PART IV: DISCUSSION 



74. Before examining results of the data collected, certain confidence 

 checks were performed. A rigorous statistical analysis could be performed in 

 subsequent reports, but direct observations of trends and selected samples 

 will provide adequate evaluation of the measurements. General characteristics 

 of the observed harbor circulation patterns will be discussed. 



Tidal Data 



Predicted tide 



75. Perhaps the most basic concern is the overall shape of the tide 

 curves over the deployment. Classic semidiurnal behavior is evident in the 

 time series of TG1S (Figure 10) . Another obvious test is to compare the 

 measured tidal data with the predicted tide for the same period. Exact agree- 

 ment is never obtained, but given the proximity of the tide station and by 

 selecting periods with low atmospheric anomalies, a close agreement can be 

 expected between observed and predicted tidal elevations. 



76. Figure 11 shows the predicted tide for 7 and 8 August (average at- 

 mospheric pressure =29.8 in. Hg, average wind speed < 7.5 kt) (NOAA 1987) 

 overlaid with measurements from TGI. A shift upward of the measured data, on 

 the order of 0.3 ft, is evident, while the overall range of 8.3 ft is matched 

 exactly. The average August departure from the annual sea level for NOS tide 

 Station 33 between 1963 and 1981 is +0.20 ft. The remaining 0.1 ft of dif- 

 ference is due to some unknown combination of gage error and predicted tide 

 error. 



Residuals 



77. A better indication of the final accuracy is obtained by comparing 

 two tide gages at the same location. A residual is calculated between any two 

 gages by subtracting one time series from another and is a plot of the instan- 

 taneous hydraulic head existing between them. Since each data set was 

 de-meaned independently, the departure from zero of the residual from two 

 gages at the same location is an indication of the overall accuracy of the 

 instruments. The best approximation to this condition occurs with Sites 1, 2, 

 and 6, all located outside the harbor breakwater at approximately the same 

 depth. 



29 



