphere, has long been challenging, because the reactive species in the atmosphere are often short-lived and present in extremely low concentrations. That said, chemical processes often fractionate the stable isotopes of the reactants due to kinetical differences. That means that these short-lived trace species could leave an "isotopic fingerprint" on other major species, which are well preserved in ice cores and could be readily measured. By analyzing the isotopic composition of such major species, we could potentially infer the atmospheric chemistry in Earth history. The "clumped" isotopes of O₂ (that is, two heavy ¹⁸O atoms bonding in the form of ¹⁸O¹⁸O) are a promising proxy to reveal past atmosphere chemistry, in particular tropospheric ozone concentrations.

The basic principle of isotope clumping in O₂ is that, if the bulk composition of O₂ is known, we could infer the theoretical abundance of ¹⁸O¹⁸O. For example, if 1 in every 100 O₂ molecules is ¹⁸O¹⁶O, we would expect 1 ¹⁸O¹⁸O molecule in every 10,000 O₂ molecules if the formation of chemical bond is purely random. However, thermodynamics dictates that this is not a random process, and heavy atoms tend to form chemical bonds with each other (say, instead 1 ¹⁸O¹⁸O molecule there are 1.01 ¹⁸O¹⁸O molecules in 10,000 O₂ molecules). This excess of heavy atoms "clumped" together is what we try to measure [in the case of O₂, Δ₃₆, which equals R₃₆/(R₁₈²) - 1]. Importantly, thermodynamic theories also predict that the extent of isotope clumping scales inversely with temperature. At very high temperature, "clumping" disappears and the distribution conforms to stochastic statistics.

For Δ₃₆ values of O₂ in the lower troposphere, two more realistic considerations are worth mentioning here. First, Δ₃₆ values are also affected by kinetics: namely, how fast could Δ₃₆ respond to a change in temperature. Photochemistry involving ozone (O₃) plays a key role here. In the simplest language, a higher level of O₃ allows the Δ₃₆ to reach the equilibrium value (set by temperature) faster. Second, stratosphere-to-troposphere transport brings O₂ with heavy Δ₃₆ values into the lower troposphere. Overall, this means that the Δ₃₆ value of lower-tropospheric O₂ reflects mixing of two end-members, each of which have its own characteristic Δ₃₆ value depending on temperature and reaction rates (O₃), shows Figure 1.

To first order, stratospheric Δ₃₆ quickly reach equilibrium values at a given temperature due to high flux of UV light and high O₃ concentrations. In the troposphere, however, Δ₃₆ is largely kinetically limited. Thus, O₃ level has a larger influence on the tropospheric Δ₃₆ values. If both troposphere and stratosphere temperature can be constrained, changes in Δ₃₆ values should primarily reflect the level of tropospheric ozone (Yeung et al, 2019).

I used the S27 ice core drilled from Allan Hills, East Antarctica to measure the Δ₃₆ of trapped O₂. This core provides a continuous deglacial record from Penultimate Glacial Maximum (PGM) to the Last Interglacial (LIG) (Spaulding et al, 2013; Yan et al, 2021; also check out the entry Cryosphere changes). Since we know the temperature difference between LIG and the Pre-Industrial Period is probably within 2 degree C (Otto-Bliensner et al, 2021), the measured Δ₃₆ gives us perhaps the oldest record of tropospheric ozone.

Surprisingly, despite the similarity in atmospheric temperature, the measured LIG Δ₃₆ value is 0.03 ± 0.02‰ (95% CI) higher than the late Holocene/pre-industrial (PI; 1,590–1,850 CE) value. In an atmospheric chemistry-transport model (GEOS-Chem), this difference corresponds to a 9% reduction in LIG tropospheric O₃ burden (95% CI: 3%–15%) and can be caused by a ~70% reduction in biomass burning emissions during the LIG relative to the PI.

This work has been published on Geophysical Research Letters:

Yan, Y., Banerjee, A., Murray, L.T., Tie, X. and Yeung, L.Y., 2022. Tropospheric ozone during the Last Interglacial. Geophysical Research Letters, 49(23), p.e2022GL101113.Yan, Y., Banerjee, A., Murray, L.T., Tie, X. and Yeung, L.Y., 2022. Tropospheric ozone during the Last Interglacial. Geophysical Research Letters, 49(23), p.e2022GL101113.

Otto-Bliesner, B.L., Brady, E.C., Zhao, A., Brierley, C.M., Axford, Y., Capron, E., Govin, A., Hoffman, J.S., Isaacs, E., Kageyama, M. and Scussolini, P., 2021. Large-scale features of Last Interglacial climate: results from evaluating the lig127k simulations for the Coupled Model Intercomparison Project (CMIP6)–Paleoclimate Modeling Intercomparison Project (PMIP4). Climate of the Past, 17(1), pp.63-94.

Spaulding, N.E., Higgins, J.A., Kurbatov, A.V., Bender, M.L., Arcone, S.A., Campbell, S., Dunbar, N.W., Chimiak, L.M., Introne, D.S. and Mayewski, P.A., 2013. Climate archives from 90 to 250 ka in horizontal and vertical ice cores from the Allan Hills Blue Ice Area, Antarctica. Quaternary Research, 80(3), pp.562-574.

Yan, Y., Spaulding, N.E., Bender, M.L., Brook, E.J., Higgins, J.A., Kurbatov, A.V. and Mayewski, P.A., 2021. Enhanced Moisture Delivery into Victoria Land, East Antarctica During the Early Last Interglacial: Implications for West Antarctic Ice Sheet Stability. Climate of the Past, 17, pp.1841-1855.

Yeung, L.Y., Murray, L.T., Ash, J.L., Young, E.D., Boering, K.A., Atlas, E.L., Schauffler, S.M., Lueb, R.A., Langenfelds, R.L., Krummel, P.B. and Steele, L.P., 2016. Isotopic ordering in atmospheric O₂ as a tracer of ozone photochemistry and the tropical atmosphere. Journal of Geophysical Research: Atmospheres, 121(20), pp.12-541.

Yeung, L.Y., Murray, L.T., Martinerie, P., Witrant, E., Hu, H., Banerjee, A., Orsi, A. and Chappellaz, J., 2019. Isotopic constraint on the twentieth-century increase in tropospheric ozone. Nature, 570(7760), pp.224-227.