Session 4: Broad Scale Modeling-Climate and Geography
(Moderated by B. Borders)
Contents
(Guofan Shao and Herman H. Shugart)
Linking
Growth and Yield Models to Aerial Information
On
the Interrelatedness of All Forest Growth and Yield Models
Searching For Robust Silviculture Options In The
Context Of Uncertain Climate: Using FVS And Its Fire And Fuels Extension
Albert R. Stage
Mensurationist (Retired)
Rocky Mountain Research Station
USDA Forest Service
1221 S. Main Street Moscow 83843 USA
Abstract
While the prospect of a static climate is no longer tenable, the direction of change for particular localities is not yet clear. This study examines management options under four scenarios: warmer-dryer, warmer-wetter, cooler-dryer, and cooler-wetter for forests east of the Cascade Range in central Washington State. Rates and species composition of regeneration and ingrowth, and rates of mortality and accretion are expected to respond differently under the four scenarios. Stand development in this locale is strongly influenced by fire and conversely, fire behavior is influenced by stand structure. We illustrate the interplay of these hypotheses using the Forest Vegetation Simulator (FVS) and its Fire and Fuels Extension (FFE). The FVS component of this system has regional variants calibrated for most of the United States and portions of Canada. FFE has been calibrated for the inland Northwest, with additional regional variants being prepared by the Rocky Mountain Research Station with support from the Interagency Joint Fire Sciences Program.
*Presented by Albert R. Stage.
Toward Precise Forest Growth Modeling For Applications
In Full Ranges Of Geographic And Climatic Domains
Guofan Shao
Department of Forestry and Natural Resources
Purdue University
West
Lafayette 47907 USA
Herman H. Shugart
Department of Environmental Sciences
Clark Hall, University of Virginia
Charlottesville, VA 22903, USA
Abstract
The past experience has showed that some forest growth models cannot be used for extrapolations, meaning that their applications have to be restricted in the only conditions under which they were developed. In order to make a forest growth model more useful in a wider range of geographic and climate conditions, specific considerations can be incorporated into a modeling procedure, such as choosing accurate equations. It is common practice to use the coefficient of determination to test equations, either empirical or theoretical, or in the between. In addition to the conventional technique, three other different test methods are proposed in this paper. Based on theoretical concerns, extreme boundary conditions and critical point conditions can be evaluated with an extreme condition test and a general condition test. By using independent observation data, an equation should be evaluated further with an independent test if necessary. All these tests can be quantitatively performed. By comparing the evaluation results, one will be able to make a better use of the existing equations or a new one. Performing such an evaluation is the essential step toward precise modeling of forest-environment interactions. The gap forest growth models are used as examples in this paper.
*Presented
by Guofan Shao.
Incorporating monthly climatic data into growth
modeling of periodically measured, forested plots: an application of the DISP
model of Zahner and Grier
Paul F. Doruska
University of Arkansas
Arkansas Forest Resources Center
PO Box 3468, 110 University Ct Monticello
71656 USA
Abstract
Forested plots are often measured on a periodic basis as opposed to an annual basis. However, annual radial increments are often sought using such data. Climatic effects have always been considered essential in making such estimations. The Drought Index for Southern Pines (DISP) model, created by Zahner and Grier and published as a chapter in the text: "Process modeling of forest growth responses to environmental stress" (Dixon, R.K. et al. [eds.] Timber Press, Portland Oregon. 1990), provides an opportunity to make these estimations using climatic effects. The DISP model relates fluctuations in annual radial increment of southern pines to seasonal fluctuations in the Palmer Drought Severity Index in conjunction with typical monthly growth trends of southern pines. This paper examines the use of the DISP model to estimate annual radial increment within 3-year time periods and compares the results to loblolly pine (Pinus taeda L.) plantations, across the Sou theastern U.S., measured periodically thereby forming those 3-year time intervals.
Key words: loblolly pine, drought index, radial increment.
*Presented by Paul Doruska.
Linking Growth And Yield Models To Aerial Information
H.
Temesgen and V. LeMay
Dept. of Forest Resources Mgmt.
University of British Columbia
2045-2424 Main
Mall Vancouver V6T IZ4 Canada
Abstract
The province of British Columbia (BC) manages 59 million ha (145.8 million acres) forestland similar in area to the entire national forests of the US (77.3 million ha, 191 million acres). This forestland is divided into various strata (polygons) based on forest cover types for forest inventory purposes. The province has developed several growth and yield models for use in various parts of the province. Among these models, PrognosisBC (adapted from the northern Idaho variant) is used in the Southeastern part of the province. However, the estimates from this model have not been linked to the existing inventory data base nor have they been used to update the polygon level estimates.
This presentation discusses how growth and yield models can be adapted to a new geographical region and later linked to aerial information and inventory databases when there are very large and widely varied land bases and insufficient ground data. The presentation also discusses the types of data required for linking growth models to aerial information.
*Presented
by Hailemariam Temesgen.
On the interrelatedness of all forest growth and yield
models
J.C.G. Goelz
Southern Research Station
USDA Forest Service
Stoneville 38776 USA
Abstract
From a certain perspective, all forest growth and yield models are diameterdistribution models, regardless if they are called whole stand models orindividual tree models. Among the different models, the diameterdistribution and the procedure to project future diameter distributionsvaries. I catalog the inherent diameter distribution for the differenttypes of models and the procedures used to project future diameterdistributions.
Although this perspective is useful in comparing existing model forms, it isalso effective in suggesting several model forms that typically have notbeen utilized in forest modeling. (At least) two novel forms will bedescribed, ³individual tree kernel² model and ³distribution aware² standtable projection model.
Related to the discussion of choice of model form is how to test models ofgreatly different structures. If a forest stand, rather than the individualtrees in a single plot in that stand, is the unit of interest, we shouldtest models with data that represent independent samples of that stand: datafrom stands that have been remeasured using temporary sample plots ratherthan permanent sample plots. If such data is used, the more aggregatedmodels could possibly be more accurate predictors than individual treemodels.
*Presented by Jeffery Goelz.
Last revised: April 6, 2000