Plotting E3SMv2 Output with UXarray

This notebook showcases the Department of Energy (DOE)’s global climate model - the Energy Exascale Earth System Model (E3SM)’s output, that is on an unstructured grid, with the use of UXarray’s plotting functions, including the basic .plot(), .rasterize(), and projections.

• E3SM Grid (ne30pg2)

Variables showcased:

• TS (Surface Temperature)

• FLUT (Upwelling longwave flux at top of model)

• QFLX (Precipitation Flux at Surface)

Imports

This notebook requires the following packages to be installed in the notebook environment. UXarray comes with dependency packages including xarray, matplotlib, cartopy, and geoviews.

mamba install -c conda-forge uxarray

import uxarray as ux
import cartopy.crs as ccrs
import geoviews as gv
import geoviews.feature as gf
import warnings

Data

Data loaded in this notebook is the simulated output from the Department of Energy (DOE) Energy Exascale Earth System Model (E3SM) version 2. The case is set up as an atmosphere-only (AMIP) simulation with present-day control forcing (F2010) at a 1-degree horizontal resolution (ne30pg2), where sea surface temperatures and sea ice set as default as in the E3SMv2 model. The case is run for 6 years.

In the following section, this notebook will showcase the Decemeber output in the 6th simulated year.

data_path = "/glade/campaign/cisl/vast/uxarray/data/e3sm_keeling/ENSO_ctl_1std/unstructured/20231220.F2010.ENSO_ctl.lagreg.ne30pg2_EC30to60E2r2.keeling.eam.h0.0006-12.nc"
grid_file = (
    "/glade/campaign/cisl/vast/uxarray/data/e3sm_keeling/E3SM_grid/ne30pg2_grd.nc"
)
uxds_e3sm = ux.open_dataset(grid_file, data_path)

Check Grid Property

uxgrid = uxds_e3sm.uxgrid
uxgrid

<uxarray.Grid>
Original Grid Type: Scrip
Grid Dimensions:
  * n_node: 21727
  * n_edge: 43326
  * n_face: 21600
  * n_max_face_nodes: 4
  * two: 2
  * n_nodes_per_face: (21600,)
Grid Coordinates (Spherical):
  * node_lon: (21727,)
  * node_lat: (21727,)
  * face_lon: (21600,)
  * face_lat: (21600,)
Grid Coordinates (Cartesian):
Grid Connectivity Variables:
  * face_node_connectivity: (21600, 4)
  * edge_node_connectivity: (43326, 2)
Grid Descriptor Variables:
  * n_nodes_per_face: (21600,)

uxarray.Grid

• Dimensions:
  • n_node: 21727
  • n_face: 21600
  • n_max_face_nodes: 4
  • n_edge: 43326
  • two: 2
• Spherical Coordinates: (4)
  • node_lon
    (n_node)
    float64
    0.0 0.0 0.0 0.0 ... 0.0 0.0 0.0 0.0

    standard_name :

    longitude

    long name :

    Longitude of the corner nodes of each face

    units :

    degrees_east

    array([0., 0., 0., ..., 0., 0., 0.])

  • node_lat
    (n_node)
    float64
    -88.5 -87.0 -85.5 ... 87.0 88.5

    standard_name :

    latitude

    long name :

    Latitude of the corner nodes of each face

    units :

    degrees_north

    array([-88.5, -87. , -85.5, ...,  85.5,  87. ,  88.5])

  • face_lon
    (n_face)
    float64
    -44.26 -42.76 ... 136.5 135.0

    standard_name :

    longitude

    long name :

    Longitude of the center of each face

    units :

    degrees_east

    array([-44.25928396, -42.75932668, -44.25864811, ..., 133.49916029,
           136.50083971, 135.        ])

  • face_lat
    (n_face)
    float64
    -34.91 -35.58 ... 36.66 35.97

    standard_name :

    latitude

    long name :

    Latitude of the center of each face

    units :

    degrees_north

    array([-34.91143916, -35.58196515, -33.51497835, ...,  36.66411307,
            36.66411307,  35.96577012])

• Cartesian Coordinates: (0)
• Connectivity: (2)
  • face_node_connectivity
    (n_face, n_max_face_nodes)
    int64
    18937 19027 19029 ... 8191 8224

    cf_role :

    face_node_connectivity

    long name :

    Maps every face to its corner nodes.

    start_index :

    0

    _FillValue :

    -9223372036854775808

    array([[18937, 19027, 19029, 18938],
           [19027, 19126, 19127, 19029],
           [18938, 19029, 19031, 18939],
           ...,
           [ 8038,  8002,  8101,  8192],
           [ 8287,  8192,  8224,  8382],
           [ 8192,  8101,  8191,  8224]])

  • edge_node_connectivity
    (n_edge, two)
    int64
    0 1 0 2817 ... 21725 21725 21726

    cf_role :

    edge_node_connectivity

    long name :

    Maps every edge to the two nodes that it connects.

    start_index :

    0

    inverse_indices :

    [37898 38075 37900 ... 16230 16408 16409]

    fill_value_mask :

    [False False False ... False False False]

    array([[    0,     1],
           [    0,  2817],
           [    0,  2819],
           ...,
           [21723, 21724],
           [21724, 21725],
           [21725, 21726]])

• Descriptors: (1)
  • n_nodes_per_face
    (n_face)
    int64
    4 4 4 4 4 4 4 4 ... 4 4 4 4 4 4 4 4

    cf_role :

    n_nodes_per_face

    long name :

    Number of nodes per face

    array([4, 4, 4, ..., 4, 4, 4])

• Attributes: (0)

E3SM ne30pg2 (1 degree) Grid

%%time
uxgrid.plot(
    height=500,
    width=750,
    title="Grid Plot Accessor",
    backend="matplotlib",
    projection=ccrs.Orthographic(central_latitude=0, central_longitude=0),
)

CPU times: user 947 ms, sys: 51.9 ms, total: 999 ms
Wall time: 1e+03 ms

Surface Temperature

Plotting requires the UXDataAarray to have the spatial geometries as the only dimension (n_faces, n_nodes, or n_edges) and therefore .isel(time=0) is called.

To set the limits of the colorbar, include the arugment clim=(lowlim,uplim). By calling the default .plot, the plots are rasterized.

Brighter regions imply warmer surface temperature.

%%time
uxds_e3sm.TS.isel(time=0).plot(
    height=200,
    width=450,
    clim=(230, 320),
    clabel="TS (K)",
    backend="matplotlib",
    cmap="plasma",
    projection=ccrs.Robinson(),
    features=["borders", "coastline"],
    title="Surface Temperature",
)

CPU times: user 60.1 ms, sys: 4.29 ms, total: 64.4 ms
Wall time: 63.6 ms

Upwelling longwave flux at top of model (FLUT)

Regions with brighter color indicate stronger upwelling longwave flux.

%%time
uxds_e3sm.FLUT.isel(time=0).plot(
    height=200,
    width=450,
    clabel="Upwelling Longwave Flux ($W m^{-2}$)",
    cmap="plasma",
    backend="matplotlib",
    projection=ccrs.Robinson(),
    features=["borders", "coastline"],
    title="Upwelling longwave flux at Top of Model",
)

CPU times: user 21.4 ms, sys: 94 μs, total: 21.5 ms
Wall time: 20.6 ms