extRemes (version 1.61)

rlplot: Create a return level plot for a fitted object of an extreme-value distribution.

Description

Plots several return levels against the return period for a fitted object from one of the ismev functions: gev.fit and gpd.fit.

Usage

rlplot(z, ci = 0.05, add.ci = FALSE)

Arguments

z

A list object as returned by one of gev.fit or gpd.fit (with appropriate class attribute added).

ci

The (1-ci)*100 confidence value.

add.ci

logical if true will add confidence bounds to plot.

Value

• A plot is created. If assigned to an object, a list will be returned with the following items.
• period
  The return periods used for calculating the return levels.
• level
  The estimated return level for each return period.
• lower
  If add.ci is TRUE, then this is a vector of lower 1-ci confidence bounds. Otherwise the value is NULL.
• upper
  If add.ci is TRUE, then this is a vector of upper 1-ci confidence bounds. Otherwise the value is NULL.

Details

Given a fitted list object from gev.fit or gpd.fit--attributed with the class "gev.fit" or "gpd.fit", respectively--the return level plot is generated. Confidence bounds, if included, are found by the delta method, which is generally appropriate for shorter return periods, but not for longer return periods because the return level distribution is generally skewed. Therefore, if a plot with better estimates of the confidence bounds are desired, use add.ci=FALSE, and use the R function lines to add different bounds (e.g., using values obtained from the gev.parameterCI or gpd.parameterCI functions).

This function is simply a modification of Stuart Coles' functions gpd.rl and gev.rl (Coles, 2001).

References

Coles, Stuart. "An introduction to statistical modeling of extreme values", Springer-Verlag (London), 2001.

Gilleland, Eric and Katz, Richard W. Tutorial for the 'Extremes Toolkit: Weather and Climate Applications of Extreme Value Statistics.' http://www.assessment.ucar.edu/toolkit, 2005.

See Also

gev.parameterCI, gpd.parameterCI, gev.diag, gpd.diag

Examples

data(ftcanmax)
fit <- gev.fit( ftcanmax[,"Prec"])
class( fit) <- "gev.fit"
rlplot( fit)