survey in short segments, causing the coordinate axes to be reset often. 



70. Laboratory testing also discovered inaccuracies in the digitizer's 

 axis rotation algorithms. Axis rotations of greater than 2 to 3 deg resulted 

 in unacceptable measurements over extended distances. To compensate, the 

 mylar grids remained fixed relative to the digitizer until all surveys for a 

 particular map were completed, thereby keeping the angle of axis rotation to a 

 minimum. 



71. Actual tracing of the survey line with the digitizer cursor is a 

 third potential source of error. At the 1:24,000 scale, errors resulting from 

 tracing and actual line width are approximately 3 to 4 m. However, assuming 

 such tracking errors to be random, they are dampened when averaged over finite 

 distances of shoreline. 



53 



