Forest-Management Modelling - Environmental Modelling: Finding Simplicity in Complexity - page 372

for calibrating total height (e.g. Temesgen et al ., 2008)
and stem taper (Trincado and Burkhart, 2006) but can
be extended to any equation when it is estimated with
a mixed-effects approach. Regardless of how it is done,
validation and calibration are important steps to ensuring
model predictions are reliable.

with most modern PCs (SmartForest, 2011; Uusitalo
and Kivinen, 2000). Two additional landscape visualiza-
tion tools are L-VIS (Pretzsch et al ., 2008) and SILVISO
(2011). Like ENVISION, these are very highly detailed
visualization tools but are unique in that they are tightly
coupled with a forest-simulation model (Pretzsch et al .,
2008).
Regardless of the scale, Pretzsch et al . (2008) identified
four tenets that all visualization tools should embody,
namely: (i) they should cover temporal and spatial
scales that are suited to human perception capabili-
ties; (ii) they should be data-driven; (iii) they should
be as realistic as possible; and (iv) they should allow free
choice of perspective. Most of the described visualization
tools address these tenets, but in different ways. Future
efforts are focused on providing more realistic real-time
visualizations.

23.4.2 Visualization

Many people tend to respond to visual images, leading to
the adage, 'a picture is worth a thousand words.' Much
information generated by forest models is in the form of
data tables, which are intelligible to the well initiated, but
meaningless to many, including public stakeholders and
many forest managers. Photographs of a forest may be
nearly as good at conveying an image of the conditions
as actually visiting a site, but models are used to project
conditions that do not yet exist. The best that is available
to provide an image of potential future conditions is a
computer representation of the data. One such system,
the Stand Visualization System (SVS) (McGaughey, 1997)
generates graphic images depicting stand conditions rep-
resented by a list of individual stand components, for
example trees, shrubs, and down material (SVS, 2011). It
is in wide use as a secondary tool, connected to growth
models such as FVS (2011), Landscape Management
System (LMS; McCarter et al ., 1999) and NED (Twery
et al ., 2005). Besides SVS, several other stand-level visu-
alization tools exist, such as TREEVIEW (Pretzsch et al .,
2008), Sylview (Scott, 2006), and the Visible Forest (2011;
Hanus and Hann, 1997).
At the landscape level, there are several tools avail-
able for visualization. These tools are particularly useful
for maintaining or protecting views, visualizing the land-
scape under alternative management regimes, and harvest
scheduling. The Environmental Visualization tool (ENVI-
SION, 2011) is a very powerful and realistic landscape-
level visualization tool. ENVISION uses an algorithm
that allows simulated scenes to be matched with real
photographs taken from known locations. UTOOLS and
UVIEW are geographic analysis and visualization soft-
ware for watershed-level planning (Agar and McGaughey,
1997). The system uses a database to store spatial infor-
mation and displays landscape conditions of a forested
watershed in a flexible framework (UTOOLS, 2011).
Another similar visualization tool is SmartForest (Orland,
1995), which is also an interactive program to display
forest data for the purposes of visualizing the effects of
various alternative treatments before actually implement-
ing them. The tool has been developed to be compatible

23.4.3 Integrationwithother software

23.4.3.1 Habitat models
Providing wildlife habitat has long been one of the objec-
tives of forest management. Often the availability of
habitat has been assumed if the forest is managed to max-
imize timber. Controversies such as those over spotted
owl and salmon habitat in the Pacific Northwest have
shown that sometimes forest-management practices need
to be altered to meet multiple objectives, and sometimes
objectives other than timber are of overriding impor-
tance. Habitat-suitability models have been a common
technique for formulating descriptions of the conditions
needed to provide habitat for individual species. These
models are typically generated from expert knowledge and
expressed in terms of ranges and thresholds of suitability
for several important habitat characteristics. Models that
use such techniques lend themselves to adaptation to the
use of fuzzy logic in a knowledge-based computer system.
Recent developments using general habitat informa-
tion in a geographic information system coupled with
other techniques have produced a number of promis-
ing approaches to integrating timber and wildlife habitat
modelling in a spatially explicit context. Hof and Joyce
(1992, 1993) were some of the first to describe the use
of mixed linear and integer programming techniques to
optimize wildlife habitat and timber in the context of
the Rocky Mountain region of the western United States.
Ortigosa et al . (2000) present a software tool called VVF,
which accomplishes an integration of habitat suitability
models into a GIS to evaluate territories as habitat for
particular species. Simons (2009) demonstrated a rather