Triple isotope composition of oxygen in atmospheric water vapor

Recently, an excess of 17O (17O‐excess) has been demonstrated in meteoric water and ice cores. Based on theory and experiments, it has been suggested that this excess originates from evaporation of ocean water into under‐saturated air. However, there has never been direct demonstration of this excess in marine vapor. Here, we present results of the first measurements of δ17O and δ18O in vapor samples collected over the South Indian and the Southern Oceans. Our data show the existence of 17O‐excess in marine vapor and also clear negative correlation between 17O‐excess and relative humidity. Thus, 17O‐excess is useful for constraining oceanic humidity in hydrological and climatic models. Using the obtained values of 17O‐excess, we estimated the fractionation factor between H218O and H216O for diffusion in air above the ocean (18αdiff). The new estimation of 18αdiff (1.008) is larger than the widely accepted value in hydrological studies.