DS                 package:clim.pact                 R Documentation

_D_o_w_n_s_c_a_l_i_n_g _o_f _m_o_n_t_h_l_y _m_e_a_n_s

_D_e_s_c_r_i_p_t_i_o_n:

     Identifies statistical relationships between large-scale spatial
     climate patterns and local climate variations for monthly and
     daily data series. Calibrates a linear regression model using
     step-wise screening and common EOFs (`EOF') as basis functions.
     Evaluates the statistical relationship. Predicts local climate
     parameter from predictor fields. Works with ordinary EOFs, common
     EOFs (`catFields') and mixed-common EOFs (`mixFields'). The
     rationale for using mixed-common EOFs is that the coupled
     structures described by the mixed-field EOFs may have a more
     physical meaning than EOFs of single fields [Benestad et al.
     (2002), "Empirically downscaled temperature scenarios for
     Svalbard", Atm. Sci. Lett., doi.10.1006/asle.2002.0051].

     The downscaling analysis returns a time series representing the
     local climate, patterns of large-scale anomalies associated with
     this, ANOVA, and analysis of residuals. Care must be taken when
     using this routine to infer local scenarios: check the R2 and
     p-values to check wether the calibration yielded an appropriate
     model. It is also important to examine the spatial structures of
     the large-scale anomalies assocaiated with the variations in the
     local climate: do these patterns make physical sense? Experiment
     with both single and mixed fields. It is also a good idea to check
     whether there are any structure in the residuals: if so, then a
     linear model for the relationship between the large and
     small-scale structures may not be appropriate. It is furthermore
     important to experiment with predictors covering different regions
     [ref: Benestad (2001), "A comparison between two empirical
     downscaling strategies", Int. J. Climatology, vol 21, Issue 13,
     pp.1645-1668. DOI 10.1002/joc.703]. 

     The function `ds()' is a generic routine which in principle works
     for when there is any real statistical relationship between the
     predictor and predictand. The predictand is therefore not limited
     to a climate variable, but may also be any quantity affected by
     the regional climate. It is important to stress that the
     downscaling model must reflect a well-understood (physical)
     relationship. 

     The trend-estimation uses regression to fit a 5th-order polynomial
     (in time) to fit the observed time series. The rate-of-change is
     estimated by taking the time-derivative of this equation. If 

          y= c0 + c1 x + c2 x^2 + c3 x^3 + c4 x^4 + c5 x^5,

     where x is the time, then the rate-of-change is: 

           y= c1 + 2 c2 x + 3 c3 x^2 + 4 c4 x^3 + 5 c5 x^5.

     [ref: Benestad (2002), What can present climate models tell us
     about climate change?, Climatic Change, accepted.] 

     The routine uses a step-wise regression (step) using the leading
     EOFs. The calibration is by default carried out on de-trended data
     [ref: Benestad (2001), "The cause of warming over Norway in the
     ECHAM4/OPYC3 GHG integration", Int. J. Clim., 15 March, vol 21,
     p.371-387.].

     The downscaled scenario is saved in a text file in the output
     directory (default: 'output').

     The course notes from Environmental statistics for climate
     researchers <URL:
     http://www.gfi.uib.no/~nilsg/kurs/notes/course.html> is a useful
     reference to statistical modelling and regression.

_U_s_a_g_e:

     DS(dat,preds,mon=NULL,direc="output/",cal.id=NULL,
                   ldetrnd=TRUE,i.eofs=seq(1,8,by=1),ex.tag="",
                   method="lm",plot=TRUE,leps=FALSE,param="t2m",
                   plot.res=FALSE,plot.rate=FALSE,xtr.args="",
                   swsm="step",predm="predict",lsave=TRUE,rmac=TRUE,
                    silent=FALSE)

_A_r_g_u_m_e_n_t_s:

     dat: A climate.station object (`station.obj' or 
          `station.obj.dm').  [e.g. from `getnacd', `getnordklim' or
          `station.obj'].

   preds: The predictor `EOF'.

     mon: month or season to downscale, this is automatically  changes
          if predictor only contains a different month (this is
          normally a redundant feature).

   direc: name of directory inwhich the output is dumped (e.g. figures,
          tables).

  cal.id: ID tag used for calibration. By default use the  first field
          (`catFields') for calibration.

 ldetrnd: F for no detrending; T for removing linear trends before
          model calibration.

  i.eofs: select which EOFs to include in the setp-wise screening.

  ex.tag: Extra labelling tag for file names for experiments.

  method: Sets the method to use for regression. Method is set to "lm"
          by default, but "anm" allows the incorporation of an analog
          model, see `anm'. "anm.weight" weights the principal
          components according to the eigenvalues, whereas "anm" uses
          unweighted series.

    plot: 'TRUE' produces figures.

    leps: 'TRUE' produces EPS figures (files).

   param: Name of parameter (for plot labels).

plot.res: 'TRUE' shows statistics for residuals.

plot.rate: 'TRUE' shows analysis of rate-of-change.

xtr.args: Extra/additional arguments in the formula.

    swsm: Step-wise screening method, default=='step'; 'none' skips
          stepwise sceeening.

   predm: Prediction method, default is "predict"

   lsave: TRUE -> saves the result on disc

    rmac: TRUE -> subtracts (removes) the annual cycle in station data.

  silent: TRUE -> no output to screen.

     .

_V_a_l_u_e:

     A 'ds' object -  a list of following elements:

         X.1 .. X.n  1..nth predictor pattern for n fields (mixFields).
           lon.1 ..  Longitude coordinate of spatial fields (a vector).
           lat.1 ..  Latitude coordinate of spatial fields (a vector).
              n.fld  Number of fields (different types of predictors, mixFields).
               unit  Unit of quantity in station series.
          pred.name  Name of predictor.
            lon.loc  Longitude of predictand location.
            lat.loc  Latitude of predictand location
             yy.gcm  Years corresponding to scenario (GCM).
             mm.gcm  Months corresponding to scenario (GCM).
             dd.gcm  Days corresponding to scenario (GCM).
             yy.cal  Years corresponding to observation (Calibration).
             mm.cal  Months corresponding to observation (Calibration).
             dd.cal  Days corresponding to observation (Calibration).
               yy.o  Years corresponding to station series (obs.).
               mm.o  Months corresponding to station series (obs.).
               dd.o  Days corresponding to station series (obs.).
            rate.ds  Estimated linear rate of change of downscaled scenario.
           rate.err  Error estimate for rate.ds.
         gcm.trnd.p  P-value of linear trend in downscaled scenario.
              fit.p  ANOVA p-value for fit between large-scale and small-scale
                     variability(from regression analysis).
             fit.r2  ANOVA R2 for fit between large-scale and small-scale
                     variability (from regression analysis).
            pre.gcm  The downscaled scenario (a vector).
              pre.y  The downscaled results using the calibration data.
           location  Nsme of location of predictor.
           gcm.stat  ANOVA of linear trend fit to scrnario.
              month  Month of study (0-> all months).
             v.name  Name of downscaled element.
             region  Region used for downscaling.
            pre.fit  Linear fit to prediction (downscaled scenario) (a vector).
          pre.p.fit  Polynomial fit to the downscaled scenario.
       tr.est.p.fit  Rate of change derived from a fifth-order polynomial
                     trend-fit to prediction (downscaled scenario) (a vector).
         id.1, id.2  IDs labelling which data was used for calibration (id.1).

_A_u_t_h_o_r(_s):

     R.E. Benestad

_E_x_a_m_p_l_e_s:

     library(clim.pact)
     data("oslo.t2m")
     data("eof.mc")
     a<-DS(dat=oslo.t2m,preds=eof.mc,plot=FALSE)

