Overview How STIM Works Growth Relationships Database Model Components (Interface) Input Output Applications and Limitations Applications Limitations

Overview

• STIM is an empirical model that integrates a stand-level sub-model with an individual-tree, distance-independent sub-model.  This integration allows the two sub-models to interact and reconcile with one another.
• It was designed to produce potential growth and yield tables for even-aged stands of coastal western hemlock and trembling aspen.

See Future Development

How STIM Works

Growth Relationships

The growth processes included in stand growth model, the individual tree model, and in the reconciliation adjustment are described below.

Stand Growth Model

This component is the focal point model of the whole model system. It is similar to that described by Hyink et al. (1988): top height increment is first derived from height-age curves; survivor basal area is next predicted as a function of top height increment and other known variables; mortality and ingrowth are predicted subsequently. Predictions are made in annual increments as follows:

• top height is incremented by applying the one-year increment to an appropriate height-age (ie. Site index) curve;
• survivor basal area increment in predicted from top-height increment, treatment history, quadratic mean diameter, age, geographic region, basal area and management options. an individual distance independent model that height growth (i.e. site index or height-age curves);
• mortality is predicted using two alternative strategies. In the first one, basal area mortality and the average size of the dead trees are predicted. In the second one, the trajectory angle defined by the net change in average stem diameter and trees per hectare of a density management diagram is predicted and used instead of basal area mortality. In both strategies, the predictions are used to calculate mortality in terms of basal area and trees per hectare; and finally
• ingrowth is predicted in terms of diameter, height and number of trees per hectare. This method is used for both the stand and the tree model.

Tree Growth Model

This component serves several purposes: it provides data on the stem diameter distribution of the grown stand, it serves as an independent check of the stand model projections, and it can be used on a stand-alone basis. The model operates by making projections of class averages. Typically, these averages are input by the user. If not, they will be generated by the model, in which case they will comprise 12 classes with an equal number of trees in each class. Apart from trees that die, a tree will thus remain within a class for the duration on the projection period. Similarly, apart from ingrowth, the number of classes will remain constant. The steps of the tree model prediction are as follows:

• the diameter growth of the trees in each class is predicted in terms of basal area, average tree diameter, basal area of the largest trees in the stand, total stand basal area, and top height increment;
• the number of trees that die in each diameter class is predicted as the probability that a tree in each class will die (which is equivalent to the proportion of trees in each class that will die. This proportion is then multiplied by the number of trees per hectare in each class to give the desired value; and finally
• the height growth of the trees in the class is predicted.

Reconciliaton

The reconciliation process is one of adjusting the output form one model to correspond with the output from the other. Generally, the tree model output is adjusted to that of the stand model, under the assumption that the later gives more robust projections at the stand level. The adjustments consists of:

• revising the grown tree list (sum of class totals) so that the number of trees per hectare agrees with the stand model projections of trees per hectare;
• by a pro-rating method, the adjustment are then applied to the individual tree classes so that the quadratic mean diameter of the two models are equal; and finally
• pro-rating is used to adjust the height estimates of the revised tree list so that the top height of the two models are equal

More information about STIM development can be found in the publication by Bonnor et al. (1995).

Database

Western hemlock: most of the data used to develop STIM is a product of the Western Hemlock Cooperative Database Project, completed by the Stand Management Cooperative in 1985. It is calibrated for both natural and thinned western hemlock stands in coastal B.C., Oregon, and Washington. The database has an age range from 7 to 110 years (breast height age), top heights from 5.5 to 50 m, and site index from 21 to 42 m. The database summary by ownership is presented in Table 1.

Cooperator	Number of Plots	Number of tree records
B.C. Ministry of Forests	369	162 538
Crown Zellerbach	146	31 978
Dept. of Natural Resources	30	35 449
ITT Rayonier	208	39 510
MacMillan Bloedel	145	67 391
U.S. Forest Service	9	3539
University of Washington	302	164 287
Western Forest Products	207	36 989
Weyerhaeuser	246	76 593
Total	1662	617 274

Table 1. Western Hemlock Database Summary by Ownership

After the validation and editing procedures the original 1662 plots were reduced to 1339 plots containing 5304 measurement sets with an average of 3.96 measurement sets per plot.

Trembling aspen: this version calibrated for both natural and thinned trembling aspen stands based on data from across Canada. The database has an age range from ? to ? years (breast height age), top heights from 12 to ? m, and site index from ? to ? m.

Model Components (Interface)

Input

Each stand scheduled for projection is either a young (seedling) stand or an older (tree) stand. The three options of the "Stand Input" selection are:

1. Generate a sapling stand: the minimum input required is site index and, optionally, top height, trees per hectare and quadratic mean diameter. The minimum value for top height of 5.5 m, and for quadratic mean diameter is 5.08 cm.
2. Generate a tree stand: the minimum input data are top height or age at breast height, and site index. Stand basal area, quadratic mean diameter, trees per hectare, diameter at the 10^th percentile, and coefficient of variation are optional.
3. Enter tree list: the minimum input required is a full set of tree and stand data including: plot size, site index, breast height age, and a stand list.

Data can also be input through the "Command" line box. Any missing data are generated STIM to ensure that a full set of data is available for input into the predicting equations.

Growth Options: A stand can be grown using three approaches:

1. under the "Run" the run option of the menu bar, select "Growth a Stand". A screen is presented for input of the number of years to growth the stand, or the size to which the stand should be grown;
2. the "Command" line box; and
3. the "Grow 5" or "Grow 10" options

Silvicultural treatments: the only silvicultural treatments available is the thinning option. The user should enter amount and type of thinning to be done.

Other options: On the menu bar, the "File" option includes the capability of accessing a filed set of stand and tree data, and of deleting one or more stands. The "Options" include model control choices:

• to run both tree and stand models, or one only;
• to reconcile to the tree or stand model;
• in selecting tree volume equations;
• in calculating mortality and generating diameter distributions; and
• in specifying ingrowth height.

It also includes a capability of setting and changing default values, and entering batch commands to process batch data.

Output

Model output can be in the form of tables or graphs.  Link to the following examples.

• Standard stand table yield report
• Tree list report
• Growth projection graphic example
• Tree list histogram graphic example

Applications and Limitations

Applications

STIM generates growth and yield information for even-aged stands of coastal western hemlock and trembling aspen. It is mainly used for:

• stand level crop planning;
• inventory updates for up to 110 years
• stand level silvicultural treatment decision-making (e.g. espacement, pre-commercial and commercial thinning); and
• forest level planning including short-term timber supply projections for natural and managed existing stands.

Limitations

Some of the limitations of STIM are:

• it does not predict the yield of complex stands (i.e. mixed-species and/or uneven-aged stands);
• it does not predict yield for stands under 5.5 m for western hemlock and 12 m for trembling aspen;
• extrapolation beyond age 110 are not recommended; and
• it does not use adjustment factors to reflect operational conditions.