You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

Proceedings of the Symposium on Global and Environmental Monitoring

the "old" deforestation, and the region of study extended south of 16S in Mato
Grosso. Re-analysis of the 1978 data set, to make the results compatible with
the 1989 survey resulted in a best estimate of 54,130sq.km of deforested area
by 1978, in addition to the "old" deforestation.
The average rate of clearing over the 11.6 year period from the 1978
to the 1989 data sets was therefore 21,218 sq.km/year, .¿10%.
The above uncertainty range results from 5% uncertainty from human
inconsistency in measuring overlays of 1:250,000 scale images (tested by
Tardin et aL 1990), and the rest from the uncertainty associated with the
reanalysis of the 1978 data set.
By 1988, deforestation in the forested portion of the region was
estimated to have reached 343,972sq.km (Tardin and da Cunha, 1989),
including old secondary forest area as deforested. This estimate was originally
done with 101 LANDSAT-TM 1:250.000 images. Tardin et aL (1990) reported
the value of 358,743sq.km, the difference being attributed to the inclusion of 26
additional such images. Both estimates include 3,861 sq.km of hydroelectric
dams and 92,546sq.km of "old" secondary forest. Cloud cover over the region
in 1988 forced the interpretation team at Brazil’s Institute for Space Research
(INPE) to use (in addition to 77 images from 1988), 38 images from 1987, 6
from 1986, 5 from 1985 and 1 from 1984. The average date for the 1988 data
set is April 1988.
The rationale for not analysing the remaining 222-127 = 95 images for
1988 was the fact that a mosaic of black-and-white LANDSAT TM images in the
scale of 1:1,000,000 showed no evidence of deforestation. The possible
remaining doubt has now been eliminated with the extension of the analysis to
the 222 images for 1989, as reported above.
In order to derive now a best estimate for the deforested area by
1988, the small deforested areas previously disregarded were computed by
assuming that their extent increased linearly from 0 in 1978 to the value
measured in 1989 (of the order of 2,000sq.km), and interpolating for 1988.
The new estimate for the deforested area by 1988 is thus
362,051 sq.km ¿5%, and it corresponds to the average date April 1988.
Another deforestation rate can be also determined from the 1988 and
1989 data sets, resulting in 26,664sq.km ¿5%. This rate corresponds to a