An improved inlet for precisely measuring the atmospheric Ar/N 2 ratio

The atmospheric Ar/N2 ratio is expected to be useful as a tracer of air-sea heat exchange, but this application has been hindered in part due to sampling artifacts. Here we show that the variability in δ(Ar/N2) due to thermal fractionation at the inlet can be on the order of 40-80 per meg, and we introduce the use of an aspirated solar shield that successfully minimizes such fractionation. The data collected using this new inlet have a mean diurnal cycle of 1.0 per meg or less, suggesting that any residual thermal fractionation effect is reduced to this level.