eobs

This module contains functionality to download and load E-OBS data.

Fetches complete grid from https://surfobs.climate.copernicus.eu/dataaccess/access_eobs.php.

Example: Example: Get elevation of whole E-OBS grid

```python
from springtime.datasets import EOBS
datasource = EOBS(product_type='elevation',
                  variables=['land_surface_elevation'],
                  years=[2000, 2002]
                  )
ds = datasource.load()
```

Example: Example: Get all variables for a single point

```python
from springtime.datasets import EOBS
datasource = EOBS(points=[5, 50],
                  product_type='ensemble_mean',
                  grid_resolution='0.25deg',
                  years=[2000,2002])
df = datasource.load()
```

Example: Example: Get elevation

```python
from springtime.datasets import EOBS
datasource = EOBS(points=[5, 50],
                  product_type='elevation',
                  variables=['land_surface_elevation'],
                  years=[2000, 2002]
                  )
df = datasource.load()
```

Examples: Example: Load all variables for a selection of points.

```python
from springtime.datasets import EOBS
datasource = EOBS(points=[[5, 50], [5, 55]],
                  product_type='ensemble_mean',
                  grid_resolution='0.25deg',
                  years=[2000,2002])
df = datasource.load()
```

Example: Example: Load coarse mean temperature around amsterdam from 2002 till 2002

```python
from springtime.datasets import EOBS
dataset = EOBS(years=[2000,2002],
               area={
                   'name': 'amsterdam',
                   'bbox': [4, 50, 5, 55]
               },
               grid_resolution='0.25deg')
df = dataset.load()
```

EOBS

Bases: Dataset

E-OBS dataset.

Attributes:

Name	Type	Description
years		timerange. For example years=[2000, 2002] downloads data for three years.
variables	Sequence[Variable]	Some variables are specific for a certain product type.
points	Point | Points | None	(Sequence of) point(s) as (longitude, latitude) in WGS84 projection.
area	NamedArea | None	A dictionary of the form {"name": "yourname", "bbox": [xmin, ymin, xmax, ymax]}. If both area and points are defined, will first crop the area before extracting points, so points outside area will be lost. If points is None, will return the full dataset as xarray object; this cannot be joined with other datasets.
resample	Optional[dict[str, Any]]	Resample the dataset to a different time resolution. If None, no resampling. Else, should be a dictonary of the form {frequency: 'M', operator: 'mean', **other_options}. Currently supported operators are 'mean', 'min', 'max', 'sum', 'median'. For valid frequencies see [1]. Other options will be passed directly to xr.resample [2]
product_type	Literal['ensemble_mean', 'ensemble_spread', 'elevation']	one of "ensemble_mean", "ensemble_spread", "elevation".
grid_resolution	Literal['0.25deg', '0.1deg']	either "0.25deg" or "0.1deg"
version	Literal['26.0e']	currently only possible value is "26.0e"
keep_grid_location	bool	If True, keep the eobs_longitude and eobs_latitude columns. If False, keep input points instead.
minimize_cache	bool	if True, only store the data for the selected years and optionally area. This saves disk space, but it requires re-downloading for new years/areas. Default is False, i.e. keep the full EOBS grid.
https	bool	//pandas.pydata.org/pandas-docs/stable/user_guide/timeseries.html#offset-aliases
https	bool	//docs.xarray.dev/en/stable/generated/xarray.Dataset.resample.html

to_recipe()

Print out a recipe to reproduce this dataset.

download()

Download the data if necessary and return the file paths.

raw_load()

Load the dataset from disk into memory without further processing.

load()

Load and pre-process the data.

extract_time(ds, start, end)

Extract time range from xarray dataset.

extract_area(ds, bbox)

Extract bounding box from xarray dataset.

monthly_agg(ds, operator=np.mean)

Return monthly aggregates based on DOY dimension.

monthly_gdd(ds, t_base=5)

Return monthly growing degree days based on DOY dimension.