Skip to contents

Score a simulated growth curve against its references

growth_score_fit.Rd

Ranks on SHAPE alone. Each reference series is compared against the simulated curve rescaled to that series' own plateau, so a candidate is never rewarded for landing at the right level with the wrong trajectory, nor penalized for the reverse – the level is recovered separately and exactly.

Usage

growth_score_fit(
  curve,
  ref,
  level_source = NA_character_,
  weights = c(sortie = 1, tipsy = 1, vdyp = 1, plots = 1),
  biomass_max_scale = 200
)

Arguments

curve

A tibble with age, aboveground_c_mg_ha, anpp_max and biomass_max for one combination.

ref

A list from growth_reference_curves().

level_source

Character. Which series' plateau to report a biomass_max recommendation against; NA picks the first available of SORTIE, TIPSY, plots.

weights

Named numeric. Relative weight per reference series in the ranking.

biomass_max_scale

Numeric. Passed to growth_inflation_factor().

Value

A one-row tibble of fit statistics.

Details

Errors are normalized by each series' level (nrmse_*) before being averaged across series, since a hemlock curve plateauing near 240 Mg C ha^-1 and a pine curve near 90 would otherwise contribute incomparable residuals.

The two reference kinds answer different questions, so their relative weight is a judgement the calibration must not make silently. SORTIE and TIPSY are potential yield curves for fully stocked, pure, undisturbed stands, which is exactly what a single-cohort calibration cell is. Ground plots are realized stands, spanning every stocking level and disturbance history, with whole-plot volume attributed to a leading species that holds a median of 69% of the stand. Weight them with weight_sortie / weight_plots in growth_scoring.csv; setting one to 0 drops it from the ranking while leaving it on the review figures.

See also

Other growth calibration helpers: extract_landscape_cohort_structures(), growth_add_objective(), growth_auto_window(), growth_best_candidates(), growth_bin_observations(), growth_calibration_design(), growth_calibration_partition(), growth_calibration_work_root(), growth_climatic_distance(), growth_climatic_weight(), growth_expand_over_pseudo_species(), growth_factorial_ratio_grid(), growth_fitting_windows(), growth_identifiability(), growth_inflation_factor(), growth_mortality_onset_frac(), growth_plot_palette(), growth_pseudo_species_name(), growth_reference_curves(), growth_reference_inflection(), growth_scoring_for(), growth_smooth_observations(), growth_structure_cell_curves(), growth_structure_cohort_table(), growth_structure_design(), growth_structure_summary(), growth_window_for(), plot_growth_calibration(), plot_growth_candidate(), plot_growth_factorial_sensitivity(), plot_growth_structures(), read_growth_scoring(), read_landscape_cohort_structures(), scale_linetype_growth_reference(), write_growth_review_bundle()