Isotopic equilibrium between raindrops and water vapor during the onset and the termination of the 2005-2006 wet season in the Bolivian Andes

摘要

The isotopic equilibrium state between precipitation and low-level water vapor is a common assumption in numerous paleoclimate and atmospheric studies based on water stable isotopes. However, the paucity of field observations limits the validation of this assumption. This study examines the isotopic equilibrium state from event-based precipitation and daily near-surface water vapor samples collected during the onset and the termination of the 2005-2006 wet season in the Bolivian Andes (Zongo valley, 16°09' S, 68°07' W). Our observations show that the observed isotopic composition of precipitation (δD_p) deviates from the theoretical isotopic composition of precipitation at equilibrium with water vapor (δD_p_eq). Disequilibriums (ΔD_(p_eq) = δD_p -δD_p_eq) are mostly negative (73), indicating that precipitation is more depleted than a condensate that would have been formed from surface water vapor, and half of them are between -10 and + 10‰. They are significantly correlated to δD_p (r~2 = 0.30, n = 70, p < 0.001) suggesting that controls on δD_p also impact ΔD_(p_eq). Although equilibrium state does not prevail at the individual rain event scale, a strong relationship is observed between δD_p and δD_p_eq over the whole period of field samplings (r~2 = 0.86, n = 70, p < 0.001). The review of possible causes to explain the disequilibriums shows that below-cloud rain evaporation and diffusive exchanges are little involved. Other local processes such as rain type, condensation conditions and surface water recycling appear as better candidates to explain ΔD_(p_eq). Lastly, we explore how local processes affect δD_p. We show that large-scale dynamic along air masses history is dominant (nearly 80) to explain δD_p whereas local effects are dominant to explain deuterium excess in precipitation. In consequence, we conclude that δD_p is a correct candidate to examine and to reconstruct large-scale atmospheric processes from past to present time scales.