NetCDF reading and writing examples

A self-contained example for EarthInversion’s NetCDF tutorials. All temperature values are synthetic teaching data, not observations or a physical model.

Run

Use Python 3.10 or newer. Create and activate a virtual environment, then run:

python -m pip install netCDF4 numpy
python write_netcdf.py
python read_netcdf.py example.nc

If you downloaded example.nc already, skip the writing command or choose a different output filename. The writer refuses to overwrite existing files.

Optional plotting and appending:

python -m pip install matplotlib
python plot_netcdf.py example.nc
python append_netcdf.py example.nc example-appended.nc
python read_netcdf.py example-appended.nc

Appending creates a copy, leaving the source unchanged. This teaching script is not an atomic transaction system for production acquisition.

Run python test_examples.py for round-trip checks. Install xarray to include its optional independent readback check. Exact tested versions are in requirements-tested.txt.

The sample has temperature(time, lat, lon), shape (3, 5, 6), units K, three six-hourly times starting January 1, 2026 at 00:00 UTC, latitudes 35 to 39 degrees north, and longitudes -124 to -119 degrees east. Cell [0, 0, 0] is missing. The deterministic formula is in the writing script. No external data is required or redistributed.

The scripts use CF-style coordinate metadata without claiming formal compliance with a particular version of CF.

Documentation: netCDF4-Python and CF conventions.