Skip to content

gridMET

What this dataset is

gridMET is a gridded surface meteorology product for the contiguous United States at ~4 km (1/24°) resolution with daily observations back to 1979. Variables include precipitation, maximum/minimum temperature, humidity, vapor pressure deficit, and wind, arranged on a regular latitude–longitude grid.

Quickstart

Get the stream (CubeDynamics grammar)

import cubedynamics as cd
from cubedynamics import pipe, verbs as v

cube = cd.gridmet(
    lat=40.0,
    lon=-105.25,
    start="2020-06-01",
    end="2020-06-30",
    variable="tmmx",
)

pipe(cube) | v.mean(dim="time") | v.plot()

Preview plot

gridMET preview

!!! note Image placeholder — after running the code below locally, save a screenshot to docs/assets/datasets/gridmet-preview.png.

Regenerate this plot

  1. Run the Quickstart code block in a local Python session or notebook.

  2. Capture the viewer returned by the pipe:

    python viewer = (pipe(cube) | v.mean(dim="time") | v.plot()).unwrap() viewer.save("docs/assets/datasets/gridmet-preview.html")

  3. Open docs/assets/datasets/gridmet-preview.html in a browser and take a 1200×700 px PNG screenshot saved to docs/assets/datasets/gridmet-preview.png.

Who collects it and why

The dataset is produced by John Abatzoglou and collaborators at the University of Idaho to support ecological, hydrological, and fire-weather applications across CONUS. It blends PRISM climatology with NLDAS reanalysis to provide spatially consistent daily meteorology widely used in ecological forecasting and climate impact studies.

How CubeDynamics accesses it

load_gridmet_cube attempts a streaming backend first, opening yearly NetCDF files over HTTP and subsetting the requested area and time range before chunking into a VirtualCube-like Dask structure. When streaming is unavailable it falls back to a small cached download while preserving the same (time, y, x) interface. Users request AOIs by point, bounding box, or GeoJSON, enabling fast analysis without retrieving the full continental archive.

!!! important "Temporal frequency and safety" - Daily (freq="D") is recommended for fire/event windows. Monthly start ("MS") requests over short ranges can produce an empty time axis; the loader now raises with guidance instead of silently returning NaNs. - Set allow_synthetic=False (default) to require real data. When True, the loader fabricates data and records provenance (source, is_synthetic, backend_error, freq, requested_start, requested_end).

Important variables and dimensions

Field Meaning Units
time Daily observation timestamp ISO date
y / x (lat / lon) Grid cell centers in geographic coordinates degrees
pr Precipitation mm day⁻¹
tmmx / tmmn Daily maximum / minimum temperature K
vpd Vapor pressure deficit Pa
vs / erc Wind speed / energy release component m s⁻¹ / index

Citation

Abatzoglou, J. T. (2013). Development of gridded surface meteorological data for ecological applications. International Journal of Climatology, 33(1), 121–131. https://doi.org/10.1002/joc.3413

See also: Fire event vase + climate merge (fire_plot)

Back to Datasets Overview Next recommended page: Which dataset should I use?