Clark and Yarnall 



spectra of LCT-1 data. This is apparently because they are of insuf- 

 ficient total duration in the data. They are evident to visual ins- 

 pection, however, appearing as wave trains with characteristic periods 

 in the range of 4 to 6 hours; Figure 19 presents one clearly defined 

 period of a wave train that was visually evident for 2 days. These 

 signatures have been manifest in the spectra of shorter test periods 

 which preceded LCT-1. ^^ 



The shallow water tidal signatures appear transiently in the 

 data, as do the signatures of short period internal waves, but they 

 are associated with much larger phase excursions. In general the peak 

 to peak range of all phase fluctuations on the tidal scale are one to 

 two orders of magnitude greater than the phase fluctuations associated 

 with surface waves and internal waves. On any record approaching a 

 day in length they easily dominate the data display. As might be 

 expected, the peak to peak fluctuation is progressively larger as 

 path length to the receiving hydrophone is increased-'-" e.g., sub- 

 stantially larger at H43 than at H3 . Evidence exists that the tidal 

 signatures are not a direct consequence of local tides as would be 

 measured, for example, by the Miami Beach tide gauge. The reasons for 

 this will be developed in the course of the following discussion, 

 Steinberg2 has suggested that these signatures are, in part, related 

 to tidal frequency variations of the Florida Current transport. This 

 seems a likely hypothesis but no specific attempt will be made here 

 to identify the driving mechanisms. The remainder of this section is 

 a discussion of direct environmental changes as possible causal agents 

 of the tidal signatures in acoustic phase. Three possibilities will 

 be considered: a) tidal variations in water depth (local tides) , b) 

 disturbances in the stratification of the medium which are spatially 

 coherent over a large portion of the acoustic path (we might, for 

 example, consider the possibility of internal tides, or a tidally 

 resonant internal seiche) . c) cross-channel (parallel to the acous- 

 tic path) current components which are oscillating at tidal frequencies, 

 possibly associated with tidal modulations of the main stream Florida 

 Current flow. 



The consideration of local tides as a direct cause of the 

 tidally related phase fluctuations starts with the presentation of a 

 spectral analysis of the last 4 months of LCT-1 data. The H43') phase 

 curve (Fig. 13) , was first digitally high-pass filtered with a cut- 

 off at a period of approximately 3 6 hours. The spectrum of the fil- 

 tered data is presented in Figure 20. The spectrum is still under 

 study, but the peaks marked "a" and "b" are considered reasonably 

 reliable and are associated respectively with K][, the dominant diurnal 

 tidal component at Miami, and with M2, the dominant semidiurnal com- 

 ponent. Note that in the phase spectrum, the diurnal component is 

 dominant with respect to the semidiurnal component. A ratio of about 

 5:1 is obtained for peaks a and b. This is in contrast with the 

 spectrum of local tides which, on an amplitude squared basis, exhibits 

 a 0.01:1 ratio between Kj^and M2. This gross difference between the 

 two spectra is taken as evidence that the tidally related phase signa- 

 tures at H43 are not predominantly of local tidal origin. 



326 



