Table B.6 Precision and Accuracy of 5 1 N data. 
5 15 N 
Precision, % 
99.1 (n=23) 
Accuracy, % 
96.6 (n=63) 
Aerial Mapping of Seagrass and Macroalgae 
The remote sensing procedure used in this study to map the intertidal distribution of 
eelgrass and benthic green macroalgae (conducted from 1997 - present) utilizes aerial 
photography with false-color near-infrared (color infrared, OR) film. This allows an aerial 
survey to be conducted during daylight low tide (typical tide level about -0.5 m MLLW) when 
the majority of the eelgrass habitat in the Yaquina Estuary is exposed. CIR film has been found 
to provide substantially better spectral resolution of exposed intertidal vegetation than has true 
color film (Young et al., 1999). The mapping method is able to detect inundated and submerged 
Z. marina to a depth of about 1 m below water level at the time of the aerial photograph. To map 
perennial eelgrass habitat, the surveys are conducted in late spring or early summer before the 
summer bloom of benthic green macroalgae that can interfere with the classification of eelgrass 
habitat. Mid-summer surveys are used to map macroalgal distributions upslope of the eelgrass 
meadows. Photoscales utilized range from about 1:6,000 to 1:20,000. The aerial photographs 
are digitally scanned and georectified while correcting for terrain and camera distortions to 
produced digital orthophotos. The spatial accuracy of the photomap for this estuary (photoscale: 
1:10,000) was assessed by comparing 14 Root Mean Square Error (RMSE) offset values for 
positions of photovisible objects obtained from the photomap, referenced to published National 
Geodetic Survey (NGS) positions. The mean offset was 0.72 m + 0.27 m (95% Cl; Clinton et 
al., in review). The digital orthophotos are classified into eelgrass and bare substrate habitats, 
defined as > 10% cover or < 10% cover, respectively. On-the-ground resolution of 0.25 m is 
obtained in this process. A hybrid technique using both unsupervised and supervised 
classification steps has been developed for this habitat mapping project (Clinton et al., in 
review). The technique requires training data from ground truth surveys, with station positioning 
accomplished by a differential-corrected global positioning system (GPS). The RMSE of GPS 
positions obtained at an NGS first-order monument in Yaquina Estuary was 0.62 m. 
142 
