Forward modeling of δ18O in Andean ice cores

Abstract

Tropical ice core archives are among the best dated and highest resolution from the tropics, but a thorough understanding of processes that shape their isotope signature as well as the simulation of observed variability remain incomplete. To address this, we develop a tropical Andean ice core isotope forward model from in situ hydrologic observations and satellite water vapor isotope measurements. A control simulation of snow δ18O captures the mean and seasonal trend but underestimates the observed intraseasonal variability. The simulation of observed variability is improved by including amount effects associated with South American cold air incursions, linking synoptic-scale disturbances and monsoon dynamics to tropical ice core δ18O. The forward model was calibrated with and run under present-day conditions but can also be driven with past climate forcings to reconstruct paleomonsoon variability. The model is transferable and may be used to render a (paleo)climatic context at other ice core locations.