floodcast_ar: Forecasting future coastal flood frequency based on time...
In ccb60/SLRSIM: Tools to Analyse Sea Level Rise

Bootstrap estimates of future annual flood frequency by adding simulated deviations to predicted tides.

1	floodcast_ar(.data, .dt, .wl, .slr, .fldlvl, .wl_offset = 0)

`.data`	A source data frame. can be NULL if .dt and .wl are defined in the enclosing environment.
`.dt`	Dates and times. A data column in .data or a data vector defined in the enclosing environment that contains date-time information. Must inherit from POSIXct.
`.wl`	Water level. A numerical data column in .data or a numerical vector defined in the enclosing environment.
`.slr`	Sea level rise estimate. A numeric vector containing one or more values. What sea level rise scenarios should the analysis be based on? Selection of appropriate sea level rise estimates should be based on your time horizon and risk tolerance. NOAA forecasts of SLR for 2100 vary from 0.3 (almost certain) to 2.5 (highly unlikely) meters.
`.fldlvl`	Flood level. A single numeric value, that defines what constitutes a flood event. Must be specified in `floodcast_tub()`. In `floodcast_tub_lookup()`, will default to the "Highest Astronomical Tide" elevation, or HAT for the selected tide station.

The function simulates "random" deviations from tidal predictions in a manner that creates deviations with statistical properties similar to those of the real deviations observed in the past. We bootstrap the simulations to produce estimates and standard errors of future annual flood frequencies.

The simulations are based on an ARMA process derived from historic data. An ARMA process is a specific flavor of time series model. (The acronym reflects the fact that the model includes both "autoregressive" and "moving average" model components.)

By definition, the observed deviations during the tidal epoch have mean zero (or very close to zero). Deviations not during the tidal epoch may have either positive or negative mean.

Ideally, they would also be stationary, but in practice they often are not. As we explain elsewhere, the deviations often have a small positive slope, as the tidal predictions did not take into account sea level rise. The effect is small, and we chose to overlook it for modeling purposes.

ccb60/SLRSIM documentation built on Jan. 21, 2022, 1:31 a.m.