metobs_toolkit.modeltimeseries.ModelTimeSeries

class ModelTimeSeries(site, datarecords: ~numpy.ndarray, timestamps: ~numpy.ndarray, modelobstype: ~metobs_toolkit.obstypes.ModelObstype, datadtype: type = <class 'numpy.float32'>, timezone: str = 'UTC', modelname: str | None = None, modelvariable: str | None = None, _convert_to_standard_units: bool = True)

Class for model-based timeseries at one location.

This class stores model data at a specific location and automatically converts the data from the model’s native units to standard units as defined by the observation type. The unit conversion is performed during initialization.

Parameters:

• site (object) – The site object representing the location.

• datarecords (np.ndarray) – Array of data records in the model’s native units. These will be automatically converted to standard units during initialization.

• timestamps (np.ndarray) – Array of timestamps corresponding to the data records.

• modelobstype (ModelObstype) – The observation type containing unit information. Must have a model_unit attribute set for unit conversion to work properly.

• datadtype (type, optional) – Data type for the records, by default np.float32.

• timezone (str, optional) – Timezone for the timestamps, by default “UTC”.

• modelname (str, optional) – Name of the model, by default None.

• modelvariable (str, optional) – Name of the model variable, by default None.

Notes

The stored data in the series attribute will be in standard units as defined by modelobstype.std_unit, not in the original model units. The conversion is performed automatically.

Raises:

MetObsUnitUnknown – If the modelobstype does not have a model_unit set, which is required for unit conversion.

__init__(site, datarecords: ~numpy.ndarray, timestamps: ~numpy.ndarray, modelobstype: ~metobs_toolkit.obstypes.ModelObstype, datadtype: type = <class 'numpy.float32'>, timezone: str = 'UTC', modelname: str | None = None, modelvariable: str | None = None, _convert_to_standard_units: bool = True)

Initialize a ModelTimeSeries with data and model metadata.

Parameters:

• site (Site) – The site object representing the measurement location.

• datarecords (numpy.ndarray) – Observation values in the model’s native units. Automatically converted to standard units unless _convert_to_standard_units is False.

• timestamps (numpy.ndarray) – Timestamps corresponding to datarecords.

• modelobstype (ModelObstype) – Observation type that carries unit-conversion and band information.

• datadtype (type, optional) – Numeric dtype for the stored series. Default is numpy.float32.

• timezone (str, optional) – Timezone of timestamps. Default is 'UTC'.

• modelname (str, optional) – Human-readable name of the model (e.g. 'ERA5_land').

• modelvariable (str, optional) – Model variable / band name.

• _convert_to_standard_units (bool, optional) – If True (default), converts datarecords from model units to standard units. Set to False when data are already in standard units (e.g. during __add__).

Raises:

• TypeError – If modelobstype is not a ModelObstype instance.

• MetObsUnitUnknown – If the model_unit of modelobstype is None.

Methods

__init__(site, datarecords, timestamps, ...)	Initialize a ModelTimeSeries with data and model metadata.
get_info([printout])	Print or return information about the ModelTimeSeries.
make_plot([linecolor, ax, figkwargs, title])	Create a plot of the model time series.
pd_plot(**pdplotkwargs)	A wrapper on the pandas.series.plot function for ModelTimeSeries.
to_xr()	Convert a Modelimeseries object to an xarray.Dataset.

Attributes

df	ModelTimeSeries DataFrame constructor.
end_datetime	Return the end datetime of the series.
freq	Return the frequency of the series.
start_datetime	Return the start datetime of the series.
stationname	Return the name of the station this SensorData belongs to.
tz	Return the timezone of the stored timestamps.