Multiyear Arctic Sea ice is a heterogeneous material whose acoustic properties are functions of time and space. Over the past few years a number of investigators have studied the reflectivity of sound at the ice-water interface for sound incident from the water. The studies have considered reflections of low-and high-frequency sound for fixed times of the year. From data collected in a recent year-long field experiment conducted in the Arctic, the reflection coefficient at the ice-water interface for sound incident from the water column has been determined as a function of time. The purpose of this paper is to study the temporal evolution of the reflection coefficient with a view to explain the observed fluctuations in the magnitude of the reflection coefficient. A thermodynamic model of ice growth was used to obtain the temperature and salinity structure of the ice for each month of the year and an elastic model was used to calculate the corresponding compressional and shear wave speeds. The reflection coefficient was calculated and the results were compared for various changes in the input parameters to the models. Correlation exists between the model and field measurements of reflection coefficient during the months of September-April. The model does not predict the behavior observed during June-August. It is believed that the reason for the discrepancy is associated with a change in water column properties during this time.

Abstract. An acoustic ice tomography experiment which extended over a 12-month period was conducted in the Sabine Bay area of the Canadian Archipelago. Acoustic transmitter and receiver arrays were deployed in the ice for this experiment. The vertical arrays penetrated through the ice and extended into the water column. The acoustic data collected from the transmitter/receiver pairs in the water column are analyzed to investigate the changes in the sound speed structure of the water column. Analysis of the acoustic data shows significant changes in the sound speed structure in the upper 2 rn of the water column during the summer melt phase. We associate these changes in sound speed with the influx of fresh water caused by the melting of ice cover, snow, and runoff from rivers. Other features of the acoustic data suggest (1) that the fresh water remained stagnant and the ice growth during the winter months was due to the freezing of the fresh water and (2) that ice formed around the transducers as the fresh water came into contact with it. As the study presented in this paper was not one of the goals of the tomography experiment, support measurements to validate the conclusions from the acoustic data, such as measurement of temperature and salinity in the water column, were not made. However, the analysis presented is a demonstration of the usefulness of this technique to monitor the changes that take place in the region beneath the ice cover.

In recent years, crosshole tomography has been used to infer the velocity structure in the earth’s interior, ocean sediments, and in sea ice. In all cases, velocity estimates were obtained from travel time data for the earliest arrival. In many instances, the field measurements contain not only the information about this direct path but also about paths that have interacted with the boundaries. If one were to use the information contained in such boundary reflected paths, the resolution of the estimates could be considerably improved. This, however, implies that one knows the boundary geometry, which is generally not the case. An approach that one can take is to jointly estimate the boundary geometry and the velocity structure in the material. Different approaches that can be used to implement this idea and evaluate the method using synthetic data and field data are presented. [Work supported by ONR.]

A crosshole tomography experiment was conducted in the Arctic between April 1992 and March 1993 to determine the time evolution of the acoustic/elastic properties of sea ice. Data were taken every third day for 9 months using an array of vertical and horizontal transducers located in the ice and water column below the ice. A method to monitor the growth of the underside of the ice consisted of comparing the arrival times of both the direct path and reflected path from transmitters in the water column to the receivers in the water column. The differences in arrival times over the 9 months indicates as much as 1-m growth of ice. In addition to the ice growth estimates, the reflection coefficient was determined as a function of time and angle. Acoustic characteristics of the ice in the skeletal layer at the ice/water interface was also estimated.

The scattering of sound in ice is a complex problem which depends upon the material properties of the ice as well as the frequency of the sound source. Because of the complexity of the problem, a simplified approach to acoustic wave scattering is taken. The ice is modeled as a two-phase medium in which fluid-filled cylindrical cavities (brine channels) are embedded in an attenuating elastic matrix. An analytical solution is obtained for the acoustic scattering of longitudinal waves from an infinitely long cylinder. The solution is extended for low-ka wave propagation in order to determine effective moduli (Lamé parameters and density) and effective wave speeds. An experiment is conducted in the lab to determine the scattering cross section versus angle for a single cylindrical scatterer for frequencies of 160 and 200 kHz. In addition, effective moduli are determined as a function of cavity concentration. The experimental and theoretical results are presented. [Work supported by ONR.]