is the focus of this module?
The purpose of "Tropical Poison" is to have students
address questions of habitat destruction in the Amazon Rainforest
and think about biodiversity worldwide.
will students gain from this module? Students should
develop critical-thinking and problem-solving skills appropriate
for complex situations. They should come to realize that environmental
issues are multi-faceted and that solutions most often involve compromise.
is the compelling problem that students will face in this module?
Students must decide, "Is there a problem?" Students
will examine opposing views on the Amazon Rainforest and will take
a position on land-use and species conservation in one of the last
areas of biodiversity on earth.
issues will students encounter as they work through this module?
encounter a variety of issues related to decision-making and the
rainforest, all of which may be explored with remote sensing technology:
- species extinction
- global warming
- water quality
of indigenous peoples
- sustained environments
in a global economy
is the role of remote sensing in this module? The Remote
Sensing Activities in this module will demonstrate how this technology
can increase our ability to make informed decisions about habitat
and biodiversity preservation.
of the remote sensing images of the Rio Branco area gives students
a chance to get some real "hands-on" experience and to
come to know the area well. The activities are written assuming
the students know some of the basics of image processing (IP): importing
files, what a gray-scale image is, etc. For students unfamiliar
with IP, there are hot-links within the activities to existing training
activities. Herein lies an important trade-off for you, the teacher:
using IP involves the students much more deeply in the learning
process and gives them "ownership" of the IP products
and their results. However, learning IP takes some time. From experience,
we find that many students pick up IP quickly, but to learn the
skills required in these activities will take several class periods.
Requirements: Macintosh users must download NIH Image. (Instructions
for Windows users are currently under construction.)
For classes 1) with no
access to NIH Image, or 2) insufficient time to learn basic IP,
processed images for each of the activities requiring processing
are also included in this teacher's section. Using the processed
images plus the images in the student module primarily intended
for inspection only, students can obtain nearly all of the information
relevant to the rainforest module by simply downloading and carefully
inspecting the images.
From inspection of the full-scale images, the enlarged sections
shown on the "Brazil Finder map," and the color composites
given below, students should be able to recognize
1. roads (bright
line with striaght sections in Area CE)
(medium to britht irregularly curving lines in Area C)
lake (medium gray horseshoe-shaped areas in Area CW)
(large bright polygonal areas in Area NC)
farms (small irregular bright areas in Area NE)
(brilliant blue-white patches in the true color composite), and
jungle itself (green in the 3-2-1 true color composites, red in
the 5-3-2 near-IR color composites).
Differences in the apearance
of vegetation in Rain Forest 4-5-7 are partially due to the variety
of plants and partially due to differences in the availability of
water. Note the prominence of stream courses.
Here are some color composites:
the true color composite made from Brazil 1-3, 1-2, and
the deep green of the old forest, the lighter greens and
browns of the open fields, the tan of the city, and the
the IR composite made from Brazil 1-5, 1-3, and 1-2.
the deep red of the forest and the lighter reds and tans
of the farms and ranches.
The changes between the
1986 and 1992 images are obvious. The increase in area of the large
ranches, the small farms, and the city of Rio Branco is significant.
Note the decrease in dark vegetative cover in Area CE. Note also
that the cloud patterns are different between the two dates, which
can be confusing to students unless they recognize the clouds for
what they are. The best way to see the differences is to animate
or flicker between Brazil 1-2 and 2-2, or to place them side by
side on the computer monitor screen. Another way is to ratio them,
in which case an image like this one will be obtained:
In this false-color rendition, the 1986 agricultural areas are
colored red and yellow, while the 1992 areas are dark blue.
Note that the same color scheme applies to the 1986 and 1992
that the farms and ranches have been built outward from the
roads and pre-existing open areas so that farmers could avoid
the difficulty of traveling through virgin rainforest. One
can assume that the greatest growth in the near future will
be outward from the existing open areas.
Linear and area
measurements on Brazil 1-2 and 2-2 can be made on the computer using
NIH Image or from printed copies of these images. The width of the
imaged area is 185 kilometers. The most likely location for a biodiversity
preserve or national park is in the large untouched area at lower
left center of Brazil 2-2, NW of Xapuri. As the students should
find from the Brazil Remote Sensing web site, rubber trees are found
along the river south of Rio Branco. The political seat of the rubber
tappers is Xapuri in the south, so the migratory rubber tappers
themselves will tend to be between Rio Branco and Xapuri.
The increase in agricultural
area between the two images can be measured fairly easily using
the density slice command as described in the student section. The
two images below show how Brazil 1-2 and 2-2 will appear with the
appropriate density slice applied.
image has pixels highlighted in red and squares outlined in
green and blue. The location of the squares is the same (to
the pixel) in each image. The sizes of the squares were chosen
to make calculation of the areas simple. The Analyze/Measure
command will calculate the area of the highlighted (red) pixels
within the selected square. The results from measurement on
these squares are;
The difference in percent
areas between the blue square and the green square in the same image
is simply due to each square's size and location. Obviously the
percentages are not exact. Students should get similar values unless
they choose small squares centered on Rio Branco or in the middle
of the rain forest. Students should be encouraged to pick a square
representative of the whole image (ignoring the clouds). Note that
the red pixels in the middle of the forest at lower left in each
image are due to lighting changes across the image, not to farms.
As seen from the measurements
in the table above, the increase in area of the developed lands
is about a factor of 2 in just 6 years. While this may seem unrealistically
large, this increase can be compared with increases in other areas
in Brazil given in the Brazil Remote Sensing web site. The changes
can also be compared with the increase in population of Rio Branco
and the state of Acre of which Rio Branco is the capital:
Finally, the sizes of
Brazilian ranches and farms can be compared with the sizes of American
farms and ranches using:
scaled and pasted Brazilian sections are Areas NC 2 (large
ranches) and NE 2 (small farms). As can be seen from the
image, the Brazilian ranches are typically ten times larger
in area than the American farms, and the small Brazilian
farms are about ten times smaller. You might suggest that
students think about what the difference in the sizes
of farms and ranches in Brazil implies about the wealth
of farmers verses ranchers, and how the difference might
affect local politics.
Checklist--have you thought of everything?