Ensemble-Based Characterization of Historical Storm Types in ERA5 over CONUS

Ensemble-Based Characterization of Historical Storm Types in ERA5 over CONUS Deeksha Rastogi1, Kevin Hodges2, Moetasim Ashfaq1 and Shih-Chieh Kao3 1Computational Sciences and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, TN 2The University of Reading, Reading, UK 3Environmental Sciences Division, Oak Ridge National Laboratory, Oak Ridge, TN We present Storm-Type Labeled Precipitation, a gridded dataset that assigns storm types to ERA5 precipitation across the conterminous United States (CONUS). We classify precipitation broadly into five storm types: extratropical cyclone (ETC), hurricane (HUR) mesoscale convective system (MCS), atmospheric river (AR), and other convective. Each grid cell and time step is labeled as one of five classes: 1 = MCS, 2 = ETC, 3 = HUR, 4 = AR, 5 = Other convective, with 0 = Unidentified reserved for cases with no detected type. We provide two products: (i) TempestExtremes (TE)-only, where storm types are derived using TE applied to ERA5 fields (Ulrich et al., 2017 and 2021); and (ii) Hybrid TRACK–TE, which uses TRACK (Hodges, 1995 and 1999) for ETCs and HURs, and TE for MCSs and ARs. Both products are provided on the ERA5 grid over the CONUS (20°N–50°N, 66°W–125°W) at a 6hourly timestep. ETCs, HURs, and ARs are detected at 6-hourly resolution, whereas MCSs are identified at ERA5’s native hourly resolution and subsequently aggregated to 6-hourly fields for consistency. Candidate features and tracks are identified using different criteria. MCS detection requires a minimum duration criteria and spatial overlap thresholds between consecutive timesteps (Reed et al., 2023). AR detection requires a minimum area filter and an Integrated Vapor Transport threshold (Ulrich et al., 2021) and is limited to the western US. ETCs and HURs are identified using SLP minima in TE and using vorticity in TRACK, with an additional warm-core criterion applied to hurricanes. ETC are further filtered based on minimum duration, sufficient cumulative path length, limits on retrograde motion, and net eastward displacement. Hurricane are filtered using minimum wind speed and duration requirements, along with geographic filters to avoid land-locked or mid-latitude impostors (e.g., requiring ocean interaction and excluding tracks confined to far-western or northern synoptic domains). For MCSs and ARs, grid cells within the detected blobs are assigned to the respective storm types. For ETC and HURs, precipitation within fixed great-circle distances—1000 km (~9°) around ETC centers and 667 km (~6°) around hurricane centers—is assigned to the storms. Finally, if precipitation at a grid cell/time step is initially labeled as Unidentified (0) but ERA5 convective precipitation > 0, it is relabeled as 5 (Other convective). Outputs are integer masks aligned with ERA5 precipitation, enabling storm-type attribution of rainfall, event compositing, trend analysis, and model evaluation. Future efforts will extend storm-type detection to high-resolution, downscaled Earth system model projections to support evaluations of projected changes in hydrologic hazards that may threaten critical water and energy infrastructures in watersheds of hydropower importance. References- • Hersbach, H., Bell, B., Berrisford, P., Hirahara, S., Horányi, A., Muñoz‐Sabater, J., ... & Thépaut, J. N. (2020). The ERA5 global reanalysis. Quarterly journal of the royal meteorological society, 146(730), 1999-2049. • Hodges, K. I. (1995). Feature tracking on the unit-sphere. Monthly Weather Review, 123(12), 3458–3465. https://doi.org/10.1175/15200493(1995)123<3458:Ftotus>2.0.Co;2 • Hodges, K. I. (1999). Adaptive constraints for feature tracking. Monthly Weather Review, 127(6), 1362–1373. https://doi.org/10.1175/15200493(1999)127<1362:Acfft>2.0.Co;2 • Reed, K. A., Stansfield, A. M., Hsu, W. C., Kooperman, G. J., Akinsanola, A. A., Hannah, W. M., ... & Medeiros, B. (2023). Evaluating the simulation of CONUS precipitation by storm type in E3SM. Geophysical Research Letters, 50(12), e2022GL102409. • Ullrich, P. A., Zarzycki, C. M., McClenny, E. E., Pinheiro, M. C., Stansfield, A. M., & Reed, K. A.(2021). TempestExtremes v2. 1: a community framework for feature detection, tracking, and analysis in large datasets. Geoscientific Model Development, 14(8), 5023-5048. • Ullrich, P. A., & Zarzycki, C. M. (2017). TempestExtremes: A framework for scaleinsensitive pointwise feature tracking on unstructured grids. Geoscientific Model Development, 10(3), 1069-1090. Acknowledgment This work is supported by the US Department of Energy (DOE) Water Power Technologies Office. Support for the simulations, data storage, and analysis is provided by the Oak Ridge Leadership Computing Facility, which is a DOE Office of Science User Facility.