recomputing the data points based on a reanalysis of the vertical angle 
correction data for each day (any out-of-level errors were not corrected for). 
Data with nearby coverage from the same or different days were overlaid and 
compared. The plan view and the cross-section view of each profile line were 
examined and suspect points, indicated by wide variations in depth between 
adjacent points, were removed. Where data from two surveys overlapped and 
didn't match well, data that best fit the rest of the data were retained. 
Confidence in a particular survey was determined by the overall smoothness of 
the data (no Geodimeter oscillation) and from the frequency and quality of the 
vertical angle error checks. Most of the data from the 1 May 1990 and the 27 
August 1990 surveys were kept. Data were deleted if there was evidence of 
sled tilt. Finally, the large gap in the data at 1,280 m longshore and 900 m 
cross-shore was filled with data from 1,100 m in the longshore. To reduce the 
total number of points only the median depth value of every three data points 
along a line was kept (limited by a maximum allowable distance and vertical 
change). Still, the raw data file includes over 10,000 points. It should be 
noted that all of the errors described in the 13-m-deep survey section result in 
less than 30 cm vertical error at the outer edge of the survey region. 
The 13-m deep survey data are shown in Figures A2 to A4. Figure A2 
shows the tracklines of all the data points used with profile numbers marked. 
Waele ee | S| | | K 
ee ell LL 
ors Ps ( ' r laa 
jah sae ‘ ' p . 7 
VARY Bass aa 9 
1500 
1300 
vi 
len] 
a 4 
wee a8 ' 
ESSA Saecel e 
EAC : 
pO zee a 
1100 
Longshore distance, m 
70 00 
Hy | 
ies 
PLE 
tue 
i 
at | 
th OU Qe 
850 1050 1250 1450 1650 1850 
Cross-shore distance, m 
Figure A2. Sled tracklines from 13-m-deep survey. Numbers indicate 
profile lines surveyed; bold rectangles indicate end points 
Appendix A Surveying 
A5 
