In aukkola/FluxnetLSM: Quality Control and Data Conversion for 'FLUXNET' Data for Land Surface Modelling

Example data conversion using the Harvard site using La Thuile synthesis.

See FLUXNET2015/example_conversion_single_site.R for the corresponding FLUXNE2015 example.

Converts useful variables from a La Thuile spreatsheet format into two netcdf files, one for fluxes, and one for met forcings.

The user must provide the input directory path, output directory path and site code. For La Thuile datasetname must be set to "LaThuile" as the code defaults to FLUXNET2015. All other settings are optional. This example extracts La Thuile years complying with "Fair Use" policy. All other options are set to their default values.

Single site

library(FluxnetLSM) 

#clear R environment
rm(list=ls(all=TRUE))


#############################
###--- Required inputs ---###
#############################

#--- User must define these ---#

#Fluxnet site ID (see http://fluxnet.fluxdata.org/sites/site-list-and-pages/)
site_code   <- "US-Ha1"
datasetname <- "LaThuile"

# This directory should contain appropriate data from 
# http://fluxnet.fluxdata.org/data/la-thuile-dataset/
in_path <- "./Inputs"

#Outputs will be saved to this directory
out_path <- "./Outputs"


#--- Automatically retrieve all Fluxnet files in input directory ---#

# Input Fluxnet data file (using FULLSET in this example, see R/Helpers.R for details)
infile <- get_fluxnet_files(in_path, site_code, datasetname=datasetname)


##########################
###--- Run analysis ---###
##########################

convert_fluxnet_to_netcdf(site_code = site_code, infile = infile, 
                          out_path = out_path, datasetname=datasetname)

Multiple sites

#--- Automatically retrieve all Fluxnet files in input directory ---#

# Input Fluxnet data files
infile <- get_fluxnet_files(in_path, datasetname=datasetname)

#Get site codes
site_codes <- unique(get_path_site_code(infile))

#Reorganise input files by site
infiles <- lapply(site_codes, function(site) get_fluxnet_files(in_path, site_code=site,
                                                              datasetname=datasetname))



##########################
###--- Run analysis ---###
##########################


#Loops through sites
mapply(function(site_code, infile) {
  try(
    convert_fluxnet_to_netcdf(site_code = site_code,
                              infile = infile,
                              out_path = out_path,
                              datasetname = datasetname
    )
  )},
  site_code = site_codes,
  infile = infiles)

Multiple sites (parallel)

#--- Automatically retrieve all Fluxnet files in input directory ---#

# Input Fluxnet data files
infile <- get_fluxnet_files(in_path, datasetname=datasetname)

#Get site codes
site_codes <- unique(get_path_site_code(infile))

#Reorganise input files by site
infiles <- lapply(site_codes, function(site) get_fluxnet_files(in_path, site_code=site,
                                                               datasetname=datasetname))



##########################
###--- Run analysis ---###
##########################


#Initialise clusters (using 2 cores here)
cl <- makeCluster(getOption('cl.cores', 2))

#Import variables to cluster
clusterExport(cl, "out_path")
if(exists("conv_opts"))  {clusterExport(cl, "conv_opts")}
if(exists("datasetversion"))   {clusterExport(cl, "datasetversion")}
if(exists("datasetname"))   {clusterExport(cl, "datasetname")}
if(exists("plot"))             {clusterExport(cl, "plot")}


#Loops through sites
clusterMap(cl = cl, function(site_code, infile) {
  library(FluxnetLSM)
  tryCatch(
    convert_fluxnet_to_netcdf(
      site_code = site_code,
      infile = infile,
      out_path = out_path,
      datasetname = datasetname
    ),
    error = function(e) NULL)
},
site_code = site_codes,
infile = infiles)


stopCluster(cl)

aukkola/FluxnetLSM documentation built on July 4, 2023, 12:04 p.m.