Geoscience Reference
In-Depth Information
Primary PI
Matches
MRLC
Prim or Alt
PI Matches
MRLC
Primary PI
Is Mode of
3
¥
3
Primary PI
Matches
Any 3
¥
3
Prim or Alt
PI Is Mode
of 3
¥
3
Prim or Alt
PI Matches
Any 3
¥
3
Class
Percentage
1.1
100.0
80.6
88.0
77.8
85.2
87.0
92.6
2.1
100.0
30.3
44.5
38.7
52.3
80.0
87.1
2.2
100.0
52.6
57.9
42.1
57.9
78.9
84.2
2.3
100.0
46.4
56.5
50.7
59.4
63.8
71.0
3.1
100.0
47.8
50.7
39.1
43.5
49.3
60.9
3.2
100.0
89.5
94.7
89.5
94.7
92.1
97.4
3.3
100.0
42.3
56.4
42.3
53.8
51.3
66.7
4.1
100.0
46.5
55.6
60.6
68.7
79.8
83.8
4.2
100.0
55.7
63.9
72.1
78.7
85.2
88.5
4.3
100.0
27.2
43.0
29.8
48.2
64.9
82.0
8.1
100.0
27.2
37.9
26.2
36.9
44.7
62.1
8.2
100.0
44.5
64.1
43.8
64.8
64.8
79.7
8.5
100.0
31.3
46.6
25.2
40.5
42.7
69.5
9.1
100.0
44.8
55.2
49.0
61.5
70.8
87.5
9.2
100.0
66.3
73.9
63.0
72.8
72.8
80.4
Total
Percentage
100.0
44.0
55.9
45.7
58.3
66.8
79.4
7.3.2
Issues and Problems
7.3.2.1
Heterogeneity
The heterogeneity of many areas caused confusion in assigning an exact class label to the sites.
Since the spatial resolution of the Landsat TM data was 30 ¥ 30 m, pixel heterogeneity was a
common problem (Plate 7.1a). For example, a site on the image frequently contained a mixture of
trees, grassland, and several houses. Thus, the reflectance of the pixel was actually a combination
of different reflectance classes within that pixel. This factor contributed to confusion between
evergreen forest and mixed forest, deciduous forest and mixed forest, low-intensity residential and
other grassland, and transitional and several classes.
7.3.2.2
Acquisition Dates
Temporal discrepancies between photograph and image acquisition dates, if not reconciled,
would negatively affect the classification accuracy (Plate 7.1b). For example, to interpret early
forest growth areas, the interpreter had to decide whether the site was a transitional or a forested
area. If the photograph was acquired before the image (e.g., as much as 6 years earlier), it was
clear that those early forest growth sites would show up as forest cover on the satellite image. In
this case, the interpreters decided the appropriate cover class based on satellite imagery.
7.3.2.3
Location Errors
Locating the reference site on the photo was sometimes problematic. This frequently occurred
when: (1) the LC had changed between the image and photo acquisition dates, (2) there were few
clearly identifiable features for positional reference, and (3) the reference site was on the border
of two or more classes (boundary pixel problem). When the LC had changed between acquisition
dates, locating reference sites was difficult because the features surrounding the reference site were
also changed. Similarly, when a reference site fell in an area with few identifiable features for
positional reference, the interpreter had to approximate the location of the reference site. For
