elevation of 754 dolosse, about 57 percent of the 1,326 units placed on the jetties 

 during the rehabilitation. The remaining 43 percent were not mapped since they 

 were either underwater or beneath the top layer of dolosse and not visible in the 

 photography. A portion of the initial south jetty map is shown in Figure 12. 

 Subsequent photogrammetric maps included smaller samples, thus reducing 

 costs of map compilation while still obtaining representative coverage of the 

 armor units on the two jetties. 



A comparison of photogrammetry and standard ground-leveling data (ground 

 truthing) for the initial photography suggested that the accuracy of the photo- 

 grammetrically derived elevations was on the order of ±0.09 m (±0.3 ft). Two 

 factors were identified; however, that could have contributed to these differ- 

 ences. The first was that the time frames between ground truthing and photo- 

 graphy differed by as much as 3 months. It was possible that dolos movement 

 could have occurred during these periods contributing to the apparent differences 

 between photogrammetric and leveling measurements of the same point. The 

 second factor was that there were no visual targets on the dolosse to ensure that 

 the survey crew and the photogrammetrist were observing exactly the same point 

 when measuring an elevation. Features such as "center of face of vertical fluke" 

 were the nominal targets used by the surveyors and photogrammetrist for identi- 

 fying locations of spot elevations. 



Prior to the September 1983 survey, 0.3-m (1.0-ft) black crosses were painted 

 on 111 dolosse distributed over the two jetties, ensuring that both the field crew 

 and the photogrammetrist would determine elevations at the same points on the 

 units. Comparisons of the data demonstrated that 84 percent of the photogram- 

 metric values were within ±0.03 m (±0.1 ft) of the elevations determined by 

 ground truthing, and 98 percent were with ±0.06 m (±0.2 ft). These findings 

 showed that photogrammetry was capable of accurately resolving slight move- 

 ments of individual armor units that would permit a detailed evaluation of 

 stability. 



Ground truthing data were essential in verifying the accuracy of the photo- 

 grammetric elevations. However, these data do not provide any information on 

 horizontal displacement, where both elevation and planimetric information are 

 provided by photogrammetry. 



As previously discussed, photogrammetric maps prior to September 1983 did 

 not achieve as high a degree of accuracy in measuring dolosse movement as did 

 later maps. However, an analysis of photogrammetric displacement data prior to 

 September 1983 indicated that 65 percent of the observed points were within 

 0.09 m (0.3 ft) and 91 percent were within 0.3 m (1.0 ft) of their initial eleva- 

 tions. The maximum vertical change detected was a drop of 1.3 m (4.2 ft) on a 

 dolos at the head of the south jetty. Ninety percent of the vertical displacements 

 that exceeded 0.3 m (1.0 ft) occurred on dolosse at the heads of the two struc- 

 tures. The largest horizontal displacement detected was nearly 1.8 m (6.0 ft) on 

 a dolos on the channel side of the south jetty. The next largest horizontal dis- 

 placement was only 1.1 m (3.5 ft), occurring on the head of the south jetty. The 

 mean horizontal movement of all monitored dolosse prior to September 1983 



Chapter 2 Prior Monitoring of the Site 



17 



