Carbon and hydrogen isotopic compositions of stratospheric methane: 1. High‐precision observations from the NASA ER‐2 aircraft

Measurements of δ 13 C and δD of atmospheric CH 4 from whole air samples collected in the upper troposphere and lower stratosphere aboard the NASA ER‐2 aircraft during the SOLVE (2000), POLARIS (1997), and STRAT (1996) campaigns are reported. Samples cover latitudes from 1°S to 89°N and altitudes from 11 to 21 km, providing CH 4 mixing ratios that range from 1744 to 716 ppbv. Measurements of isotope ratios were made by continuous‐flow gas chromatography isotope ratio mass spectrometry which provides high‐precision analyses on 60 ml aliquots of air. These measurements comprise the first upper atmosphere isotopic CH 4 data set to date using this technique and the most extensive with respect to latitude and season in any case. Values of δ 13 C‐CH 4 on the V‐PDB scale range from −47.28‰ near the tropical tropopause to −34.05‰ in the high northern latitude stratosphere. Values of δD on the V‐SMOW scale range from −90.9‰ to +26.4‰. Correlations of isotope ratios with CH 4 mixing ratios show enrichment in the heavy isotopes as CH 4 mixing ratios decrease due to kinetic isotope effects associated with oxidation by reaction with OH, Cl, and O( 1 D). Empirical fractionation factors are found to be highly dependent on the range of CH 4 mixing ratio considered, increasing with decreasing mixing ratio. Systematic nonlinearity in a Rayleigh fractionation model suggests a range of stratospheric fractionation factors, α C strat = 1.0108 ± 0.0004 to 1.0204 ± 0.0004 (2σ) and α H strat = 1.115 ± 0.008 to 1.198 ± 0.008 (2σ), from high to low CH 4 mixing ratio, respectively. The variation in α over the range in mixing ratios reflect changes in partitioning between CH 4 sink reactions in different regions of the stratosphere. In Part 1, these new high‐precision observations are discussed and compared with other stratospheric and tropospheric isotope measurements. In Part 2 [ McCarthy et al. , 2003 ] the observations are compared with 2‐D model results, and implications for the kinetic isotope effects for reactions with OH, Cl, and O( 1 D) are discussed.