features such as road intersections and shoreline features, including points 

 of marsh, were scaled from the map compiled from aerial photography. Where 

 these features were then located by field traverse and the geodetic coordinate 

 values compared, the check revealed a maximum error of ±3.0 m. This accuracy 

 is not claimed for all surveys, but it does serve as an indicator of the ac- 

 curacy of surveys conducted within NOS. 



68. The last source of potential error is the conversion of digitized 

 values to GP's. Digitizing equipment automatically recorded 1,000 coordinate 

 values for every inch of shoreline traced, which values were then corrected 

 to true latitude and longitude positions, as previously discussed. The 

 GP0LYT2 program printout provided a final error column each for "Latitude Y" 

 and "Longitude X," which were examined on each printout. In the event any of 

 the figures exceeded 0.5 mm (at map scale), the digitizing effort was rejected 

 and the original sheet was redigitized. Although the maximum allowable error 

 from this source was 4.99 m on the ground for a 1: 10,000-scale map and 9.99 m 

 on the ground for a l:20,000-scale map, rarely were the error column values as 

 high as 0.5 mm; in most cases, they were 0.2 mm or smaller. As such, the 

 possible errors from this source were more likely to be in the vicinity of 

 1.99 m on the ground for a 1: 10,000-scale map and 3.99 m on the ground for a 

 l:20,000-scale map. Since most data were finally portrayed at a scale smaller 

 than the map being digitized, the shoreline movement maps produced are well 

 within map accuracy standards. 



54 



