So, I have some difficulties on the statistics and on programming in R.

Just to the moment the nearest graph (g2) that I've found is this one:

...............................................................................................

require(car)

x<-seq(1:10)

y<-c(16,16,48.4,49.9,56,57,57,51,54,48.9)

plot(y~x)

a<-lm(y~x)

qq.plot(a,simulate=TRUE,dist="norm",envelope=0.9)

................................................................................................

On the code, y are annual maximum speeds of winds on a meteorological station in Brazil.

The problem is that on the graph of the link indicated "g1" it exists a straight on full line, and teh legend shows that it corresponds to the fitting on GEV-I or Gumbel.

It comes the following doubt: the straight on "g2" is of a normal distribution (?) 'cause the term dist="norm", so how can I change this straight on code to the correct fitting,

and how to make the vaules on ordinates of the "g2" graph assuming the y values of speed (like "g1") and not a reduced variable?

Thanks for the help.