The CO2 isotopic composition in the upper troposphere: Results of the CARIBIC aircraft project.

CARIBIC (Civil Aircraft for the Regular Investigation of the atmosphere Based on an Instrumented Container) used a Boeing 767 to investigate atmospheric composition at cruising altitude, and resumes fall 2003 using an Airbus 340-600 of Lufthansa. Starting from 1998, whole air samples collected (12 per flight) were analysed for isotopic composition of CO_2 and CO. Before January 2001, some canisters demonstrated systematic depletion of ¹⁸O in CO_2 due to CO_2-water exchange; thereafter this canister effect has disappeared and all subsequent CO_2 isotope data are considered to be reliable. The δ¹³C values obtained demonstrate a seasonal cycle similar to ground stations. Furthermore, there are δ¹³C deviations due to sampling location (NH/SH hemisphere gradient) and due to intercepting stratospheric air. The latter is well documented by concurrent increases in O_3, decreases of SF_6, N_2O and CO_2 and by changes in CO isotopic composition as well. Here we focus on several samples collected between 38 to 48^oN on May 13 and 19, 2001 that have most prominent stratospheric signals. The O_3 content is 200 to 400 ppb above the chemical tropopause level (defined in Zahn et al., 2002), ¹⁴CO is up to 93 molecules/cc STP (against 9 to 12 in tropospheric air). Considering SF_6 data obtained on tropospheric CARIBIC samples from the NH as a typical range for the remote troposphere and correcting this line for 1/2 of a mean difference between the NH and SH (≈0.15 ppt after the plot of I. Levin, http://www.iup.uni-heidelberg.de/institut/forschung/en/groups/kk/SF6.html), the samples of interest have an SF_6 age of about 1.0±0.5 y. The δ¹³C values of the samples of interest being up to 0.2 per mil above the values measured on the tropospheric air at the same time give a tight correlation with the inverse content of CO_2 as well as N_2O, SF_6 and the O_3 excess above the level of the chemical tropopause. Because CO_2 from the upper stratosphere (where the air is about 6 years old) is known to have large ¹⁸O enrichments (4 to 8 per mil, e.g. Lämmerzahl et al., 2002), δ¹⁸O(CO_2) may serve as indicator of stratospheric history. However, δ¹⁸O(CO_2) of the samples of interest are within ±0.2 per mil equal to the values of tropospheric air. Therefore, we exclude significant admixture of old stratospheric air and consider air from the lowermost stratosphere as the most likely case.