Reads MPD or MCMC estimates and data from iSCAM file structure into an operating model

Functions for importing an iSCAM assessment. From a fitted model, iSCAM2OM populates the various slots of an operating model and iSCAM2Data generates a Data object. These functions rely on several functions written by Chris Grandin (DFO PBS).

Usage

iSCAM2OM(
  iSCAMdir,
  nsim = 48,
  proyears = 50,
  mcmc = FALSE,
  spawn_time_frac = 0,
  Name = "iSCAM model",
  Source = "No source provided",
  length_timestep = 1,
  nyr_par_mu = 2,
  Author = "No author provided",
  report = FALSE,
  silent = FALSE
)

iSCAM2Data(
  iSCAMdir,
  Name = "iSCAM assessment",
  Source = "No source provided",
  length_timestep = 1,
  Author = "No author provided"
)

Arguments

iSCAMdir: A folder with iSCAM input and output files in it. Alternatively, a list returned by load.iscam.files.
nsim: The number of simulations to take for parameters with uncertainty (for OM@cpars custom parameters)
proyears: The number of MSE projection years
mcmc: Logical, whether to use mcmc samples to create custom parameters cpars. Alternatively, a list returned by read.mcmc. Set the seed for the function to sub-sample the mcmc samples.
spawn_time_frac: Numeric between 0-1 indicating when spawning occurs within the time step
Name: The name of the operating model
Source: Reference to assessment documentation e.g. a url
length_timestep: How long is a model time step in years (e.g. a quarterly model is 0.25, a monthly model 1/12) (currently only uses annual time step)
nyr_par_mu: integer, the number of recent years to estimate vulnerability over for future projections
Author: Who did the assessment
report: logical should a numbers at age reconstruction plot be produced?
silent: logical should progress reporting be printed to the console?

Biological parameters

The function calls model <- load.iscam.files(iSCAMdir) and grabs the following matrices:

model$mpd$d3_weight_mat - fecundity (product of weight and maturity at age)
model$mpd$ma - maturity at age

MPD historical reconstruction

The function calls model <- load.iscam.files(iSCAMdir) and then grabs the following matrices:

model$mpd$N - abundance at age
model$mpd$F - fishing mortality at age
model$mpd$M - natural mortality at age

If a delay-difference model is recognized, then the following is used instead:

model$mpd$F_dd - fishing mortality at age
model$mpd$M_dd - natural mortality at age

Abundance at age is reconstructed using model$mpd$rt (recruitment) and projected with F_dd and M_dd to match model$mpd$numbers.

MCMC historical reconstruction

If mcmc = TRUE the function calls mcmc_model <- read.mcmc(iSCAMdir), and grabs nsim sub-samples of the MCMC output through the following arrays:

mcmc_model$params and mcmc_model$ft - fishing mortality at age from the fleet selectivity parameters and apical F's
mcmc_model$m - year-specific natural mortality at age
mcmc_model$params$rinit and mcmc_model$rt - recruitment and abundance

Start age

While the iSCAM start age can be greater than zero, abundance at age is back-calculated to age zero with M, maturity, growth = 0. In this way, the stock-recruit dynamics from iSCAM are preserved.

These arrays are then passed to Assess2OM to generate the operating model.

Reference points

iSCAM calculates the stock-recruit relationship and subsequently a single set of MSY and unfished reference points using R0, steepness, and unfished spawners per recruit from the mean M, fecundity, and growth (mean with respect to time).

R0 and h are recalculated for the operating model by obtaining the stock-recruit alpha and beta from the iSCAM parameters and the mean unfished spawners per recruit in the first ageM (age of 50% maturity) years.

Author

T. Carruthers, Q. Huynh