Arguments yaml file.

This procedure needs many datasets which tend to be large. Therefore, It is recommended to use meticulous directory structure, such as, project directory, observation directory, CMIP5 directory and so on. Before explaining those directories, see below yaml formatted file.

prjdir: D:/"Your project name"/foo
dbdir: D:/"Your project name"/Database
stndir: $(prjdir)/Observed/"station or regional name recommened"
bnddir: $(prjdir)/gis-boundary
qmapdir: $(prjdir)/Downscale/SQM
syear_obs: 1976     # Starting year of observed data
eyear_obs: 2005     # Ending year of observed data
syear_his: 1976     # Starting year of historical period (GCM)
eyear_his: 2005     # Ending year of historical period (GCM)
syear_scn:         
  - 2010
  - 2040
eyear_scn:
  - 2039
  - 2069
SimAll: FALSE       # Option for simulation all (GCM model, Variable, RCPs) combinations
ModelNames:
  - bcc-csm1-1-m    # Beijing Climate Center,  China Meteorological Administration (128x64)
  - CanESM2         # Canadian Centre for Climate Modelling and Analysis (128x64)
  - CMCC-CMS        # Centro Euro-Mediterraneo per I Cambiamenti Climatici (192x96)
  - CSIRO-Mk3-6-0   # Commonwealth Scientific and Industrial Research Organisation in  collaboration with the Queensland Climate Change Centre of Excellence (192x96)
  - FGOALS-g2       # LASG, Institute of Atmospheric Physics, Chinese Academy of  Sciences; and CESS, Tsinghua University (128x60)
  - HadGEM2-AO      # National Institute of Meteorological Research, Korea Meteorological Administration (192x145)
  - inmcm4          # Institute for Numerical Mathematics (180x120)
  - IPSL-CM5A-LR    # Institut Pierre-Simon Laplace (96x96)
  - MIROC-ESM       # Japan Agency for Marine-Earth Science and Technology, Atmosphere and Ocean Research Institute, and  National Institute for Environmental Studies (128x64)
  - MPI-ESM-LR      # Max Planck Institute for Meteorology (MPI-M) (192x96)
  - NorESM1-M       # Norwegian Climate Centre (144x96)
RcpNames:
  - rcp85           # Representative Concentration Pathway (RCP) 8.5 Scenarios
VarNames:
  - pr              #Precipitation (mm)
  - tasmax          #Max. temperature (C)
  - tasmin          #Min. temperature (C)
NtlCode: KR
stnfile: Station-Info.csv   # Station meta file, name it dishtinguibly in case many regions involved.
bndfile: Korea.shp
OWrite: TRUE
SRadiation: FALSE

You are expected to have some exposure to those arguments, now see each of them one by one. At first, you create a super directory at large memory available path with name distinguishable, date and region are really good things to be written in the name, making your job path distinguishable. In this vignette, I name the name “APCC”(APEC Climate Center). D:/APCC.

prjdir: D:/APCC/project # This is your project directory where the downscaled results would be filed up.
dbdir: D:/APCC/Database # This is your database directory where the CMIP5 data needed for the work would be saved.
stndir: $(prjdir)/Observed/Korea # This is the directory to be filed up with observation data. I name it "Korea" in this tutorial

Above three directories(prjdir, dbdir, stndir) must be prepared(created) in advance. That’s because we assume you have your own observation data beforehand. You need to store station csv file and observation csv file in stndir(station directory). Station file(stnfile) is described in detail below. Observation file should be csv formatted and look like this. Each file name must contain the station ID(eg, ID108).

Note : Day is month day not Julian format, that is, 2017/2/1 works but 2017/2/32 does not.
Header names are not much critical, but the order is. Year, Month, Day, Precipitation, Tasmax, Tasmin, Wind Speed, Relative Humidity, and Solar Radiation should be in this order. Of course, the unit matters too.

bnddir: $(prjdir)/gis-boundary # Under development, providing shp. files for further work.
qmapdir: $(prjdir)/Downscale/SQM # This directory will contain final result passed through SQM(Simple Quantile Mapping)
syear_obs: 1976     # Starting year of observed data
eyear_obs: 2005     # Ending year of observed data
syear_his: 1976     # Starting year of historical period (GCM)
eyear_his: 2005     # Ending year of historical period (GCM)
syear_scn:         
  - 2010
  - 2040
eyear_scn:
  - 2039
  - 2069            # Start years and End years of climate change scenario.
SimAll: FALSE       # Option for simulation all (GCM model, Variable, RCPs) combinations

If you put TRUE to SimAll argument, your process runs over all the models including GCMs, RCMs and RCPs. Obviously, takes a long time.

ModelNames:
  - bcc-csm1-1-m    # Beijing Climate Center,  China Meteorological Administration (128x64)
  - CanESM2         # Canadian Centre for Climate Modelling and Analysis (128x64)
  - CMCC-CMS        # Centro Euro-Mediterraneo per I Cambiamenti Climatici (192x96)
  - CSIRO-Mk3-6-0   # Commonwealth Scientific and Industrial Research Organisation in  collaboration with the Queensland Climate Change Centre of Excellence (192x96)
  - FGOALS-g2       # LASG, Institute of Atmospheric Physics, Chinese Academy of  Sciences; and CESS, Tsinghua University (128x60)
  - HadGEM2-AO      # National Institute of Meteorological Research, Korea Meteorological Administration (192x145)
  - inmcm4          # Institute for Numerical Mathematics (180x120)
  - IPSL-CM5A-LR    # Institut Pierre-Simon Laplace (96x96)
  - MIROC-ESM       # Japan Agency for Marine-Earth Science and Technology, Atmosphere and Ocean Research Institute, and  National Institute for Environmental Studies (128x64)
  - MPI-ESM-LR      # Max Planck Institute for Meteorology (MPI-M) (192x96)
  - NorESM1-M       # Norwegian Climate Centre (144x96)
RcpNames:
  - rcp85           # Representative Concentration Pathway (RCP) 8.5 Scenarios

Otherwise, FALSE on SimAll, and specify model names you want to use in simulation.

VarNames:
  - pr              #Precipitation (mm)
  - tasmax          #Max. temperature (C)
  - tasmin          #Min. temperature (C)

Variable names, pr(precipitation inmm), tasmax/tasmin(max/min temperature in Celcius degree), sfcWind(wind speed in m/s), rhs(relative humidity in fraction, not percentage), rsds(solar radiation in Mega Joule per square meter)

NtlCode: KR         

National Code used when downloading clipped CMIP5 data from ADSS(APEC Data Service Syetem). See below tables.

table1. Available national-level data based on clipped CMIP5 climate change scenario data.

table2. Available United State data based on clipped CMIP5 climate change scenario data.

(The countries in above tables are currently supported by APCC through ADSS. We plan to expand the list of supported countries through future updates. If you wish to see support for your country, please place a request by contacting us at climate.service@apcc21.org)

stnfile: Station-Info.csv

‘stnfile’ is a csv(comma-seperated values) file. This is a kind of meta file over all the station information. One picture is worth a thousand words, see below.

Lon, Lat, and Elev are Longitude, Latitude, and Elevation of observatory respectively. And ID is the ID of obvervatory and Ename is the region where it is. SYear is the year the observation starts. There is one more thing you need to be aware of. If you lack of your own observation data so use GHCN observation data, name the stnfile ‘Station-Info.csv’. We know it seems to be little awkward. But, the procedure is designed to run with that name over GHCN data.

bndfile: Korea.shp  # Under development, the shape file for further work.
OWrite: TRUE

Downscaling process is heavy work. That means sometimes you need pause it and go again. Then you put TRUE on OWrite(Overwrite) which make things continue.

SRadiation: FALSE

SRadiation(Solar Radiation) is a variable not likely to be. If it is luckily, put TRUE on it.

Workflow

Now we complete writing yaml file, that is, necessary arguments are prepared. Before going on, let’s review the directories and corresponing data.

Your Working Directory(recommended to be in where large amount of disk memory available)
        |
        |  In your working directory, the yaml file which has necessary arguments must be in.
      -------------------
      |                 |
  Database            Project Diretory
      |                           |
 CMIP5 Directory          ---------------------------------------
      |                   |                   |                 |
 CMIP5 scenario data   gis-boundary       Observed          Downscale
downloaded from ADSS      |                   |                 |(final downscaled results)
must have 4073 files.   shape file          KOREA               -------
                       (Under dev.)   (or the region you        |     |
                                       are interested in.)     OBS   SQM
                                              |
                                station meta file and observation csv files
                              for each stations(meteorological observatories)

(If you have any trouble understanding directory structure and observation path, just run rSQMSampleProject() function and see what happens.)

>library(rSQM)
>setwd("D:/sampleProject")

>EnvList <- SetWorkingEnvironment(envfile = "rSQM.yaml")

>EnvList
$prjdir
[1] "D:/rSQMsample/prj"
$dbdir
[1] "D:/rSQMsample/Database"
$qmapdir
[1] "D:/rSQMsample/prj/Downscale/SQM"
$bnddir
[1] "D:/rSQMsample/prj/gis-boundary"
$stndir
[1] "D:/rSQMsample/prj/Observed/Korea"
$syear_obs
[1] 1976
$eyear_obs
[1] 2005
$syear_his
[1] 1976
$eyear_his
[1] 2005
$syear_scn
[1] 2010 2040
$eyear_scn
[1] 2039 2069
$SimAll
[1] FALSE
$ModelNames
 [1] "bcc-csm1-1-m"  "CanESM2"       "CMCC-CMS"      "CSIRO-Mk3-6-0"
 [5] "FGOALS-g2"     "HadGEM2-AO"    "inmcm4"        "IPSL-CM5A-LR" 
 [9] "MIROC-ESM"     "MPI-ESM-LR"    "NorESM1-M"    
$RcpNames
[1] "rcp85"
$VarNames
[1] "pr"     "tasmax" "tasmin"
$NtlCode
[1] "KR"
$stndir
[1] "D:/rSQMsample/prj/Observed/Korea"
$stnfile
[1] "Station-Info.csv"
$bndfile
[1] "Korea.shp"
$OWrite
[1] TRUE
$SRadiation
[1] FALSE
$cmip5dir
[1] "D:/rSQMsample/Database/cmip5_daily_KR"

LoadCmip5DataFromAdss(dbdir = EnvList$dbdir, NtlCode = EnvList$NtlCode)

do.call(LoadCmip5DataFromAdss, EnvList)

 GhcnDailyUpdate(
   NtlCode = EnvList$NtlCode,
   stndir = EnvList$stndir,
   syear_obs = EnvList$syear_obs,
   eyear_obs = EnvList$eyear_obs)

do.call(GhcnDailyUpdate, EnvList)

DailyExtractAll(
  cmip5dir = EnvList$cmip5dir,
  stndir = EnvList$stndir,
  stnfile = EnvList$stnfile,
  qmapdir = EnvList$qmapdir,
  SimAll = EnvList$SimAll,
  ModelNames = EnvList$ModelNames,
  RcpNames = EnvList$RcpNames,
  VarNames = EnvList$VarNames,
  OWrite = EnvList$OWrite)

do.call(DailyExtractAll, EnvList)

DailyQMapAll(
  stndir = EnvList$stndir,
  stnfile = EnvList$stnfile,
  qmapdir = EnvList$qmapdir,
  prjdir = EnvList$prjdir,
  SimAll = EnvList$SimAll,
  RcpNames = EnvList$RcpNames,
  VarNames = EnvList$VarNames,
  syear_obs = EnvList$syear_obs,
  eyear_obs = EnvList$eyear_obs,
  syear_his = EnvList$syear_his,
  eyear_his = EnvList$eyear_his,
  syear_scn = EnvList$syear_scn,
  eyear_scn = EnvList$eyear_scn,
  OWrite = EnvList$OWrite,
  SRadiation = EnvList$SRadiation)

do.call(DailyQMapAll, EnvList)

ID108_SQM_CanESM2_historical.csv
ID108_SQM_CanESM2_historical_original.csv
ID108_SQM_CanESM2_rcp45.csv
ID108_SQM_CanESM2_rcp45_original.csv

Additional Remarks

1. A process over this package needs to connect to web server to download required data. This means that unstable internet connectivity fails works.
2. If you have any problem, contact climate.service@apcc21.org

Hope this package would be useful. :)