The NCEI Historical Data Record for Atmospheric Rivers: Visualization over the Western United States

December 15, 2025

CW3E staff, Emily Slinskey and Jon Rutz, along with collaborator Bin Guan (UCLA) and CW3E director Marty Ralph, recently published an article in the Bulletin of the American Meteorological Society (BAMS), titled “The NCEI Historical Data Record for Atmospheric Rivers: Visualization over the Western United States.” The research was supported by the U.S. National Centers for Environmental Information (NCEI). The work supports the Atmospheric Rivers and Extreme Precipitation Research, Prediction, and Applications Priority identified in CW3E’s Strategic Plan by documenting an important new long-term dataset.

The study documents the development of a new atmospheric river (AR) dataset that combines an AR detection tool (ARDT) with the AR scale, producing an AR climatology that both characterizes AR strength via the scale, and differentiates between ARs and other long-lived, high-IVT events (e.g., tropical cyclones) via the ARDT. The dataset is calculated at high spatiotemporal resolution over a long period of record (1940-present) and is updated monthly. Atmospheric rivers (ARs) are identified through the implementation of the intensity, geometric, and directional thresholds of the Tracking Atmospheric Rivers Globally as Elongated Targets version 3 (tARget v3) algorithm, and then subsequently categorized based on their strength using the AR scale. The resulting database, hereafter referred to as the AR Historical Data Record (HDR), therefore differentiates between AR and non-AR storm types, while retaining the AR-scale based categorical rankings on a scale of 1–5 (from weak to strong). The AR HDR is offered globally based on the fifth-generation global reanalysis dataset from the European Centre for Medium-Range Weather Forecasts (ERA5) from 1940–present at 6-h intervals and a 0.25° × 0.25° horizontal resolution.

The resulting data for both the AR scale and combined ARDT-AR scale are compared through selected cases (Fig. 1), long-term climatology, and interannual variability over the western U.S. between 1940–2022. Results show that the combined ARDT-AR scale usefully differentiates between AR and non-AR events in most cases. During individual events, the spatial extent of AR scale and ARDT-AR scale features can differ dramatically if the ARDT removes a large portion (or entirety) of the object.

The study demonstrates how the AR HDR can serve as a valuable resource for the scientific community and users who rely upon historical information to plan their operations. NCEI will host all AR HDR output online.

Figure 1. AR scale categories for three case study events (columns), including (a–c) Hurricane Kay, (d–f) a cutoff low, and (g–i) a landfalling AR. Results are shown for the AR scale (top row) and ARDT-AR scale (middle row) output. Gray shading indicates an AR0 designation. Instantaneous IVT (bottom row), based on ERA5 reanalysis, is shown for added context. IVT contours reflect 100 kg m⁻¹ s⁻¹ (black), 250 kg m⁻¹ s⁻¹ (red), and 250 kg m⁻¹ s⁻¹ intervals greater than 250 kg m⁻¹ s⁻¹ (gray dashed). (Fig. 1 from Slinskey et al., 2025)

Slinskey, E. A., Rutz, J. J., Guan, B., & Ralph, F. M. (2025). The NCEI Historical Data Record for Atmospheric Rivers: Visualization over the Western United States. Bulletin of the American Meteorological Society, (published online ahead of print 2025), BAMS-D-24-0342.1. https://doi.org/10.1175/BAMS-D-24-0342.1