LightResponseCurveFitter_optimLRCOnAdjustedPrior {REddyProc}

LightResponseCurveFitter optimLRCOnAdjustedPrior

Description

Lower bound flux uncertainty and adjust prior uncertainty before calling optimLRC

Usage

LightResponseCurveFitter_optimLRCOnAdjustedPrior(theta, 
    iOpt, dsDay, parameterPrior, ctrl, ...)

Arguments

theta

numeric vector of starting values

iOpt

integer vector: positions of subset of parameters that are optimized

dsDay

dataframe of NEE, sdNEE and predictors Rg, VPD and Temp

parameterPrior

numeric vector of prior parameter estimates (corresponding to theta) # TODO rename to thetaPrior

ctrl

list of further controls

...

further arguments to LightResponseCurveFitter_optimLRC (passed to LightResponseCurveFitter_computeCost)

Details

Only those records are used for optimization where both NEE and sdNEE are finite. In larger settings, already filtered at

Optimization of LRC parameters takes into account the uncertainty of the flux values. In order to avoid very strong leverage, values with a very low uncertainty (< a lower quantile) are assigned the lower quantile is assigned. This procedure downweighs records with a high uncertainty, but does not apply a large leverage for records with a very low uncertainty. Avoid this correction by setting ctrl$isBoundLowerNEEUncertainty = FALSE

The uncertainty of the prior, that maybe derived from fluxes) is allowed to adapt to the uncertainty of the fluxes. This is done in link{LightResponseCurveFitter_getPriorScale}

Value

result of LightResponseCurveFitter_optimLRC with items theta, iOpt and convergence

Author(s)

Department for Biogeochemical Integration at MPI-BGC, Jena, Germany <[e-mail]> [cph], Thomas Wutzler <[e-mail]> [aut, cre], Markus Reichstein <[e-mail]> [aut], Antje Maria Moffat <[e-mail]> [aut, trl], Olaf Menzer <[e-mail]> [ctb], Mirco Migliavacca <[e-mail]> [aut], Kerstin Sickel <[e-mail]> [ctb, trl], Ladislav <U+0160>igut <[e-mail]> [ctb]