RadKlim - Precipitation Climatology over Germany

This notebook describes how to access and use the zarr-version of the RadKlim dataset. The dataset includes hourly accumulated rainfall and five-minute rain-rate data over Germany, which is derived from the German Weather Service (DWD) radar network. The RadKlim dataset has been extended annually since the first version released in 2017 (covering 2001-2017). The the v0.1.1 zarr version of the dataset includes up to Extension 6, covering 2001-2023.

The dataset was created by processing the netCDF files for the hourly and five-minute precipitation data into two separate zarr datasets which are described below. The processing was done with code hosted on mlcast-community/mlcast-dataset-radklim.

import matplotlib.pyplot as plt
import cartopy.crs as ccrs

import mlcast_datasets

mlcast_datasets.__version__

'0.3.0.post1+g40465eeaa'

# open the mlcast datasets intake catalog
cat = mlcast_datasets.open_catalog()
list(cat.precipitation)

['radklim_hourly',
 'radklim_5_minutes',
 'dmi_10_minutes',
 'it_dpc_sri_5min',
 'uk_metoffice_5min',
 'be_rmi_radclim_mfb_5min']

5-minute rainfall amount

One part of RadKlim is the hourly rainfall amount which is available in the intake catalog as radklim_hourly.

ds = cat.precipitation.radklim_5_minutes.to_dask()
ds

<xarray.Dataset> Size: 10TB
Dimensions:          (time: 2419200, y: 1100, x: 900)
Coordinates:
  * time             (time) datetime64[ns] 19MB 2001-01-01 ... 2023-12-31T23:...
  * y                (y) float64 9kB -4.758e+03 -4.757e+03 ... -3.659e+03
  * x                (x) float64 7kB -443.0 -442.0 -441.0 ... 454.0 455.0 456.0
    lat              (y, x) float64 8MB dask.array<chunksize=(1100, 900), meta=np.ndarray>
    lon              (y, x) float64 8MB dask.array<chunksize=(1100, 900), meta=np.ndarray>
Data variables:
    rainfall_amount  (time, y, x) float32 10TB dask.array<chunksize=(1, 1100, 900), meta=np.ndarray>
    crs              float64 8B ...
Attributes: (12/13)
    Author:                            Harald Rybka, Katharina Lengfeld
    Conventions:                       CF-1.6
    history:                           Created at 2021-07-09 09:10:06.385653
    institution:                       Deutscher Wetterdienst (DWD)
    reference:                         10.5676/DWD/RADKLIM_YW_V2017.002
    title:                             RADKLIM - radar-based precipitation cl...
    ...                                ...
    mlcast_created_on:                 2026-02-27T12:03:00
    mlcast_created_by:                 Leif Denby <lcd@dmi.dk>
    mlcast_created_with:               https://github.com/mlcast-community/ml...
    mlcast_dataset_version:            0.1.1
    mlcast_dataset_identifier:         DE-DWD-radar_precipitation-RADKLIM
    mlcast_dataset_identifier_format:  {country_code}-{entity}-{physical_vari...

xarray.Dataset

• Dimensions:
  • time: 2419200
  • y: 1100
  • x: 900
• Coordinates: (5)
  • time
    (time)
    datetime64[ns]
    2001-01-01 ... 2023-12-31T23:55:00

    long_name :

    Time

    standard_name :

    time

    array(['2001-01-01T00:00:00.000000000', '2001-01-01T00:05:00.000000000',
           '2001-01-01T00:10:00.000000000', ..., '2023-12-31T23:45:00.000000000',
           '2023-12-31T23:50:00.000000000', '2023-12-31T23:55:00.000000000'],
          shape=(2419200,), dtype='datetime64[ns]')

  • y
    (y)
    float64
    -4.758e+03 ... -3.659e+03

    units :

    km

    array([-4758.144724, -4757.144724, -4756.144724, ..., -3661.144724,
           -3660.144724, -3659.144724], shape=(1100,))

  • x
    (x)
    float64
    -443.0 -442.0 ... 455.0 456.0

    units :

    km

    array([-442.962167, -441.962167, -440.962167, ...,  454.037833,  455.037833,
            456.037833], shape=(900,))

  • lat
    (y, x)
    float64
    dask.array<chunksize=(1100, 900), meta=np.ndarray>

    long_name :

    Latitude

    standard_name :

    latitude

    units :

    degrees_north

    Array Chunk Bytes 7.55 MiB 7.55 MiB Shape (1100, 900) (1100, 900) Dask graph 1 chunks in 2 graph layers Data type float64 numpy.ndarray

  • lon
    (y, x)
    float64
    dask.array<chunksize=(1100, 900), meta=np.ndarray>

    long_name :

    Longitude

    standard_name :

    longitude

    units :

    degrees_east

    Array Chunk Bytes 7.55 MiB 7.55 MiB Shape (1100, 900) (1100, 900) Dask graph 1 chunks in 2 graph layers Data type float64 numpy.ndarray

• Data variables: (2)
  • rainfall_amount
    (time, y, x)
    float32
    dask.array<chunksize=(1, 1100, 900), meta=np.ndarray>

    grid_mapping :

    crs

    long_name :

    5 minute rainfall

    standard_name :

    rainfall_amount

    units :

    kg m-2

    Array Chunk Bytes 8.71 TiB 3.78 MiB Shape (2419200, 1100, 900) (1, 1100, 900) Dask graph 2419200 chunks in 2 graph layers Data type float32 numpy.ndarray

  • crs
    ()
    float64
    ...

    earth_radius :

    6370040.0

    crs_wkt :

    PROJCRS["Stereographic_North_Pole",BASEGEOGCRS["GCS_unnamed ellipse",DATUM["D_unknown",ELLIPSOID["Unknown",6370040,0,LENGTHUNIT["metre",1]]],PRIMEM["Greenwich",0,ANGLEUNIT["Degree",0.0174532925199433]]],CONVERSION["unnamed",METHOD["Polar Stereographic (variant B)"],PARAMETER["Latitude of standard parallel",60,ANGLEUNIT["Degree",0.0174532925199433]],PARAMETER["Longitude of origin",10,ANGLEUNIT["Degree",0.0174532925199433]],PARAMETER["False easting",0,LENGTHUNIT["metre",1000]],PARAMETER["False northing",0,LENGTHUNIT["metre",1000]]],CS[Cartesian,2],AXIS["x",south,MERIDIAN[90,ANGLEUNIT["degree",0.0174532925199433]],ORDER[1],LENGTHUNIT["metre",1000]],AXIS["y",south,MERIDIAN[180,ANGLEUNIT["degree",0.0174532925199433]],ORDER[2],LENGTHUNIT["metre",1000]], BBOX[46.187191009521484, 3.0975799560546875, 55.77946853637695, 17.1041202545166]]

    false_easting :

    0.0

    false_northing :

    0.0

    grid_mapping_name :

    polar_stereographic

    inverse_flattening :

    0.0

    latitude_of_projection_origin :

    90.0

    long_name :

    RADOLAN Grid

    proj4 :

    +proj=stere +lat_0=90 +lat_ts=90 +lon_0=10 +k_0=0.933012701892 +x_0=0 +y_0=0 +a=6370040 +b=6370040 +units=km +no_defs

    scale_factor_at_projection_origin :

    0.9330127019

    semi_major_axis :

    6370040.0

    semi_minor_axis :

    6370040.0

    spatial_ref :

    PROJCRS["Stereographic_North_Pole",BASEGEOGCRS["GCS_unnamed ellipse",DATUM["D_unknown",ELLIPSOID["Unknown",6370040,0,LENGTHUNIT["metre",1]]],PRIMEM["Greenwich",0,ANGLEUNIT["Degree",0.0174532925199433]]],CONVERSION["unnamed",METHOD["Polar Stereographic (variant B)"],PARAMETER["Latitude of standard parallel",60,ANGLEUNIT["Degree",0.0174532925199433]],PARAMETER["Longitude of origin",10,ANGLEUNIT["Degree",0.0174532925199433]],PARAMETER["False easting",0,LENGTHUNIT["metre",1000]],PARAMETER["False northing",0,LENGTHUNIT["metre",1000]]],CS[Cartesian,2],AXIS["x",south,MERIDIAN[90,ANGLEUNIT["degree",0.0174532925199433]],ORDER[1],LENGTHUNIT["metre",1000]],AXIS["y",south,MERIDIAN[180,ANGLEUNIT["degree",0.0174532925199433]],ORDER[2],LENGTHUNIT["metre",1000]]]

    standard_parallel :

    60.0

    straight_vertical_longitude_from_pole :

    10.0

    units :

    m

    [1 values with dtype=float64]

• Attributes: (13)

  Author :

  Harald Rybka, Katharina Lengfeld

  Conventions :

  CF-1.6

  history :

  Created at 2021-07-09 09:10:06.385653

  institution :

  Deutscher Wetterdienst (DWD)

  reference :

  10.5676/DWD/RADKLIM_YW_V2017.002

  title :

  RADKLIM - radar-based precipitation climatology

  license :

  CC-BY-4.0

  mlcast_created_on :

  2026-02-27T12:03:00

  mlcast_created_by :

  Leif Denby <lcd@dmi.dk>

  mlcast_created_with :

  https://github.com/mlcast-community/mlcast-dataset-DE-DWD-radklim@v0.1.1

  mlcast_dataset_version :

  0.1.1

  mlcast_dataset_identifier :

  DE-DWD-radar_precipitation-RADKLIM

  mlcast_dataset_identifier_format :

  {country_code}-{entity}-{physical_variable}-{common_name}

var_name = "rainfall_amount"

crs_name = ds[var_name].grid_mapping
data_crs = ccrs.Projection(ds[crs_name].crs_wkt)

g = (
    ds[var_name]
    .sel(time="2010-01-02")
    .isel(time=slice(None, 2))
    .plot(
        transform=data_crs,
        cmap="viridis",
        add_colorbar=True,
        col="time",
        robust=True,
        subplot_kws=dict(projection=data_crs),
    )
)

for ax in g.axes.flat:
    ax.coastlines()
    ax.gridlines(draw_labels=["top", "left"])

/tmp/ipykernel_2683/839515734.py:20: FutureWarning: self.axes is deprecated since 2022.11 in order to align with matplotlibs plt.subplots, use self.axs instead.
  for ax in g.axes.flat:

Hourly rainfall amount

ds = cat.precipitation.radklim_hourly.to_dask()
ds

<xarray.Dataset> Size: 798GB
Dimensions:          (time: 201600, y: 1100, x: 900)
Coordinates:
  * time             (time) datetime64[ns] 2MB 2001-01-01T00:50:00 ... 2023-1...
  * y                (y) float64 9kB -4.758e+03 -4.757e+03 ... -3.659e+03
  * x                (x) float64 7kB -443.0 -442.0 -441.0 ... 454.0 455.0 456.0
    lat              (y, x) float64 8MB dask.array<chunksize=(1100, 900), meta=np.ndarray>
    lon              (y, x) float64 8MB dask.array<chunksize=(1100, 900), meta=np.ndarray>
Data variables:
    rainfall_amount  (time, y, x) float32 798GB dask.array<chunksize=(1, 1100, 900), meta=np.ndarray>
    crs              float64 8B ...
Attributes: (12/13)
    Author:                            Harald Rybka, Katharina Lengfeld
    Conventions:                       CF-1.6
    history:                           Created at 2021-07-09 08:48:10.797384
    institution:                       Deutscher Wetterdienst (DWD)
    reference:                         10.5676/DWD/RADKLIM_RW_V2017.002
    title:                             RADKLIM - radar-based precipitation cl...
    ...                                ...
    mlcast_created_on:                 2026-02-27T12:03:00
    mlcast_created_by:                 Leif Denby <lcd@dmi.dk>
    mlcast_created_with:               https://github.com/mlcast-community/ml...
    mlcast_dataset_version:            0.1.1
    mlcast_dataset_identifier:         DE-DWD-radar_precipitation-RADKLIM
    mlcast_dataset_identifier_format:  {country_code}-{entity}-{physical_vari...

xarray.Dataset

• Dimensions:
  • time: 201600
  • y: 1100
  • x: 900
• Coordinates: (5)
  • time
    (time)
    datetime64[ns]
    2001-01-01T00:50:00 ... 2023-12-...

    long_name :

    Time

    standard_name :

    time

    array(['2001-01-01T00:50:00.000000000', '2001-01-01T01:50:00.000000000',
           '2001-01-01T02:50:00.000000000', ..., '2023-12-31T21:50:00.000000000',
           '2023-12-31T22:50:00.000000000', '2023-12-31T23:50:00.000000000'],
          shape=(201600,), dtype='datetime64[ns]')

  • y
    (y)
    float64
    -4.758e+03 ... -3.659e+03

    units :

    km

    array([-4758.144724, -4757.144724, -4756.144724, ..., -3661.144724,
           -3660.144724, -3659.144724], shape=(1100,))

  • x
    (x)
    float64
    -443.0 -442.0 ... 455.0 456.0

    units :

    km

    array([-442.962167, -441.962167, -440.962167, ...,  454.037833,  455.037833,
            456.037833], shape=(900,))

  • lat
    (y, x)
    float64
    dask.array<chunksize=(1100, 900), meta=np.ndarray>

    long_name :

    Latitude

    standard_name :

    latitude

    units :

    degrees_north

    Array Chunk Bytes 7.55 MiB 7.55 MiB Shape (1100, 900) (1100, 900) Dask graph 1 chunks in 2 graph layers Data type float64 numpy.ndarray

  • lon
    (y, x)
    float64
    dask.array<chunksize=(1100, 900), meta=np.ndarray>

    long_name :

    Longitude

    standard_name :

    longitude

    units :

    degrees_east

    Array Chunk Bytes 7.55 MiB 7.55 MiB Shape (1100, 900) (1100, 900) Dask graph 1 chunks in 2 graph layers Data type float64 numpy.ndarray

• Data variables: (2)
  • rainfall_amount
    (time, y, x)
    float32
    dask.array<chunksize=(1, 1100, 900), meta=np.ndarray>

    grid_mapping :

    crs

    long_name :

    hourly rainfall

    standard_name :

    rainfall_amount

    units :

    kg m-2

    Array Chunk Bytes 743.51 GiB 3.78 MiB Shape (201600, 1100, 900) (1, 1100, 900) Dask graph 201600 chunks in 2 graph layers Data type float32 numpy.ndarray

  • crs
    ()
    float64
    ...

    crs_wkt :

    PROJCRS["Stereographic_North_Pole",BASEGEOGCRS["GCS_unnamed ellipse",DATUM["D_unknown",ELLIPSOID["Unknown",6370040,0,LENGTHUNIT["metre",1]]],PRIMEM["Greenwich",0,ANGLEUNIT["Degree",0.0174532925199433]]],CONVERSION["unnamed",METHOD["Polar Stereographic (variant B)"],PARAMETER["Latitude of standard parallel",60,ANGLEUNIT["Degree",0.0174532925199433]],PARAMETER["Longitude of origin",10,ANGLEUNIT["Degree",0.0174532925199433]],PARAMETER["False easting",0,LENGTHUNIT["metre",1000]],PARAMETER["False northing",0,LENGTHUNIT["metre",1000]]],CS[Cartesian,2],AXIS["x",south,MERIDIAN[90,ANGLEUNIT["degree",0.0174532925199433]],ORDER[1],LENGTHUNIT["metre",1000]],AXIS["y",south,MERIDIAN[180,ANGLEUNIT["degree",0.0174532925199433]],ORDER[2],LENGTHUNIT["metre",1000]], BBOX[46.187191009521484, 3.0975799560546875, 55.77946853637695, 17.1041202545166]]

    earth_radius :

    6370040.0

    false_easting :

    0.0

    false_northing :

    0.0

    grid_mapping_name :

    polar_stereographic

    inverse_flattening :

    0.0

    latitude_of_projection_origin :

    90.0

    long_name :

    RADOLAN Grid

    proj4 :

    +proj=stere +lat_0=90 +lat_ts=90 +lon_0=10 +k_0=0.933012701892 +x_0=0 +y_0=0 +a=6370040 +b=6370040 +units=km +no_defs

    scale_factor_at_projection_origin :

    0.9330127019

    semi_major_axis :

    6370040.0

    semi_minor_axis :

    6370040.0

    spatial_ref :

    PROJCRS["Stereographic_North_Pole",BASEGEOGCRS["GCS_unnamed ellipse",DATUM["D_unknown",ELLIPSOID["Unknown",6370040,0,LENGTHUNIT["metre",1]]],PRIMEM["Greenwich",0,ANGLEUNIT["Degree",0.0174532925199433]]],CONVERSION["unnamed",METHOD["Polar Stereographic (variant B)"],PARAMETER["Latitude of standard parallel",60,ANGLEUNIT["Degree",0.0174532925199433]],PARAMETER["Longitude of origin",10,ANGLEUNIT["Degree",0.0174532925199433]],PARAMETER["False easting",0,LENGTHUNIT["metre",1000]],PARAMETER["False northing",0,LENGTHUNIT["metre",1000]]],CS[Cartesian,2],AXIS["x",south,MERIDIAN[90,ANGLEUNIT["degree",0.0174532925199433]],ORDER[1],LENGTHUNIT["metre",1000]],AXIS["y",south,MERIDIAN[180,ANGLEUNIT["degree",0.0174532925199433]],ORDER[2],LENGTHUNIT["metre",1000]]]

    standard_parallel :

    60.0

    straight_vertical_longitude_from_pole :

    10.0

    units :

    m

    [1 values with dtype=float64]

• Attributes: (13)

  Author :

  Harald Rybka, Katharina Lengfeld

  Conventions :

  CF-1.6

  history :

  Created at 2021-07-09 08:48:10.797384

  institution :

  Deutscher Wetterdienst (DWD)

  reference :

  10.5676/DWD/RADKLIM_RW_V2017.002

  title :

  RADKLIM - radar-based precipitation climatology

  license :

  CC-BY-4.0

  mlcast_created_on :

  2026-02-27T12:03:00

  mlcast_created_by :

  Leif Denby <lcd@dmi.dk>

  mlcast_created_with :

  https://github.com/mlcast-community/mlcast-dataset-DE-DWD-radklim@v0.1.1

  mlcast_dataset_version :

  0.1.1

  mlcast_dataset_identifier :

  DE-DWD-radar_precipitation-RADKLIM

  mlcast_dataset_identifier_format :

  {country_code}-{entity}-{physical_variable}-{common_name}

var_name = "rainfall_amount"

crs_name = ds[var_name].grid_mapping
data_crs = ccrs.Projection(ds[crs_name].crs_wkt)

g = (
    ds[var_name]
    .sel(time="2010-01-02")
    .isel(time=slice(None, 2))
    .plot(
        transform=data_crs,
        cmap="viridis",
        add_colorbar=True,
        col="time",
        robust=True,
        subplot_kws=dict(projection=data_crs),
    )
)

for ax in g.axes.flat:
    ax.coastlines()
    ax.gridlines(draw_labels=["top", "left"])

/tmp/ipykernel_2683/839515734.py:20: FutureWarning: self.axes is deprecated since 2022.11 in order to align with matplotlibs plt.subplots, use self.axs instead.
  for ax in g.axes.flat: