6
39
The Development of a DTM
Most digital terrain models that are produced are byproducts of
some other process. The digital data gathered in the process of
scanning profiles in the preparation of orthophotos is a common
way to collect data for the generation of a DIM.
There are a number of problems in using data so gathered to de-
velop an efficient DTM. Following are some of the compromises
that are made:
1. Prominent breaklines such as streams and ridges may be mis-
sed if they run parallel to the scan line.
2, [t is not practical to identify drain or ridge lines.
3. When scanning a stereomodel along the profile line, one is
watching a moving target and it is less likely that the
"dot" will be right on the ground under these conditions.
Better accuracy is achieved by bringing the "dot" to a stop
and carefully setting it on the ground at a specific spot.
4. There is a tendency to gather more data than is necessary
since there is bound to be a considerable redundancy of data.
How many times have we heard of the problem of thinning data
gathered in the high-speed automated systems? This cry is
because the computer, with all its power, cannot economical-
ly handle so much data.
5.. Accuracy of terrain data is always a problem but there is a
tendency to obtain irregular data in proportion to the sophis-
tication and speed of the scanning system.
The following is a procedure that is being successfully used to
define terrain by numbers.
The slope of a line between any two adjacent terrain points of
the DTM will conform to the slope of the ground between the same
two points. Essentially, what this means, is that the position
of the selected points will fall on the breaklines of the terrain.
All of the computer algorithms hereafter referred to are designed
to operate on terrain data which meets this criteria.
How does one gather data that will meet this criteria?
1. One way is to digitize existing topographic maps with an X Y
digitizer that provides a means of entering the contour ele-
vation from a keyboard. A topographic map can be digitized in
such a manner that the data will meet the criteria.
A line following technique is not generally recommended be-
cause of the complications that seem to arise when one attempts
to automate this. Chances are that an automated line following
system will result in large amounts of redundant data and will
increase the computer costs of manipulation with little gain
in accuracy.