Assessing Precipitation Isotope Variations during Atmospheric River Events to Reveal Dominant Atmospheric/Hydrologic Processes

Abstract:

Extreme precipitation events across the Western US commonly associated with atmospheric rivers (ARs), whereby extensive fluxes of moisture are transported from the subtropics, can result in major damage and are projected by most climate models to increase in frequency and severity. However, they are difficult to project beyond ~ten days and the location of landfall and topographically induced precipitation is even more uncertain. Water isotopes, often used to reconstruct past rainfall variability, are useful natural tracers of atmospheric hydrologic processes. Because of the typical tropical and sub-tropical origins, ARs can carry unique water isotope (δ¹⁸O and δ²H, d-excess) signatures that can be utilized to provide source and process information that can lead to improving AR predictions. Recent analysis of the top 10 weekly precipitation total samples from Sequoia National Park, CA, of which 9 contained AR events, shows a high variability in the isotopic values. NOAA Hysplit back trajectory analyses reveals a variety of trajectories and varying latitudinal source regions contributed to moisture delivered to this site, which may explain part of the high variability (δ²H = -150.03 to -49.52 ‰, δ¹⁸O = -19.27 to -7.20 ‰, d-excess = 4.1 to 25.8). Here we examine the top precipitation totals occurring during AR events and the associated isotopic composition of precipitation samples from several sites across the Western US. We utilize IsoGSM, an isotope-enabled atmospheric general circulation model, to characterize the hydrologic processes and physical dynamics contributing to the observed isotopic variations. We investigate isotopic influences from moisture source location, AR speed, condensation height, and associated temperature. We explore the dominant controls on spatial and temporal variations of the isotopic composition of AR precipitation which highlights different physical processes for different AR events.