RTOFS

# Install required libraries
# !pip install -q xarray_subset_grid@git+https://github.com/asascience-open/xarray-subset-grid.git
# !pip install -q s3fs cftime xarray cf-xarray fsspec dask h5netcdf

import cf_xarray # noqa
import datetime
import fsspec
import xarray as xr

import xarray_subset_grid.accessor  # noqa: F401

# open dataset as zarr object using fsspec reference file system and xarray
fs = fsspec.filesystem(
    "reference",
    fo="s3://noaa-nodd-kerchunk-pds/rtofs/rtofs.rtofs_glo_2ds_diag.best.nc.zarr",
    remote_protocol="s3",
    remote_options={"anon": True},
    target_protocol="s3",
    target_options={"anon": True},
)
m = fs.get_mapper("")

ds = xr.open_dataset(
    m,
    engine="zarr",
    backend_kwargs=dict(consolidated=False),
    chunks={},
)
ds

<xarray.Dataset> Size: 618GB
Dimensions:                           (MT: 2081, Y: 3298, X: 4500)
Coordinates:
    Date                              (MT) float64 17kB dask.array<chunksize=(512,), meta=np.ndarray>
    Latitude                          (Y, X) float32 59MB dask.array<chunksize=(825, 1125), meta=np.ndarray>
    Longitude                         (Y, X) float32 59MB dask.array<chunksize=(825, 1125), meta=np.ndarray>
  * MT                                (MT) datetime64[ns] 17kB 2024-05-09 ......
  * X                                 (X) int32 18kB 1 2 3 4 ... 4498 4499 4500
  * Y                                 (Y) int32 13kB 1 2 3 4 ... 3296 3297 3298
Data variables:
    mixed_layer_thickness             (MT, Y, X) float32 124GB dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>
    ssh                               (MT, Y, X) float32 124GB dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>
    surface_boundary_layer_thickness  (MT, Y, X) float32 124GB dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>
    u_barotropic_velocity             (MT, Y, X) float32 124GB dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>
    v_barotropic_velocity             (MT, Y, X) float32 124GB dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>
Attributes:
    Conventions:  CF-1.0
    experiment:   92.8
    history:      archv2ncdf2d
    institution:  National Centers for Environmental Prediction
    source:       HYCOM archive file
    title:        HYCOM ATLb2.00

xarray.Dataset

• Dimensions:
  • MT: 2081
  • Y: 3298
  • X: 4500
• Coordinates: (6)
  • Date
    (MT)
    float64
    dask.array<chunksize=(512,), meta=np.ndarray>

    C_format :

    %13.4f

    FORTRAN_format :

    (f13.4)

    long_name :

    date

    units :

    day as %Y%m%d.%f

    Array Chunk Bytes 16.26 kiB 4.00 kiB Shape (2081,) (512,) Dask graph 5 chunks in 2 graph layers Data type float64 numpy.ndarray

  • Latitude
    (Y, X)
    float32
    dask.array<chunksize=(825, 1125), meta=np.ndarray>

    standard_name :

    latitude

    units :

    degrees_north

    Array Chunk Bytes 56.61 MiB 3.54 MiB Shape (3298, 4500) (825, 1125) Dask graph 16 chunks in 2 graph layers Data type float32 numpy.ndarray

  • Longitude
    (Y, X)
    float32
    dask.array<chunksize=(825, 1125), meta=np.ndarray>

    modulo :

    360 degrees

    standard_name :

    longitude

    units :

    degrees_east

    Array Chunk Bytes 56.61 MiB 3.54 MiB Shape (3298, 4500) (825, 1125) Dask graph 16 chunks in 2 graph layers Data type float32 numpy.ndarray

  • MT
    (MT)
    datetime64[ns]
    2024-05-09 ... 2024-08-07

    axis :

    T

    long_name :

    time

    array(['2024-05-09T00:00:00.000000000', '2024-05-09T01:00:00.000000000',
           '2024-05-09T02:00:00.000000000', ..., '2024-08-06T18:00:00.000000000',
           '2024-08-06T21:00:00.000000000', '2024-08-07T00:00:00.000000000'],
          dtype='datetime64[ns]')

  • X
    (X)
    int32
    1 2 3 4 5 ... 4497 4498 4499 4500

    axis :

    X

    point_spacing :

    even

    array([   1,    2,    3, ..., 4498, 4499, 4500], dtype=int32)

  • Y
    (Y)
    int32
    1 2 3 4 5 ... 3295 3296 3297 3298

    axis :

    Y

    point_spacing :

    even

    array([   1,    2,    3, ..., 3296, 3297, 3298], dtype=int32)

• Data variables: (5)
  • mixed_layer_thickness
    (MT, Y, X)
    float32
    dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>

    long_name :

    mix.layr.thickness [92.8H]

    standard_name :

    ocean_mixed_layer_thickness

    units :

    m

    valid_range :

    [0.5049223899841309, 1273.3824462890625]

    Array Chunk Bytes 115.05 GiB 3.54 MiB Shape (2081, 3298, 4500) (1, 825, 1125) Dask graph 33296 chunks in 2 graph layers Data type float32 numpy.ndarray

  • ssh
    (MT, Y, X)
    float32
    dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>

    long_name :

    sea surf. height [92.8H]

    standard_name :

    sea_surface_elevation

    units :

    m

    valid_range :

    [-2.0905275344848633, 1.3661648035049438]

    Array Chunk Bytes 115.05 GiB 3.54 MiB Shape (2081, 3298, 4500) (1, 825, 1125) Dask graph 33296 chunks in 2 graph layers Data type float32 numpy.ndarray

  • surface_boundary_layer_thickness
    (MT, Y, X)
    float32
    dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>

    long_name :

    bnd.layr.thickness [92.8H]

    units :

    m

    valid_range :

    [0.9318792223930359, 766.9793090820312]

    Array Chunk Bytes 115.05 GiB 3.54 MiB Shape (2081, 3298, 4500) (1, 825, 1125) Dask graph 33296 chunks in 2 graph layers Data type float32 numpy.ndarray

  • u_barotropic_velocity
    (MT, Y, X)
    float32
    dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>

    long_name :

    baro. u-vel. [92.8H]

    standard_name :

    barotropic_eastward_sea_water_velocity

    units :

    m/s

    valid_range :

    [-1.0019874572753906, 1.198006272315979]

    Array Chunk Bytes 115.05 GiB 3.54 MiB Shape (2081, 3298, 4500) (1, 825, 1125) Dask graph 33296 chunks in 2 graph layers Data type float32 numpy.ndarray

  • v_barotropic_velocity
    (MT, Y, X)
    float32
    dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>

    long_name :

    baro. v-vel. [92.8H]

    standard_name :

    barotropic_northward_sea_water_velocity

    units :

    m/s

    valid_range :

    [-1.2001447677612305, 1.0958094596862793]

    Array Chunk Bytes 115.05 GiB 3.54 MiB Shape (2081, 3298, 4500) (1, 825, 1125) Dask graph 33296 chunks in 2 graph layers Data type float32 numpy.ndarray

• Indexes: (3)
  • MT
    PandasIndex

    PandasIndex(DatetimeIndex(['2024-05-09 00:00:00', '2024-05-09 01:00:00',
                   '2024-05-09 02:00:00', '2024-05-09 03:00:00',
                   '2024-05-09 04:00:00', '2024-05-09 05:00:00',
                   '2024-05-09 06:00:00', '2024-05-09 07:00:00',
                   '2024-05-09 08:00:00', '2024-05-09 09:00:00',
                   ...
                   '2024-08-05 21:00:00', '2024-08-06 00:00:00',
                   '2024-08-06 03:00:00', '2024-08-06 06:00:00',
                   '2024-08-06 09:00:00', '2024-08-06 12:00:00',
                   '2024-08-06 15:00:00', '2024-08-06 18:00:00',
                   '2024-08-06 21:00:00', '2024-08-07 00:00:00'],
                  dtype='datetime64[ns]', name='MT', length=2081, freq=None))

  • X
    PandasIndex

    PandasIndex(Index([   1,    2,    3,    4,    5,    6,    7,    8,    9,   10,
           ...
           4491, 4492, 4493, 4494, 4495, 4496, 4497, 4498, 4499, 4500],
          dtype='int32', name='X', length=4500))

  • Y
    PandasIndex

    PandasIndex(Index([   1,    2,    3,    4,    5,    6,    7,    8,    9,   10,
           ...
           3289, 3290, 3291, 3292, 3293, 3294, 3295, 3296, 3297, 3298],
          dtype='int32', name='Y', length=3298))

• Attributes: (6)

  Conventions :

  CF-1.0

  experiment :

  92.8

  history :

  archv2ncdf2d

  institution :

  National Centers for Environmental Prediction

  source :

  HYCOM archive file

  title :

  HYCOM ATLb2.00

Reprojected HYCOM has some garbage data in the last X column, so we just get rid of it for this example

ds = ds.isel(X=slice(0, -1))
ds

<xarray.Dataset> Size: 618GB
Dimensions:                           (MT: 2081, Y: 3298, X: 4499)
Coordinates:
    Date                              (MT) float64 17kB dask.array<chunksize=(512,), meta=np.ndarray>
    Latitude                          (Y, X) float32 59MB dask.array<chunksize=(825, 1125), meta=np.ndarray>
    Longitude                         (Y, X) float32 59MB dask.array<chunksize=(825, 1125), meta=np.ndarray>
  * MT                                (MT) datetime64[ns] 17kB 2024-05-09 ......
  * X                                 (X) int32 18kB 1 2 3 4 ... 4497 4498 4499
  * Y                                 (Y) int32 13kB 1 2 3 4 ... 3296 3297 3298
Data variables:
    mixed_layer_thickness             (MT, Y, X) float32 124GB dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>
    ssh                               (MT, Y, X) float32 124GB dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>
    surface_boundary_layer_thickness  (MT, Y, X) float32 124GB dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>
    u_barotropic_velocity             (MT, Y, X) float32 124GB dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>
    v_barotropic_velocity             (MT, Y, X) float32 124GB dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>
Attributes:
    Conventions:  CF-1.0
    experiment:   92.8
    history:      archv2ncdf2d
    institution:  National Centers for Environmental Prediction
    source:       HYCOM archive file
    title:        HYCOM ATLb2.00

xarray.Dataset

• Dimensions:
  • MT: 2081
  • Y: 3298
  • X: 4499
• Coordinates: (6)
  • Date
    (MT)
    float64
    dask.array<chunksize=(512,), meta=np.ndarray>

    C_format :

    %13.4f

    FORTRAN_format :

    (f13.4)

    long_name :

    date

    units :

    day as %Y%m%d.%f

    Array Chunk Bytes 16.26 kiB 4.00 kiB Shape (2081,) (512,) Dask graph 5 chunks in 2 graph layers Data type float64 numpy.ndarray

  • Latitude
    (Y, X)
    float32
    dask.array<chunksize=(825, 1125), meta=np.ndarray>

    standard_name :

    latitude

    units :

    degrees_north

    Array Chunk Bytes 56.60 MiB 3.54 MiB Shape (3298, 4499) (825, 1125) Dask graph 16 chunks in 3 graph layers Data type float32 numpy.ndarray

  • Longitude
    (Y, X)
    float32
    dask.array<chunksize=(825, 1125), meta=np.ndarray>

    modulo :

    360 degrees

    standard_name :

    longitude

    units :

    degrees_east

    Array Chunk Bytes 56.60 MiB 3.54 MiB Shape (3298, 4499) (825, 1125) Dask graph 16 chunks in 3 graph layers Data type float32 numpy.ndarray

  • MT
    (MT)
    datetime64[ns]
    2024-05-09 ... 2024-08-07

    axis :

    T

    long_name :

    time

    array(['2024-05-09T00:00:00.000000000', '2024-05-09T01:00:00.000000000',
           '2024-05-09T02:00:00.000000000', ..., '2024-08-06T18:00:00.000000000',
           '2024-08-06T21:00:00.000000000', '2024-08-07T00:00:00.000000000'],
          dtype='datetime64[ns]')

  • X
    (X)
    int32
    1 2 3 4 5 ... 4496 4497 4498 4499

    axis :

    X

    point_spacing :

    even

    array([   1,    2,    3, ..., 4497, 4498, 4499], dtype=int32)

  • Y
    (Y)
    int32
    1 2 3 4 5 ... 3295 3296 3297 3298

    axis :

    Y

    point_spacing :

    even

    array([   1,    2,    3, ..., 3296, 3297, 3298], dtype=int32)

• Data variables: (5)
  • mixed_layer_thickness
    (MT, Y, X)
    float32
    dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>

    long_name :

    mix.layr.thickness [92.8H]

    standard_name :

    ocean_mixed_layer_thickness

    units :

    m

    valid_range :

    [0.5049223899841309, 1273.3824462890625]

    Array Chunk Bytes 115.03 GiB 3.54 MiB Shape (2081, 3298, 4499) (1, 825, 1125) Dask graph 33296 chunks in 3 graph layers Data type float32 numpy.ndarray

  • ssh
    (MT, Y, X)
    float32
    dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>

    long_name :

    sea surf. height [92.8H]

    standard_name :

    sea_surface_elevation

    units :

    m

    valid_range :

    [-2.0905275344848633, 1.3661648035049438]

    Array Chunk Bytes 115.03 GiB 3.54 MiB Shape (2081, 3298, 4499) (1, 825, 1125) Dask graph 33296 chunks in 3 graph layers Data type float32 numpy.ndarray

  • surface_boundary_layer_thickness
    (MT, Y, X)
    float32
    dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>

    long_name :

    bnd.layr.thickness [92.8H]

    units :

    m

    valid_range :

    [0.9318792223930359, 766.9793090820312]

    Array Chunk Bytes 115.03 GiB 3.54 MiB Shape (2081, 3298, 4499) (1, 825, 1125) Dask graph 33296 chunks in 3 graph layers Data type float32 numpy.ndarray

  • u_barotropic_velocity
    (MT, Y, X)
    float32
    dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>

    long_name :

    baro. u-vel. [92.8H]

    standard_name :

    barotropic_eastward_sea_water_velocity

    units :

    m/s

    valid_range :

    [-1.0019874572753906, 1.198006272315979]

    Array Chunk Bytes 115.03 GiB 3.54 MiB Shape (2081, 3298, 4499) (1, 825, 1125) Dask graph 33296 chunks in 3 graph layers Data type float32 numpy.ndarray

  • v_barotropic_velocity
    (MT, Y, X)
    float32
    dask.array<chunksize=(1, 825, 1125), meta=np.ndarray>

    long_name :

    baro. v-vel. [92.8H]

    standard_name :

    barotropic_northward_sea_water_velocity

    units :

    m/s

    valid_range :

    [-1.2001447677612305, 1.0958094596862793]

    Array Chunk Bytes 115.03 GiB 3.54 MiB Shape (2081, 3298, 4499) (1, 825, 1125) Dask graph 33296 chunks in 3 graph layers Data type float32 numpy.ndarray

• Indexes: (3)
  • MT
    PandasIndex

    PandasIndex(DatetimeIndex(['2024-05-09 00:00:00', '2024-05-09 01:00:00',
                   '2024-05-09 02:00:00', '2024-05-09 03:00:00',
                   '2024-05-09 04:00:00', '2024-05-09 05:00:00',
                   '2024-05-09 06:00:00', '2024-05-09 07:00:00',
                   '2024-05-09 08:00:00', '2024-05-09 09:00:00',
                   ...
                   '2024-08-05 21:00:00', '2024-08-06 00:00:00',
                   '2024-08-06 03:00:00', '2024-08-06 06:00:00',
                   '2024-08-06 09:00:00', '2024-08-06 12:00:00',
                   '2024-08-06 15:00:00', '2024-08-06 18:00:00',
                   '2024-08-06 21:00:00', '2024-08-07 00:00:00'],
                  dtype='datetime64[ns]', name='MT', length=2081, freq=None))

  • X
    PandasIndex

    PandasIndex(Index([   1,    2,    3,    4,    5,    6,    7,    8,    9,   10,
           ...
           4490, 4491, 4492, 4493, 4494, 4495, 4496, 4497, 4498, 4499],
          dtype='int32', name='X', length=4499))

  • Y
    PandasIndex

    PandasIndex(Index([   1,    2,    3,    4,    5,    6,    7,    8,    9,   10,
           ...
           3289, 3290, 3291, 3292, 3293, 3294, 3295, 3296, 3297, 3298],
          dtype='int32', name='Y', length=3298))

• Attributes: (6)

  Conventions :

  CF-1.0

  experiment :

  92.8

  history :

  archv2ncdf2d

  institution :

  National Centers for Environmental Prediction

  source :

  HYCOM archive file

  title :

  HYCOM ATLb2.00

f"Dataset size: {ds.nbytes * 1.0e-9} Gb"

'Dataset size: 617.66392334 Gb'

Check the grid type that is recognized

ds.xsg.grid.name

'regular_grid_2d'

Create our region of interest and subset

bbox = [-93.63283364104035, 16.18222316056857, -67.60864242620244, 34.02542167172069]

ds_subset = ds.xsg.subset_bbox(bbox)
ds_subset

<xarray.Dataset> Size: 3GB
Dimensions:                           (MT: 2081, Y: 247, X: 325)
Coordinates:
    Date                              (MT) float64 17kB dask.array<chunksize=(512,), meta=np.ndarray>
    Latitude                          (Y, X) float32 321kB dask.array<chunksize=(247, 325), meta=np.ndarray>
    Longitude                         (Y, X) float32 321kB dask.array<chunksize=(247, 325), meta=np.ndarray>
  * MT                                (MT) datetime64[ns] 17kB 2024-05-09 ......
  * X                                 (X) int32 1kB 2404 2405 2406 ... 2727 2728
  * Y                                 (Y) int32 988B 1711 1712 ... 1956 1957
Data variables:
    mixed_layer_thickness             (MT, Y, X) float32 668MB dask.array<chunksize=(1, 247, 325), meta=np.ndarray>
    ssh                               (MT, Y, X) float32 668MB dask.array<chunksize=(1, 247, 325), meta=np.ndarray>
    surface_boundary_layer_thickness  (MT, Y, X) float32 668MB dask.array<chunksize=(1, 247, 325), meta=np.ndarray>
    u_barotropic_velocity             (MT, Y, X) float32 668MB dask.array<chunksize=(1, 247, 325), meta=np.ndarray>
    v_barotropic_velocity             (MT, Y, X) float32 668MB dask.array<chunksize=(1, 247, 325), meta=np.ndarray>
    subset_mask                       (Y, X) float64 642kB 1.0 1.0 ... 1.0 1.0
Attributes:
    Conventions:  CF-1.0
    experiment:   92.8
    history:      archv2ncdf2d
    institution:  National Centers for Environmental Prediction
    source:       HYCOM archive file
    title:        HYCOM ATLb2.00

xarray.Dataset

• Dimensions:
  • MT: 2081
  • Y: 247
  • X: 325
• Coordinates: (6)
  • Date
    (MT)
    float64
    dask.array<chunksize=(512,), meta=np.ndarray>

    C_format :

    %13.4f

    FORTRAN_format :

    (f13.4)

    long_name :

    date

    units :

    day as %Y%m%d.%f

    Array Chunk Bytes 16.26 kiB 4.00 kiB Shape (2081,) (512,) Dask graph 5 chunks in 2 graph layers Data type float64 numpy.ndarray

  • Latitude
    (Y, X)
    float32
    dask.array<chunksize=(247, 325), meta=np.ndarray>

    standard_name :

    latitude

    units :

    degrees_north

    Array Chunk Bytes 313.57 kiB 313.57 kiB Shape (247, 325) (247, 325) Dask graph 1 chunks in 5 graph layers Data type float32 numpy.ndarray

  • Longitude
    (Y, X)
    float32
    dask.array<chunksize=(247, 325), meta=np.ndarray>

    modulo :

    360 degrees

    standard_name :

    longitude

    units :

    degrees_east

    Array Chunk Bytes 313.57 kiB 313.57 kiB Shape (247, 325) (247, 325) Dask graph 1 chunks in 5 graph layers Data type float32 numpy.ndarray

  • MT
    (MT)
    datetime64[ns]
    2024-05-09 ... 2024-08-07

    axis :

    T

    long_name :

    time

    array(['2024-05-09T00:00:00.000000000', '2024-05-09T01:00:00.000000000',
           '2024-05-09T02:00:00.000000000', ..., '2024-08-06T18:00:00.000000000',
           '2024-08-06T21:00:00.000000000', '2024-08-07T00:00:00.000000000'],
          dtype='datetime64[ns]')

  • X
    (X)
    int32
    2404 2405 2406 ... 2726 2727 2728

    axis :

    X

    point_spacing :

    even

    array([2404, 2405, 2406, ..., 2726, 2727, 2728], dtype=int32)

  • Y
    (Y)
    int32
    1711 1712 1713 ... 1955 1956 1957

    axis :

    Y

    point_spacing :

    even

    array([1711, 1712, 1713, ..., 1955, 1956, 1957], dtype=int32)

• Data variables: (6)
  • mixed_layer_thickness
    (MT, Y, X)
    float32
    dask.array<chunksize=(1, 247, 325), meta=np.ndarray>

    long_name :

    mix.layr.thickness [92.8H]

    standard_name :

    ocean_mixed_layer_thickness

    units :

    m

    valid_range :

    [0.5049223899841309, 1273.3824462890625]

    Array Chunk Bytes 637.25 MiB 313.57 kiB Shape (2081, 247, 325) (1, 247, 325) Dask graph 2081 chunks in 7 graph layers Data type float32 numpy.ndarray

  • ssh
    (MT, Y, X)
    float32
    dask.array<chunksize=(1, 247, 325), meta=np.ndarray>

    long_name :

    sea surf. height [92.8H]

    standard_name :

    sea_surface_elevation

    units :

    m

    valid_range :

    [-2.0905275344848633, 1.3661648035049438]

    Array Chunk Bytes 637.25 MiB 313.57 kiB Shape (2081, 247, 325) (1, 247, 325) Dask graph 2081 chunks in 7 graph layers Data type float32 numpy.ndarray

  • surface_boundary_layer_thickness
    (MT, Y, X)
    float32
    dask.array<chunksize=(1, 247, 325), meta=np.ndarray>

    long_name :

    bnd.layr.thickness [92.8H]

    units :

    m

    valid_range :

    [0.9318792223930359, 766.9793090820312]

    Array Chunk Bytes 637.25 MiB 313.57 kiB Shape (2081, 247, 325) (1, 247, 325) Dask graph 2081 chunks in 7 graph layers Data type float32 numpy.ndarray

  • u_barotropic_velocity
    (MT, Y, X)
    float32
    dask.array<chunksize=(1, 247, 325), meta=np.ndarray>

    long_name :

    baro. u-vel. [92.8H]

    standard_name :

    barotropic_eastward_sea_water_velocity

    units :

    m/s

    valid_range :

    [-1.0019874572753906, 1.198006272315979]

    Array Chunk Bytes 637.25 MiB 313.57 kiB Shape (2081, 247, 325) (1, 247, 325) Dask graph 2081 chunks in 7 graph layers Data type float32 numpy.ndarray

  • v_barotropic_velocity
    (MT, Y, X)
    float32
    dask.array<chunksize=(1, 247, 325), meta=np.ndarray>

    long_name :

    baro. v-vel. [92.8H]

    standard_name :

    barotropic_northward_sea_water_velocity

    units :

    m/s

    valid_range :

    [-1.2001447677612305, 1.0958094596862793]

    Array Chunk Bytes 637.25 MiB 313.57 kiB Shape (2081, 247, 325) (1, 247, 325) Dask graph 2081 chunks in 7 graph layers Data type float32 numpy.ndarray

  • subset_mask
    (Y, X)
    float64
    1.0 1.0 1.0 1.0 ... 1.0 1.0 1.0 1.0

    array([[1., 1., 1., ..., 1., 1., 1.],
           [1., 1., 1., ..., 1., 1., 1.],
           [1., 1., 1., ..., 1., 1., 1.],
           ...,
           [1., 1., 1., ..., 1., 1., 1.],
           [1., 1., 1., ..., 1., 1., 1.],
           [1., 1., 1., ..., 1., 1., 1.]])

• Indexes: (3)
  • MT
    PandasIndex

    PandasIndex(DatetimeIndex(['2024-05-09 00:00:00', '2024-05-09 01:00:00',
                   '2024-05-09 02:00:00', '2024-05-09 03:00:00',
                   '2024-05-09 04:00:00', '2024-05-09 05:00:00',
                   '2024-05-09 06:00:00', '2024-05-09 07:00:00',
                   '2024-05-09 08:00:00', '2024-05-09 09:00:00',
                   ...
                   '2024-08-05 21:00:00', '2024-08-06 00:00:00',
                   '2024-08-06 03:00:00', '2024-08-06 06:00:00',
                   '2024-08-06 09:00:00', '2024-08-06 12:00:00',
                   '2024-08-06 15:00:00', '2024-08-06 18:00:00',
                   '2024-08-06 21:00:00', '2024-08-07 00:00:00'],
                  dtype='datetime64[ns]', name='MT', length=2081, freq=None))

  • X
    PandasIndex

    PandasIndex(Index([2404, 2405, 2406, 2407, 2408, 2409, 2410, 2411, 2412, 2413,
           ...
           2719, 2720, 2721, 2722, 2723, 2724, 2725, 2726, 2727, 2728],
          dtype='int32', name='X', length=325))

  • Y
    PandasIndex

    PandasIndex(Index([1711, 1712, 1713, 1714, 1715, 1716, 1717, 1718, 1719, 1720,
           ...
           1948, 1949, 1950, 1951, 1952, 1953, 1954, 1955, 1956, 1957],
          dtype='int32', name='Y', length=247))

• Attributes: (6)

  Conventions :

  CF-1.0

  experiment :

  92.8

  history :

  archv2ncdf2d

  institution :

  National Centers for Environmental Prediction

  source :

  HYCOM archive file

  title :

  HYCOM ATLb2.00

f"Dataset size: {ds_subset.nbytes * 1.0e-9} Gb"

'Dataset size: 3.342365484 Gb'

Lets just get the first timestep

now = datetime.datetime.now()
ds_subset = ds_subset.cf.sel(time=now, method="nearest")
ds_subset

<xarray.Dataset> Size: 3MB
Dimensions:                           (Y: 247, X: 325)
Coordinates:
    Date                              float64 8B dask.array<chunksize=(), meta=np.ndarray>
    Latitude                          (Y, X) float32 321kB dask.array<chunksize=(247, 325), meta=np.ndarray>
    Longitude                         (Y, X) float32 321kB dask.array<chunksize=(247, 325), meta=np.ndarray>
    MT                                datetime64[ns] 8B 2024-07-30T16:00:00
  * X                                 (X) int32 1kB 2404 2405 2406 ... 2727 2728
  * Y                                 (Y) int32 988B 1711 1712 ... 1956 1957
Data variables:
    mixed_layer_thickness             (Y, X) float32 321kB dask.array<chunksize=(247, 325), meta=np.ndarray>
    ssh                               (Y, X) float32 321kB dask.array<chunksize=(247, 325), meta=np.ndarray>
    surface_boundary_layer_thickness  (Y, X) float32 321kB dask.array<chunksize=(247, 325), meta=np.ndarray>
    u_barotropic_velocity             (Y, X) float32 321kB dask.array<chunksize=(247, 325), meta=np.ndarray>
    v_barotropic_velocity             (Y, X) float32 321kB dask.array<chunksize=(247, 325), meta=np.ndarray>
    subset_mask                       (Y, X) float64 642kB 1.0 1.0 ... 1.0 1.0
Attributes:
    Conventions:  CF-1.0
    experiment:   92.8
    history:      archv2ncdf2d
    institution:  National Centers for Environmental Prediction
    source:       HYCOM archive file
    title:        HYCOM ATLb2.00

xarray.Dataset

• Dimensions:
  • Y: 247
  • X: 325
• Coordinates: (6)
  • Date
    ()
    float64
    dask.array<chunksize=(), meta=np.ndarray>

    C_format :

    %13.4f

    FORTRAN_format :

    (f13.4)

    long_name :

    date

    units :

    day as %Y%m%d.%f

    Array Chunk Bytes 8 B 8 B Shape () () Dask graph 1 chunks in 3 graph layers Data type float64 numpy.ndarray

  • Latitude
    (Y, X)
    float32
    dask.array<chunksize=(247, 325), meta=np.ndarray>

    standard_name :

    latitude

    units :

    degrees_north

    Array Chunk Bytes 313.57 kiB 313.57 kiB Shape (247, 325) (247, 325) Dask graph 1 chunks in 5 graph layers Data type float32 numpy.ndarray

  • Longitude
    (Y, X)
    float32
    dask.array<chunksize=(247, 325), meta=np.ndarray>

    modulo :

    360 degrees

    standard_name :

    longitude

    units :

    degrees_east

    Array Chunk Bytes 313.57 kiB 313.57 kiB Shape (247, 325) (247, 325) Dask graph 1 chunks in 5 graph layers Data type float32 numpy.ndarray

  • MT
    ()
    datetime64[ns]
    2024-07-30T16:00:00

    axis :

    T

    long_name :

    time

    array('2024-07-30T16:00:00.000000000', dtype='datetime64[ns]')

  • X
    (X)
    int32
    2404 2405 2406 ... 2726 2727 2728

    axis :

    X

    point_spacing :

    even

    array([2404, 2405, 2406, ..., 2726, 2727, 2728], dtype=int32)

  • Y
    (Y)
    int32
    1711 1712 1713 ... 1955 1956 1957

    axis :

    Y

    point_spacing :

    even

    array([1711, 1712, 1713, ..., 1955, 1956, 1957], dtype=int32)

• Data variables: (6)
  • mixed_layer_thickness
    (Y, X)
    float32
    dask.array<chunksize=(247, 325), meta=np.ndarray>

    long_name :

    mix.layr.thickness [92.8H]

    standard_name :

    ocean_mixed_layer_thickness

    units :

    m

    valid_range :

    [0.5049223899841309, 1273.3824462890625]

    Array Chunk Bytes 313.57 kiB 313.57 kiB Shape (247, 325) (247, 325) Dask graph 1 chunks in 8 graph layers Data type float32 numpy.ndarray

  • ssh
    (Y, X)
    float32
    dask.array<chunksize=(247, 325), meta=np.ndarray>

    long_name :

    sea surf. height [92.8H]

    standard_name :

    sea_surface_elevation

    units :

    m

    valid_range :

    [-2.0905275344848633, 1.3661648035049438]

    Array Chunk Bytes 313.57 kiB 313.57 kiB Shape (247, 325) (247, 325) Dask graph 1 chunks in 8 graph layers Data type float32 numpy.ndarray

  • surface_boundary_layer_thickness
    (Y, X)
    float32
    dask.array<chunksize=(247, 325), meta=np.ndarray>

    long_name :

    bnd.layr.thickness [92.8H]

    units :

    m

    valid_range :

    [0.9318792223930359, 766.9793090820312]

    Array Chunk Bytes 313.57 kiB 313.57 kiB Shape (247, 325) (247, 325) Dask graph 1 chunks in 8 graph layers Data type float32 numpy.ndarray

  • u_barotropic_velocity
    (Y, X)
    float32
    dask.array<chunksize=(247, 325), meta=np.ndarray>

    long_name :

    baro. u-vel. [92.8H]

    standard_name :

    barotropic_eastward_sea_water_velocity

    units :

    m/s

    valid_range :

    [-1.0019874572753906, 1.198006272315979]

    Array Chunk Bytes 313.57 kiB 313.57 kiB Shape (247, 325) (247, 325) Dask graph 1 chunks in 8 graph layers Data type float32 numpy.ndarray

  • v_barotropic_velocity
    (Y, X)
    float32
    dask.array<chunksize=(247, 325), meta=np.ndarray>

    long_name :

    baro. v-vel. [92.8H]

    standard_name :

    barotropic_northward_sea_water_velocity

    units :

    m/s

    valid_range :

    [-1.2001447677612305, 1.0958094596862793]

    Array Chunk Bytes 313.57 kiB 313.57 kiB Shape (247, 325) (247, 325) Dask graph 1 chunks in 8 graph layers Data type float32 numpy.ndarray

  • subset_mask
    (Y, X)
    float64
    1.0 1.0 1.0 1.0 ... 1.0 1.0 1.0 1.0

    array([[1., 1., 1., ..., 1., 1., 1.],
           [1., 1., 1., ..., 1., 1., 1.],
           [1., 1., 1., ..., 1., 1., 1.],
           ...,
           [1., 1., 1., ..., 1., 1., 1.],
           [1., 1., 1., ..., 1., 1., 1.],
           [1., 1., 1., ..., 1., 1., 1.]])

• Indexes: (2)
  • X
    PandasIndex

    PandasIndex(Index([2404, 2405, 2406, 2407, 2408, 2409, 2410, 2411, 2412, 2413,
           ...
           2719, 2720, 2721, 2722, 2723, 2724, 2725, 2726, 2727, 2728],
          dtype='int32', name='X', length=325))

  • Y
    PandasIndex

    PandasIndex(Index([1711, 1712, 1713, 1714, 1715, 1716, 1717, 1718, 1719, 1720,
           ...
           1948, 1949, 1950, 1951, 1952, 1953, 1954, 1955, 1956, 1957],
          dtype='int32', name='Y', length=247))

• Attributes: (6)

  Conventions :

  CF-1.0

  experiment :

  92.8

  history :

  archv2ncdf2d

  institution :

  National Centers for Environmental Prediction

  source :

  HYCOM archive file

  title :

  HYCOM ATLb2.00

f"Dataset size: {ds_subset.nbytes * 1.0e-6} Mb"

'Dataset size: 2.892204 Mb'

Lets plot the sea surface height

ds_subset.ssh.cf.plot(x='longitude', y='latitude', cmap='viridis')

<matplotlib.collections.QuadMesh at 0x36247d4e0>