Bchron (version 3.2)

BchronRSL: Function to estimate rates of relative sea level change

Description

This function takes output from a Bchron model run and fit an errors-in-variables regression of the defined order. The results can be used to estimate rates of change.

Usage

BchronRSL(Bchrondata, RSLmean, RSLsd, degree = 1, iter = 10000, burnin = 2000, thin = 8, reportevery = 1000, BP = FALSE)

Arguments

Bchrondata

The output from a Bchron model run. See BchronMCMC

RSLmean

A vector containing the mean RSL estimates. The vector should be of the same length as the number of 'output depths' of the Bchron run.

RSLsd

A vector containing the standard deviation RSL estimates. The vector should be of the same length as the number of 'output depths' of the Bchron run.

degree

The polynomial degree for the linear regression. 1 = linear, 2 = quadratic, 3 = cubic, etc

iter

The number of MCMC iterations to compute

burnin

The number of initial MCMC iterations to disgard

thin

Keep every (thin) MCMC iterations

reportevery

How often to report the results to the screen

BP

Whether to report the results in years BP (default) or years BC/AD

Value

• A list with the following elements
• samples
  The posterior samples of the regression coefficients
• degree
  The degree of the regression (see above)
• BP
  Whether run in years before present or BC/AD
• RSLmean
  The relative sea level means
• RSLsds
  The relative sea level standard deviations
• chrons
  The posterior sample of chronologies from Bchron
• const
  The mean of the posterior samples of the chronology

Details

This function runs a non-Gaussian errors-in-variables regression on the relative sea level data taking into account the uncertainty in the chronology. The output includes error intervals for the rates, as well as the full posterior distribution in case further analysis is required.

References

Forthcoming

See Also

BchronRSLplot to plot output, BchronMCMC to create Bchron chronologies.

Examples

## See http://mathsci.ucd.ie/~parnell_a/