General temporal aggregation methods

[1]:

# If first time running, uncomment the line below to install any additional dependencies
# !bash requirements-for-notebooks.sh

[2]:

from earthkit.transforms import aggregate as ek_aggregate
from earthkit import data as ek_data

from earthkit.data.testing import earthkit_remote_test_data_file
ek_data.settings.set("cache-policy", "user")

Load some test data

All earthkit-transforms methods can be called with earthkit-data objects (Readers and Wrappers) or with the pre-loaded xarray.

In this example we will use hourly ERA5 2m temperature data on a 0.5x0.5 spatial grid for the year 2015 as our physical data.

First we download (if not already cached) lazily load the ERA5 data (please see tutorials in earthkit-data for more details in cache management).

We inspect the data using the describe method and see we have some 2m air temperature data. For a more detailed representation of the data you can use the to_xarray method.

[3]:

# Get some demonstration ERA5 data, this could be any url or path to an ERA5 grib or netCDF file.
remote_era5_file = earthkit_remote_test_data_file("test-data", "era5_temperature_europe_2015.grib")
era5_data = ek_data.from_source("url", remote_era5_file)
era5_data.describe()
# era5_data.to_xarray()

[3]:

		level	date	time	step	paramId	class	stream	type	experimentVersionNumber
shortName	typeOfLevel
2t	surface	0	20150301,20150302,...	0,1800,...	0	167	ea	oper	an	0001

Reduce the ERA5 data over the time dimension

The default reduction method is mean, other methods can be applied using the how kwarg.

Note that we do not need to worry about the data format of the input array, earthkit will convert it to the required xarray format internally.

The returned object is an xarray dataset, however this may change in future version of the package.

Calculate a rolling mean with a 50 timestep window

There is no temporal specific method for a rolling reduction. The general rolling_reduce method can do this calculation by specifying the dimension over which you would like to reduce.

[9]:

era5_rolling = ek_aggregate.rolling_reduce(
    era5_data, time=50, center=True,
)
era5_rolling

[9]:

<xarray.Dataset> Size: 330MB
Dimensions:     (number: 1, time: 1460, step: 1, surface: 1, latitude: 201,
                 longitude: 281)
Coordinates:
  * number      (number) int64 8B 0
  * time        (time) datetime64[ns] 12kB 2015-01-01 ... 2015-12-31T18:00:00
  * step        (step) timedelta64[ns] 8B 00:00:00
  * surface     (surface) float64 8B 0.0
  * latitude    (latitude) float64 2kB 80.0 79.75 79.5 79.25 ... 30.5 30.25 30.0
  * longitude   (longitude) float64 2kB -10.0 -9.75 -9.5 ... 59.5 59.75 60.0
    valid_time  (time, step) datetime64[ns] 12kB dask.array<chunksize=(1460, 1), meta=np.ndarray>
Data variables:
    t2m         (number, time, step, surface, latitude, longitude) float32 330MB dask.array<chunksize=(1, 1459, 1, 1, 201, 281), meta=np.ndarray>
Attributes:
    GRIB_edition:            1
    GRIB_centre:             ecmf
    GRIB_centreDescription:  European Centre for Medium-Range Weather Forecasts
    GRIB_subCentre:          0
    Conventions:             CF-1.7
    institution:             European Centre for Medium-Range Weather Forecasts
    history:                 2024-07-29T08:52 GRIB to CDM+CF via cfgrib-0.9.1...

[ ]:

General temporal aggregation methods

Load some test data

Reduce the ERA5 data over the time dimension

The mean over the time dimension

The median over the time dimension

Calling the temporal reduce method with an arbitary function

Calculate a rolling mean with a 50 timestep window