Fidelity_test: Fidelity testing
In timokelder/UNSEEN: Evaluate Large ensemble simulations

Description Usage Arguments Details Value Source

Test the statistical consistency between large-ensemble simulations and shorter observed records

1	fidelity_test(obs, ensemble, units = "mm/day", fontsize = 11, biascor = FALSE)

`obs`	The observations. This function expects a vector, i.e. dataframe$variable.
`ensemble`	The UNSEEN ensemble. This function expects a vector, i.e. dataframe$variable
`units`	units label. Defaults to "mm/day".
`fontsize`	The font size. Defaults to 11.
`biascor`	Boolean. Do you want to apply a mean-bias correction?

Large-ensemble climate model simulations have a much greater sample size than observed records, complicating the evaluation of such ensembles. The fidelity test bootstraps the large ensemble into sample sizes of length equal to the observed record. It plots histograms of the distribution characteristics (mean, standard deviation, skewness and kurtosis) of the bootstrapped timeseries, inlcuding 95 intervals. Distribution characteristic of the observed records are plotted on top.

plots showing the bootstrapped tests of the mean, sd, skewness and kurtosis

Evaluation explaned in more detail in Kelder et al. 2020

timokelder/UNSEEN documentation built on June 1, 2021, 2:30 a.m.