Surface Reflection Phase in Two Way Acoustic Signal in Oceanic Crustal Deformation Measurement

We are developing a geodetic method of monitoring crustal deformation under the ocean using kinematic GPS and acoustic ranging. The measurements are done by measuring two way traveltime of supersonic signal between a vessel, whose position is precisely determined by kinematic GPS, and transponders array (benchmark) on the ocean bottom. The goal of our research is to achieve sub-centimeter accuracy in measuring position of the benchmark by a very short-time measurement like 10 hours. In this study, we focused the under-water acoustic part of the system to improve data acquisition rate and then number of observation equations to solve the position of the benchmark with better accuracy. The measurements have started in Suruga Bay in 2003 and in Kumano Basin in 2004, which have been repeated a few times in a year. The accuracy of the benchmark positioning depends on the quality and quantity of the acoustic signal data. We are using M-sequence signal because of its robustness against ambient noises (The signal length is 14.322ms, Carrier frequency is 12.987kHz). We calculate cross-correlation between emitted and received signal and then accept the signal with cross correlation coefficient higher than a threshold. However, we often failed to achieve well correlated signals and then obtain very few traveltime data through one cruise. Sometimes in the cruise of good condition, 70 % of acoustic data have correlation coefficient above 0.7, on the other hand, only 10 % of all the data have correlation coefficient of 0.7 in bad condition cruise. We found that increase of ambient noise and contamination of later phase resembling to the main signal occurs independently each other. The ambient noise should be due to screw noise of the vessel because the noise grew up when sailing against the wind and current. On the other hand the later phases have following features: 1. Arrive in between 1 and 2 ms after the main signal arrival 2. The cross-correlation coefficient sometimes higher than that of the main signal. The vale is 0.5 at a maximum. 3. They appear more often when the benchmark located in the left side of the vessel where onboard transducer is equipped. 4. Observed more clearly in calm condition The later phases should be the signals reflected by the surface judged from the latter two features. We manually corrected the miss reading of this reflecting phase on the basis of the amplitude of the first arrival. The data acquisition rate increased from 30 % to 70 % at the maximum. In this study we manually picked up the first arrival because of high error rate of automatic pickup algorithm. Both improvement of automatically picking algorithm and hard reduction of reflection phase should be important in the first place.