ProbabilitySpace: Creating synthetic field data by running the SWIFT model

In HannesDeDeurwaerder/iSWIFT: iSWIFT: Inverse method to derive the relative distribution of the absorbing root surface using the SWIFT model

Description

In this function, all required parameters of the SWIFT model are declared to perform a user-defined number of SWIFT runs, creating synthetic xylem water isotope composition (D2H and D18O).

NOTE that in this specific configuration, the function apeals to the 'VarMatrix' and the 'SoilHeterogeneity' functions in order to retreive all SWIFT variables. The user is encouraged to alter both functions according to the requirements of their study.

IMPORTANT: Note that for this function to properly work, all variables of the SWIFT model should be declared locally or globally.

Usage

 ProbabilitySpace( beta.i=NULL, 
                          BiophysVariables=NULL,
                          PSIprofiles=NULL,
                          soilDeuterium=NULL, 
                          soilOxygen=NULL)

Arguments

beta.i	Description: Allocation parameter that describes the exponential decay of the distribution of absorbing root surfaces for the focal plant; Functionality: used in the iSWIFT loglikelihood optimization Structure: Randomly assigned numeric value within defined ranges as provided in the 'optim' function.
BiophysVariables	Description: Matrix of plant-trait and sampling strategy required variables required to declare all SWIFT model variables; Structure: Matrix of SWIFT input variables and/or variables required to derive the SWIFT input variables (i.e. DBH, lumen fraction,...) Creation: Via 'VarMatrix' function
PSIprofiles	Description: Soil water potential values with soil depth over the discritized soil layers Structure: Matrix of distinct soil water potetnial profiles with depth. Each column presents a soil water potential profile per planned SWIFT run, rows correspond to the discritized soil layers each individual SWIFT run planned; Creation: Via 'SoilHeterogeneity' function
soilDeuterium	Description: Soil water deuterium composition profiles with soil depth over the discritized soil layers Structure: Matrix of distinct soil water deuterium composition with depth. Each column presents a soil water deuterium composition profile per planned SWIFT run, rows correspond to the discritized soil layers each individual SWIFT run planned; Creation: Via 'SoilHeterogeneity' function
soilOxygen	Description: Soil water 18 Oxygen composition profiles with soil depth over the discritized soil layers Structure: Matrix of distinct soil water 18 Oxygen composition with depth. Each column presents a soil water 18 Oxygen composition profile per planned SWIFT run, rows correspond to the discritized soil layers each individual SWIFT run planned; Creation: Via 'SoilHeterogeneity' function

Value

Returning values are SWIFT generated xylem water istopic composition, i.e., D2H and/or D18O. Hence providing synthetic field data corresponding the input provided by the reader, but without consideration of potential extraction error

Author(s)

Hannes De Deurwaerder, Marco D. Visser, Matteo Detto, Pascal Boeckx, Felicien Meunier, Pedro Herve Fernandez and Hans Verbeeck

Examples

  ProbabilitySpace <- function(beta.i, BiophysVariables, PSIprofiles, 
                               soilDeuterium, soilOxygen){
      #--------------------------------------------------------------------#
      # NOTE that for this function to properly work, all variables of the #
      # SWIFT model should be declared locally or globally.                #
      #--------------------------------------------------------------------#

    Iso_H <- rep(NA, SWIFTitterations)   # decalre empty vector of isotopic data

    # which isotope is considered
    if(ConsideredIsotopes=='D2H'){Isotopes <- soilDeuterium}  # deuterium
    if(ConsideredIsotopes!='D2H'){Isotopes <- soilOxygen} 
                                    # oxygen or dual using the LMWL approach
    for (ii in 1:SWIFTitterations){
      Ax <- BiophysVariables[ii,'LA']*BiophysVariables[ii,'Asapwood'] # lumen area
      SF <- relSF*BiophysVariables[ii,'SFdaily']* 3600/tF 
                                        # Sapflow in m3/time-frequency considered
      k  <- SoilRootCond(betaCom, BiophysVariables[ii,'Kr'], PSIprofiles[,ii], 
                         Z, 'Sandy Clay') # Total soil to root conductivity
      tmp <- beta.i^(100*Z)*(1-beta.i^(100*dZ))
      ARi <- BiophysVariables[ii,'ARtot']*tmp/sum(tmp) # root area distribution

      # run SWIFT for stem base and sampling at specific height    
      Iso_SB <- SWIFT_SB(ARi, Isotopes[,ii], k, PSIprofiles[,ii], SF, tt, Z)
      Iso_H[ii] <- SWIFT_H( Ax, Iso_SB, BiophysVariables[ii,'hos'], SF, 
                            BiophysVariables[ii,'tstud'], tF)
      # report progress of one run
      cat("\rbeta = ",beta.i,"; progress = ",round((ii/SWIFTitterations)*100,2))
    }
    return(Iso_H)
   }

HannesDeDeurwaerder/iSWIFT documentation built on Dec. 17, 2021, 10:29 p.m.