cdtCrossValidationClimDataCMD: Cross-validation, merging stations observation and gridded...

In rijaf-iri/CDT: Climate Data Tools

Cross-validation, merging stations observation and gridded climate data.

Description

Function to perform a Leave-One-Out Cross-Validation for climate data merging.

Usage

cdtCrossValidationClimDataCMD(
  variable = "temp",
  time.step = "dekadal",
  minhour = 1,
  dates = list(from = "range", pars = list(start = "2018011", end = "2018123")),
  station.data = list(file = "", sep = ",", na.strings = "-99"),
  netcdf.data = list(dir = "", format = "tmax_adj_%s%s%s.nc", varid = "temp", ilon =
    1, ilat = 2),
  merge.method = list(method = "SBA", nrun = 3, pass = c(1, 0.75, 0.5)),
  interp.method = list(method = "idw", nmin = 8, nmax = 16, maxdist = 2.5, use.block =
    TRUE, vargrd = FALSE, vgm.model = c("Sph", "Exp", "Gau", "Pen")),
  crossv.station = list(from = "all", pars = NULL),
  auxvar = list(dem = FALSE, slope = FALSE, aspect = FALSE, lon = FALSE, lat = FALSE),
  dem.data = list(file = "", varid = "dem", ilon = 1, ilat = 2),
  RnoR = list(use = FALSE, wet = 1, smooth = FALSE),
  output.dir = "",
  GUI = FALSE
)

Arguments

variable	character, name of the climate variable to merge. Available options: "rain", "temp", "rh", "pres", "prmsl", "rad", "wspd", "ugrd", "vgrd". "rain": rainfall data "temp": temperature data "rh": relative humidity data "pres": surface pressure data "prmsl": pressure at mean sea level "rad": radiation data "wspd": wind speed data "ugrd": zonal wind components, E-W speed "vgrd": meridional wind components, N-S speed
time.step	character, the time step of the data. Available options: "minute", "hourly", "daily", "pentad", "dekadal", "monthly".
minhour	integer, the step for "minute" and "hourly". "minute": can have the value 1, 5, 10, 15, 20 or 30 "hourly": can take the value 1, 3, 6 or 12
dates	named list, providing the dates to merge. The list includes an element from with available options "range", "file" or "dates", and an element pars which is a named list specifying the parameters related to from: "range": pars specifies the start and end dates to merge. Example: pars = list(start = "2018011", end = "2018123") "file": pars specifies the full path to the file containing the dates to merge. Example: pars = list(file = "/home/data/files/dates.txt") The contents of the file are as follows: ## cat /home/data/files/dates.txt 2020011 2020012 2020013 ...... "dates": pars specifies a vector containing the dates to merge. Example: pars = list(dates = c("2020011", "2020012", 2020091, 2020113))
station.data	named list, providing the station data to be used in CDT format. file: character, full path to the file containing the stations data sep: character, column separator of the data na.strings: character, missing values flag
netcdf.data	named list, providing the input netCDF dataset to be used. dir: character, full path to the directory containing the netCDF files. format: character, format of the netCDF file names varid: character, name of the variable to read from the netCDF data ilon: integer, order for the longitude dimension of the variable. Example: if the variable "precip" has the dimension order [Lat, Lon] then ilon must be 2 ilat: integer, order for the latitude dimension of the variable.
merge.method	named list, indicating the merging method. "method": character, the merging method. Valid options: "CSc", "BSc", "SBA" or "RK". "CSc": Cressman Scheme "BSc": Barnes Scheme "SBA": Simple Bias Adjustment "RK": Regression Kriging "nrun": integer, number of the nested run to be performed "pass": numeric vector giving the fraction of nmin, nmax and maxdist to be used for each pass.
interp.method	named list, indicating the interpolation method and parameters to be used for "SBA" and "RK". method: character, the interpolation method to be used. Valid options: "idw", "shepard", "sphere" or "okr". "idw": Inverse distance weighted "shepard": Modified Shepard interpolation "sphere": Spheremap interpolation method "okr": Ordiranry kriging nmin: integer, minimum number of stations to be used to interpolate a grid point nmax: integer, maximum number of stations to be used to interpolate a grid point maxdist: numeric, maximum radius of influence in decimal degree use.block: logical, use block mean values to interpolate a grid point vargrd: logical, use a variable radius of influence vgm.model: character vector of variogram model to be used if method is "okr". Default is c("Exp", "Gau", "Sph", "Pen")
crossv.station	named list, selecting the stations to use for the cross-validation. The list includes an element from with available options "all", "file" or "cdt", and an element pars which is a named list specifying the parameters related to from: "all": all the stations from station.data will be used for cross-validation, pars can be omitted "file": the list of stations to be used for the cross-validation comes from a file. type: character, the type of the data, valid options are: "cdtstation": the stations come from a CDT stations data, "cdtcoords": the stations come from a CDT coordinates file file: character, full path to the file containing the stations sep: character, column separator of the data na.strings: character, missing values flag header: logical, in case of "cdtcoords", set TRUE if the data has a header "cdt": the list of stations to be used for the cross-validation will be selected from station.data by providing the percent of minimum available data for each station. Example: pars = list(min.perc = 40)
auxvar	named list, specifying the auxiliary variables to use when the merging method is "RK". dem: logical, include elevation data as auxiliary variable slope: logical, include slope as auxiliary variable aspect: logical, include aspect as auxiliary variable lon: logical, include longitude as auxiliary variable lat: logical, include latitude as auxiliary variable
dem.data	named list, providing the Digital Elevation Model (in netCDF format) when using regression kriging method with elevation related data as auxiliary variable. file: character, full path to the netCDF file containing the elevation data. varid: character, name of the variable to read from the netCDF data ilon: integer, order for the longitude dimension of the variable. ilat: integer, order for the latitude dimension of the variable.
RnoR	named list, specifying the rain-no-rain mask parameters in case of var.clim = "rain". use: logical, apply rain-no-rain mask wet: numeric, threshold to be use to define the wet/dry event smooth: logical, smooth the rain-no-rain mask after interpolation
output.dir	character, full path to the directory to save the output.
GUI	logical, indicating whether or not the output message should be displayed on CDT GUI. If TRUE, CDT GUI must be open.

rijaf-iri/CDT documentation built on July 3, 2024, 2:54 a.m.