tsreg: Monitor post-disturbance regrowth

In bendv/rgrowth: Post-disturbance regrowth monitoring with dense LTS

Description

Structural monitoring method for post-disturbance forest regrowth in a time series

Usage

tsreg(x, change, h, dates = NULL, type = c("irregular", "16-day"),
  startOffset = 0, formula = response ~ harmon, order = 1,
  history = "BP", level = 0.05, w = 3, s = 1, plot = FALSE,
  ylabs = c("data", "|MOSUM|"))

Arguments

x	Time series. Can be an object of type numeric (vector), zoo or ts
change	Numeric. Date of original disturbance. See bfastmonitor for a potential method to determine this on a time series.
dates	Date. Vector with length length(x) with observation dates. If omitted, x will simply be an ordered series with arbitrary (regularly spaced) dates
startOffset	Numeric. Number of (decimal) years by which to offset the start of the regrowth monitoring process. Can also be set to "floor", in which case the floor of the change date will be taken as the start of the monitoring period.
formula	Formula to be fit to stable history period (see bfastmonitor)
order	Numeric. Order of the harmonic term of formula (see bfastmonitor)
history	Method or pre-determined period to use for history model (see bfastmonitor)
level	see bfastmonitor
w	Numeric. Number of years before a regrowth flag is allowed to be assigned.
plot	Logical. Plot the results?
ylabs	Character. Vector of length 2 with y labels of the data time series and MOSUM plots, respectively

Value

object of class tsreg

Author(s)

Ben DeVries

See Also

regSpatial

Examples

data(ndmi)
plot(ndmi, type='b', cex=0.5)

# breakpoint detection with bfastmonitor
x <- bfastts(ndmi, time(ndmi), type="irregular")
bfm <- bfastmonitor(x, start=c(2005, 1), formula=response~harmon, order=1, plot=TRUE)

reg <- tsreg(ndmi, change = bfm$breakpoint, h = 0.5, plot = TRUE)
print(reg)

reg2 <- tsreg(ndmi, change = bfm$breakpoint, startOffset = "floor", h = 0.5, plot = TRUE)
print(reg2)

reg3 <- tsreg(ndmi, change = bfm$breakpoint, startOffset = "floor", h = 0.5, history='all', plot=TRUE)
print(reg3)

reg4 <- tsreg(ndmi, change = bfm$breakpoint, startOffset = "floor", h = 0.5, history='all', s=0, plot=TRUE)
print(reg4)

bendv/rgrowth documentation built on May 12, 2019, 10:58 a.m.