Storm250-L2 : A Storm-Centric 250 m 250\text{\,}\mathrm{m} NEXRAD Level-II Dataset for High-Resolution ML Nowcasting
Andy Shi · 2025
Read Storm250-L2 : A Storm-Centric 250 m 250\text{\,}\mathrm{m} NEXRAD Level-II Dataset for High-Resolution ML Nowcasting by Andy Shi (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-16031.txt.
Description
Andy Shi, as sole author, releases a storm-centred radar data set at quarter-kilometre resolution for machine learning nowcasting. Nowcasting predicts a storm's next moves from radar reflectivity sequences. Public data sets at one to two kilometre resolution smooth away the fine structures that matter most for extreme weather, which has held model quality back.
Storm250-L2 is derived from NEXRAD Level-II data and GridRad-Severe storm tracks. An algorithm crops a fixed high-resolution window around each tracked storm, keeps the native polar geometry, and supplies consistent sequences of single-tilt sweeps plus a pseudo-composite reflectivity product. Thousands of storm events across the continental United States are packaged as HDF5 tensors with rich context metadata and reproducible file lists. The data set aims to let nowcasting models learn the sharp edges of convective storms.
Catalog record
| CREATOR | Andy Shi |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 13 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-16031