Spatial-seasonal patterns reveal large-scale atmospheric controls on Asian Monsoon precipitation water isotope ratios

Correlated seasonal variation in precipitation amount and its oxygen isotope ratios (δ18O) has long been observed and frequently invoked in the interpretation of paleo-δ18O records from the Asian Monsoon (AM) region. However, the underlying cause of the observed seasonal δ18O variation is still under debate. Precipitation δ18O values show a single, consistent seasonal pattern across the region, with high values in pre-monsoon (late-spring to early-summer) and low values in the monsoon mature (mid- to late-summer) periods. We tested three hypotheses that may give rise to the measured precipitation δ18O pattern, involving variation in: 1) local precipitation, 2) moisture sources, and 3) convection in moisture source regions. We show that seasonal precipitation amounts across the AM region exhibit two pattern types: pre-monsoon peak or monsoon mature peak, and thus do not provide a consistent explanation for the isotopic pattern. We test the hypotheses that changing moisture sources or moisture source isotope ratios drive the seasonal isotopic pattern using a combination of Lagrangian moisture source attribution and output from an isotope-enabled General Circulation Model (LMDZ4), but find little seasonal variation in moisture sources; that which does exist is not consistent with the observed isotopic change. Instead, we show that the precipitation δ18O transition from pre-monsoon to monsoon mature stage is correlated with intensification of convective activity in moisture source regions, which is the result of monsoon establishment and northward migration of the Intertropical Convergence Zone (ITCZ). We propose that light water vapor, transferred to the lower-troposphere through downdrafts and evaporation of rain in the vicinity of intense convection, is advected to the AM region and labels precipitation during the monsoon mature period. These results demonstrate that precipitation δ18O in the AM region more strongly reflect large-scale atmospheric dynamics than local precipitation amount or moisture source, guiding the interpretation of paleo-isotope data from this region.