-20.71 + 3.09 = -17.62 (4) 



84. The difference of 0.02 ft is on the same order as the difference 

 between NOS and CERC calculations of the datum, but considering the less than 

 ideal surveying conditions, is not significant. 



Current Data 



Overview 



85. Illustration of the velocity and direction time series of the en- 

 tire data set in one plot provides an overview of the data quality and trends. 

 Some spikes remain in the velocity after removal of the solitary spikes by the 

 error checking, or editing, routine. Before arbitrarily eliminating values, 

 each instance should be evaluated to discriminate between random signal errors 

 and hydrodynamic phenomena. Several high velocities appear in the record at 

 CM2S (Plate 11), but given the proximity of the gage to ship traffic, they 

 could be attributed to wake- induced turbulence. Most, however, are the result 

 of noisy signals, such as CM3B,, which had 103 spikes over 2.0 fps before 

 editing, approximately half of which could be removed by editing. Considering 

 that a whole data set contains about 30,000 points, this is not a significant 

 number of points. Gage CM6S recorded spikes nearly 2 percent of the time, so 

 additional filtering/editing will be required before it can be completely 

 used. 



86. Anomalies can also indicate physical events. At site 2, all three 

 gages clearly show the occurrence of the buoy failure on 30 August (Plates 11- 

 16). Note the slow revolution of meter CM2S (Plate 12) over the next 2-week 

 interval while lying on the channel bottom, most likely tangled in its own 

 mooring line. 



87 . Semimonthly spring tidal currents are quite evident in all of the 

 plots and correspond well to the spring tides measured in the specified tide 

 data plots. Daily floods and ebbs are visible as diurnal peaks in velocity 

 and as reversals in direction. 



Statistics 



88. The rose plots provide the most condensed display of the current 

 data for statistical purposes. The average velocity in each sector is less 

 than 0.5 fps, as expected. The site farthest away from boundaries, Site 4 



32 



