Simulate Forest Stand with Simulated Annealing

Run complete stand simulation to match empirical targets using simulated annealing optimization. Optimizes spatial pattern, species composition, size structure, and fire behavior metrics.

Usage

simulate_stand(
  targets,
  weights = NULL,
  plot_size = 100,
  max_iterations = 1e+05,
  initial_temp = 0.01,
  cooling_rate = 0.9999,
  energy_threshold = 1e-06,
  verbose = TRUE,
  print_every = 1000,
  plot_interval = 1000,
  save_plots = FALSE,
  nurse_distance = 3,
  use_nurse_effect = TRUE,
  mortality_prop = 0
)

Arguments

targets

List of target values (density_ha, species_props, mean_dbh, etc.)

weights

List of optimization weights (0-100 scale)

plot_size

Plot dimension (m), creates plot_size x plot_size area

max_iterations

Maximum annealing iterations

initial_temp

Initial temperature for annealing

cooling_rate

Temperature cooling rate per iteration

energy_threshold

Stop if energy below this threshold

verbose

Print progress messages

print_every

Print status every N iterations

plot_interval

Update plots every N iterations (NULL = no plotting)

save_plots

Save intermediate plot images to files

nurse_distance

Target distance for PIED trees to nearest juniper (m)

use_nurse_effect

Include nurse tree effect in optimization

mortality_prop

Simulate this proportion of dead trees after optimization (0-1)

Value

List containing trees, metrics, history, and final energy

Examples

if (FALSE) { # \dontrun{
targets <- list(density_ha = 960, species_props = c(0.55, 0.45),
                mean_dbh = 18, sd_dbh = 8, canopy_cover = 0.40, cfl = 1.16)
result <- simulate_stand(targets, max_iterations = 50000)
} # }