The final
hazard assessment map (figure 9), shows
areas of low, medium and high susceptibility to damage by MPB. As
can be seen by the map, much of this part of the Vanderhoof forest
region could be susceptible to MPB damage; thus preventative measures
could be implemented in order to minimize the damage caused or completely
forestall an advancement of MPB infestation. The hazard assessment
is not an analysis of the risk but only a measure of the susceptibility
of a given stand to an infestation; thus it does not indicate when
or where an infestation may occur. A risk assessment would include
the analysis of the proximity of each stand to an existing infestation. This would provide a clearer understanding of where the MPB may
be moving next. A risk assessment was beyond the scope of this particular
project.
SOURCES OF ERROR
One of the major sources of error in this analysis was the large
amount of unknown data in all forestry datasets. Unknowns in the
forest cover data are not unusual since often, as a result of
accessibility, time or money constraints, not every stand can
be ground truthed or surveyed. An improvement to this analysis
may be the employment of some form of statistical interpolation
analysis to estimate the values of the missing data, based on
the theory of spatial autocorrelation. The main outcome of statistically
manipulating the data would still be to provide a general overview
of the susceptibility of stands in a given region, however, it
would provide a more complete hazard assessment map. Having said
this however, one would have to be wary of the outcomes as more
data manipulation would be involved.
One other source of error could be with the elevation values used to calculate
the location factor. The Trim data were contour lines and in ArcView these could not be
converted to Grid. The only method available in ArcView was to convert the contours to Tin then convert the
Tin to grid in IDRISI. Fortunately the accuracy of the elevation data was not imperitive as the location
factor only provides a general idea of whether or not the Vanderhoof area is susceptible to Mountain
Pine Beetle infestation. Since it is already understood that the North of BC is a susceptible
area, this was not too much of a concern.
Other difficulties were encountered when joining tables in ArcView
and subsequently attempting to import the files into IDRISI. It
was quickly realized that the joined shapefile had to be either
summarized to create a new shapefile or new fields from the joined
table had to be created prior to exporting. This was a function
of the join operation in ArcView.
Fortunately not many difficulties were encountered for this project.
The major consumer of time was simply gathering the data. As can
be seen from the data sources page, a variety of contacts were
required to locate all the data necessary to complete the analysis.
