CORAL (MBR2)

The polarised 35.5 GHz Doppler radar is part of the Combined Radar And Lidar instrument (CORAL) at the Barbados Cloud Observatory (BCO). It has a sensitivity of -48 dBZ at an altitude of 5 km and -70 dBZ at an altitude of 500 m. The radar has a range gating of 31.18 m and measures at ranges between 150 m and 18.9 km. It operates in a zenith looking mode, so that range gating effectively measures distance in the vertical. To measure the vertical velocity the radar uses the Doppler technique with an FFT of 256 samples, giving it a Doppler resolution of < 0.02 m s-1 between -10 and 10 m s-1. The radar is able to provide the linear depolarisation ratio (LDR), which can be used for a target classification based on their shape. Data processing and radar calibration are done by following [Görsdorf et al., 2015], which leads to an uncertainty of 1.3 dB.

Configurations

The radar has been operated in slightly different configurations over time, the different available configurations are listed here:

dataset	timestep	range	start date	end date
BCO.radar_MBR2_c1_v1	10s	155m-14km	2015-06-05	2016-06-01
BCO.radar_MBR2_c2_v1	10s	155m-25km	2016-06-02	2018-04-18
BCO.radar_MBR2_c3_v1	10s	155m-18km	2018-04-19	2018-04-29
BCO.radar_MBR2_c4_v1	2s	155m-18km	2018-04-30	today

Examples

La Soufrière Eruption in April 2021

The La Soufrière Eruption started on 9 April 2021 at 12:41 UTC. The dispersion and aging of the emitted volcanic aerosols have been investigated by [Bruckert et al., 2023], where data from the MBR2 radar has been used. We’ll show how to recreate a similar plot to the radar plot from Figure 2 of that publication:

import intake
import numpy as np

# this is the time-period investigated in the paper
simulation_start = np.datetime64("2021-04-09T12:00")
simulation_end = simulation_start + np.timedelta64(96, "h")

# get an appropriate radar dataset from tcodata
cat = intake.open_catalog("https://tcodata.mpimet.mpg.de/catalog.yaml")
mbr2 = cat.BCO.radar_MBR2_c4_v1.to_dask()

# cut, resample and plot the data
reflectivity_in_simulation_period = mbr2.Ze.sel(time=slice(simulation_start, simulation_end))
reflectivity_10min_avg = reflectivity_in_simulation_period.resample(time="10min").mean()
reflectivity_10min_avg.plot(x="time", y="alt", vmin=-75, vmax=40, figsize=(12, 4), cmap="turbo");

/builds/tco/bco/docs/.venv/lib/python3.14/site-packages/intake_xarray/base.py:21: FutureWarning: The return type of `Dataset.dims` will be changed to return a set of dimension names in future, in order to be more consistent with `DataArray.dims`. To access a mapping from dimension names to lengths, please use `Dataset.sizes`.
  'dims': dict(self._ds.dims),

The entire dataset:

mbr2

<xarray.Dataset> Size: 2TB
Dimensions:     (time: 62718216, alt: 605)
Coordinates:
  * time        (time) datetime64[ns] 502MB 2018-04-30T00:00:17 ... 2023-03-1...
  * alt         (alt) float64 5kB 174.4 205.6 236.8 ... 1.898e+04 1.901e+04
    lat         float64 8B ...
    lon         float64 8B ...
Data variables: (12/23)
    HSDco       (time, alt) float32 152GB dask.array<chunksize=(259200, 55), meta=np.ndarray>
    HSDcx       (time, alt) float32 152GB dask.array<chunksize=(259200, 55), meta=np.ndarray>
    LDR         (time, alt) float32 152GB dask.array<chunksize=(259200, 55), meta=np.ndarray>
    LDRg        (time, alt) float32 152GB dask.array<chunksize=(259200, 55), meta=np.ndarray>
    MeltHei     (time) float32 251MB dask.array<chunksize=(2592000,), meta=np.ndarray>
    RMS         (time, alt) float32 152GB dask.array<chunksize=(259200, 55), meta=np.ndarray>
    ...          ...
    mask        (time, alt) bool 38GB dask.array<chunksize=(259200, 55), meta=np.ndarray>
    northangle  (time) float32 251MB dask.array<chunksize=(2592000,), meta=np.ndarray>
    sensor_alt  float64 8B ...
    status      (time) int16 125MB dask.array<chunksize=(2592000,), meta=np.ndarray>
    tpow        (time) float32 251MB dask.array<chunksize=(2592000,), meta=np.ndarray>
    zenith      float64 8B ...
Attributes:
    Conventions:      CF-1.7
    bcoproc_version:  2.0.1.dev221+g18013fd
    created_with:     bcoproc (kband.py)
    creation_date:    Sat Mar 18 12:01:17 2023
    institution:      Max Planck Institute for Meteorology, Hamburg
    instrument:       MBR2 cloud radar
    location:         The Barbados Cloud Observatory, Deebles Point, Barbados...
    reference:        Ka Band Cloud Radar MIRA-3x, METEK GmbH www.metek.de
    system:           MIRA36
    title:            MIRA-3x Cloud Radar Data
    tool_versions:    {"Python": "3.10.6 | packaged by conda-forge | (main, A...

xarray.Dataset

• Dimensions:
  • time: 62718216
  • alt: 605
• Coordinates: (4)
  • time
    (time)
    datetime64[ns]
    2018-04-30T00:00:17 ... 2023-03-...

    axis :

    T

    standard_name :

    time

    array(['2018-04-30T00:00:17.000000000', '2018-04-30T00:00:28.000000000',
           '2018-04-30T00:00:38.000000000', ..., '2023-03-14T23:59:57.000000000',
           '2023-03-14T23:59:59.000000000', '2023-03-15T00:00:01.000000000'],
          shape=(62718216,), dtype='datetime64[ns]')

  • alt
    (alt)
    float64
    174.4 205.6 ... 1.898e+04 1.901e+04

    axis :

    Z

    long_name :

    height above mean sea level

    standard_name :

    altitude

    units :

    m

    array([  174.4 ,   205.58,   236.75, ..., 18944.28, 18975.45, 19006.63],
          shape=(605,))

  • lat
    ()
    float64
    ...

    [1 values with dtype=float64]

  • lon
    ()
    float64
    ...

    [1 values with dtype=float64]

• Data variables: (23)
  • HSDco
    (time, alt)
    float32
    dask.array<chunksize=(259200, 55), meta=np.ndarray>

    db :

    1

    long_name :

    co-channel HSdiv noise power. The arbitrary unit is DSP.

    units :

    1

    yrange :

    [-20.0, 40.0]

    Array Chunk Bytes 141.35 GiB 54.38 MiB Shape (62718216, 605) (259200, 55) Dask graph 2662 chunks in 2 graph layers Data type float32 numpy.ndarray

  • HSDcx
    (time, alt)
    float32
    dask.array<chunksize=(259200, 55), meta=np.ndarray>

    db :

    1

    long_name :

    cross-channel HSdiv noise power. The arbitrary unit is DSP.

    units :

    1

    yrange :

    [-20.0, 40.0]

    Array Chunk Bytes 141.35 GiB 54.38 MiB Shape (62718216, 605) (259200, 55) Dask graph 2662 chunks in 2 graph layers Data type float32 numpy.ndarray

  • LDR
    (time, alt)
    float32
    dask.array<chunksize=(259200, 55), meta=np.ndarray>

    long_name :

    linear de-polarization ratio LDR of all hydrometeors

    units :

    dBZ

    Array Chunk Bytes 141.35 GiB 54.38 MiB Shape (62718216, 605) (259200, 55) Dask graph 2662 chunks in 2 graph layers Data type float32 numpy.ndarray

  • LDRg
    (time, alt)
    float32
    dask.array<chunksize=(259200, 55), meta=np.ndarray>

    long_name :

    linear de-polarization ratio LDR of all targets (global)

    units :

    dBZ

    Array Chunk Bytes 141.35 GiB 54.38 MiB Shape (62718216, 605) (259200, 55) Dask graph 2662 chunks in 2 graph layers Data type float32 numpy.ndarray

  • MeltHei
    (time)
    float32
    dask.array<chunksize=(2592000,), meta=np.ndarray>

    db :

    0

    long_name :

    melting layer height

    units :

    m

    yrange :

    [0, 14000]

    Array Chunk Bytes 239.25 MiB 9.89 MiB Shape (62718216,) (2592000,) Dask graph 25 chunks in 2 graph layers Data type float32 numpy.ndarray

  • RMS
    (time, alt)
    float32
    dask.array<chunksize=(259200, 55), meta=np.ndarray>

    db :

    0

    long_name :

    peak width RMS of all hydrometeors

    units :

    m s-1

    yrange :

    [0.0, 3.0]

    Array Chunk Bytes 141.35 GiB 54.38 MiB Shape (62718216, 605) (259200, 55) Dask graph 2662 chunks in 2 graph layers Data type float32 numpy.ndarray

  • RMSg
    (time, alt)
    float32
    dask.array<chunksize=(259200, 55), meta=np.ndarray>

    db :

    0

    long_name :

    peak width RMS of all targets (global)

    units :

    m s-1

    yrange :

    [0.0, 3.0]

    Array Chunk Bytes 141.35 GiB 54.38 MiB Shape (62718216, 605) (259200, 55) Dask graph 2662 chunks in 2 graph layers Data type float32 numpy.ndarray

  • RadarConst
    (time)
    float32
    dask.array<chunksize=(2592000,), meta=np.ndarray>

    db :

    1

    long_name :

    radar constant related to 5 km height

    units :

    mm6 m-3

    yrange :

    [-35.0, -20.0]

    Array Chunk Bytes 239.25 MiB 9.89 MiB Shape (62718216,) (2592000,) Dask graph 25 chunks in 2 graph layers Data type float32 numpy.ndarray

  • SNR
    (time, alt)
    float32
    dask.array<chunksize=(259200, 55), meta=np.ndarray>

    long_name :

    reflectivity SNR

    units :

    dBZ

    Array Chunk Bytes 141.35 GiB 54.38 MiB Shape (62718216, 605) (259200, 55) Dask graph 2662 chunks in 2 graph layers Data type float32 numpy.ndarray

  • SNRg
    (time, alt)
    float32
    dask.array<chunksize=(259200, 55), meta=np.ndarray>

    long_name :

    reflectivity SNRg

    units :

    dBZ

    Array Chunk Bytes 141.35 GiB 54.38 MiB Shape (62718216, 605) (259200, 55) Dask graph 2662 chunks in 2 graph layers Data type float32 numpy.ndarray

  • SNRplank
    (time, alt)
    float32
    dask.array<chunksize=(259200, 55), meta=np.ndarray>

    long_name :

    reflectivity SNRplank

    units :

    dBZ

    Array Chunk Bytes 141.35 GiB 54.38 MiB Shape (62718216, 605) (259200, 55) Dask graph 2662 chunks in 2 graph layers Data type float32 numpy.ndarray

  • VEL
    (time, alt)
    float32
    dask.array<chunksize=(259200, 55), meta=np.ndarray>

    db :

    0

    long_name :

    doppler velocity VEL of all hydrometeors

    units :

    m s-1

    yrange :

    [-10.661450386047363, 10.661450386047363]

    Array Chunk Bytes 141.35 GiB 54.38 MiB Shape (62718216, 605) (259200, 55) Dask graph 2662 chunks in 2 graph layers Data type float32 numpy.ndarray

  • VELg
    (time, alt)
    float32
    dask.array<chunksize=(259200, 55), meta=np.ndarray>

    db :

    0

    long_name :

    doppler velocity VEL of all targets (global)

    units :

    m s-1

    yrange :

    [-10.661450386047363, 10.661450386047363]

    Array Chunk Bytes 141.35 GiB 54.38 MiB Shape (62718216, 605) (259200, 55) Dask graph 2662 chunks in 2 graph layers Data type float32 numpy.ndarray

  • Ze
    (time, alt)
    float32
    dask.array<chunksize=(259200, 55), meta=np.ndarray>

    long_name :

    equivalent radar reflectivity of all hydrometeors

    units :

    dBZ

    Array Chunk Bytes 141.35 GiB 54.38 MiB Shape (62718216, 605) (259200, 55) Dask graph 2662 chunks in 2 graph layers Data type float32 numpy.ndarray

  • Zf
    (time, alt)
    float32
    dask.array<chunksize=(259200, 55), meta=np.ndarray>

    applied_thresholds :

    dBZ < -35 and VELg < -2.0 or VELg > 0.5 between 1000m and 8000m

    details :

    artifacts (rabbit-ears), caused by side lobes, are removed

    long_name :

    rabbit-ear filtered radar reflectivity of all hydrometeors

    standard_name :

    equivalent_reflectivity_factor

    units :

    dBZ

    Array Chunk Bytes 141.35 GiB 54.38 MiB Shape (62718216, 605) (259200, 55) Dask graph 2662 chunks in 2 graph layers Data type float32 numpy.ndarray

  • Zg
    (time, alt)
    float32
    dask.array<chunksize=(259200, 55), meta=np.ndarray>

    long_name :

    equivalent radar reflectivity of all targets (global)

    units :

    dBZ

    Array Chunk Bytes 141.35 GiB 54.38 MiB Shape (62718216, 605) (259200, 55) Dask graph 2662 chunks in 2 graph layers Data type float32 numpy.ndarray

  • azimuth
    ()
    float64
    ...

    long_name :

    horizontal angle offset to north (0 degrees).

    standard_name :

    sensor_azimuth_angle

    units :

    degree

    [1 values with dtype=float64]

  • mask
    (time, alt)
    bool
    dask.array<chunksize=(259200, 55), meta=np.ndarray>

    flag_meanings :

    rabbit_ear_retrieval no_rabbit_ear_retrieval

    flag_values :

    False True

    long_name :

    rabbit ear mask

    units :

    1

    usage :

    multiply radar products with this mask to reduce false signals by rabbit-ear retrievals

    Array Chunk Bytes 35.34 GiB 13.60 MiB Shape (62718216, 605) (259200, 55) Dask graph 2662 chunks in 2 graph layers Data type bool numpy.ndarray

  • northangle
    (time)
    float32
    dask.array<chunksize=(2592000,), meta=np.ndarray>

    long_name :

    north angle

    units :

    degree

    Array Chunk Bytes 239.25 MiB 9.89 MiB Shape (62718216,) (2592000,) Dask graph 25 chunks in 2 graph layers Data type float32 numpy.ndarray

  • sensor_alt
    ()
    float64
    ...

    long_name :

    antenna height above mean sea level

    standard_name :

    altitude

    units :

    m

    [1 values with dtype=float64]

  • status
    (time)
    int16
    dask.array<chunksize=(2592000,), meta=np.ndarray>

    flag_meanings :

    radar_off radar_on

    flag_values :

    0 1

    long_name :

    radar power flag

    units :

    1

    Array Chunk Bytes 119.63 MiB 4.94 MiB Shape (62718216,) (2592000,) Dask graph 25 chunks in 2 graph layers Data type int16 numpy.ndarray

  • tpow
    (time)
    float32
    dask.array<chunksize=(2592000,), meta=np.ndarray>

    long_name :

    average transmit power

    units :

    W

    Array Chunk Bytes 239.25 MiB 9.89 MiB Shape (62718216,) (2592000,) Dask graph 25 chunks in 2 graph layers Data type float32 numpy.ndarray

  • zenith
    ()
    float64
    ...

    long_name :

    zenith angle

    standard_name :

    sensor_zenith_angle

    units :

    degree

    [1 values with dtype=float64]

• Attributes: (11)

  Conventions :

  CF-1.7

  bcoproc_version :

  2.0.1.dev221+g18013fd

  created_with :

  bcoproc (kband.py)

  creation_date :

  Sat Mar 18 12:01:17 2023

  institution :

  Max Planck Institute for Meteorology, Hamburg

  instrument :

  MBR2 cloud radar

  location :

  The Barbados Cloud Observatory, Deebles Point, Barbados, West Indies

  reference :

  Ka Band Cloud Radar MIRA-3x, METEK GmbH www.metek.de

  system :

  MIRA36

  title :

  MIRA-3x Cloud Radar Data

  tool_versions :

  {"Python": "3.10.6 | packaged by conda-forge | (main, Aug 22 2022, 20:36:39) [GCC 10.4.0]", "cftime": "1.6.2", "pyyaml": "6.0", "pyzstd": "0.15.3", "python-dateutil": "2.8.2", "locket": "1.0.0", "netCDF4": "1.6.1", "xarray": "2022.6.0", "bcoproc": "2.0.1.dev221+g18013fd", "cloudpickle": "2.2.0", "toolz": "0.12.0", "packaging": "21.3", "pandas": "1.5.0", "numpy": "1.23.3", "pyparsing": "3.0.9", "fsspec": "2022.8.2", "pytz": "2022.2.1", "partd": "1.3.0", "six": "1.16.0", "dask": "2022.9.1"}

Note

The dataset we used for the plot comes with 2 seconds temporal sampling, which can be a lot of data. For this example, we resampled this to 10 minutes, because otherwise plotting would take much longer.