A model for the relationships between precipitation D/H ratios and precipitation intensity

Abstract

A simple quantitative model has been developed which predicts that a parametric correlation between the isotopic composition of precipitation and precipitation intensity can exist as a consequence of the condensation process when certain conditions are met. These conditions include the following: (1) approximate constancy, from one sampling interval to the next, of the initial isotopic composition of the water vapor in precipitation‐producing air masses; (2) approximate constancy of the initial condensation temperature; (3) approximate constancy of the rate of vertical ascent of the precipitating air mass; and (4) approximate constancy of the ratio of condensed water to water vapor. Such conditions are probably rare in nature for the short collection intervals represented by existing samples. However, with the probable exception of condition 3, tropical oceanic islands may satisfy the preceding conditions on the month‐to‐month time scale for which precipitation samples have generally been collected for isotopic analysis. Precipitation samples from Wake and Johnston islands that plot on the meteoric water line were used to compare monthly δD values with corresponding precipitation intensity. In general, there appears to be a convergence of the data at high values of δD (low intensity) and a divergence at low values of δD (high intensity). Such behavior is predicted by the isotopeintensity model largely in terms of variation in the vertical rate of ascent. For Wake Island this divergence of the data appears to be bimodal in character. However, for Johnston Island, there is considerable scatter in the data. This scatter is roughly constrained by the envelope defined by two vertical ascent rates that approximately characterize the presumably bimodal Wake Island trends. Thus much of the scatter in the Johnston Island data may be explained by month‐to‐variability in vertical ascent rates. The data for these two islands suggest that a simple precipitation δD versus intensity correlation is unlikely for monthly sample intervals. However, the issue remains unresolved for longer time scales.