Categorical gridded data

In this example, we will visualise “type of vegetation” data from the ERA5 reanalysis dataset in the C3S Climate Data Store.

import earthkit.data as ekd
import earthkit.plots as ekp

To retrieve data from the CDS, we use the "cds" source type with earthkit.data.from_source. For more information, see the earthkit-data documentation page on retrieving data from the CDS.

NOTE: To maximise compatibility without requiring a CDS account, the CDS request is commented out and a sample data request is run instead. If you have a CDS account, you can uncomment the CDS request.

# data = ekd.from_source(
#     "cds",
#     "reanalysis-era5-single-levels-monthly-means",
#     {
#         "product_type": "monthly_averaged_reanalysis",
#         "year": "2018",
#         "month": "05",
#         "time": "00:00",
#         "data_format": "grib",
#         "download_format": "unarchived",
#         "variable": "type_of_high_vegetation",
#     },
# )

data = ekd.from_source("sample", "era5-vegetation-2018.grib")

From the ERA5 documentation, we know that the types of high vegetation are categorised with the following values:

• 0: No vegetation

• 3: Evergreen needle

• 4: Deciduous needle

• 5: Deciduous broad

• 6: Evergreen broad

• 18: Mixed forest/wood

• 19: Interrupted forest

Using a special Categorical style from earthkit plots, we can easily associate each value in the data with its category by passing the levels argument as a dictionary.

style = ekp.styles.Categorical(
    levels={
        3: "Evergreen needle",
        4: "Deciduous needle",
        5: "Deciduous broad",
        6: "Evergreen broad",
        18: "Mixed forest/wood",
        19: "Interrupted forest",
    },
    colors=[
        "#33cc33",
        "#ffd700",
        "#ff9900",
        "#009933",
        "#0066cc",
        "#cc6600",
    ],
)

Now we can plot our data with this style:

chart = ekp.Map(crs="Robinson")

chart.grid_cells(data, style=style)

chart.title("ERA5 {variable_name!l} - {time:%B %Y}")
chart.coastlines(resolution="low")
chart.gridlines()
chart.legend(location="bottom", label="{variable_name}", ncols=3)

chart.show()