Steps to use RClimDex and Rhtest:

Document Sample
Steps to use RClimDex and Rhtest: Powered By Docstoc
					                     Quick Guide for the RClimDex and RHtests Users

                                        Prepared by
                                     Xiaolan L. Wang
                    Climate Research Division, STB, Environment Canada
                                       18 May 2010

Steps to use the RClimDex and RHtests combined package:
    (1) Enter source(RClimDex.r) at the R prompt to use RClimDex to perform QA/QC of the
        daily tmax, tmin and prcp.
    (2) Enter source(RClimDex_RHtest.r) at the R prompt to use RClimDex2RHtest1, or
        enter source(RHtestsV3.r) and StartGUI() to use the Transform Data button in the GUI
        mode, to produce the input files for homogeneity testing and adjustment. Nine output
        files2 are produced: for tmax, tmin and prpc, monthly (MLY) and daily (DLY), plus a
        series of monthly totals of daily prcp ≥ 1mm (*_prcpMLY1mm), and the log-transformed
        series of the two monthly total precipitation series, respectively (*_LogprcpMLY and
        *_LogprcpMLY1mm). Note that if the series to be log-transformed includes a zero value,
        each of its value plus 1 is transformed, i.e., log(P_i +1) is the transformed series of P_i.
    (3) For temperature, use one of the RHtestsV3 functions to detect the changepoints in the
        monthly data series and to determine those that need to be adjusted. Use
        QMadj.GaussianDLY3 (with this list of changepoints and the original daily temperature
        series) to obtain an QM-adjusted4 daily temperature series. The 5th column in the output
        file *_QMadjDLY_F.dat is the QM-adjusted daily series.
    (4) For daily precipitation, enter source(RHtests_dlyPrcp.r) at the R prompt to use
        FindU.dlyPrcp, FindUD.dlyPrcp, and StepSize.dlyPrcp to detect and adjust shifts in
        daily precipitation series (using pthr=0.99 mm). This is done because daily prcp > pthr is
        tested in these functions, while RClimDex uses daily prcp ≥ 1.0 mm to calculate the
        indices. The procedure for using these functions5 are the same as using the FindU and
        StepSize functions in the RHtestsV3. The QM-adjusted daily precipitation data series is
        given in the 5th column of the output file *_QMadjDLY_UC.dat or *_QMadjDLY_UDfinal.dat.
    (5) Enter source(RClimDex_RHtest.r) at the R prompt to use RHtest2RClimDex6 to
        produce the input file of adjusted daily tmax, tmin and prcp for RClimDex.
    (6) Enter source(RClimDex.r) at the R prompt to use RClimDex with the adjusted daily
        data to calculate climate indices.
1
  The command line is like this:
        RClimDex2RHtest(InputFileName, MissingStr=”-99.9”)
2
  A monthly value is assigned to missing if there are more than 3 days of missing values in the month.
3
  The command line is like this:
        QMadj.GaussianDLY(InSeries=“D:\\....”, InCs=“D:\\...._mCs.txt”, output=“D:\\....”,
                              MissingValueCode=”-999.99”, GUI=FALSE, Iadj=10000, Mq=10, Ny4a=0)
4
  “QM-adjusted” means “adjusted using the Quantile Matching algorithm”.
5
  The command lines are like these:
         FindU.dlyPrcp(InSeries=“D:\\....”, output=“D:\\...”,
                  MissingValueCode=“-999.99”, GUI=FALSE, pthr=0.99, Mq=10, Ny4a=0, p.lev=0.95, Iadj=10000)
         FindUD.dlyPrcp(InSeries=“D:\\....”, InCs=“ D:\\..._mCs.txt”, output=“D:\\...”,
                  MissingValueCode=“-999.99”, GUI=FALSE, pthr=0.99, Mq=10, Ny4a=0, p.lev=0.95, Iadj=10000)
         StepSize.dlyPrcp(InSeries=“D:\\...”, InCs=“D:\\..._mCs.txt”, output=“D:\\...”,
                  MissingValueCode=“-999.99”, GUI=FALSE, pthr=0.99, Mq=10, Ny4a=0, p.lev=0.95, Iadj=10000)
6
  The command line is like this:
         RHtest2RClimDex(PrcpFileName, column# of PrcpSeries, TmaxFileName, column# of TmaxSeries,
                            TminFileName, column# of TminSeries, OutputFileName, MissingStr=”-99.9”)

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:388
posted:2/9/2012
language:English
pages:1