Downscale PDF {clim.pact} | R Documentation |
Functions handy for downscaling PFDs.
DSpdf.exp() is used to downscale expeoential distributions for daily rainfall according to Benestad, R.E., C. Achberger, & E. Fernandez (2005) Empirical-statistical downscaling of distribution functions for daily precipitation, met-no report, 12-2005, Climate, pp.43 (http://met.no/english/r_and_d_activities/publications/2005/12_2005/abstract_12_2005.html). The dataset exp.law1 is used for this downscaling.
CDFtransfer() is a function that facilitates local quantile transformation.
empiricalRanking() uses a formula for estimating the cumulative probability P corresponding to rank m References: Jenkinson, A.F., 1977, U.K. Met.Office Synoptic Clim. Branch Memo 58; Beard, L.R., 1943, Trans. Amer. Meteor. Soc. Civ. Eng., 108, 1110-1160; Chegodaev, N.N., 1953 (in Russian) State Rail Transport Publishing House; Folland, C. and Anderson, C. (2002), J. Clim. 15, 2954-2960, equation (1)
DSpdf.exp(obs=NULL,dT=0,dP=0,plot=TRUE,year=NULL,month=NULL) data(exp.law1) CDFtransfer(Y,CDF.2,CDF.1=NULL,method="empiricalRanking",plot=FALSE,silent=FALSE,smooth=TRUE) empiricalRanking(x)
x |
A field object. |
plot |
TRUE: plot. |
obs |
daily station object |
dT |
Projected change in annual mean T(2m), e.g. from objDS (unit deg C). |
dP |
Projected change in annual mean precipitation, e.g. from objDS (units: mm/day!). |
year |
Year to extract |
month |
Month to extract |
Y |
A data series or daily station object. |
CDF.1 |
If provided, assume this CDF to describe the cumulative distribution of Y. |
CDF.2 |
cumulative distribution predicted for the future. |
method |
Method for estimating emprical distribution function (EDF). |
silent |
TRUE: less clutter on the screen. |
smooth |
Uses a 9th-order polynomial fit to provide a smoother fit for the local quantile transformation. |
A field object.
R.E. Benestad
data(exp.law1) data(oslo.dm) a<-DSpdf.exp(oslo.dm,dT=3,dP=1) F1<- list(x=a$x,P=a$Fx.obs) F2<- list(x=a$x,P=a$Fx.chg) y<-CDFtransfer(Y=oslo.dm$precip,CDF.2=F2,CDF.1=F1,plot=TRUE) plot(oslo.dm$precip,y)