Calculate the spatial mean of gridded data over geometry data

This notebook demonstrates how to compute the spatial mean of some sample 2m temperature in ERA5 data for a given geometry (such as an administrative region) using earthkit-data and earthkit-transforms. This is useful for generating regional climate statistics.

from earthkit import data as ekd
from earthkit import transforms as ekt
from earthkit.transforms._tools import earthkit_remote_test_data_file

remote_era5_file = earthkit_remote_test_data_file("era5-Europe-sfc-2m-temperature-3deg-2015-2017.grib")
era5_data = ekd.from_source("url", remote_era5_file)
ds = era5_data.to_xarray()

remote_geometry_url = earthkit_remote_test_data_file("NUTS_RG_60M_2021_4326_LEVL_0.geojson")
geometry_data = ekd.from_source("url", remote_geometry_url).to_geopandas()

# Calculate area mean
area_mean = ekt.spatial.reduce(ds, geometry_data, how="mean", weights="latitude")

area_mean

<xarray.Dataset> Size: 1MB
Dimensions:                  (forecast_reference_time: 4384, index: 37)
Coordinates:
  * forecast_reference_time  (forecast_reference_time) datetime64[us] 35kB 20...
  * index                    (index) int64 296B 0 1 2 3 4 5 ... 32 33 34 35 36
Data variables:
    2t                       (index, forecast_reference_time) float64 1MB nan...
Attributes:
    Conventions:  CF-1.8
    institution:  ECMWF

xarray.Dataset

• Dimensions:
  • forecast_reference_time: 4384
  • index: 37
• Coordinates: (2)
  • forecast_reference_time
    (forecast_reference_time)
    datetime64[us]
    2015-01-01 ... 2017-12-31T18:00:00

    standard_name :

    forecast_reference_time

    long_name :

    initial time of forecast

    array(['2015-01-01T00:00:00.000000', '2015-01-01T06:00:00.000000',
           '2015-01-01T12:00:00.000000', ..., '2017-12-31T06:00:00.000000',
           '2017-12-31T12:00:00.000000', '2017-12-31T18:00:00.000000'],
          shape=(4384,), dtype='datetime64[us]')

  • index
    (index)
    int64
    0 1 2 3 4 5 6 ... 31 32 33 34 35 36

    [37 values with dtype=int64]

• Data variables: (1)
  • 2t
    (index, forecast_reference_time)
    float64
    nan nan nan ... 264.7 264.9 264.8

    standard_name :

    unknown

    long_name :

    mean 2 metre temperature

    units :

    kelvin

    level_type :

    surface

    _earthkit :

    {"message": {"__bytes_b64__": "R1JJQgAAbAEAADSAYv//gKcBAAAPAQEAAAEAAAAAAAAVAAAAAAAAAAAAAAAAAAAAARcCBAEwMDAxAAAAAAAgAP8AAB0AEwE0mIBhqIAAYagA5ngLuAu4AAAAAAAAAAwAgAJEuX1uAAA3Nzc3"}, "bitsPerValue": 16, "grid_spec": {"area": [79, -25, 25, 59], "grid": [3, 3], "reference": [2, 1]}}

    reduce_dims :

    ['latitude', 'longitude']

    array([[         nan,          nan,          nan, ...,          nan,
                     nan,          nan],
           [256.40283203, 257.93721008, 264.29278564, ..., 274.8478241 ,
            280.21913147, 279.15505981],
           [275.28173828, 275.33174133, 275.81036377, ..., 272.0743866 ,
            272.67811584, 272.80935669],
           ...,
           [269.46728516, 270.1842804 , 273.04083252, ..., 278.32145691,
            281.48866272, 279.431427  ],
           [275.30685043, 274.37511372, 276.32913995, ..., 281.86877755,
            284.66424221, 283.50551419],
           [273.55405956, 273.4085369 , 274.4432307 , ..., 264.69922437,
            264.92087831, 264.80226706]], shape=(37, 4384))

• Attributes: (2)

  Conventions :

  CF-1.8

  institution :

  ECMWF