MODEL

Provides access to the underlying stand model object implemented by all Variants. Core model properties (settings) than can be altered via OSM commands include:

1) IngrowthOn True | False

If true, ingrowth is predicted by the model.

2) IsRandom True | False

If true, predictions will be random where possible.

3) SetRandomSeed Integer (seed number)

If stated, the model random number generator will be immediately reinitialized with a deterministic seed. This can be used to generate random results that are repeatable.

4) LimitStocking True | False

If true, the maximum stocking boundary is enforced.

5) ALine Percent

Percent of maximum relative density where stand self-thinning is to be enforced; e.g., 85%

6) Calibrator Object interface

Model calibrator object interface. Provides access to a number of properties and routines for validating and calibrating Variant diameter and height models against local observations.

7) TreeAgitation Object interface

Tree record partitioning rules interface. Settings alter the frequency and conditions when OSM will automatically partition tree records that represent a large number of individual trees per hectare.

8) HeightModel Object interface

Height model interface. Can be used to modify height model local calibration properties or other model settings specific to the Variant model.

9) SnagModel Object interface

Standing Snag model interface. Can be used to modify snag model behaviour and any model settings specific to the Variant model. Some Variants may not implement a snag model.

10) GeneticGains Object interface

Genetic Gain rules interface. Represents a collection of user-declared genetic gain rules. Can be used to increase the size and growth rate of trees before and during simulation, respectively.

11) GradeModels Object interface

Grade Models interface. Represents a collection of user and/or Variant tree grade models. Can be used to predict tree grades and grade transitions over time.

12) SBW Object interface

Spruce Budworm disturbance extension. Specific to Acadian Variant.

Because Model is inherited by all Variants, Variant developers have a great deal of leeway to introduce new model properties and methods specific to their Variant. Variant developers can allow users to change certain model properties through Variant specific commands. For example, the Acadian Variant permits switching between different growth and mortality models.

Consult Variant model documentation for Variant specific commands. Where documentation may not be available or is outdated, you could try the HELP command (OSM can spy into the Variant model DLL to search for any model properties that have been exposed to the OSM command model).

Validation and Local Calibration

Variant models aim to provide robust models of growth and mortality for a region. Locally, predictions may be biased for certain climates, sites, silviculture, or when influenced by natural disturbance or climate change.

OSM provides a few ways to evaluate and modify base model diameter and height predictions if local observations are available:

1) For single stand tree height calibration with simple least square species modifiers, see HeightModel calibration properties.

2) For more rigorous statistical evaluation and calibration of diameter increment, height increment, and height models for one or multiple stands, see Calibrator.

It is also possible to modify predictions, including mortality rates, through Amend commands. Amend commands can be automatically generated by the Calibrator from local observations of height or growth. Amend commands can be conditionally activated under certain stand conditions, treatments, or disturbances.

OSM manages prediction adjustments amongst these various calibration means during simulation. Amend commands take precedent over the Calibrator, and the Calibrator takes precedent over the HeightModel. In other words, within a simulation cycle, if height predictions are modified through Amend statements, then the HeightModel and Calibrator calibration (if ether is active) will both be ignored for that cycle. Likewise, if the Calibrator contains height model predictions for cycle 0, then the HeightModel calibration (if active) will be ignored.