Skip to content

ncep_grib

NCEP GRIB Reader

NCEPGribReader

Bases: GriddedReader

Reader for NCEP GRIB files.

Source code in monetio/readers/ncep_grib.py 
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
@register_reader("ncep_grib")
class NCEPGribReader(GriddedReader):
    """
    Reader for NCEP GRIB files.
    """

    def open_dataset(self, files: str | list[str], **kwargs: Any) -> xr.Dataset:
        """
        Reads NCEP GRIB files.

        Parameters
        ----------
        files : Union[str, List[str]]
            File path, list of paths, or glob pattern.
        **kwargs : Any
            Additional arguments passed to xarray.open_mfdataset or the driver.

        Returns
        -------
        xarray.Dataset
            The processed NCEP GRIB dataset.

        Examples
        --------
        >>> from monetio.readers.ncep_grib import NCEPGribReader
        >>> reader = NCEPGribReader()
        >>> ds = reader.open_dataset("gfs.*.grib2", engine="pynio")
        """
        # Ensure we have engine='pynio' if not specified
        # Note: pynio is often used for these files but might be hard to install.
        if "engine" not in kwargs:
            kwargs["engine"] = "pynio"

        # Also supports open_mfdataset logic
        if "concat_dim" not in kwargs:
            kwargs["concat_dim"] = "time"

        if "preprocess" not in kwargs:
            kwargs["preprocess"] = ncep_grib_preprocess

        ds = self.driver.open(files, **kwargs)

        # Update history
        ds = update_history(ds, "Read NCEP GRIB data.")

        return ds

open_dataset(files, **kwargs)

Reads NCEP GRIB files.

Parameters:

Name Type Description Default
files Union[str, List[str]]

File path, list of paths, or glob pattern.

 required
**kwargs Any

Additional arguments passed to xarray.open_mfdataset or the driver.

 {}

Returns:

Type Description
Dataset

The processed NCEP GRIB dataset.

Examples:

>>> from monetio.readers.ncep_grib import NCEPGribReader
>>> reader = NCEPGribReader()
>>> ds = reader.open_dataset("gfs.*.grib2", engine="pynio")
Source code in monetio/readers/ncep_grib.py 
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
def open_dataset(self, files: str | list[str], **kwargs: Any) -> xr.Dataset:
    """
    Reads NCEP GRIB files.

    Parameters
    ----------
    files : Union[str, List[str]]
        File path, list of paths, or glob pattern.
    **kwargs : Any
        Additional arguments passed to xarray.open_mfdataset or the driver.

    Returns
    -------
    xarray.Dataset
        The processed NCEP GRIB dataset.

    Examples
    --------
    >>> from monetio.readers.ncep_grib import NCEPGribReader
    >>> reader = NCEPGribReader()
    >>> ds = reader.open_dataset("gfs.*.grib2", engine="pynio")
    """
    # Ensure we have engine='pynio' if not specified
    # Note: pynio is often used for these files but might be hard to install.
    if "engine" not in kwargs:
        kwargs["engine"] = "pynio"

    # Also supports open_mfdataset logic
    if "concat_dim" not in kwargs:
        kwargs["concat_dim"] = "time"

    if "preprocess" not in kwargs:
        kwargs["preprocess"] = ncep_grib_preprocess

    ds = self.driver.open(files, **kwargs)

    # Update history
    ds = update_history(ds, "Read NCEP GRIB data.")

    return ds

ncep_grib_preprocess(ds)

Preprocess function for a single NCEP GRIB file.

Converts 1D latitude/longitude to 2D coordinates lazily and applies scientific hygiene.

Parameters:

Name Type Description Default
ds Dataset

Input NCEP GRIB dataset.

 required

Returns:

Type Description
Dataset

Processed dataset with 'latitude' and 'longitude' coordinates on (y, x) dims.

Examples:

>>> ds = ncep_grib_preprocess(ds)
Source code in monetio/readers/ncep_grib.py 
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
def ncep_grib_preprocess(ds: xr.Dataset) -> xr.Dataset:
    """
    Preprocess function for a single NCEP GRIB file.

    Converts 1D latitude/longitude to 2D coordinates lazily and applies
    scientific hygiene.

    Parameters
    ----------
    ds : xarray.Dataset
        Input NCEP GRIB dataset.

    Returns
    -------
    xarray.Dataset
        Processed dataset with 'latitude' and 'longitude' coordinates on (y, x) dims.

    Examples
    --------
    >>> ds = ncep_grib_preprocess(ds)
    """
    # 1. Coordinate Renaming & Promotion
    # Some backends might have them as variables but not coords
    rename_dict = {}
    if "lat_0" in ds.variables:
        rename_dict["lat_0"] = "latitude"
    if "lon_0" in ds.variables:
        rename_dict["lon_0"] = "longitude"

    if rename_dict:
        ds = ds.rename(rename_dict)
        # Promote renamed variables to coordinates
        to_coord = [v for v in rename_dict.values() if v in ds.variables and v not in ds.coords]
        if to_coord:
            ds = ds.set_coords(to_coord)

    # 2. Generate 2D Latitude and Longitude lazily
    if "latitude" in ds.coords and "longitude" in ds.coords:
        # Check if they are 1D
        if ds.latitude.ndim == 1 and ds.longitude.ndim == 1:
            lat_dim = ds.latitude.dims[0]
            lon_dim = ds.longitude.dims[0]

            # Extract data and attributes to create new DataArrays for broadcast.
            # This avoids alignment issues when re-assigning to the dataset.
            lon_data = ds.longitude.data
            lat_data = ds.latitude.data

            # If the dataset is chunked but coordinates are not, wrap them in dask
            # to maintain laziness throughout the broadcast.
            if ds.chunks:
                try:
                    import dask.array as da

                    if not hasattr(lon_data, "dask"):
                        lon_data = da.from_array(lon_data, chunks=ds.chunks.get(lon_dim, -1))
                    if not hasattr(lat_data, "dask"):
                        lat_data = da.from_array(lat_data, chunks=ds.chunks.get(lat_dim, -1))
                except ImportError:
                    pass

            lon1d = xr.DataArray(lon_data, dims="x", attrs=ds.longitude.attrs)
            lat1d = xr.DataArray(lat_data, dims="y", attrs=ds.latitude.attrs)

            # Broadcast to 2D
            # xr.broadcast will handle both NumPy and Dask lazily
            lon2d, lat2d = xr.broadcast(lon1d, lat1d)

            # Ensure dimension order is (y, x) to match original meshgrid behavior
            # We use transpose to ensure the order is correct for the new dimensions
            if "y" in lon2d.dims and "x" in lon2d.dims:
                lon2d = lon2d.transpose("y", "x")
                lat2d = lat2d.transpose("y", "x")

            # Replace 1D coords in the dataset with index ranges to avoid dim/coord conflict
            ds = ds.assign_coords(
                **{
                    lat_dim: np.arange(ds.sizes[lat_dim]),
                    lon_dim: np.arange(ds.sizes[lon_dim]),
                }
            )
            # Rename dims to y, x
            ds = ds.rename({lat_dim: "y", lon_dim: "x"})

            # Assign 2D coordinates
            ds = ds.assign_coords(
                longitude=lon2d.assign_attrs(
                    {"long_name": "Longitude", "units": "degree_east", "standard_name": "longitude"}
                ),
                latitude=lat2d.assign_attrs(
                    {"long_name": "Latitude", "units": "degree_north", "standard_name": "latitude"}
                ),
            )

            ds = ds.set_coords(["latitude", "longitude"])
            ds = update_history(ds, "Generated 2D latitude/longitude coordinates lazily.")

    # 3. Scientific Hygiene
    ds = _scientific_hygiene(ds)

    # Update history
    ds = update_history(ds, "Preprocessed NCEP GRIB data.")

    return ds