740
A key component in the OCRS process is its Applicon
printing system which has the capability to print
high quality black and white or colour maps at a wide
variety of scales. The Applicon system and
supplementary software provide a rapid and economical
method of producing hard-copy land use maps. Colour
separation masters for map publication can be
produced within one hour. The inherent flexibility
and relative economy of the printing system provided
a chance to experiment with a large range of map
formats and scales in the course of the study. The
Applicon system allows the operator to interact with
data to produce generalized or customized hard-copy
maps. The ability of the system to isolate and map
only those portions of an image within the river
basin was a highly desirable feature. Opportunities
to superimpose digitized information in both raster
and vector format on a classified image (e.g.,
administrative and reservoir boundaries) also proved
to be an effective tool.
not yet been undertaken due to budget restrictions,
and due to the remoteness of the study area. Based
on previous experience with this type of mapping,
OCRS has estimated an accuracy of 80-95 percent for
most cover types for the LJR project (Pala et al
1981). Detailed field investigations to be conducted
to support current planning and development studies
on the LJR should provide opportunities to better
evaluate the accuracy of data derived in this pilot
project.
KEY
ACTIVIT'
Inventory
of
Hydroelecti
Potential
(River Basin I
3 APPLICATION TO PLANNING AND DEVELOPMENT STUDIES
Based on results of the LJR pilot project, the
following sections discuss how remote sensing has or
could potentially be utilized to support or
complement planning and development studies for
future hydroelectric projects (Figure 2). Potential
problems or limitations related to certain
applications are also identified.
River Syst«
Feasibility
Studies
Economi!
Engineering
Environmer
Studies
2.1 Mapping results
The output of the LJR pilot project included eleven
hard-copy map sets, at scales ranging from 1:50,000
to 1:500,000. The entire drainage basin was mapped
according to generalized land cover types using a
geometrically corrected ground resolution (picture
element or 'pixel' size) of 50 m x 50 m (0.5 ha).
All maps produced are geo-referenced and can
therefore be used to complement, or in conjunction
with, existing data and topographic maps.
Test applications were run on a smaller part of the
basin (about 1000 km 2 ) which would be directly
influenced by hydroelectric development, (i.e., the
LJR between Mojikit Lake and Ombabika Bay - Figure 1)
in order to test the capabilities of the technology,
and to obtain more information relevant to assessing
environmental effects. The following outputs were
obtained:
1. 1:50,000 scale map of Ombabika Bay turbidity
(suspended sediments);
2. 1:50,000 and 1:100,000 scale classified (and
unclassified) theme maps of LJR, with and without
elevation contours and flooded (proposed reservoir)
area, including hard copies and one transparent
overlay;
3. 1:50,000 scale black and white map of LJR only,
with one theme (deciduous forest) highlighted in
colour;
4. 1:50,000 scale map showing only forest areas,
combined and segregated into themes; and
5. 1:50,000 scale map showing correlated land cover
types and potential moose habitat over a black and
white background.
2.2 Costs and accuracy
The project was evaluated in terms of its overall
costs based only on what it cost to produce
generalized land cover maps of the entire drainage
basin at 1:250,000 and 1:100,000 scales (e.g., not to
perform the test applications). Based on a total
study area of about 15,000 km 2 (excluding Ombabika
Bay), the approximate total cost to produce the maps
was $40,000 or $2.67 per km 2 (CDN). Sears (1985)
compared costs of other selected LANDSAT mapping
projects and found that typical operational costs can
be in the $1.50 to $2.50 km 2 range. Costs
associated with conventional data collection methods
(e.g., ground surveys and air photos) are
significantly higher, and can range from $8 to
$44/km 2 (Illinois EPA 1978) (Still and Shih 1985).
A detailed assessment of the accuracy of the
resulting classified maps (e.g., ground-truthing) has
3.1 River system planning
A major difficulty in carrying out broad-based
assessments of large study areas, particularly in
northern remote regions of Ontario, is obtaining
complete and consistent data coverage for the area in
question. The collection and management of baseline
environmental and resource use information in Ontario
is often highly fragmented and based primarily on
administrative districts established by regulatory
authorities. The scope, vintage and format of
available data can vary substantially among
districts. Data quality is often a function of
administrative policy (e.g., district land use
priorities) and/or budget-time constraints within a
district. Some regional and provincial data bases do
exist for certain land and resource use parameters in
Ontario, but these tend to be relatively dated,
one-time inventories (e.g., Forest Resource Inventory
done in the 1940's).
The land cover mapping produced for the LJR is
well-suited to broad river system level studies. The
maps provide useful, up-to-date, generalized
information on primary land uses, vegetative cover
types, wetlands, forest cutovers and burns, and
land-water ratios. These maps are useful in
providing the planner with a feel for the "context"
within which development will take place. Tests
conducted during the pilot project suggest that these
primary land cover data can be potentially extended
to provide considerably more detail with regard to
certain other resource uses within a river basin
(e.g., wildlife habitat). A 1:250,000 scale was
found to be the most appropriate format for mapping
information for the LJR basin. The appropriateness
of scale will be a function of river basin size.
Application of remote sensing represents an
acceptable intermediate level of detail and cost
between the extremes of "quick and dirty"
evaluations, which try to piece together fragmented
and diverse data sets, and detailed land use mapping
which often involves expensive and time-consuming
field data collection and air photo interpretation.
A key benefit of developing information from remote
sensing is that the data base derived becomes a
dynamic entity, that can be manipulated to assess and
highlight certain attributes; and that can be readily
updated in the future. The ability to input
digitized information (e.g., roads, park boundaries)
to a classified image adds important context to the
generalized maps produced.
An added bonus with the remote sensing based system
is the availability of statistical summaries for the
land cover attributes classified. Conventional data
Reviews
Hearings
Approval:
Design an
Constructic
(Pre-Operai
Monitorini
Compliance
Monitoring
Figure 2 - Remo
process for hydi
collection
efforts the
extensive a
successes i
(GIS) and i
(DIA) syste
utility of
planning ap
Based on
exercise, c
OCRS to pre
River basir
Mapping her
river basir
hydroelecti
portion of
3.2 Concept
Conceptual
iterative p
are any maj
hydroelectr
These studi
alternative
projects fc
environment
Environment
qualitative
readily acc
field recon
type and la