|
Forest
Stand Data Visualization with POVRAY
and VRML
|
|
The
SFF uses the ray tracing program POVRAY and the Virtual
Reality Modelling Language (VRML) as data visualization
tools.
This
page explains how and why we are processing field infomration
with these visualization tools, and presents a series of
POVRAY images. The VRML page provides
a gateway to VRML forests we have created.
POVRAY
is 3D rendering software which uses scene description files
written in the POVRAY language and emits beautifully rendered,
photorealistic images. Our usage of POVRAY is somewhat unconventional.
All POVRAY output we know of is oriented towards artistic
expression. We used the program to solve a communication
problem using data visualization.
|
|
The
SFF foresters collect a large amount of data about forest stands
when we do field work. Like all foresters, we record the species,
height and diameter of trees within our sample plots, to enable
us to calculate timber volume. However, we also measure crown
class, the amount of living crown, the width of the tree crown,
tree age and radial growth rates. This is because our timber management
prescriptions are for partial cutting. We are interested in the
way each tree is growing, and its relationship with the rest of
the forest stand. Unlike clearcutters, we need to decide which
trees to cut and which trees to leave, based on the dynamics of
the stand and each tree. Only healthy trees with good growth potential
should be left on a site after logging -- partial cutting must
not equate to highgrading.
Over
the years, Peter Offerman of Lakeside Software in Nelson B.C.
has developed a custom stand information summary program for us.
The program produces tabular summaries of our field data, like
the one shown below. Now, Daniel Mack of Cave Creek Systems has
developed a new version of this summary program. Because there
wasn't enough funding for him to completely finish the project,
the SFF resident programmer is currently completing the small
bits, and writing a manual. This new version will be available
for distribution, probably by the end of August. This summary
is for a patchy, old growth forest in the Rolling Stone Creek
watershed in Clayoquot Sound. We studied this stand while preparing
a presentation for the Clayoquot
Symposium on Alternative Silvicultural Systems in March 1996.
Stand
Profile for Rolling Stone Block R20, South Portion
Height
Class |
Species |
Stems/
Ha |
Net
Usable
Volume
(m3/ha) |
 |
Height
(m) |
DBH
(cm) |
Age |
Rings
Last
2 cm |
%
Live
Crown |
Crown
Width
(m) |
%
Good
Vigor |
%
Fair
Vigor |
%
Poor
Vigor |
| Dominant |
Hemlock |
7 |
161 |
Avg |
 |
125 |
 |
35 |
35 |
8.0 |
 |
100% |
 |
| Codominant |
Hemlock |
59 |
727 |
Min
Avg
Max |
43.5
49.0
55.0 |
66
95
131 |
102
235
432 |
8
14
18 |
25
39
50 |
5.0
6.0
8.0 |
38% |
50% |
12% |
| Codominant |
Balsam |
74 |
618 |
Min
Avg
Max |
44.9
49.7
53.6 |
72
85
106 |
116
207
261 |
5
16
26 |
20
33
50 |
2.5
5.4
7.0 |
10% |
60% |
30% |
| Intermediate |
Hemlock |
22 |
25 |
Min
Avg
Max |
21.3
32.1
42.8 |
23
34
43 |
90 |
28 |
40
50
60 |
4.0
4.3
5.0 |
33% |
67% |
 |
| Intermediate |
Balsam |
37 |
80 |
Min
Avg
Max |
37.4 |
34
47
56 |
 |
 |
10
21
30 |
2.5
3.0
4.0 |
 |
20% |
80% |
| Suppressed |
Hemlock |
156 |
9 |
Min
Avg
Max |
7.0
12.3
23.4 |
8
17
24 |
 |
 |
5
24
50 |
1.5
3.0
4.0 |
14% |
24% |
62% |
| Understory |
Hemlock |
169 |
0 |
Min
Avg
Max |
2.0
5.0
8.0 |
2.0
4.0
6.0 |
 |
 |
5
60
80 |
 |
 |
30% |
70% |
| Regeneration |
Hemlock |
236 |
0 |
Min
Avg
Max |
1.0 |
1.0 |
 |
 |
80
85
90 |
1.0 |
 |
 |
100% |
| TOTAL |
ALL |
760 |
1,620 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
|
|
Our problem
was that while the Silva staff have learned to read these complex
tables fluently, we have found that the tables often communicate
less effectively to our clients.
While studying
POVRAY for recreational purposes, we hatched the idea: Why not
use POVRAY to make diagrams of forest stands? Instead of dry columns
of numbers, we could show our clients “pictures” of
the forest in question, from any angle. Also, we could “cut”
certain trees in the stand, and have before and after views of
logging prescriptions. (And of course, after we got the .POV files
working, we just had to go for full motion using VRML.)
In order to
do 3-D modelling, we added one more measurement to our field data
collection process: the location of each tree on a sample plot
in relation to the plot center. We gather this information using
a simple sketch map while doing the plot, and convert it to X,Y
coordinates in the office. We wrote a dBase program which reads
the species, diameter, height, live crown height, live crown width
and X,Y coordinates of each tree in our plots and creates a POVRAY
description of each tree. To create a forest effect, we grouped
images of our sample plots on a tight grid, and developed an algorithm
to fill the small spaces between the grouped circular plots with
randomly selected tree data, at the appropriate density. The dBase
program creates the scene description files which POVRAY uses
to create the images below. We hope to eventually distribute this
program for other foresters to use, along with our Silvacultural
Assessment program (SILVASS). Our POVRAY program will hopefully
be finished by October of 2001, and posted for distribution on
this website.
While the
scenes certainly have an acrylic feel to them, they enable us
to communicate and highlight a wealth of information about a forest
to a general audience, without the benefit of a half hour opening
lecture on data summaries and interpretations. Each small image
below is a different view of the Clayoquot forest described in
the table above, and is linked to a larger image. These pictures
are not intended to replace photographs. They are graphical presentations
of quantitative information, accurate in all spatial dimensions
and internally consistent.
These two
frames show the Rolling Stone forest at current stand density,
prior to any logging. Note the variey of stem sizes and crown
sizes. Hemlock trees are shown with dark green crowns and red-brown
stem; amabalis fir (balsam) trees have light green crowns and
gray stems.
|
| This
image shows the Rolling Stone forest after a hypothetical
uniform partial cut which removed 61% of the standing timber
volume. The proposed cutting concentrated on individual stems
with low amounts of live crown and/or poor vigor, well distributed
through the stand. This type of logging could only be carried
out using a helicopter and full lift logging. |
|
|
|
This
image shows an aerial view of the Rolling Stone forest, with
33% of standing volume removed using a patch cut system. There
is somewhat of a fisheye lens effect in the image, but it
is not unlike the view from a helicopter at low altitude. |
| The final
image shows the same patch cut, but from a lower camera angle.
This view shows how the patch cuts become less visible as
the viewer approaches sea level. |
|
For more information
on POVRAY, and examples of POV art, visit:
•
The
full report on Ecologically
Responsible Timber Management in Clayoquot Sound is in our
archives. It contains further POVRAY graphics and explanation
of the forestry issues in Clayoquot, or
•
Pop
over to the VRML page to see the VRML option.

©
2002-2003 Silva Forest Foundation
|
|