Dataset Details

HRRR ML Storm Mode Analyses

Storms extracted from the HRRR and associated modes predicted by machine learning. The dataset consists of storms extracted from the operational HRRR runs with the Hagelslag package. Multiple storm mode machine learning models are applied to the storms to predict the mode (supercell, QLCS, or disorganized). The data are available in an AWS cloud bucket to support interactive visualization and analysis but could be used for other applications.

185 kt tropical cyclone

Data Type

Numerical model output

Dataset Size

445.3 GB and growing

File Type

csv, netCDF, geojson

Time Range (YYYY-MM-DD)

2022-04-03 to present

Geographic Scope

Continental U.S.

Spatial Resolution

3 km grid spacing

Time Resolution


Grid Size

Continental U.S.

Input Data Source

High Resolution Rapid Refresh (HRRR)

Application / Use Case

Surface and subsurface ocean conditions, marine weather, marine fisheries, particle tracking, etc. 

Zambon, J. B., R. He, and J. C. Warner (2014) Investigation of Hurricane Ivan using the Coupled Ocean-Atmosphere-Wave-Sediment Transport (COAWST) Model, Ocean Dynamics, 64(11), 1535-1554, doi: 10.1007/s10236-014-0777-7

Dataset and/or GitHub URL

Point of Contact

David John Gagne (dgagne [at] ucar [dot] edu)