The involvement of slab-derived fluids in non-volcanogenic hot springs in SW Japan

The main aim of this study is to contribute to better understanding of fluid processes occurring in subduction zones with a comprehensive framework involving slab-derived fluids to near-surface fluids such as seawater, meteoric water and hot spring waters. In the non-volcanic or forearc regions, evidences that show an involvement of slab-derived fluids have been hardly found. One exception could be the so-called ‘Arima-type brine,’ welling up in the Kinki district, SW Japan. According to the previous geochemical and hydrological studies (Matsubaya et al., 1974; Sano and Wakita, 1985; Morikawa et al., 2008), their oxygen, hydrogen, and helium isotopic ratios are anomalously high and similar to those of magmatic fluids or the upper mantle, despite the absence of Quaternary volcanic activities or underlying magma. It is speculated that the Arima-type brine might be relevant to NaCl-CO2-rich aqueous fluids originated from a deep part of the forearc region, which is possibly slab-derived fluids. The purposes of this study are to examine the origin of the Arima-type brine and to test its possible relationship to slab-derived fluids. In order to resolve various processes and sources relevant to the origin of spring solutions, a multi-elemental and isotopic data set for each sample was obtained from the hot springs in the Arima area, and a robust multivariate database by accumulating such a dataset was constructed. In this study, the concentrations of major ions, total alkalinity, dD, d18O, d13C, the concentrations of noble gases, 3He/4He, 87Sr/86Sr, and Pb isotopes were measured. The observed variations of the spring waters can be explained by mixing of the two end-member waters, i.e., meteoric water and the possible ‘original deep brine’. The binary mixing has already been proposed for the origin of the Arima-type brine in the oxygen-hydrogen stable isotopic space (e.g., Masuda et al., 1985), and for the first time this study identifies and characterizes the original deep brine in the multi-elemental/isotopic space. The estimated original deep brine has the Sr concentration of 86 ppm and the 87Sr/86Sr ratio of 0.70873, which is close to the predicted ratio of the Philippine Sea slab-derived fluid, i.e., 0.709492, and is distinct from the Pacific slab-derived fluid, i.e., 0.704762 (calculated values are taken from Nakamura et al., 2008). Furthermore, a numerical model to predict geochemistry of slab-derived fluids was constructed in a framework related to dehydration processes from subducting slabs and the corresponding isotopic fractionations between generated fluids and minerals. It suggests that hydrogen and oxygen isotopic ratios of slab-derived fluids are similar to observationally estimated values of the original deep brine at about 50 km depth, which coincides with the depth of Philippine Sea plate beneath the studied area. These results may explain the similarities of the Arima-type brine to magmatic fluids as the involvement of the Philippine Sea slab-derived fluids without being involved by magma formation processes.