Net2Cog fails on converting PREFIRE SAT1 3 SFC data due to the absent of Coordinate value

[vutrannasa]

Problem:
Net2Cog fails on converting PREFIRE SAT1 3 SFC data due to the absent of Coordinate value.

Error Output:

rioxarray.exceptions.MissingSpatialDimensionError: y dimension not found. 'rio.set_spatial_dims()' or using 'rename()' to change the dimension name to 'y' can address this. Data variable: emis_mean

net2cog.utilities.Net2CogError: Variable Sfc-Sorted/emis_mean cannot be converted to tif: 'DataTree' object has no attribute 'lat'

Analysis:
I believe the MissingSpatialDimensionError exception is cause by variables coordinates with empty values
Ex. Sfc-Sorted/emis_mean

variable.dims: ('xtrack', 'sfc_type', 'lat', 'lon', 'spectral')
variable.coords: Coordinates:
    *empty*
variable: <xarray.DataArray 'emis_mean' (xtrack: 8, sfc_type: 9, lat: 168, lon: 360,
                               spectral: 63)> Size: 1GB
[274337280 values with dtype=float32]
Dimensions without coordinates: xtrack, sfc_type, lat, lon, spectral
Attributes:
    long_name:    mean surface emissivity
    description:  mean surface emissivity per latitude/longitude grid cell, c... 

Reproduce Problem:

• Download PREFIRE SAT1 3 SFC granule
• Verify current test pass

$ git clone https://github.com/podaac/net2cog.git
$ cd net2cog
$ ./run_tests

• Add the following code to end of tests/test_netcdf_convert.py

def test_PREFIRE_SAT1_3_SFC_SORTED_ALLSKY(logger):
    """Verify a SPL2SMP nested variable in a hierarchical granule can be converted."""
    temp_dir2 = pathlib.Path('<path to downloaded PREFIRE SAT1 3 SFC granule')
    test_file = pathlib.Path(temp_dir2, 'PREFIRE_SAT1_3-SFC-SORTED-ALLSKY_R01_P00_20250601000000_20250630235959.nc')

    # Process test file:
    results = netcdf_converter(
        test_file,
        pathlib.Path(temp_dir2),
        ['Sfc-Sorted/emis_mean'],
        logger
    )

    # Check results are as expected:
    assert len(results) == 1, 'Incorrect number of output file names.'

    assert pathlib.Path(results[0]).is_file(), 'No file created.'
            
     # Verify correct CRS from .tif
    assert cog_validate(pathlib.Path(results[0])) == (True, [], [])

• Rerun tests

$ source .venv/bin/activate
(.venv) $ poetry run pytest -s poetry run pytest -s tests/test_netcdf_convert.py::test_PREFIRE_SAT1_3_SFC_SORTED_ALLSKY

• Verify Error

E                   net2cog.utilities.Net2CogError: Variable Sfc-Sorted/emis_mean cannot be converted to tif: 'DataTree' object has no attribute 'lat'

net2cog/netcdf_convert.py:131: Net2CogError
============================================================================ warnings summary =============================================================================
tests/test_netcdf_convert.py::test_PREFIRE_SAT1_3_SFC_SORTED_ALLSKY
  <frozen importlib._bootstrap>:488: RuntimeWarning: numpy.ndarray size changed, may indicate binary incompatibility. Expected 16 from C header, got 96 from PyObject

-- Docs: https://docs.pytest.org/en/stable/how-to/capture-warnings.html
========================================================================= short test summary info =========================================================================
FAILED tests/test_netcdf_convert.py::test_PREFIRE_SAT1_3_SFC_SORTED_ALLSKY - net2cog.utilities.Net2CogError: Variable Sfc-Sorted/emis_mean cannot be converted to tif: 'DataTree' object has no attribute 'lat'