Load DWD data

Support provided for ICON via the DWD Open Data Server. This includes interfaces for accessing global forecasts and high-resolution regional forecasts for Europe (ICON-EU) and Germany (ICON-D2).

Note that the current operational version of ICON uses a single-moment microphysics scheme that underestimates relative humidity in ice-supersaturated regions. This notebook will be updated when the double-moment microphysics scheme currently under development becomes operational.

No credentials are required to use the DWD Open Data Server.

Tools for accessing DWD data are not throughly tested in CI because they are vulnerable to upstream changes to the DWD Open Data Server. If you think you have found a problem please open an issue!

Forecast availability

The DWD Open Data Server provides access to recent forecasts (typically those initialized within the last 24 hours) but does not provide a long-term forecast archive.

from pycontrails.datalib.dwd import ods

Global forecasts are initialized at 0z, 6z, 12z, and 18z. They provide hourly data out to 78 hours and three-hourly data thereafter (out to 180 hours for the 0z and 12z runs and 120 hours for the 6z and 18z runs).

ods.list_forecasts("global")

[datetime.datetime(2025, 11, 14, 0, 0),
 datetime.datetime(2025, 11, 14, 6, 0),
 datetime.datetime(2025, 11, 14, 12, 0),
 datetime.datetime(2025, 11, 14, 18, 0)]

Regional forecasts are initialized every three hours. ICON-EU forecasts initialized and 0z, 6z, 12z, and 18z provide hourly data out to 78 hours and three-hourly data out to 120 hours. Other ICON-EU forecasts provide hourly data out to 30 hours. ICON-D2 forecasts provide hourly data out to 48 hours.

ods.list_forecasts("europe")

[datetime.datetime(2025, 11, 13, 21, 0),
 datetime.datetime(2025, 11, 14, 0, 0),
 datetime.datetime(2025, 11, 14, 3, 0),
 datetime.datetime(2025, 11, 14, 6, 0),
 datetime.datetime(2025, 11, 14, 9, 0),
 datetime.datetime(2025, 11, 14, 12, 0),
 datetime.datetime(2025, 11, 14, 15, 0),
 datetime.datetime(2025, 11, 14, 18, 0)]

ods.list_forecasts("germany")

[datetime.datetime(2025, 11, 14, 0, 0),
 datetime.datetime(2025, 11, 14, 3, 0),
 datetime.datetime(2025, 11, 14, 6, 0),
 datetime.datetime(2025, 11, 14, 9, 0),
 datetime.datetime(2025, 11, 14, 12, 0),
 datetime.datetime(2025, 11, 14, 15, 0),
 datetime.datetime(2025, 11, 14, 18, 0),
 datetime.datetime(2025, 11, 14, 21, 0)]

Global ICON

Global ICON output is retrieved on the model’s native icosahedral grid and remapped after retrieval to a regular latitude-longitude grid using a nearest-neighbor algorithm. (Support for other remapping algorithms may be added in future releases). The resolution of the latitude-longitude grid is set to 0.25 degrees by default, which provides resolution comparable to the model’s native icosahedral grid at midlatitudes. This default can be overridden using the grid parameter.

Three-dimensional fields are retrieved on model levels and interpolated after retrieval to an arbitrary set of pressure levels. By default, fields are interpolated to standard flight levels between barometric altitudes of 20,000 and 50,000 ft.

A specific forecast cycle can be selected using the forecast_time parameter:

from pycontrails.datalib.dwd import ICON

icon = ICON(
    time=("2025-11-14 12:00", "2025-11-14 15:00"),
    forecast_time="2025-11-14 00:00",
    variables=("t", "q"),
)
icon

ICON
        Timesteps: ['2025-11-14 12', '2025-11-14 13', '2025-11-14 14', '2025-11-14 15']
        Variables: ['t', 'q']
        Pressure levels: [116, 122, 128, 134, 141, 147, 155, 162, 170, 179, 188, 197, 206, 217, 227, 238, 250, 262, 274, 287, 301, 315, 329, 344, 360, 376, 393, 410, 428, 446, 466]
        Grid: 0.25
        Domain: global
        Forecast time: 2025-11-14 00Z
        Steps: [12, 13, 14, 15]

Otherwise, the forecast cycle that provides the shortest lead time is selected automatically:

icon = ICON(
    time=("2025-11-14 12:00", "2025-11-14 15:00"),
    variables=("t", "q"),
)
icon

ICON
        Timesteps: ['2025-11-14 12', '2025-11-14 13', '2025-11-14 14', '2025-11-14 15']
        Variables: ['t', 'q']
        Pressure levels: [116, 122, 128, 134, 141, 147, 155, 162, 170, 179, 188, 197, 206, 217, 227, 238, 250, 262, 274, 287, 301, 315, 329, 344, 360, 376, 393, 410, 428, 446, 466]
        Grid: 0.25
        Domain: global
        Forecast time: 2025-11-14 12Z
        Steps: [0, 1, 2, 3]

mds = icon.open_metdataset()
mds

MetDataset with data:

<xarray.Dataset> Size: 1GB
Dimensions:            (longitude: 1441, latitude: 721, level: 31, time: 4)
Coordinates:
  * longitude          (longitude) float64 12kB -180.0 -179.8 ... 179.8 180.0
  * latitude           (latitude) float64 6kB -90.0 -89.75 -89.5 ... 89.75 90.0
  * level              (level) float64 248B 116.0 122.0 128.0 ... 446.0 466.0
  * time               (time) datetime64[ns] 32B 2025-11-14T12:00:00 ... 2025...
    air_pressure       (level) float32 124B 1.16e+04 1.22e+04 ... 4.66e+04
    altitude           (level) float32 124B 1.524e+04 1.492e+04 ... 6.09e+03
Data variables:
    air_temperature    (longitude, latitude, level, time) float32 515MB dask.array<chunksize=(1441, 721, 31, 1), meta=np.ndarray>
    specific_humidity  (longitude, latitude, level, time) float32 515MB dask.array<chunksize=(1441, 721, 31, 1), meta=np.ndarray>
Attributes:
    GRIB_edition:            2
    GRIB_centre:             edzw
    GRIB_centreDescription:  Offenbach
    GRIB_subCentre:          255
    Conventions:             CF-1.7
    institution:             Offenbach
    history:                 2025-11-14T22:34 GRIB to CDM+CF via cfgrib-0.9.1...
    pycontrails_version:     0.59.1.dev2
    provider:                DWD
    dataset:                 ICON
    product:                 forecast

xarray.Dataset

• Dimensions:
  • longitude: 1441
  • latitude: 721
  • level: 31
  • time: 4
• Coordinates: (6)
  • longitude
    (longitude)
    float64
    -180.0 -179.8 ... 179.8 180.0

    array([-180.  , -179.75, -179.5 , ...,  179.5 ,  179.75,  180.  ],
          shape=(1441,))

  • latitude
    (latitude)
    float64
    -90.0 -89.75 -89.5 ... 89.75 90.0

    array([-90.  , -89.75, -89.5 , ...,  89.5 ,  89.75,  90.  ], shape=(721,))

  • level
    (level)
    float64
    116.0 122.0 128.0 ... 446.0 466.0

    units :

    hPa

    long_name :

    Pressure

    positive :

    down

    array([116., 122., 128., 134., 141., 147., 155., 162., 170., 179., 188., 197.,
           206., 217., 227., 238., 250., 262., 274., 287., 301., 315., 329., 344.,
           360., 376., 393., 410., 428., 446., 466.])

  • time
    (time)
    datetime64[ns]
    2025-11-14T12:00:00 ... 2025-11-...

    standard_name :

    time

    long_name :

    time

    array(['2025-11-14T12:00:00.000000000', '2025-11-14T13:00:00.000000000',
           '2025-11-14T14:00:00.000000000', '2025-11-14T15:00:00.000000000'],
          dtype='datetime64[ns]')

  • air_pressure
    (level)
    float32
    1.16e+04 1.22e+04 ... 4.66e+04

    standard_name :

    air_pressure

    long_name :

    Air pressure

    units :

    Pa

    array([11600., 12200., 12800., 13400., 14100., 14700., 15500., 16200.,
           17000., 17900., 18800., 19700., 20600., 21700., 22700., 23800.,
           25000., 26200., 27400., 28700., 30100., 31500., 32900., 34400.,
           36000., 37600., 39300., 41000., 42800., 44600., 46600.],
          dtype=float32)

  • altitude
    (level)
    float32
    1.524e+04 1.492e+04 ... 6.09e+03

    standard_name :

    altitude

    long_name :

    Altitude

    units :

    m

    array([15238.3545, 14918.544 , 14614.092 , 14323.588 , 14000.676 ,
           13736.406 , 13400.351 , 13120.237 , 12814.561 , 12487.417 ,
           12176.324 , 11879.782 , 11596.49  , 11266.595 , 10980.886 ,
           10679.271 , 10362.849 , 10058.495 ,  9765.228 ,  9459.032 ,
            9141.597 ,  8835.9   ,  8541.014 ,  8236.133 ,  7922.57  ,
            7620.096 ,  7309.93  ,  7010.445 ,  6704.1094,  6408.0347,
            6090.202 ], dtype=float32)

• Data variables: (2)
  • air_temperature
    (longitude, latitude, level, time)
    float32
    dask.array<chunksize=(1441, 721, 31, 1), meta=np.ndarray>

    Array Chunk Bytes 491.45 MiB 122.86 MiB Shape (1441, 721, 31, 4) (1441, 721, 31, 1) Dask graph 4 chunks in 14 graph layers Data type float32 numpy.ndarray

  • specific_humidity
    (longitude, latitude, level, time)
    float32
    dask.array<chunksize=(1441, 721, 31, 1), meta=np.ndarray>

    Array Chunk Bytes 491.45 MiB 122.86 MiB Shape (1441, 721, 31, 4) (1441, 721, 31, 1) Dask graph 4 chunks in 14 graph layers Data type float32 numpy.ndarray

• Attributes: (11)

  GRIB_edition :

  2

  GRIB_centre :

  edzw

  GRIB_centreDescription :

  Offenbach

  GRIB_subCentre :

  255

  Conventions :

  CF-1.7

  institution :

  Offenbach

  history :

  2025-11-14T22:34 GRIB to CDM+CF via cfgrib-0.9.15.1/ecCodes-2.44.0 with {"source": "../../../../../../../../tmp/7137646a62bd042fbdf6ff3ea08fd66cd7797b2973f28e3c", "filter_by_keys": {}, "encode_cf": ["parameter", "time", "geography", "vertical"]}

  pycontrails_version :

  0.59.1.dev2

  provider :

  DWD

  dataset :

  ICON

  product :

  forecast

import cartopy.crs as ccrs

mds["specific_humidity"].data[..., 14, -1].T.plot(
    figsize=(10, 4),
    cmap="Blues",
    subplot_kws=dict(projection=ccrs.PlateCarree(), transform=ccrs.PlateCarree()),
).axes.coastlines();

Top-of-atmosphere radiative fluxes can be retrieved by setting pressure_levels=-1. Fluxes are provided as averages over the previous hour, so the time coordinate is shifted backward by 30 minutes relative to three-dimensional fields.

icon = ICON(
    time=("2025-11-14 12:00", "2025-11-14 15:00"),
    variables=("rlut", "rst"),
    pressure_levels=-1,
)
icon

ICON
        Timesteps: ['2025-11-14 12', '2025-11-14 13', '2025-11-14 14', '2025-11-14 15']
        Variables: ['rlut', 'rst']
        Pressure levels: [-1]
        Grid: 0.25
        Domain: global
        Forecast time: 2025-11-14 12Z
        Steps: [0, 1, 2, 3]

mds = icon.open_metdataset()
mds

MetDataset with data:

<xarray.Dataset> Size: 25MB
Dimensions:                          (level: 1, time: 3, latitude: 721,
                                      longitude: 1441)
Coordinates:
  * level                            (level) float64 8B -1.0
  * time                             (time) datetime64[ns] 24B 2025-11-14T12:...
  * latitude                         (latitude) float64 6kB -90.0 ... 90.0
  * longitude                        (longitude) float64 12kB -180.0 ... 180.0
Data variables:
    toa_net_downward_shortwave_flux  (longitude, latitude, level, time) float32 12MB dask.array<chunksize=(1441, 721, 1, 1), meta=np.ndarray>
    toa_outgoing_longwave_flux       (longitude, latitude, level, time) float32 12MB dask.array<chunksize=(1441, 721, 1, 1), meta=np.ndarray>
Attributes:
    GRIB_edition:            2
    GRIB_centre:             edzw
    GRIB_centreDescription:  Offenbach
    GRIB_subCentre:          255
    Conventions:             CF-1.7
    institution:             Offenbach
    history:                 2025-11-14T22:39 GRIB to CDM+CF via cfgrib-0.9.1...
    pycontrails_version:     0.59.1.dev2
    provider:                DWD
    dataset:                 ICON
    product:                 forecast

xarray.Dataset

• Dimensions:
  • level: 1
  • time: 3
  • latitude: 721
  • longitude: 1441
• Coordinates: (4)
  • level
    (level)
    float64
    -1.0

    units :

    long_name :

    Single Level

    array([-1.])

  • time
    (time)
    datetime64[ns]
    2025-11-14T12:30:00 ... 2025-11-...

    standard_name :

    time

    long_name :

    time

    shift_radiation_time :

    -30 minutes

    array(['2025-11-14T12:30:00.000000000', '2025-11-14T13:30:00.000000000',
           '2025-11-14T14:30:00.000000000'], dtype='datetime64[ns]')

  • latitude
    (latitude)
    float64
    -90.0 -89.75 -89.5 ... 89.75 90.0

    array([-90.  , -89.75, -89.5 , ...,  89.5 ,  89.75,  90.  ], shape=(721,))

  • longitude
    (longitude)
    float64
    -180.0 -179.8 ... 179.8 180.0

    array([-180.  , -179.75, -179.5 , ...,  179.5 ,  179.75,  180.  ],
          shape=(1441,))

• Data variables: (2)
  • toa_net_downward_shortwave_flux
    (longitude, latitude, level, time)
    float32
    dask.array<chunksize=(1441, 721, 1, 1), meta=np.ndarray>

    Array Chunk Bytes 11.89 MiB 3.96 MiB Shape (1441, 721, 1, 3) (1441, 721, 1, 1) Dask graph 3 chunks in 15 graph layers Data type float32 numpy.ndarray

  • toa_outgoing_longwave_flux
    (longitude, latitude, level, time)
    float32
    dask.array<chunksize=(1441, 721, 1, 1), meta=np.ndarray>

    Array Chunk Bytes 11.89 MiB 3.96 MiB Shape (1441, 721, 1, 3) (1441, 721, 1, 1) Dask graph 3 chunks in 16 graph layers Data type float32 numpy.ndarray

• Attributes: (11)

  GRIB_edition :

  2

  GRIB_centre :

  edzw

  GRIB_centreDescription :

  Offenbach

  GRIB_subCentre :

  255

  Conventions :

  CF-1.7

  institution :

  Offenbach

  history :

  2025-11-14T22:39 GRIB to CDM+CF via cfgrib-0.9.15.1/ecCodes-2.44.0 with {"source": "../../../../../../../../tmp/47036215eded9daa4f877f579d260dd1f1d9138bb603e411", "filter_by_keys": {}, "encode_cf": ["parameter", "time", "geography", "vertical"]}

  pycontrails_version :

  0.59.1.dev2

  provider :

  DWD

  dataset :

  ICON

  product :

  forecast

mds["toa_outgoing_longwave_flux"].data[..., 0, -1].T.plot(
    figsize=(10, 4),
    cmap="magma",
    subplot_kws=dict(projection=ccrs.PlateCarree(), transform=ccrs.PlateCarree()),
).axes.coastlines();

Regional ICON-EU

ICON-EU output is retrieved on a regular latitude-longitude grid, so no remapping is required and a warning is issued if a value is provided for grid. The interface is otherwise identical to the interface for global ICON forecasts.

icon = ICON(
    domain="europe",
    time=("2025-11-14 12:00", "2025-11-14 15:00"),
    variables=("t", "q"),
)
icon

ICON
        Timesteps: ['2025-11-14 12', '2025-11-14 13', '2025-11-14 14', '2025-11-14 15']
        Variables: ['t', 'q']
        Pressure levels: [116, 122, 128, 134, 141, 147, 155, 162, 170, 179, 188, 197, 206, 217, 227, 238, 250, 262, 274, 287, 301, 315, 329, 344, 360, 376, 393, 410, 428, 446, 466]
        Grid: None
        Domain: europe
        Forecast time: 2025-11-14 12Z
        Steps: [0, 1, 2, 3]

mds = icon.open_metdataset()
mds

MetDataset with data:

<xarray.Dataset> Size: 897MB
Dimensions:            (longitude: 1377, latitude: 657, level: 31, time: 4)
Coordinates:
  * longitude          (longitude) float64 11kB -23.5 -23.44 ... 62.44 62.5
  * latitude           (latitude) float64 5kB 29.5 29.56 29.62 ... 70.44 70.5
  * level              (level) float64 248B 116.0 122.0 128.0 ... 446.0 466.0
  * time               (time) datetime64[ns] 32B 2025-11-14T12:00:00 ... 2025...
    air_pressure       (level) float32 124B 1.16e+04 1.22e+04 ... 4.66e+04
    altitude           (level) float32 124B 1.524e+04 1.492e+04 ... 6.09e+03
Data variables:
    air_temperature    (longitude, latitude, level, time) float32 449MB dask.array<chunksize=(1377, 657, 31, 1), meta=np.ndarray>
    specific_humidity  (longitude, latitude, level, time) float32 449MB dask.array<chunksize=(1377, 657, 31, 1), meta=np.ndarray>
Attributes:
    GRIB_edition:            2
    GRIB_centre:             edzw
    GRIB_centreDescription:  Offenbach
    GRIB_subCentre:          255
    Conventions:             CF-1.7
    institution:             Offenbach
    history:                 2025-11-14T22:41 GRIB to CDM+CF via cfgrib-0.9.1...
    pycontrails_version:     0.59.1.dev2
    provider:                DWD
    dataset:                 ICON-EU
    product:                 forecast

xarray.Dataset

• Dimensions:
  • longitude: 1377
  • latitude: 657
  • level: 31
  • time: 4
• Coordinates: (6)
  • longitude
    (longitude)
    float64
    -23.5 -23.44 -23.38 ... 62.44 62.5

    units :

    degrees_east

    standard_name :

    longitude

    long_name :

    longitude

    array([-23.5   , -23.4375, -23.375 , ...,  62.375 ,  62.4375,  62.5   ],
          shape=(1377,))

  • latitude
    (latitude)
    float64
    29.5 29.56 29.62 ... 70.44 70.5

    units :

    degrees_north

    standard_name :

    latitude

    long_name :

    latitude

    array([29.5   , 29.5625, 29.625 , ..., 70.375 , 70.4375, 70.5   ], shape=(657,))

  • level
    (level)
    float64
    116.0 122.0 128.0 ... 446.0 466.0

    units :

    hPa

    long_name :

    Pressure

    positive :

    down

    array([116., 122., 128., 134., 141., 147., 155., 162., 170., 179., 188., 197.,
           206., 217., 227., 238., 250., 262., 274., 287., 301., 315., 329., 344.,
           360., 376., 393., 410., 428., 446., 466.])

  • time
    (time)
    datetime64[ns]
    2025-11-14T12:00:00 ... 2025-11-...

    standard_name :

    time

    long_name :

    time

    array(['2025-11-14T12:00:00.000000000', '2025-11-14T13:00:00.000000000',
           '2025-11-14T14:00:00.000000000', '2025-11-14T15:00:00.000000000'],
          dtype='datetime64[ns]')

  • air_pressure
    (level)
    float32
    1.16e+04 1.22e+04 ... 4.66e+04

    standard_name :

    air_pressure

    long_name :

    Air pressure

    units :

    Pa

    array([11600., 12200., 12800., 13400., 14100., 14700., 15500., 16200.,
           17000., 17900., 18800., 19700., 20600., 21700., 22700., 23800.,
           25000., 26200., 27400., 28700., 30100., 31500., 32900., 34400.,
           36000., 37600., 39300., 41000., 42800., 44600., 46600.],
          dtype=float32)

  • altitude
    (level)
    float32
    1.524e+04 1.492e+04 ... 6.09e+03

    standard_name :

    altitude

    long_name :

    Altitude

    units :

    m

    array([15238.3545, 14918.544 , 14614.092 , 14323.588 , 14000.676 ,
           13736.406 , 13400.351 , 13120.237 , 12814.561 , 12487.417 ,
           12176.324 , 11879.782 , 11596.49  , 11266.595 , 10980.886 ,
           10679.271 , 10362.849 , 10058.495 ,  9765.228 ,  9459.032 ,
            9141.597 ,  8835.9   ,  8541.014 ,  8236.133 ,  7922.57  ,
            7620.096 ,  7309.93  ,  7010.445 ,  6704.1094,  6408.0347,
            6090.202 ], dtype=float32)

• Data variables: (2)
  • air_temperature
    (longitude, latitude, level, time)
    float32
    dask.array<chunksize=(1377, 657, 31, 1), meta=np.ndarray>

    Array Chunk Bytes 427.94 MiB 106.98 MiB Shape (1377, 657, 31, 4) (1377, 657, 31, 1) Dask graph 4 chunks in 15 graph layers Data type float32 numpy.ndarray

  • specific_humidity
    (longitude, latitude, level, time)
    float32
    dask.array<chunksize=(1377, 657, 31, 1), meta=np.ndarray>

    Array Chunk Bytes 427.94 MiB 106.98 MiB Shape (1377, 657, 31, 4) (1377, 657, 31, 1) Dask graph 4 chunks in 15 graph layers Data type float32 numpy.ndarray

• Attributes: (11)

  GRIB_edition :

  2

  GRIB_centre :

  edzw

  GRIB_centreDescription :

  Offenbach

  GRIB_subCentre :

  255

  Conventions :

  CF-1.7

  institution :

  Offenbach

  history :

  2025-11-14T22:41 GRIB to CDM+CF via cfgrib-0.9.15.1/ecCodes-2.44.0 with {"source": "../../../../../../../../tmp/f59eeff7285e8d288b357ef4c84576369d16c0d29afe8aa6", "filter_by_keys": {}, "encode_cf": ["parameter", "time", "geography", "vertical"]}

  pycontrails_version :

  0.59.1.dev2

  provider :

  DWD

  dataset :

  ICON-EU

  product :

  forecast

mds["specific_humidity"].data[..., 14, -1].T.plot(
    figsize=(10, 4),
    cmap="Blues",
    subplot_kws=dict(projection=ccrs.PlateCarree(), transform=ccrs.PlateCarree()),
).axes.coastlines();

icon = ICON(
    domain="europe",
    time=("2025-11-14 12:00", "2025-11-14 15:00"),
    variables=("rlut", "rst"),
    pressure_levels=-1,
)
icon

ICON
        Timesteps: ['2025-11-14 12', '2025-11-14 13', '2025-11-14 14', '2025-11-14 15']
        Variables: ['rlut', 'rst']
        Pressure levels: [-1]
        Grid: None
        Domain: europe
        Forecast time: 2025-11-14 12Z
        Steps: [0, 1, 2, 3]

mds = icon.open_metdataset()
mds

MetDataset with data:

<xarray.Dataset> Size: 22MB
Dimensions:                          (level: 1, time: 3, latitude: 657,
                                      longitude: 1377)
Coordinates:
  * level                            (level) float64 8B -1.0
  * time                             (time) datetime64[ns] 24B 2025-11-14T12:...
  * latitude                         (latitude) float64 5kB 29.5 29.56 ... 70.5
  * longitude                        (longitude) float64 11kB -23.5 ... 62.5
Data variables:
    toa_outgoing_longwave_flux       (longitude, latitude, level, time) float32 11MB dask.array<chunksize=(1377, 657, 1, 1), meta=np.ndarray>
    toa_net_downward_shortwave_flux  (longitude, latitude, level, time) float32 11MB dask.array<chunksize=(1377, 657, 1, 1), meta=np.ndarray>
Attributes:
    GRIB_edition:            2
    GRIB_centre:             edzw
    GRIB_centreDescription:  Offenbach
    GRIB_subCentre:          255
    Conventions:             CF-1.7
    institution:             Offenbach
    history:                 2025-11-14T22:43 GRIB to CDM+CF via cfgrib-0.9.1...
    pycontrails_version:     0.59.1.dev2
    provider:                DWD
    dataset:                 ICON-EU
    product:                 forecast

xarray.Dataset

• Dimensions:
  • level: 1
  • time: 3
  • latitude: 657
  • longitude: 1377
• Coordinates: (4)
  • level
    (level)
    float64
    -1.0

    units :

    long_name :

    Single Level

    array([-1.])

  • time
    (time)
    datetime64[ns]
    2025-11-14T12:30:00 ... 2025-11-...

    standard_name :

    time

    long_name :

    time

    shift_radiation_time :

    -30 minutes

    array(['2025-11-14T12:30:00.000000000', '2025-11-14T13:30:00.000000000',
           '2025-11-14T14:30:00.000000000'], dtype='datetime64[ns]')

  • latitude
    (latitude)
    float64
    29.5 29.56 29.62 ... 70.44 70.5

    units :

    degrees_north

    standard_name :

    latitude

    long_name :

    latitude

    array([29.5   , 29.5625, 29.625 , ..., 70.375 , 70.4375, 70.5   ], shape=(657,))

  • longitude
    (longitude)
    float64
    -23.5 -23.44 -23.38 ... 62.44 62.5

    units :

    degrees_east

    standard_name :

    longitude

    long_name :

    longitude

    array([-23.5   , -23.4375, -23.375 , ...,  62.375 ,  62.4375,  62.5   ],
          shape=(1377,))

• Data variables: (2)
  • toa_outgoing_longwave_flux
    (longitude, latitude, level, time)
    float32
    dask.array<chunksize=(1377, 657, 1, 1), meta=np.ndarray>

    GRIB_paramId :

    235040

    GRIB_dataType :

    fc

    GRIB_numberOfPoints :

    904689

    GRIB_typeOfLevel :

    nominalTop

    GRIB_stepUnits :

    1

    GRIB_stepType :

    avg

    GRIB_gridType :

    regular_ll

    GRIB_uvRelativeToGrid :

    0

    GRIB_NV :

    0

    GRIB_Nx :

    1377

    GRIB_Ny :

    657

    GRIB_cfName :

    unknown

    GRIB_cfVarName :

    avg_tnlwrf

    GRIB_gridDefinitionDescription :

    Latitude/longitude

    GRIB_iDirectionIncrementInDegrees :

    0.0625

    GRIB_iScansNegatively :

    0

    GRIB_jDirectionIncrementInDegrees :

    0.0625

    GRIB_jPointsAreConsecutive :

    0

    GRIB_jScansPositively :

    1

    GRIB_latitudeOfFirstGridPointInDegrees :

    29.5

    GRIB_latitudeOfLastGridPointInDegrees :

    70.5

    GRIB_longitudeOfFirstGridPointInDegrees :

    336.5

    GRIB_longitudeOfLastGridPointInDegrees :

    62.5

    GRIB_missingValue :

    3.4028234663852886e+38

    GRIB_name :

    Time-mean top net long-wave radiation flux

    GRIB_shortName :

    avg_tnlwrf

    GRIB_units :

    W m**-2

    long_name :

    Time-mean top net long-wave radiation flux

    units :

    W m**-2

    standard_name :

    unknown

    Array Chunk Bytes 10.35 MiB 3.45 MiB Shape (1377, 657, 1, 3) (1377, 657, 1, 1) Dask graph 3 chunks in 17 graph layers Data type float32 numpy.ndarray

  • toa_net_downward_shortwave_flux
    (longitude, latitude, level, time)
    float32
    dask.array<chunksize=(1377, 657, 1, 1), meta=np.ndarray>

    GRIB_paramId :

    235039

    GRIB_dataType :

    fc

    GRIB_numberOfPoints :

    904689

    GRIB_typeOfLevel :

    nominalTop

    GRIB_stepUnits :

    1

    GRIB_stepType :

    avg

    GRIB_gridType :

    regular_ll

    GRIB_uvRelativeToGrid :

    0

    GRIB_NV :

    0

    GRIB_Nx :

    1377

    GRIB_Ny :

    657

    GRIB_cfName :

    unknown

    GRIB_cfVarName :

    avg_tnswrf

    GRIB_gridDefinitionDescription :

    Latitude/longitude

    GRIB_iDirectionIncrementInDegrees :

    0.0625

    GRIB_iScansNegatively :

    0

    GRIB_jDirectionIncrementInDegrees :

    0.0625

    GRIB_jPointsAreConsecutive :

    0

    GRIB_jScansPositively :

    1

    GRIB_latitudeOfFirstGridPointInDegrees :

    29.5

    GRIB_latitudeOfLastGridPointInDegrees :

    70.5

    GRIB_longitudeOfFirstGridPointInDegrees :

    336.5

    GRIB_longitudeOfLastGridPointInDegrees :

    62.5

    GRIB_missingValue :

    3.4028234663852886e+38

    GRIB_name :

    Time-mean top net short-wave radiation flux

    GRIB_shortName :

    avg_tnswrf

    GRIB_units :

    W m**-2

    long_name :

    Time-mean top net short-wave radiation flux

    units :

    W m**-2

    standard_name :

    unknown

    Array Chunk Bytes 10.35 MiB 3.45 MiB Shape (1377, 657, 1, 3) (1377, 657, 1, 1) Dask graph 3 chunks in 16 graph layers Data type float32 numpy.ndarray

• Attributes: (11)

  GRIB_edition :

  2

  GRIB_centre :

  edzw

  GRIB_centreDescription :

  Offenbach

  GRIB_subCentre :

  255

  Conventions :

  CF-1.7

  institution :

  Offenbach

  history :

  2025-11-14T22:43 GRIB to CDM+CF via cfgrib-0.9.15.1/ecCodes-2.44.0 with {"source": "../../../../../../../../tmp/17d9464319fd375a2264d287d468ff93b424df3d68b3b25d", "filter_by_keys": {}, "encode_cf": ["parameter", "time", "geography", "vertical"]}

  pycontrails_version :

  0.59.1.dev2

  provider :

  DWD

  dataset :

  ICON-EU

  product :

  forecast

mds["toa_outgoing_longwave_flux"].data[..., 0, -1].T.plot(
    figsize=(10, 4),
    cmap="magma",
    subplot_kws=dict(projection=ccrs.PlateCarree(), transform=ccrs.PlateCarree()),
).axes.coastlines();

Regional ICON-D2

The interface for ICON-D2 is identical to ICON-EU.

icon = ICON(
    domain="germany",
    time=("2025-11-14 12:00", "2025-11-14 15:00"),
    variables=("t", "q"),
)
icon

ICON
        Timesteps: ['2025-11-14 12', '2025-11-14 13', '2025-11-14 14', '2025-11-14 15']
        Variables: ['t', 'q']
        Pressure levels: [116, 122, 128, 134, 141, 147, 155, 162, 170, 179, 188, 197, 206, 217, 227, 238, 250, 262, 274, 287, 301, 315, 329, 344, 360, 376, 393, 410, 428, 446, 466]
        Grid: None
        Domain: germany
        Forecast time: 2025-11-14 12Z
        Steps: [0, 1, 2, 3]

mds = icon.open_metdataset()
mds

MetDataset with data:

<xarray.Dataset> Size: 899MB
Dimensions:            (longitude: 1215, latitude: 746, level: 31, time: 4)
Coordinates:
  * longitude          (longitude) float64 10kB -3.94 -3.92 -3.9 ... 20.32 20.34
  * latitude           (latitude) float64 6kB 43.18 43.2 43.22 ... 58.06 58.08
  * level              (level) float64 248B 116.0 122.0 128.0 ... 446.0 466.0
  * time               (time) datetime64[ns] 32B 2025-11-14T12:00:00 ... 2025...
    air_pressure       (level) float32 124B 1.16e+04 1.22e+04 ... 4.66e+04
    altitude           (level) float32 124B 1.524e+04 1.492e+04 ... 6.09e+03
Data variables:
    air_temperature    (longitude, latitude, level, time) float32 450MB dask.array<chunksize=(1215, 746, 31, 1), meta=np.ndarray>
    specific_humidity  (longitude, latitude, level, time) float32 450MB dask.array<chunksize=(1215, 746, 31, 1), meta=np.ndarray>
Attributes:
    GRIB_edition:            2
    GRIB_centre:             edzw
    GRIB_centreDescription:  Offenbach
    GRIB_subCentre:          255
    Conventions:             CF-1.7
    institution:             Offenbach
    history:                 2025-11-14T22:43 GRIB to CDM+CF via cfgrib-0.9.1...
    pycontrails_version:     0.59.1.dev2
    provider:                DWD
    dataset:                 ICON-D2
    product:                 forecast

xarray.Dataset

• Dimensions:
  • longitude: 1215
  • latitude: 746
  • level: 31
  • time: 4
• Coordinates: (6)
  • longitude
    (longitude)
    float64
    -3.94 -3.92 -3.9 ... 20.32 20.34

    units :

    degrees_east

    standard_name :

    longitude

    long_name :

    longitude

    array([-3.94, -3.92, -3.9 , ..., 20.3 , 20.32, 20.34], shape=(1215,))

  • latitude
    (latitude)
    float64
    43.18 43.2 43.22 ... 58.06 58.08

    units :

    degrees_north

    standard_name :

    latitude

    long_name :

    latitude

    array([43.18, 43.2 , 43.22, ..., 58.04, 58.06, 58.08], shape=(746,))

  • level
    (level)
    float64
    116.0 122.0 128.0 ... 446.0 466.0

    units :

    hPa

    long_name :

    Pressure

    positive :

    down

    array([116., 122., 128., 134., 141., 147., 155., 162., 170., 179., 188., 197.,
           206., 217., 227., 238., 250., 262., 274., 287., 301., 315., 329., 344.,
           360., 376., 393., 410., 428., 446., 466.])

  • time
    (time)
    datetime64[ns]
    2025-11-14T12:00:00 ... 2025-11-...

    standard_name :

    time

    long_name :

    time

    array(['2025-11-14T12:00:00.000000000', '2025-11-14T13:00:00.000000000',
           '2025-11-14T14:00:00.000000000', '2025-11-14T15:00:00.000000000'],
          dtype='datetime64[ns]')

  • air_pressure
    (level)
    float32
    1.16e+04 1.22e+04 ... 4.66e+04

    standard_name :

    air_pressure

    long_name :

    Air pressure

    units :

    Pa

    array([11600., 12200., 12800., 13400., 14100., 14700., 15500., 16200.,
           17000., 17900., 18800., 19700., 20600., 21700., 22700., 23800.,
           25000., 26200., 27400., 28700., 30100., 31500., 32900., 34400.,
           36000., 37600., 39300., 41000., 42800., 44600., 46600.],
          dtype=float32)

  • altitude
    (level)
    float32
    1.524e+04 1.492e+04 ... 6.09e+03

    standard_name :

    altitude

    long_name :

    Altitude

    units :

    m

    array([15238.3545, 14918.544 , 14614.092 , 14323.588 , 14000.676 ,
           13736.406 , 13400.351 , 13120.237 , 12814.561 , 12487.417 ,
           12176.324 , 11879.782 , 11596.49  , 11266.595 , 10980.886 ,
           10679.271 , 10362.849 , 10058.495 ,  9765.228 ,  9459.032 ,
            9141.597 ,  8835.9   ,  8541.014 ,  8236.133 ,  7922.57  ,
            7620.096 ,  7309.93  ,  7010.445 ,  6704.1094,  6408.0347,
            6090.202 ], dtype=float32)

• Data variables: (2)
  • air_temperature
    (longitude, latitude, level, time)
    float32
    dask.array<chunksize=(1215, 746, 31, 1), meta=np.ndarray>

    Array Chunk Bytes 428.74 MiB 107.19 MiB Shape (1215, 746, 31, 4) (1215, 746, 31, 1) Dask graph 4 chunks in 15 graph layers Data type float32 numpy.ndarray

  • specific_humidity
    (longitude, latitude, level, time)
    float32
    dask.array<chunksize=(1215, 746, 31, 1), meta=np.ndarray>

    Array Chunk Bytes 428.74 MiB 107.19 MiB Shape (1215, 746, 31, 4) (1215, 746, 31, 1) Dask graph 4 chunks in 15 graph layers Data type float32 numpy.ndarray

• Attributes: (11)

  GRIB_edition :

  2

  GRIB_centre :

  edzw

  GRIB_centreDescription :

  Offenbach

  GRIB_subCentre :

  255

  Conventions :

  CF-1.7

  institution :

  Offenbach

  history :

  2025-11-14T22:43 GRIB to CDM+CF via cfgrib-0.9.15.1/ecCodes-2.44.0 with {"source": "../../../../../../../../tmp/a6c5e5e32649310181215f0e69108ef8327dcf43dbbdd2a0", "filter_by_keys": {}, "encode_cf": ["parameter", "time", "geography", "vertical"]}

  pycontrails_version :

  0.59.1.dev2

  provider :

  DWD

  dataset :

  ICON-D2

  product :

  forecast

mds["specific_humidity"].data[..., 14, 0].T.plot(
    figsize=(10, 4),
    cmap="Blues",
    subplot_kws=dict(projection=ccrs.PlateCarree(), transform=ccrs.PlateCarree()),
).axes.coastlines();

icon = ICON(
    domain="germany",
    time=("2025-11-14 12:00", "2025-11-14 15:00"),
    variables=("rlut", "rst"),
    pressure_levels=-1,
)
icon

ICON
        Timesteps: ['2025-11-14 12', '2025-11-14 13', '2025-11-14 14', '2025-11-14 15']
        Variables: ['rlut', 'rst']
        Pressure levels: [-1]
        Grid: None
        Domain: germany
        Forecast time: 2025-11-14 12Z
        Steps: [0, 1, 2, 3]

mds = icon.open_metdataset()
mds

MetDataset with data:

<xarray.Dataset> Size: 22MB
Dimensions:                          (level: 1, time: 3, latitude: 746,
                                      longitude: 1215)
Coordinates:
  * level                            (level) float64 8B -1.0
  * time                             (time) datetime64[ns] 24B 2025-11-14T12:...
  * latitude                         (latitude) float64 6kB 43.18 43.2 ... 58.08
  * longitude                        (longitude) float64 10kB -3.94 ... 20.34
Data variables:
    toa_outgoing_longwave_flux       (longitude, latitude, level, time) float32 11MB dask.array<chunksize=(1215, 746, 1, 1), meta=np.ndarray>
    toa_net_downward_shortwave_flux  (longitude, latitude, level, time) float32 11MB dask.array<chunksize=(1215, 746, 1, 1), meta=np.ndarray>
Attributes:
    GRIB_edition:            2
    GRIB_centre:             edzw
    GRIB_centreDescription:  Offenbach
    GRIB_subCentre:          255
    Conventions:             CF-1.7
    institution:             Offenbach
    history:                 2025-11-14T22:45 GRIB to CDM+CF via cfgrib-0.9.1...
    pycontrails_version:     0.59.1.dev2
    provider:                DWD
    dataset:                 ICON-D2
    product:                 forecast

xarray.Dataset

• Dimensions:
  • level: 1
  • time: 3
  • latitude: 746
  • longitude: 1215
• Coordinates: (4)
  • level
    (level)
    float64
    -1.0

    units :

    long_name :

    Single Level

    array([-1.])

  • time
    (time)
    datetime64[ns]
    2025-11-14T12:30:00 ... 2025-11-...

    standard_name :

    time

    long_name :

    time

    shift_radiation_time :

    -30 minutes

    array(['2025-11-14T12:30:00.000000000', '2025-11-14T13:30:00.000000000',
           '2025-11-14T14:30:00.000000000'], dtype='datetime64[ns]')

  • latitude
    (latitude)
    float64
    43.18 43.2 43.22 ... 58.06 58.08

    units :

    degrees_north

    standard_name :

    latitude

    long_name :

    latitude

    array([43.18, 43.2 , 43.22, ..., 58.04, 58.06, 58.08], shape=(746,))

  • longitude
    (longitude)
    float64
    -3.94 -3.92 -3.9 ... 20.32 20.34

    units :

    degrees_east

    standard_name :

    longitude

    long_name :

    longitude

    array([-3.94, -3.92, -3.9 , ..., 20.3 , 20.32, 20.34], shape=(1215,))

• Data variables: (2)
  • toa_outgoing_longwave_flux
    (longitude, latitude, level, time)
    float32
    dask.array<chunksize=(1215, 746, 1, 1), meta=np.ndarray>

    GRIB_paramId :

    235040

    GRIB_dataType :

    fc

    GRIB_numberOfPoints :

    906390

    GRIB_typeOfLevel :

    nominalTop

    GRIB_stepUnits :

    1

    GRIB_stepType :

    avg

    GRIB_gridType :

    regular_ll

    GRIB_uvRelativeToGrid :

    0

    GRIB_NV :

    0

    GRIB_Nx :

    1215

    GRIB_Ny :

    746

    GRIB_cfName :

    unknown

    GRIB_cfVarName :

    avg_tnlwrf

    GRIB_gridDefinitionDescription :

    Latitude/longitude

    GRIB_iDirectionIncrementInDegrees :

    0.02

    GRIB_iScansNegatively :

    0

    GRIB_jDirectionIncrementInDegrees :

    0.02

    GRIB_jPointsAreConsecutive :

    0

    GRIB_jScansPositively :

    1

    GRIB_latitudeOfFirstGridPointInDegrees :

    43.18

    GRIB_latitudeOfLastGridPointInDegrees :

    58.08

    GRIB_longitudeOfFirstGridPointInDegrees :

    356.06

    GRIB_longitudeOfLastGridPointInDegrees :

    20.34

    GRIB_missingValue :

    3.4028234663852886e+38

    GRIB_name :

    Time-mean top net long-wave radiation flux

    GRIB_shortName :

    avg_tnlwrf

    GRIB_units :

    W m**-2

    long_name :

    Time-mean top net long-wave radiation flux

    units :

    W m**-2

    standard_name :

    unknown

    Array Chunk Bytes 10.37 MiB 3.46 MiB Shape (1215, 746, 1, 3) (1215, 746, 1, 1) Dask graph 3 chunks in 17 graph layers Data type float32 numpy.ndarray

  • toa_net_downward_shortwave_flux
    (longitude, latitude, level, time)
    float32
    dask.array<chunksize=(1215, 746, 1, 1), meta=np.ndarray>

    GRIB_paramId :

    235039

    GRIB_dataType :

    fc

    GRIB_numberOfPoints :

    906390

    GRIB_typeOfLevel :

    nominalTop

    GRIB_stepUnits :

    1

    GRIB_stepType :

    avg

    GRIB_gridType :

    regular_ll

    GRIB_uvRelativeToGrid :

    0

    GRIB_NV :

    0

    GRIB_Nx :

    1215

    GRIB_Ny :

    746

    GRIB_cfName :

    unknown

    GRIB_cfVarName :

    avg_tnswrf

    GRIB_gridDefinitionDescription :

    Latitude/longitude

    GRIB_iDirectionIncrementInDegrees :

    0.02

    GRIB_iScansNegatively :

    0

    GRIB_jDirectionIncrementInDegrees :

    0.02

    GRIB_jPointsAreConsecutive :

    0

    GRIB_jScansPositively :

    1

    GRIB_latitudeOfFirstGridPointInDegrees :

    43.18

    GRIB_latitudeOfLastGridPointInDegrees :

    58.08

    GRIB_longitudeOfFirstGridPointInDegrees :

    356.06

    GRIB_longitudeOfLastGridPointInDegrees :

    20.34

    GRIB_missingValue :

    3.4028234663852886e+38

    GRIB_name :

    Time-mean top net short-wave radiation flux

    GRIB_shortName :

    avg_tnswrf

    GRIB_units :

    W m**-2

    long_name :

    Time-mean top net short-wave radiation flux

    units :

    W m**-2

    standard_name :

    unknown

    Array Chunk Bytes 10.37 MiB 3.46 MiB Shape (1215, 746, 1, 3) (1215, 746, 1, 1) Dask graph 3 chunks in 16 graph layers Data type float32 numpy.ndarray

• Attributes: (11)

  GRIB_edition :

  2

  GRIB_centre :

  edzw

  GRIB_centreDescription :

  Offenbach

  GRIB_subCentre :

  255

  Conventions :

  CF-1.7

  institution :

  Offenbach

  history :

  2025-11-14T22:45 GRIB to CDM+CF via cfgrib-0.9.15.1/ecCodes-2.44.0 with {"source": "../../../../../../../../tmp/fa3e25436804407c899b45029d8a48ab16bb4c90d7f5c880", "filter_by_keys": {}, "encode_cf": ["parameter", "time", "geography", "vertical"]}

  pycontrails_version :

  0.59.1.dev2

  provider :

  DWD

  dataset :

  ICON-D2

  product :

  forecast

mds["toa_outgoing_longwave_flux"].data[..., 0, -1].T.plot(
    figsize=(10, 4),
    cmap="magma",
    subplot_kws=dict(projection=ccrs.PlateCarree(), transform=ccrs.PlateCarree()),
).axes.coastlines();

Using DWD data with CoCiP

The DWD Open Data Server provides almost all the fields requires to run CoCiP. The exception is vertical velocity: CoCiP expects vertical velocity to be provided in pressure coordinates (units of Pa/s), while the Open Data Server provides vertical velocities in height coordinates (units of m/s). As a workaround, height-coordinate vertical velocity can be converted approximately to pressure-coordinate vertical velocity using the hydrostatic relation. (This conversion is most accurate at large scales; see e.g. section 3.5 of An Introduction to Dynamic Meteorology for details.)

Retrieve required fields, including height-coordinate vertical velocity:

icon = ICON(
    time=("2025-11-14 12:00", "2025-11-15 00:00"),
    variables=("t", "q", "u", "v", "wz", "cli"),
)
met = icon.open_metdataset()

icon = ICON(
    time=("2025-11-14 12:00", "2025-11-15 00:00"), variables=("rlut", "rst"), pressure_levels=-1
)
rad = icon.open_metdataset()

Convert approximately to pressure-coordinate vertical velocity:

\[\omega \approx -\rho g w = -\frac{p g w}{RT}\]

from pycontrails.physics import constants

w = met["upward_air_velocity"].data
p = met["air_pressure"].data.broadcast_like(w)
T = met["air_temperature"].data
met["lagrangian_tendency_of_air_pressure"] = -p * constants.g * w / (constants.R_d * T)

Run gridded CoCiP

import numpy as np

from pycontrails.core import MetDataset
from pycontrails.models.cocipgrid import CocipGrid
from pycontrails.models.humidity_scaling import ConstantHumidityScaling
from pycontrails.models.ps_model import PSGrid
from pycontrails.physics import units

params = {
    "dt_integration": np.timedelta64(5, "m"),
    "max_age": np.timedelta64(10, "h"),
    "humidity_scaling": ConstantHumidityScaling(rhi_adj=1.0),
    "aircraft_performance": PSGrid(),
}

source = MetDataset.from_coords(
    level=units.ft_to_pl(35_000),
    time="2025-11-14 13:00",
    longitude=met["longitude"].values,
    latitude=met["latitude"].values,
)

cocip_grid = CocipGrid(met=met, rad=rad, params=params)
result = cocip_grid.eval(source)

Plot results

import matplotlib.colors

result["ef_per_m"].data.T.plot(
    figsize=(10, 4),
    cmap="RdBu_r",
    norm=matplotlib.colors.SymLogNorm(linthresh=1e7),
    subplot_kws=dict(projection=ccrs.PlateCarree(), transform=ccrs.PlateCarree()),
).axes.coastlines();