Skip to content

CubeDynamics: Climate Cube Math

PRISM JJA variance

PRISM JJA variance

This recipe computes June–July–August precipitation variance from PRISM for a Boulder, CO point. It extends the original "PRISM precipitation anomaly / z-score cube" guide and adds Lexcube-ready context.

Story

PRISM cubes complement gridMET by offering higher-resolution precipitation records. We stream a monthly cube, compute anomalies, filter to summer months, and collapse the time dimension with v.variance for a single summary layer.

Code

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

cube = cd.load_prism_cube(
    lat=40.0,
    lon=-105.25,
    start="2000-01-01",
    end="2020-12-31",
    variable="ppt",
)

jja_var = (
    pipe(cube)
    | v.anomaly(dim="time")
    | v.month_filter([6, 7, 8])
    | v.variance(dim="time")
).unwrap()

v.variance defaults to keep_dim=True, so the time dimension remains in the result (length 1) and stays Lexcube-ready. Set keep_dim=False when you intentionally want a 2D map:

jja_map = pipe(cube) | v.month_filter([6, 7, 8]) | v.variance(dim="time", keep_dim=False)

Because the cube stays in xarray, you can coarsen it, clip to focus on extremes, or export to NetCDF with v.to_netcdf. Reuse the same grammar for other PRISM variables (tmax, tdmean) by switching the variable parameter.

Optional: Lexcube visualization

pipe(jja_var) | v.show_cube_lexcube(title="PRISM JJA precipitation variance")

Lexcube previews help verify that AOI and time selections are correct before you compute downstream stats or share results. For larger regions, stride or coarsen the cube prior to visualization.