100'
be more economical to use in relation to aerial
photography.
Budworm infested white spruce
iHealthy white spruce
* Black spruce
Q Water
,63 - .69 micrometer intensity
Figure 6. Scattergram outlines of MEIS-II bands 4 and 6
for five classes at 10 m resolution.
COMPARISON OF MEIS-II AND TM DATA
The MEIS and TM data were both geometrically cor
rected and resampled to 10 m in order to compare the
data sets. MEIS-II bands 2, 4, and 5 and TM bands 1,
2, and 3 were each combined to give natural colour
composites. MEIS, with its higher radiometric reso
lution illustrated the budworm infestation more
clearly and with less confusion with other vegetation
types than TM. The TM colour composite with its
higher noise levels and wider spectral band ranges
demonstrated budworm infestation but with greater
confusion with other vegetation types, especially in
the wetland areas.
Spruce budworm infestation was detectable with the
MEIS-II data due to the superior spatial and radio
metric resolution of the sensor. The major factor
is, however, the narrow spectral bands, particularly
the red band which is well positioned relative to the
chlorophyll absorption minimum at.675 m of vegetation
reflectance. There is some overlap between the bud
worm infested spruce and healthy spruce but this is
minimal, especially in the drier areas.
ACKNOWLEDGEMENTS
The authors would like to thank J.D. Heyland,
F.J. Ahern and D.G. leckie for their critical review
of this paper.
REFERENCES
Ahern, F.J., W.J. Bennett and E.G. Kettela, 1986.
Surveying spruce budworm defoliation with an air
borne pushbroom scanner. Journal of Photogram-
metric Engineering and Remote Sensing (in press).
Kucera, D.B. and R.G. Taylor, 1984. Spruce budworm
situation in North America 1983. U.S. Dept, of
Agriculture - Environment Canada, Mise. Publ.
No. 1441.
McColl, W.D., R.A. Neville and S.M. Till, 1984.
Multi-detector electro-optical imaging scanner
MEIS-II. Proceedings of the 8th Canadian Symposium
on Remote Sensing and 4th Conference of
l'Association québécoise de télédétection,
Montreal, Quebec, May 1983, pp. 71-79.
McLean, D.A. and T.A. Erdle, 1984. A method to
determine effects of spruce budworm on stand yield
and wood supply projections for New Brunswick. The
Forestry Chronicle, June 1984, pp. 167-173.
Stanley, L.W. and R.J. Reed, 1986. Report on the
spruce budworm infestation situation in
Saskatchewan as of September 1985, Saskatchewan
Parks and Renewable Resources Internal Report,
6 pg.
Taylor, M.M., 1973. Principal components colour
display of ERTS imagery. Proceedings of the 3rd
ERTS Symposium, Vol. 1, Section B, NASA,
Washington, D.C., pp. 1877-1879.
CONCLUSION
The results from this study indicate that it is
possible to map current year spruce budworm infesta
tions from natural colour images. Severe spruce
budworm infestation appears as a reddish-brown dis
colouration in data from an electro-optical pushbroom
scanner and Thematic Mapper data. This suggests that
more accurate maps could be obtained by this method
than by traditional aerial sketch mapping. This
capability could improve program planning and effec
tive assessment of budworm infestation for cutting.
It could ultimately improve the wood supply through
better growth-loss predictions for harvest
scheduling.
Detection, of varying degrees of severe current
year budworm infestation was possible by using data
from the pushbroom scanner and TM data. The ability
to detect budworm infestation with the TM sensor is
diminished due to its spectral band placement,
greater band widths, and lesser signal-to-noise
ratio.
In Saskatchewan the areas affected by spruce bud
worm are quite small. With the confusion between
budworm infested areas and other vegetation types,
especially in the wetland areas, no improvements in
mapping accuracy are obtained using either TM data or
MEIS-II data over aerial photography with its
subsequent ground check. Should, in the future,
larger areas be affected then MEIS-II or TM data may