The NCEI Climate Data Record for Atmospheric Rivers: Initial Results over the Western United States

The National Centers for Environmental Information (NCEI) is sponsoring the development of an atmospheric river (AR) climate data record (CDR) to serve as a valuable resource for the scientific community, water management, and decision making across the Western US (and soon, globally). While most CDRs are satellite-based, this CDR is based on atmospheric reanalysis with a long period of record (1940-present). The method chosen to identify ARs for the CDR uses the AR Scale, which broadly characterizes the strength of ARs using categorical rankings from 1-5 based on the peak integrated water vapor transport (IVT) and duration of AR conditions (i.e., IVT ≥ 250 kg m^-1 s^-1) at a given location. Since the AR scale does not account for object geometry, it detects some events associated with strong and persistent IVT that are not considered ARs (e.g., tropical storms and monsoon surges). We apply “tARget v3”, a well-known AR detection tool that uses climatological, geometric, and directional thresholds, to “filter” the AR scale and differentiate between AR and non-AR events. This study explores the resulting data and effects of this “filtering” through selected cases, long-term climatology, and interannual variability. NCEI will host both the unfiltered and filtered version online. During individual events, the spatial extent of “unfiltered” and “filtered” objects can differ dramatically (i.e., the spatial footprint of “filtered” objects is smaller), with greater differences apparent across the weaker, lower-category ARs compared to the higher-category ARs. The climatological AR frequencies for each of the 5 AR categories for both versions reveal higher-category ARs occur less frequently than lower-category ARs, with the greatest disagreement confined to the North Pacific. Over the Western U.S., disagreement is less, but a large fraction of lower-category events is removed further inland. Future work will include geographic expansion to a global data set, attribution of precipitation to identified events, and assessment of long-term trends.