Sensitivity of atmospheric oxidation chemistry to climate inferred from Greenland ice core records of Δ17O(NO3-)

Abstract:

The oxygen-17 excess of atmospheric nitrate (Δ¹⁷O(NO₃^-)) is indicative of the relative abundances of oxidants (O₃/HO_x, where HO_x = OH + HO₂ + RO₂) in the troposphere at the time of nitrate formation, with larger O₃/HO_x ratios leading to larger values of Δ¹⁷O(NO₃^-). The preservation of nitrate in polar ice sheets can thus be used to retrieve information related to past changes in local and/or regional tropospheric oxidation chemistry, contributing to the understanding of atmospheric chemistry and climate interactions. We measured Δ¹⁷O(NO₃^-) in ice core samples from the Greenland Ice Sheet Project 2 (GISP2), covering the Holocene and the last glacial periods, as well as two Dansgaard–Oeschger (DO) events in the last glacial period. The values of Δ¹⁷O(NO₃^-) are higher in the glacial period compared to during the Holocene. In addition, Δ¹⁷O(NO₃^-) changes abruptly by as much as 3 ‰ in ~100 years during these two DO events, suggesting a high sensitivity of O₃/HO_x to climate in the mid- to high northern latitudes. The observations suggest a general relationship between tropospheric O₃/HO_x ratios and climate, with larger O₃/HO_x ratios in colder climates compared to during warmer climates. High resolution measurements during the two DO events further indicate a second regime that in relatively warm climates, the O₃/HO_xratios decrease with decreasing temperature.

We hypothesize that the switch between the two regimes is due to changes in the relative importance of 1) surface emissions and chemistry, and 2) stratosphere-troposphere exchange (STE) for tropospheric O₃/HO_x in the mid- to high northern latitudes. Surface emissions and chemistry dominate variability in tropospheric O₃/HO_x in relatively warm climates. In colder climates, changes in STE become dominant for the observed variability in tropospheric O₃/HO_x in the mid- to high northern latitudes. Prior calculations from global chemistry-climate models are qualitatively consistent with this hypothesis.