herbie-data

What it is and what it does

Herbie is a Python interface to numerical weather prediction model data from NOAA (HRRR, GFS, RAP, GEFS, NAM, and others), ECMWF (IFS, AIFS), and additional sources. It abstracts away the complexity of locating and downloading GRIB2 files from multiple archives (NOMADS, AWS, Google Cloud, Azure) by automatically searching available sources and downloading either full files or subsets by variable. The package integrates tightly with xarray and pandas, allowing you to load downloaded data directly into analysis-ready formats.

Typical workflows involve creating a Herbie object with a date, model name, and product type, then either downloading raw GRIB2 files or reading specific variables directly into xarray Datasets. It includes a command-line interface for scripting and batch operations, plus built-in helpers for Cartopy-based mapping. The package is designed for researchers, meteorologists, and data scientists working with atmospheric forecasts—both for real-time operational use and historical analysis.

Use it for:

• Download high-resolution HRRR surface forecasts for regional weather analysis or nowcasting applications.
• Retrieve specific atmospheric variables (e.g., temperature at 850 mb) from global GFS forecasts for ensemble analysis.
• Batch-download multiple forecast hours across a date range for machine learning training on weather patterns.
• Access ECMWF IFS data for comparison with NOAA models in research or operational workflows.
• Extract point-specific data (e.g., at a weather station location) from GRIB2 files without processing entire grids.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Herbie downloads numerical weather prediction model data (GRIB2 files) from NOAA, ECMWF, and other sources, then loads it into xarray for analysis and visualization.

Yes. Herbie is actively maintained, has no known vulnerabilities, uses a permissive MIT license, and solves a real problem for weather data workflows. Install friction is low and dependencies are standard scientific Python packages. Suitable for research, operational meteorology, and data science projects involving atmospheric forecasts.

Install

herbie-data on PyPI

pip install herbie-data

uv add herbie-data

poetry add herbie-data

Installing herbie-data

Before you install

Low friction install with a pure-Python wheel. Active maintenance (last commit 2026-06-07) and 780 repository stars. Eight runtime dependencies including numpy, pandas, xarray, and cfgrib—all standard scientific Python packages with broad ecosystem support.

License in practice

MIT License permits commercial and private use, modification, and redistribution with minimal restrictions. You may use this in proprietary projects provided you include the license text and copyright notice.

Quickstart

pip install herbie-data

from herbie import Herbie

H = Herbie('2021-01-01 12:00', model='hrrr', product='sfc', fxx=6)
temperature = H.xarray('TMP:2 m')

Requires Python 3.11 or later. Optional features (wgrib2 integration) require manual system installation of wgrib2.

Verify before relying

• Whether all 15+ supported weather models are equally well-maintained and documented.
• Performance characteristics when downloading large subsets or multiple forecast hours.
• Specific data latency and availability guarantees across different source providers.
• Whether Cartopy integration requires additional system dependencies beyond those listed.

Package facts

Evidence: herbie_data-2026.3.0-py3-none-any.whl

Keywords: GRIB2, HRRR, atmosphere, forecast, meteorology, weather, xarray

Tags

More Atmospheric Science packages