evalSens: Evaluating the SCAPE performance

In dpabon/ecofunr: Deriving Ecosystem Functional Properties

Description

Evaluating the SCAPE performance

Usage

evalSens(SCAPE_res, Rb = NA)

Arguments

SCAPE_res	list: ouput from a successful run of a getQ10, getArrhenius or getLloydTaylor function
Rb	vector: in case it applies: an Rb time series

Details

Function to properly evaluating the model based on the SCAPE estimated sensitivities. A straight forward evaluation of the SCAPE predictions e.g. via the RMSE or other error metrics are possible, these could be overoptimistic. The reason is that the time varying basal respiration is extracted as part of the original observations. Hence, a model that includes this part of the signal is hence comparing a fraction of the signal with itself. The evaluation wrapper uses the output of the get*** model and performs the evaluation based on the spectrally decomposed signals (i.e. in frequency ranges where Rb does not play a direct role), using the same spectral method, parameterization, and surrogate setting. Metrics used here are

Author(s)

Fabian Gans, Miguel D. Mahecha, MPI BGC Jena, Germany, [e-mail] [e-mail]

See Also

getQ10, getLloydTaylor, getArrhenius

dpabon/ecofunr documentation built on July 15, 2020, 12:58 p.m.