already
A
cover
can be
nap.
ealthy
valley
on the
| also
of the
erised
paired
1976
oarser
three
ed, as
Then
ed, in
ges in
nages
ion of
on of
ludes
edles
ot be
data.
edles
forest
es of
local
storey
anopy
, the
rds a
le to
edles
h lost
were
class
ls) at
> and
h are
lands
As
sified
of the
high
nplex
Jude,
ns of
ACY.
S -
1990
rest.
The primary changes were as follows (Zawila-
Niedzwiecki, 1994):
® increase in the area comprising moderately impaired
spruce stands by 30%;
® decrease in the area of dying and dead tress from
15% to 5%;
® increase in the area of openings and clear-cuts from
8% to 15% with domination of vegetated openings in
1990.
The comparison of the classified image with forest maps
revealed good consistency between the classes
determined from satellite data and stand characteristics
contained at forest maps. Thus, the characteristics of
stands, obtained from classification of Landsat TM image
is suitable for large-area description of health conditions
and structure of stands in this region.
Comparative analysis of 1976 and 1990 classifications
enabled evaluation of extent of deforestation within a 14-
year period, as well as assessment of the changes of
other forest categories. The following conclusions were
drawn from this analysis:
= the area of healthy/slightly impaired stands was two
times larger in 1976 than in 1990:
* there was 2.5% more moderately impaired stands in
1990;
» heavily impaired stands covered 5.5% less area in
1990;
Table 1
Comparison of the area of forest classes in western Sudety
= the most significant difference exists for
deforestations. They cover 3.5 times more area in
1990 than in 1976.
The deforestation category is spectrally homogeneous on
the 1976 MSS classification image, while it is
differentiated on 1990 TM image. Hence we decided to
distinguish three classes within this category
characterised by various vegetation cover. Detailed
comparison of 1976 and 1990 classifications was
presented in Table 1.
An analysis of forest quality with the use of Landsat
satellite data has been also done for the eastern Sudety -
Sowie and Stolowe Ranges. In case of this region it was
found that heavy and moderate forest damages occur
only on a relative small areas - forests are less degraded
than in western Sudety. Hence the analysis was
concentrated on differentiating coniferous and deciduous
stands, located on various forest sites.
In summary we conclude, that this research conducted at
two test areas demonstrated the usefulness of Landsat
TM satellite images for large-area classification of forests.
They also permitted an appraisal of the health conditions
and structure of stands, enabling the evaluation of
degradation processes, that occurred between 1976 and
1990.
Forest category
Area (in ha)
MSS 1976
Area (in ha)
TM 1990
Healthy/slightly impaired spruce stands
18 529 - 32.696
9 919 - 17,5%
Moderately impaired spruce stands
10 380 - 18.3%
11 792 - 20.8%
Heavily impaired spruce stands
9 938 - 17.596
6 831 - 12.096
Mixed stands
8 636 - 15.296
10 104 - 17.896
Deciduous stands 2497 - 4.4% 435- 0.8%
Vegetated openings 5528-9.7%
Openings 4 894 - 8.8% 6 752 - 11.996
Clear-cuts 4 434 - 7.8%
REFERENCES
Ciolkosz A., lracka M., Zawila-Niedzwiecki T.. 1987
Monitoring of forest decline in Poland by remote sensing.
Proceedings of the Seminar on Remote Sensing and
Forest Pollutants, held in Laxenburg, Austria, March 11-
12, 1987, IIASA-EPRI, EA-575, Chapter 7.
Ciolkosz A., Zawila-Niedzwiecki T., 1990. Remotely
sensed data and limitation of forest productivity in Poland.
Nature & Resources, Vol. 26, No 1; 41-44.
Horler D.N.H., Ahern F.J., 1986: Forestry information
content of Thematic Mapper data. Int. J. Remote Sensing,
Vol. 7, No 3.
Intemational Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 7, Budapest, 1998 803
