Table 1. The NF1 and FBD estimates of some mean characteristics in an area of 52000
hectares.
Mean tree stem volume (m 3 /ha)
NFI
FBD
- of pine
47.4
37.7
- of spruce
43.0
56.8
- of broad leave trees
18.7
13.7
All tree species
109.1
108.2
Development classes ( % )
- open area
2.4
3.2
- seed tree or shelterwood stand
1.0
0.9
- small seedling stand
9.2
10.3
- advanced seedling stand
17.6
21.0
- young thinning stand
32.6
24.4
- advanced thinning stand
25.0
24.2
- mature stand
12.3
13.1
- seedling stand with seed trees
-
1.6
- low yielding stand
-
1.9
The estimates of the total mean tree stem volume are about the same with both
methods. Some differences occur however in tree species proportions. The percentages
of different development classes in both data are fairly close to each other except the
groups ’advanced seedling stand’ and ’young thinning stand’. One reason, in addition to
the possible confusion, may be that the definitional boundary between these groups is
not same in different inventory systems. One should also remember that FBD-estimates
are based on visual assesment..
However, this preliminary test shows that the satellite image-aided inventory method
may work fairly well at the communal level. Differences in tree species proportions are
rather high. The reasons must be traced in the future. The testing will continue with
other variables and other data. If systematic errors occur, the applicability of new
features and perhaps new classifiction techniques will be tested. The final conclusions
can be derived after thoroughly-measured test data are available.
6. THE ADVANTAGES OF THE METHOD
The satellite image-aided NFI method is expected to make it possible to:
1 ) estimate all the variables of the national forest survey lor each point of the country;
I'he variables include the properties of site type, soil, growing stock, mortality, cuffing,
and damages and diseases. If is anticipated that some improvement in estimates of
all variables can be expected compared with the current situation. The estimates of
variables concerning a single pixel may be unreliable. However, the communal level
estimates, and in the case of some variables possibly also the forest holding and stand
estimates, are expected to be applicable.
2) keep the data up-to-date in the whole country; Clouds prevent obtaining imagery