In high-temperature volcanic fumaroles (>400°C), the isotopic composition of molecular hydrogen (H 2 ) reaches equilibrium with that of the fumarolic H 2 O. In this study, we used this hydrogen isotope exchange equilibrium of fumarolic H 2 as a tracer for the remote temperature at volcanic fumaroles. In this remote sensing, we deduced the hydrogen isotopic composition (dD value) of fumarolic H 2 from those in the volcanic plume. To ascertain that we can estimate the dD value of fumarolic H 2 from those in a volcanic plume, we estimated the values in three fumaroles with outlet temperatures of 630°C (Tarumae), 203°C (Kuju), and 107°C (E-san). For this we measured the concentration and dD value of H 2 in each volcanic plume, along with those determined directly at each fumarole. The average and maximum mixing ratios of fumarolic H 2 within a plume's total H 2 were 97% and 99% (at Tarumae), 89% and 96% (at Kuju), and 97% and 99% (at E-san). We found a linear relationship between the depletion in the dD values of H 2 , with the reciprocal of H 2 concentration. Furthermore, the estimated end-member dD value for each H 2 -enriched component (À260 ± 30& vs. VSMOW in Tarumae, À509 ± 23& in Kuju, and À437 ± 14& in E-san) coincided well with those observed at each fumarole (À247.0 ± 0.6& in Tarumae, À527.7 ± 10.1& in Kuju, and À432.1 ± 2.5& in E-san). Moreover, the calculated isotopic temperatures at the fumaroles agreed to within 20°C with the observed outlet temperature at Tarumae and Kuju. We deduced that the dD value of the fumarolic H 2 was quenched within the volcanic plume. This enabled us to remotely estimate these in the fumarole, and thus the outlet temperature of fumaroles, at least for those having the outlet temperatures more than 400°C. By applying this methodology to the volcanic plume emitted from the Crater 1 of Mt. Naka-dake (the volcano Aso) where direct measurement on fumaroles was impractical, we estimated that the dD value of the fumarolic H 2 to be À172 ± 16& and the outlet temperature to be 868 ± 97°C. The remote temperature sensing using hydrogen isotopes developed in this study is widely applicable to many volcanic systems.

We have developed a new automated analytical system that employs a continuous flow isotope ratio mass spectrometer to determine the stable hydrogen isotopic composition (δD) of nanomolar quantities of molecular hydrogen (H(2)) in an air sample. This method improves previous methods to attain simpler and lower-cost analyses, especially by avoiding the use of expensive or special devices, such as a Toepler pump, a cryogenic refrigerator, and a special evacuation system to keep the temperature of a coolant under reduced pressure. Instead, the system allows H(2) purification from the air matrix via automatic multi-step gas chromatographic separation using the coolants of both liquid nitrogen (77 K) and liquid nitrogen + ethanol (158 K) under 1 atm pressure. The analytical precision of the δD determination using the developed method was better than 4‰ for >5 nmol injections (250 mL STP for 500 ppbv air sample) and better than 15‰ for 1 nmol injections, regardless of the δD value, within 1 h for one sample analysis. Using the developed system, the δD values of H(2) can be quantified for atmospheric samples as well as samples of representative sources and sinks including those containing small quantities of H(2) , such as H(2) in soil pores or aqueous environments, for which there is currently little δD data available. As an example of such trace H(2) analyses, we report here the isotope fractionations during H(2) uptake by soils in a static chamber. The δD values of H(2) in these H(2)-depleted environments can be useful in constraining the budgets of atmospheric H(2) by applying an isotope mass balance model.