Nitrous oxide (N2O) is an atmospheric trace gas that contributes to the greenhouse effect. It is also involved in the catalytic destruction of ozone in the stratosphere and is increasing in concentration by about 0.25% per year. The increase is believed to result from fertilizer use, emissions from internal combustion engines, biomass burning, and industrial processes (Khalil 1995). It is naturally produced by nitrification and denitrification in soils and in the oceans, and is destroyed in the stratosphere via photolysis (90%) and reaction with excited atomic oxygen [O(1D)] (10%). Its atmospheric lifetime is between 100 and 150 years. Although the major sources and sinks of N2O are known, they are poorly quantified and inadequately balanced, both in terms of mass exchange and in their N and O isotopic composition.
Stable isotopes have been used in the past to constrain sources and sinks of other atmospheric trace gases but have yet to be successfully applied to N2O. The isotopic approach to a global N2O budget is hindered by the wide range of observed isotopic values for each of the major natural sources, making it difficult to assign a unique value to each of the source terms. Soil flux samples have been shown to be variable but consistently depleted in both 15N and 18O relative to atmospheric N2O. Oceanic samples have exhibited a trend similar to typical nutrient profiles, with slightly depleted surface waters becoming progressively enriched along the nutricline and stabilizing with depth. Early analytical methods employed infrared absorption techniques (Wahlen 1985) or required decomposition of N2O with subsequent analyses of N2 and CO2. The use of direct injection techniques was introduced in 1993 when Kim and Craig reported heavy enrichment in both the N and O isotopes in two samples of stratospheric air. They proposed that a stratosphere to troposphere return flux of heavy N2O could balance the observed isotopically light source terms, although a simple mass-balance model showed that this led to a considerable overcorrection. Direct injection of N2O was subsequently shown to result in erroneous enrichment of 15N and Delta18O when contaminated by trace amounts of CO2.
We present results for 15N and 18O of N2O obtained from samples collected in the lower stratosphere. Five samples were collected at midnorthern latitudes on board NASA's WB-57 aircraft, and two samples were collected at high northern latitude during the 1988 Juelich balloon campaign. We also measured, for comparison, the isotopic composition of tropospheric N2O sampled in La Jolla, California, under clean air conditions. Nitrous oxide mixing ratios decreased with height above the tropopause, whereas the heavy-isotope composition of the remnant N2O was found to be increasingly enriched. If the process responsible for this enrichment is an irreversible sink and if the fractionation factor remains constant, the data should obey what is known as a Rayleigh distillation, in which the resulting isotopic enrichment is related to the fraction remaining by the equation
(1) R = R0 x fAlpha-1
where R and Ro are the residual (stratospheric) and initial (tropospheric) heavy-to-light isotope ratios, respectively; f is the fraction of N2O remaining (residual concentration divided by the initial concentration); and is the ratio of the heavy-to-light reaction or photolysis rates. This relationship