⌨ Learn · Code sandbox

Run real analysis in your browser.

Real Python — xarray, numpy, pymannkendall, matplotlib — running entirely in your browser (via Pyodide/WebAssembly). No install, no NASA login, no backend. It opens synthetic data that mirrors the real product (dimensions, units, seasonality, noise) carrying a known planted signal — so when your analysis recovers it, you've verified your own method. Edit the code, hit Run.

editable · runs in your browser

import earthschool, numpy as np, pymannkendall as mk
import xarray as xr, matplotlib.pyplot as plt

# open a synthetic dataset shaped like the real one — swap "GPM_3IMERGM" for ANY short_name
ds = earthschool.dataset("GPM_3IMERGM")
var = list(ds.data_vars)[0]; da = ds[var]

# area-weight -> annual -> recover the hidden planted trend (the real workflow)
w = np.cos(np.deg2rad(ds.lat))
annual = da.groupby("time.year").sum() if var == "precipitation" else da.groupby("time.year").mean()
series = annual.weighted(w).mean(dim=["lat", "lon"])
years, vals = series["year"].values, series.values
r = mk.original_test(vals)
print(f"recovered trend: {r.slope:+.3f} {da.attrs.get('units')}/yr   (Mann-Kendall p={r.p:.3f})")
print(f"planted truth:   {earthschool.planted('GPM_3IMERGM')}   <- your analysis should be close")

plt.figure(figsize=(6, 3))
plt.plot(years, vals, "o-", label=var)
plt.plot(years, np.polyval(np.polyfit(years, vals, 1), years), "--", label="fit")
plt.title("synthetic " + var + " — annual"); plt.xlabel("year"); plt.ylabel(var); plt.legend()

Swap "GPM_3IMERGM" for any of the ~3,400 catalog short_names — MODIS NDVI (MOD13A2), LST (MOD11A2), GRACE, SMAP, Black Marble… each comes back shaped like the real product with its own planted signal. To run the real access idioms (earthaccess, open_zarr, s3://) against this layer, see Run real-data code on virtual data. Everything is synthetic and labelled not real measurements — learn the workflow here, then point the same code at the real archive.