728
DD.
U +
AA
^a
ook
| :
o 9
> e
2
223995
TRAX
oF ©
£1
oo
| 22
ooo
i
Casa D
+CC+C++®=
«CMe mem
Fa hee
rrr
aussi f p
E
over-
d
oto
-1729 =
Summary of Application to:
Forest Exploitation in the Central Nigeria Forest Reserves
Lee D. Miller Darrel L. Williams
Background and Objective of the Test: Extensive areas of the African equatorial forests are
being rapidly cleared of forest cover to provide merchantable lumber and pulpwood. The
economic drive to harvest such areas is so great that a major portion of the virgin tropical
forests of Nigeria have been harvested. Often the harvesting is an iterative process consisting
of individually locating and cutting the premium wood, such as mahogany, on a tree-by-tree
basis; next the valuable sawtimber is sought and taken; and finally the area is clearcut for the
remaining pulpwood. Current forest management practices and environmental characteristics
are such that the extensive cleared areas do not readily regrow as forest cover, but either as
a persistent brushland of low potential value or as subsistence level agricultural lands.
The need to preserve islands of natural first growth forest to act as seed, gene and wildlife
reserves prompted the Nigerian Government to establish a widespread network of approxi-
mately one hundred forest reserves under the control of a National Forest Service. Unfortu-
nately this organization is new and the forest reserves so widely distributed that they cannot
be maintained under close observation. Thus, illegal cutting and exploitation of the national
forest reserves takes place. Generally the cloud cover over these areas is so persistent that it
has not been possible to collect periodic airphotos for surveillance purposes. It has proven
very expensive to have a contract aircraft stand by waiting for that one clear day over a given
area, whereas the systematic collection of satellite imagery would probably result in at least
one usable image per year for maintaining such surveillance. The objective of this small test
effort was to determine if the analysis of available Landsat imagery could be used to provide
a mechanism for detecting and monitoring the physical boundary and internal forest cover
of these forest reserves. The original analysis efforts reported here on this topic were sup-
ported by the United Nations/FAO Forestry Division, Rome. Subsequently, the digital
analysis was repeated at NASA/GSFC to provide geometrically rectified imagery and color
display products. However, the interpretations of the exploitation process noted above are
purely those of the authors.
Study Site: The general area of this test application consists of approximately one-fifth of
that Landsat frame whose location is designated as path 204, row 55. The total frame is just
east of Lagos and Ibadan and just touches the Gulf of Guinea and the Bight of Benin on its
southern boundary. Only one frame of usable Landsat-1 imagery (I.D. # 1107-09273) has
been collected to date and it was imaged on 7 November 1972. À high plains forest inven-
tory being conducted by FAO in the general area of this and adjacent Landsat frames was
initially thought to be available to provide ground control. Unfortunately two-thirds of the
November frame was covered with clouds and was unusable, thereby negating the possible
correlation of the Landsat-derived information with the existing field data. A cloud-free
test area representing approximately 7,000 square kilometers was established in the northern
portion of the image. This area had been totally cut over except for six forest reserves rang-
ing from approximately 50 to 100 square kilometers each. These reserves had been estab-
lished for 5 to 10 years at the time of imaging and had well defined legal boundaries. Thus,
close examination of the reserves with special attention to their boundaries provided an easy
test of the utility of Landsat digital image classification procedures for monitoring forest
exploitation.