For isotropic ocean turbulence we obtain the rms travel time fluctuations, for a signal propagated along 5s ray, close to 30ms. This value is noticeably larger than signal fluctuations due to surface scattering and ice cover influences.
In the course of implementing the second approach, we obtained the broad characteristics of wave arrival groups needed to calculate the cumulative sum of signal amplitudes. The temperature dependence of the arrival-pattern duration leads to variation in the slope of the cumulative sum regression line. The analytical evaluation of the derivative of this regression line slope with respect to temperature is in agreement with the numerical results. The advantage of this method is that it is based only on relative measurements of arrivals duration. Measurement of the regression line slope does not require as sophisticated an apparatus to match the receiver and the source clock as is required for travel time measurements. The cumulative sum procedure suggested for signal processing presents the possibility of determining the average temperature along an acoustic pathway in the ocean without considering detailed arrival features (Fig. 7).
The "collective arrival time" approach demonstrate that different sensitivities can be obtained using different frequencies of the sound wave and setting different depth of the sound source. In average the "collective arrival time" processing gives the same features and sensitivity as the "cumulative sum" analysis. However, "cumulative sum" is more sensitive to the stratification variation while the "collective arrival time" is more appropriate for uniform ocean heating registration.
In the sensitivity study of acoustic propagation to current velocity changes the methods of reciprocal transmission, scintillation, and Horizontal - Refraction - Modal -Tomography (HRTM) are used. It is shown that HRMT potentially can be used in Fram Strait for transverse current monitoring. The numerical simulation shows that a measurable modal phase difference between 800 - 1600 can be expected for an averaged current of 0.3 (m/s) for acoustic frequency between 100 - 200Hz. The higher frequency the higher phase difference, but the space coherency could be lower, so it needs a trade off consideration. The computer simulations have been undertaken to demonstrate the feasibility of acoustic monitoring of current profile in cross-section of the Fram Strait. 12,000 cross-sections have been generated to simulate 3D environment of the Fram Strait. Statistical properties of the generated cross-sections were weighted with respect to real field oceanographic data of the Strait environment obtained from AMOC'97 experiment in the frame work of a‘frozen turbulence’model. Small aperture tomography scheme was considered to retrieve the spatial current distribution (K. A. Naugolnykh, et. Al., 1998b).
The method relies on the advections of small-scale inhomogeneities across the acoustic path and travel-time variations in process of signal crossing of the Strait on a number of paths to infer the intervening of fine-scale variability and transverse current. These inhomogeneities produce perturbations in the travel time of the sound, and the current can be sensed by generating a time-lagged cross-correlation of the full acoustic field. The linear four element transmission array h and the four element receiving array with equally spaced elements with total length of 15km each were used on 200km path for calculations. By combining the signals from each transmitter-receiver pair in different ways, a number of different path position were probed and profile of transverse current along the propagation path was retrieved. The ambiguity in predicting the current structure is related to the spatial scale of the turbulence and the antenna aperture. Experimental retrieval of the real current profile versus signal arrival times for different subsurface layer temperature with the prior information about the scales of the current velocity in the interval from 0.3m/s to -0.3m/s shows the ambiguity close to 16%. Current profile model takes into consideration two floes separated by the polar front which is typical for the Fram Strait cross-section along 79｡? from 6｡? to 4｡?. The eastern flow (Spitzbergen current) has a velocity directed to North (positive direction) and the western flow (Greenland current) which flows in opposite direction. See Fig. 8.
