Acceso a ECCOv4 por medio de EarthData in the Cloud de la NASA

Este tutorial demuestra acceso al producto ECCOv4 de un modelo numerico global. Para mayor informacion del producto de ECCO, puede ir aqui.

Requisitos

1. Tener una cuenta de Earth Data Login por medio de la NASA.

2. Tener un Token valido.

Tambien se puede utilizar el metodo de Nombre de Usuario / Contrasena descrito en el tutorial de Autenticacion

Objectivos

Utilizar pydap para demostrar

• Acceso to archivos cientificos de la NASA por medio del uso de tokens y EarthData como metodo de autenticacion.

Algunas variables de interes son:

• Native grid

• Temperature and Salinity

Autor: Miguel Jimenez-Urias, ‘24

import matplotlib.pyplot as plt
import numpy as np
from pydap.net import create_session
from pydap.client import open_url
import xarray as xr
import pydap

print("xarray version: ", xr.__version__)
print("pydap version: ", pydap.__version__)

xarray version:  2025.4.0
pydap version:  3.5.5

Acceso a EARTHDATA

El listado de las variables de este producto pueden encontrarse aqui. En este caso, accederemos a las variables en su malla original, el LL90.

Grid_url = 'https://opendap.earthdata.nasa.gov/providers/POCLOUD/collections/ECCO%20Geometry%20Parameters%20for%20the%20Lat-Lon-Cap%2090%20(llc90)%20Native%20Model%20Grid%20(Version%204%20Release%204)/granules/GRID_GEOMETRY_ECCO_V4r4_native_llc0090'

Autenticacion via .netrc

Las credenciales son recuperadas automaticamente por pydap.

my_session = create_session()

Alternativa: El use the tokens

session_extra = {"token": "YourToken"}

# initialize a requests.session object with the token headers. All handled by pydap.
my_session = create_session(session_kwargs=session_extra)

Accesso a los metadatos solamente, por medio de pydap

pydap aprovecha el protocolo de OPeNDAP, el cual permite la separacion de los metadatos de los valores numericos. Esto permite una inspeccion remota de los datasets.

ds_grid = open_url(Grid_url, session=my_session, protocol="dap4")

ds_grid.tree()

.GRID_GEOMETRY_ECCO_V4r4_native_llc0090.nc
├──dxC
├──PHrefF
├──XG
├──dyG
├──rA
├──hFacS
├──Zp1
├──Zl
├──rAw
├──dxG
├──maskW
├──YC
├──XC
├──maskS
├──YG
├──hFacC
├──drC
├──drF
├──XC_bnds
├──Zu
├──Z_bnds
├──YC_bnds
├──PHrefC
├──rAs
├──Depth
├──dyC
├──SN
├──rAz
├──maskC
├──CS
├──hFacW
├──Z
├──i
├──i_g
├──j
├──j_g
├──k
├──k_l
├──k_p1
├──k_u
├──nb
├──nv
└──tile

Note

PyDAP accesa solo a los metadatos de archivo en el servidor de OPeNDAP, y ningun arreglo numerico ha sido descargardo hasta este punto!

Para descargar los arreglos numericos hay que indexar la variable

Esto es, de la siguiente manera:

# this fetches remote data into a pydap object container
pydap_Array = dataset[<VarName>][:]

donde <VarName> es el nombre de una de las variables. El producto sera una representation “en memoria” de tipo pydap.model.BaseType, el cual permite el acceso a los arreglos de numpy.

Para extraer los valores de la variable remota, hay que ejecutar el siguiente comando

# The `.data` command allows direct access to the Numpy array (e.g. for manipulation)
pydap_Array.data

# lets download some data
Depth = ds_grid['Depth'][:]
print(type(Depth))

<class 'pydap.model.BaseType'>

Depth.attributes

{'_FillValue': 9.969209968e+36,
 'long_name': 'model seafloor depth below ocean surface at rest',
 'units': 'm',
 'coordinate': 'XC YC',
 'coverage_content_type': 'modelResult',
 'standard_name': 'sea_floor_depth_below_geoid',
 'comment': "Model sea surface height (SSH) of 0m corresponds to an ocean surface at rest relative to the geoid. Depth corresponds to seafloor depth below geoid. Note: the MITgcm used by ECCO V4r4 implements 'partial cells' so the actual model seafloor depth may differ from the seafloor depth provided by the input bathymetry file.",
 'coordinates': 'YC XC',
 'origname': 'Depth',
 'fullnamepath': '/Depth',
 'Maps': (),
 'checksum': array([950678499], dtype=uint32)}

Depth.shape, Depth.dims

((13, 90, 90), ['/tile', '/j', '/i'])

Visualizando el fondo oceanico Depth en la malla original del modelo

En este caso, el producto ECCO esta definido en una malla con topologia de un Cubo Esferico. De esta manera, la malla horizontal contiene una dimension extra llamada: tile o face. A continuacion visualizamos la variable en su topologia original

Variable = [Depth[i].data for i in range(13)]
clevels =  np.linspace(0, 6000, 100)
cMap = 'Greys_r'

fig, axes = plt.subplots(nrows=5, ncols=5, figsize=(8, 8), gridspec_kw={'hspace':0.01, 'wspace':0.01})
AXES = [
    axes[4, 0], axes[3, 0], axes[2, 0], axes[4, 1], axes[3, 1], axes[2, 1],
    axes[1, 1], 
    axes[1, 2], axes[1, 3], axes[1, 4], axes[0, 2], axes[0, 3], axes[0, 4],
]
for i in range(len(AXES)):
    AXES[i].contourf(Variable[i], clevels, cmap=cMap)

for ax in np.ravel(axes):
    ax.axis('off')
    plt.setp(ax.get_xticklabels(), visible=False)
    plt.setp(ax.get_yticklabels(), visible=False)

plt.show()

Fig. 1. La variable Depth visualizada en una malla horitonzal. Tiles con valores 0-5 tienen un ordenamiento de valores (indiciales) C-ordering, mientras que cualquier arreglo numerico tiles 7-13 siguen el F-ordering. Cualquier arreglo numerico en el arctic cap, tiene un comportamiento de una coordenada polar.

Visualizacion con topologia corregida

fig, axes = plt.subplots(nrows=4, ncols=4, figsize=(8, 8), gridspec_kw={'hspace':0.01, 'wspace':0.01})
AXES_NR = [
    axes[3, 0], axes[2, 0], axes[1, 0], axes[3, 1], axes[2, 1], axes[1, 1],
]
AXES_CAP = [axes[0, 0]]
AXES_R = [
    axes[1, 2], axes[2, 2], axes[3, 2], axes[1, 3], axes[2, 3], axes[3, 3],
]
for i in range(len(AXES_NR)):
    AXES_NR[i].contourf(Variable[i], clevels, cmap=cMap)

for i in range(len(AXES_CAP)):
    AXES_CAP[i].contourf(Variable[6].transpose()[:, ::-1], clevels, cmap=cMap)

for i in range(len(AXES_R)):
    AXES_R[i].contourf(Variable[7+i].transpose()[::-1, :], clevels, cmap=cMap)

for ax in np.ravel(axes):
    ax.axis('off')
    plt.setp(ax.get_xticklabels(), visible=False)
    plt.setp(ax.get_yticklabels(), visible=False)

plt.show()

Fig. 2. Visualizacion de la variable Depth en una malla horizontal que approxima una malla horizontal uniforme con ejes lat-lon. Sin embargo, cualquier arreglo numerico en el arctic cap, continua en una malla que approxima una malla en coordenadas polares

Utilizando xarray

pydap se puede llamar dentro de xarray para acceder a los archivos expuestos mediante el servidor de OPeNDAP. En particular, xarray permite aggregar multiples archivos remotos de opendap. A continuacion demostramos un pequeno ejemplo.

Un URL individual –> un archivo remoto

Below we access remote Temperature and Salinity ECCO data with xarray via (internally) pydap.

baseURL = 'https://opendap.earthdata.nasa.gov/providers/POCLOUD/collections/'
Temp_Salt = "ECCO%20Ocean%20Temperature%20and%20Salinity%20-%20Monthly%20Mean%20llc90%20Grid%20(Version%204%20Release%204)/granules/OCEAN_TEMPERATURE_SALINITY_mon_mean_"
year = '2017-'
month = '01'
end_ = '_ECCO_V4r4_native_llc0090'
Temp_2017 = baseURL + Temp_Salt +year  + month + end_

# Convertimos el URL en un DAP4 opendap url
Temp_2017 = Temp_2017.replace("https", "dap4")
Temp_2017

'dap4://opendap.earthdata.nasa.gov/providers/POCLOUD/collections/ECCO%20Ocean%20Temperature%20and%20Salinity%20-%20Monthly%20Mean%20llc90%20Grid%20(Version%204%20Release%204)/granules/OCEAN_TEMPERATURE_SALINITY_mon_mean_2017-01_ECCO_V4r4_native_llc0090'

pyds = open_url(Temp_2017, session=my_session)
pyds.tree()

.OCEAN_TEMPERATURE_SALINITY_mon_mean_2017-01_ECCO_V4r4_native_llc0090.nc
├──XG
├──Zp1
├──Zl
├──YC
├──XC
├──SALT
├──YG
├──XC_bnds
├──Zu
├──THETA
├──Z_bnds
├──YC_bnds
├──time_bnds
├──Z
├──i
├──i_g
├──j
├──j_g
├──k
├──k_l
├──k_p1
├──k_u
├──nb
├──nv
├──tile
└──time

Para evitar descargas variables que no sean necesariamente de interest, se le puede instruir al Servidor Hyrax que variables requiere uno.

Para esto utilizaremos las Expresiones de Restriccion (CE, por sus siglas en ingles). Este metodo nos ayudara a construir datasets mucho mas simples, y evitar descargar N veces, informacion que no es requerida.

A continuacion demostramos el caso de solo requerir la variable THETA y sus dimensiones.

dims = pyds['/THETA'].dims
Vars = ['/THETA'] + dims

# Below construct Contraint Expression
CE = "?dap4.ce="+(";").join(Vars)
print("Constraint Expression: ", CE)

Constraint Expression:  ?dap4.ce=/THETA;/time;/k;/tile;/j;/i

Temp_2017 = baseURL + Temp_Salt +year  + month + end_ + CE
Temp_2017

'https://opendap.earthdata.nasa.gov/providers/POCLOUD/collections/ECCO%20Ocean%20Temperature%20and%20Salinity%20-%20Monthly%20Mean%20llc90%20Grid%20(Version%204%20Release%204)/granules/OCEAN_TEMPERATURE_SALINITY_mon_mean_2017-01_ECCO_V4r4_native_llc0090?dap4.ce=/THETA;/time;/k;/tile;/j;/i'

Importante

Ahora aplicamos la CE a todos los possibles URLs the nuestro dataset.

Temp_2017 = [baseURL.replace("https", "dap4") + Temp_Salt + year + f'{i:02}' + end_ + CE for i in range(1, 13)]
Temp_2017[:3]

['dap4://opendap.earthdata.nasa.gov/providers/POCLOUD/collections/ECCO%20Ocean%20Temperature%20and%20Salinity%20-%20Monthly%20Mean%20llc90%20Grid%20(Version%204%20Release%204)/granules/OCEAN_TEMPERATURE_SALINITY_mon_mean_2017-01_ECCO_V4r4_native_llc0090?dap4.ce=/THETA;/time;/k;/tile;/j;/i',
 'dap4://opendap.earthdata.nasa.gov/providers/POCLOUD/collections/ECCO%20Ocean%20Temperature%20and%20Salinity%20-%20Monthly%20Mean%20llc90%20Grid%20(Version%204%20Release%204)/granules/OCEAN_TEMPERATURE_SALINITY_mon_mean_2017-02_ECCO_V4r4_native_llc0090?dap4.ce=/THETA;/time;/k;/tile;/j;/i',
 'dap4://opendap.earthdata.nasa.gov/providers/POCLOUD/collections/ECCO%20Ocean%20Temperature%20and%20Salinity%20-%20Monthly%20Mean%20llc90%20Grid%20(Version%204%20Release%204)/granules/OCEAN_TEMPERATURE_SALINITY_mon_mean_2017-03_ECCO_V4r4_native_llc0090?dap4.ce=/THETA;/time;/k;/tile;/j;/i']

Debajo inicializaremos una session que persista, cada vez que descargue datos de servidores de OPeNDAP. En este caso utilizaremost la libraria requests_cache. El comportamiento es casi identico excepto la siguiente celda:

cache_session = create_session(use_cache=True) # caching the session
cache_session.cache.clear() # clear all cache to demonstrate the behavior

Tip

Ahora se puede usar el metodo consolidated_metadata para que pydap descarge todos los metadatos, y reusarlos. Para usar consolidated_metadata se requiere installar una version de pydap >= 3.5.5.

from pydap.client import consolidate_metadata

%%time
consolidate_metadata(Temp_2017, cache_session)

datacube has dimensions {'j[0:1:89]', 'i[0:1:89]', 'tile[0:1:12]', 'time[0:1:0]', 'k[0:1:49]'}

CPU times: user 367 ms, sys: 207 ms, total: 575 ms
Wall time: 9.69 s

%%time
theta_salt_ds = xr.open_mfdataset(
    Temp_2017, 
    engine='pydap',
    session=cache_session, 
    parallel=True, 
    combine='nested', 
    concat_dim='time',
    decode_times=False
)

CPU times: user 392 ms, sys: 235 ms, total: 627 ms
Wall time: 468 ms

theta_salt_ds

<xarray.Dataset> Size: 253MB
Dimensions:  (time: 12, k: 50, tile: 13, j: 90, i: 90)
Coordinates:
  * i        (i) int32 360B 0 1 2 3 4 5 6 7 8 9 ... 81 82 83 84 85 86 87 88 89
  * j        (j) int32 360B 0 1 2 3 4 5 6 7 8 9 ... 81 82 83 84 85 86 87 88 89
  * k        (k) int32 200B 0 1 2 3 4 5 6 7 8 9 ... 41 42 43 44 45 46 47 48 49
  * tile     (tile) int32 52B 0 1 2 3 4 5 6 7 8 9 10 11 12
  * time     (time) int32 48B 219528 219528 219528 ... 219528 219528 219528
Data variables:
    THETA    (time, k, tile, j, i) float32 253MB dask.array<chunksize=(1, 50, 13, 90, 90), meta=np.ndarray>
Attributes: (12/62)
    acknowledgement:                 This research was carried out by the Jet...
    author:                          Ian Fenty and Ou Wang
    cdm_data_type:                   Grid
    comment:                         Fields provided on the curvilinear lat-l...
    Conventions:                     CF-1.8, ACDD-1.3
    coordinates_comment:             Note: the global 'coordinates' attribute...
    ...                              ...
    time_coverage_duration:          P1M
    time_coverage_end:               2017-12-31T12:00:00
    time_coverage_resolution:        P1M
    time_coverage_start:             2017-12-01T00:00:00
    title:                           ECCO Ocean Temperature and Salinity - Mo...
    uuid:                            fba67542-4181-11eb-97c8-0cc47a3f47df

xarray.Dataset

• Dimensions:
  • time: 12
  • k: 50
  • tile: 13
  • j: 90
  • i: 90
• Coordinates: (5)
  • i
    (i)
    int32
    0 1 2 3 4 5 6 ... 84 85 86 87 88 89

    axis :

    X

    long_name :

    grid index in x for variables at tracer and 'v' locations

    swap_dim :

    XC

    comment :

    In the Arakawa C-grid system, tracer (e.g., THETA) and 'v' variables (e.g., VVEL) have the same x coordinate on the model grid.

    coverage_content_type :

    coordinate

    origname :

    i

    fullnamepath :

    /i

    Maps :

    ()

    array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17,
           18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35,
           36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53,
           54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71,
           72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89],
          dtype=int32)

  • j
    (j)
    int32
    0 1 2 3 4 5 6 ... 84 85 86 87 88 89

    axis :

    Y

    long_name :

    grid index in y for variables at tracer and 'u' locations

    swap_dim :

    YC

    comment :

    In the Arakawa C-grid system, tracer (e.g., THETA) and 'u' variables (e.g., UVEL) have the same y coordinate on the model grid.

    coverage_content_type :

    coordinate

    origname :

    j

    fullnamepath :

    /j

    Maps :

    ()

    array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17,
           18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35,
           36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53,
           54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71,
           72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89],
          dtype=int32)

  • k
    (k)
    int32
    0 1 2 3 4 5 6 ... 44 45 46 47 48 49

    axis :

    Z

    long_name :

    grid index in z for tracer variables

    swap_dim :

    Z

    coverage_content_type :

    coordinate

    origname :

    k

    fullnamepath :

    /k

    Maps :

    ()

    array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17,
           18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35,
           36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49], dtype=int32)

  • tile
    (tile)
    int32
    0 1 2 3 4 5 6 7 8 9 10 11 12

    long_name :

    lat-lon-cap tile index

    comment :

    The ECCO V4 horizontal model grid is divided into 13 tiles of 90x90 cells for convenience.

    coverage_content_type :

    coordinate

    origname :

    tile

    fullnamepath :

    /tile

    Maps :

    ()

    array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12], dtype=int32)

  • time
    (time)
    int32
    219528 219528 ... 219528 219528

    long_name :

    center time of averaging period

    axis :

    T

    bounds :

    time_bnds

    coverage_content_type :

    coordinate

    standard_name :

    time

    units :

    hours since 1992-01-01T12:00:00

    calendar :

    proleptic_gregorian

    origname :

    time

    fullnamepath :

    /time

    Maps :

    ()

    array([219528, 219528, 219528, 219528, 219528, 219528, 219528, 219528, 219528,
           219528, 219528, 219528], dtype=int32)

• Data variables: (1)
  • THETA
    (time, k, tile, j, i)
    float32
    dask.array<chunksize=(1, 50, 13, 90, 90), meta=np.ndarray>

    long_name :

    Potential temperature

    units :

    degree_C

    coverage_content_type :

    modelResult

    standard_name :

    sea_water_potential_temperature

    comment :

    Sea water potential temperature is the temperature a parcel of sea water would have if moved adiabatically to sea level pressure. Note: the equation of state is a modified UNESCO formula by Jackett and McDougall (1995), which uses the model variable potential temperature as input assuming a horizontally and temporally constant pressure of $p_0=-g ho_{0} z$.

    valid_min :

    -2.2909388542175293

    valid_max :

    36.032955169677734

    origname :

    THETA

    fullnamepath :

    /THETA

    Maps :

    ()

    Array Chunk Bytes 241.01 MiB 20.08 MiB Shape (12, 50, 13, 90, 90) (1, 50, 13, 90, 90) Dask graph 12 chunks in 25 graph layers Data type float32 numpy.ndarray

• Indexes: (5)
  • i
    PandasIndex

    PandasIndex(Index([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17,
           18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35,
           36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53,
           54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71,
           72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89],
          dtype='int32', name='i'))

  • j
    PandasIndex

    PandasIndex(Index([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17,
           18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35,
           36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53,
           54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71,
           72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89],
          dtype='int32', name='j'))

  • k
    PandasIndex

    PandasIndex(Index([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17,
           18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35,
           36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49],
          dtype='int32', name='k'))

  • tile
    PandasIndex

    PandasIndex(Index([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12], dtype='int32', name='tile'))

  • time
    PandasIndex

    PandasIndex(Index([219528, 219528, 219528, 219528, 219528, 219528, 219528, 219528, 219528,
           219528, 219528, 219528],
          dtype='int32', name='time'))

• Attributes: (62)

  acknowledgement :

  This research was carried out by the Jet Propulsion Laboratory, managed by the California Institute of Technology under a contract with the National Aeronautics and Space Administration.

  author :

  Ian Fenty and Ou Wang

  cdm_data_type :

  Grid

  comment :

  Fields provided on the curvilinear lat-lon-cap 90 (llc90) native grid used in the ECCO model.

  Conventions :

  CF-1.8, ACDD-1.3

  coordinates_comment :

  Note: the global 'coordinates' attribute describes auxillary coordinates.

  creator_email :

  ecco-group@mit.edu

  creator_institution :

  NASA Jet Propulsion Laboratory (JPL)

  creator_name :

  ECCO Consortium

  creator_type :

  group

  creator_url :

  https://ecco-group.org

  date_created :

  2020-12-18T14:40:09

  date_issued :

  2020-12-18T14:40:09

  date_metadata_modified :

  2021-03-15T21:56:26

  date_modified :

  2021-03-15T21:56:26

  geospatial_bounds_crs :

  EPSG:4326

  geospatial_lat_max :

  90.0

  geospatial_lat_min :

  -90.0

  geospatial_lat_resolution :

  variable

  geospatial_lat_units :

  degrees_north

  geospatial_lon_max :

  180.0

  geospatial_lon_min :

  -180.0

  geospatial_lon_resolution :

  variable

  geospatial_lon_units :

  degrees_east

  geospatial_vertical_max :

  0.0

  geospatial_vertical_min :

  -6134.5

  geospatial_vertical_positive :

  up

  geospatial_vertical_resolution :

  variable

  geospatial_vertical_units :

  meter

  history :

  Inaugural release of an ECCO Central Estimate solution to PO.DAAC

  id :

  10.5067/ECL5M-OTS44

  institution :

  NASA Jet Propulsion Laboratory (JPL)

  instrument_vocabulary :

  GCMD instrument keywords

  keywords :

  EARTH SCIENCE SERVICES > MODELS > EARTH SCIENCE REANALYSES/ASSIMILATION MODELS, EARTH SCIENCE > OCEANS > SALINITY/DENSITY > SALINITY, EARTH SCIENCE > OCEANS > OCEAN TEMPERATURE > POTENTIAL TEMPERATURE

  keywords_vocabulary :

  NASA Global Change Master Directory (GCMD) Science Keywords

  license :

  Public Domain

  metadata_link :

  https://cmr.earthdata.nasa.gov/search/collections.umm_json?ShortName=ECCO_L4_TEMP_SALINITY_LLC0090GRID_MONTHLY_V4R4

  naming_authority :

  gov.nasa.jpl

  platform :

  ERS-1/2, TOPEX/Poseidon, Geosat Follow-On (GFO), ENVISAT, Jason-1, Jason-2, CryoSat-2, SARAL/AltiKa, Jason-3, AVHRR, Aquarius, SSM/I, SSMIS, GRACE, DTU17MDT, Argo, WOCE, GO-SHIP, MEOP, Ice Tethered Profilers (ITP)

  platform_vocabulary :

  GCMD platform keywords

  processing_level :

  L4

  product_name :

  OCEAN_TEMPERATURE_SALINITY_mon_mean_2017-12_ECCO_V4r4_native_llc0090.nc

  product_time_coverage_end :

  2018-01-01T00:00:00

  product_time_coverage_start :

  1992-01-01T12:00:00

  product_version :

  Version 4, Release 4

  program :

  NASA Physical Oceanography, Cryosphere, Modeling, Analysis, and Prediction (MAP)

  project :

  Estimating the Circulation and Climate of the Ocean (ECCO)

  publisher_email :

  podaac@podaac.jpl.nasa.gov

  publisher_institution :

  PO.DAAC

  publisher_name :

  Physical Oceanography Distributed Active Archive Center (PO.DAAC)

  publisher_type :

  institution

  publisher_url :

  https://podaac.jpl.nasa.gov

  references :

  ECCO Consortium, Fukumori, I., Wang, O., Fenty, I., Forget, G., Heimbach, P., & Ponte, R. M. 2020. Synopsis of the ECCO Central Production Global Ocean and Sea-Ice State Estimate (Version 4 Release 4). doi:10.5281/zenodo.3765928

  source :

  The ECCO V4r4 state estimate was produced by fitting a free-running solution of the MITgcm (checkpoint 66g) to satellite and in situ observational data in a least squares sense using the adjoint method

  standard_name_vocabulary :

  NetCDF Climate and Forecast (CF) Metadata Convention

  summary :

  This dataset provides monthly-averaged ocean potential temperature and salinity on the lat-lon-cap 90 (llc90) native model grid from the ECCO Version 4 Release 4 (V4r4) ocean and sea-ice state estimate. Estimating the Circulation and Climate of the Ocean (ECCO) state estimates are dynamically and kinematically-consistent reconstructions of the three-dimensional, time-evolving ocean, sea-ice, and surface atmospheric states. ECCO V4r4 is a free-running solution of a global, nominally 1-degree configuration of the MIT general circulation model (MITgcm) that has been fit to observations in a least-squares sense. Observational data constraints used in V4r4 include sea surface height (SSH) from satellite altimeters [ERS-1/2, TOPEX/Poseidon, GFO, ENVISAT, Jason-1,2,3, CryoSat-2, and SARAL/AltiKa]; sea surface temperature (SST) from satellite radiometers [AVHRR], sea surface salinity (SSS) from the Aquarius satellite radiometer/scatterometer, ocean bottom pressure (OBP) from the GRACE satellite gravimeter; sea-ice concentration from satellite radiometers [SSM/I and SSMIS], and in-situ ocean temperature and salinity measured with conductivity-temperature-depth (CTD) sensors and expendable bathythermographs (XBTs) from several programs [e.g., WOCE, GO-SHIP, Argo, and others] and platforms [e.g., research vessels, gliders, moorings, ice-tethered profilers, and instrumented pinnipeds]. V4r4 covers the period 1992-01-01T12:00:00 to 2018-01-01T00:00:00.

  time_coverage_duration :

  P1M

  time_coverage_end :

  2017-12-31T12:00:00

  time_coverage_resolution :

  P1M

  time_coverage_start :

  2017-12-01T00:00:00

  title :

  ECCO Ocean Temperature and Salinity - Monthly Mean llc90 Grid (Version 4 Release 4)

  uuid :

  fba67542-4181-11eb-97c8-0cc47a3f47df

Finalmente, visualizamos THETA

Variable = [theta_salt_ds['THETA'][0, 0, i, :, :] for i in range(13)]
clevels = np.linspace(-5, 30, 100)
cMap='RdBu_r'

fig, axes = plt.subplots(nrows=4, ncols=4, figsize=(8, 8), gridspec_kw={'hspace':0.01, 'wspace':0.01})
AXES_NR = [
    axes[3, 0], axes[2, 0], axes[1, 0], axes[3, 1], axes[2, 1], axes[1, 1],
]
AXES_CAP = [axes[0, 0]]
AXES_R = [
    axes[1, 2], axes[2, 2], axes[3, 2], axes[1, 3], axes[2, 3], axes[3, 3],
]
for i in range(len(AXES_NR)):
    AXES_NR[i].contourf(Variable[i], clevels, cmap=cMap)

for i in range(len(AXES_CAP)):
    AXES_CAP[i].contourf(Variable[6].transpose()[:, ::-1], clevels, cmap=cMap)

for i in range(len(AXES_R)):
    AXES_R[i].contourf(Variable[7+i].transpose()[::-1, :], clevels, cmap=cMap)

for ax in np.ravel(axes):
    ax.axis('off')
    plt.setp(ax.get_xticklabels(), visible=False)
    plt.setp(ax.get_yticklabels(), visible=False)

plt.show()

Fig. 3. Visualizacion de la variable Surface temperature, similar al metodo utilizado en la Figura 2.