SN:  Thanks for doing this with me Peter!

Peter Rennie:  You bet and thanks for the invitation.

SN:  I have had very little experience using forest models.   Most newcomers to forestry may not even know that models are used as quantitative tools of the trade.  You work with models in your present position and are an advocate of their use. Start with explaining what a model does.

Peter Rennie:  I'll give it a shot.  A model is a simplified version of reality, utilized to help us understand how reality works, to aid in the making of complex decisions and to understand the implications of decisions. Modeling is generally done  in a controlled environment.  For example, models are used in aircraft design to test flight characteristics without unnecessarily endangering people and equipment.  And modeling has been an important tool in forest management for a long time.

SN:   Expand a little on this forestry model as a forest management tool.  How is a forestry model actually used by the forest manager?

Peter Rennie:   In forestry, modeling is primarily done using computer programs to imitate individual stem growth, stand growth, fire characteristics, insect infestation, forest ecosystem dynamics, log inventory distribution, forest land valuation, wildlife population dynamics and strategic forest level planning to name a few.

A forest model works - and there are many variations according to the design of the individual model - by growing individual stands through to rotation, harvesting and then regrowing the stand. Models  may or may not determine which areas or "cutblocks" are to be harvested. For some models, the model defines the cutblock, while for other models, the forester must predefine the cutblocks. The model then calculates how much timber can be harvested each year sustainably over the planning period.

SN:  Ok.  Are you saying you can take a model and "grow" a forest, infest a forest with "insects", or "burn" a forest under various conditions to predict volumes or damage.

Peter Rennie:  Yes I am.  Forest modeling is used to simulate forest (as defined as a multiple of stands) growth and dynamics to undertake timber supply analysis, prepare strategic level plans and calculate allowable annual cut (AAC).

Where the forest management regime of a forest estate is to harvest a regular and expected volume of timber on a sustainable basis, either yearly or multi-year, then an annual allowable cut must be calculated. Planning of this type is done for on a long-term basis, usually over two or three rotations.  Models are used to do this.

SN:  You mentioned that in British Columbia the planning horizon is set at 250 – 300 years for publicly managed land. Modeling is extremely critical on large forested tracts with long rotations like you work on in B.C. is it not?

Peter Rennie:  I think that is true and let me explain.  The only way you can get a  handle on the future  timber supply on vast forest acreage is by using  a model.  The model helps  determine the optimal silvicultural investment in a forest, both in terms of prioritizing stands, timing and investment returns.  It would be impossible to do this without some method of prediction over vast tracts of timber.

SN:  What kind of stand models are available for foresters to use?

Peter Rennie:  There are two general types of forest estate models - spatial and aspatial.  These models can use one or many modeling techniques and algorithms, such as simulation, optimization, linear programming, simulated annealing etc. Each of these techniques has distinct advantages and disadvantages depending on the problem to be addressed.

Spatial estate models work with graphic forest information systems such as Geographic Information Systems (GIS), to not only calculate a sustainable annual allowable cut, but also show where that timber should be harvested from. Aspatial models will only give a calculated annual allowable cut, and it becomes the job of the forest manager to turn this strategic plan into an operational plan.

SN:  I understand that models can either grow the stand based on certain "rules" or the model itself is designed to grow a predetermined forest.  Please explain these two concepts.

Peter Rennie:  That's right.  Two analysis methodologies must also be considered - rule-based management systems and target based systems. In a rule-based system the forest owner will define a set of rules that the model follows, with a disregard to the way the forest will look at the end of the planning period.

An example of harvesting rules would be: Cutblock size less than 40 ha, cutblocks can’t be located closer than 400 m to another until the regeneration in the first block is greater than 10 m high. Rules are established to try and meet objectives, but their effect over a long term planning horizon often cannot be properly visualized.

In a target-based model, the look of the forest at the end of the planning horizon is determined and the model works to create this effect in the shortest amount of time. Often long term planning targets are defined by percentage of "old growth" or seral age and patch size distribution.

SN:  I have to ask this.  Are computer models really to be trusted and used routinely?

Peter Rennie:  Although considered by many to be a black box, computer models play an important role in understanding tree growth, nutrient cycling, stand dynamics, forest dynamics and the interactions of many non-timber forest resources. The implications of decisions made on complex forest systems become too difficult to understand on a surficial level.