lair.background.thoning#

lair.background.thoning(data, smooth_time=None, **kwargs)[source]#

Thoning curve fitting.

Wraps code published by NOAA GML: https://gml.noaa.gov/ccgg/mbl/crvfit/crvfit.html

Thoning, K.W., P.P. Tans, and W.D. Komhyr, 1989,

Atmospheric carbon dioxide at Mauna Loa Observatory, 2. Analysis of the NOAA/GMCC data, 1974 1985., J. Geophys. Res. ,94, 8549 8565.

Parameters:
  • data (pd.Series) – Time series of data to be smoothed. Must have a datetime index. NaN values are dropped before the curve is fitted.

  • smooth_time (list[datetime] | pd.Index, optional) – Times at which to evaluate the fitted curve; they become the index of the result. Default is the index of data, so rows where data is NaN get the curve’s value.

  • **kwargs – Additional keyword arguments to pass to the ccgFilter class (see thoning_filter()). sampleinterval is the step, in days, of the even grid the filter interpolates the data onto. The default (0) lets the filter estimate it from the steps of about a day or more; when the data has none, as hourly data without a gap of a day does, lair uses the median step (1/24 for hourly data). Hourly data with such gaps needs sampleinterval=1/24.

Returns:

pd.Series – Smoothed data, at smooth_time.

Return type:

Series

See also

thoning_filter

The filter itself, for its trend and harmonics.

rolling_baseline

A low quantile instead of a curve through everything.

Examples

Three years of made-up daily CO2 that grows 2.5 ppm a year, with a seasonal cycle and noise, smoothed, then read on the first of each month:

>>> import numpy as np
>>> import pandas as pd
>>> from lair.background import thoning
>>> rng = np.random.default_rng(0)
>>> days = pd.date_range("2020-01-01", "2022-12-31", freq="D")
>>> years = (days - days[0]).days / 365.25
>>> seasons = 3 * np.cos(2 * np.pi * years)
>>> noise = rng.normal(0, 0.5, days.size)
>>> co2 = pd.Series(410 + 2.5 * years + seasons + noise, index=days)
>>> smooth = thoning(co2)
>>> round(float(smooth.iloc[-1] - smooth.iloc[0]), 1)
7.6
>>> months = pd.date_range("2020-01-01", "2022-12-01", freq="MS")
>>> thoning(co2, smooth_time=months).round(1).head(3).tolist()
[413.0, 412.8, 412.1]