No abstract

In support of the proposed Comprehensive Nuclear Test Ban Treaty, a large database of hydrophone recordings including T-phases, explosions, and noise has been compiled and cross referenced with known seismic events at the Center for Monitoring Research. Using this database, an automated hydroacoustic arrival detection and classification system has been developed. Detection is accomplished with a long-term-average/short-term-average power detector operating in several passbands. Station specific tuning of SNR thresholds and passband bounds allows the detector to trigger reliably on T-phases and explosions while passing over the majority of noise events such as whale calls. For each detected arrival, features such as duration, energy moments, spectral ratio, and order statistics are measured in multiple passbands from 2–85 Hz. A neural network uses these features to classify each arrival as signal or noise. Declared signals are passed to a second-stage network which classifies them as T-phases, explosions, or unknown events. T-phases arriving within a 4-min window around the time predicted from a seismic location are associated with that seismic event. These associations reveal relationships among event parameters such as location, magnitude, depth, duration, and coupling region.

In the spring of 1994, a long-range propagation experiment was conducted in the Arctic during which large time-bandwidth product signals (M sequences) were transmitted to two receiver sites. These data were processed to extract the pulse response of the propagation channel at the vertical arrays at each site and a horizontal array at one of the sites. These data are being used for an assessment of the utility of Arctic acoustic measurements for global warming signature detection and monitoring. A critical part of this measurement is the identification of the observed paths with specific propagating modes. A method has been developed based on coupled normal modes which allows us to estimate pulse responses quickly. The results agree favorably with the experimental data when historical environment data are used as the model inputs. The modeling methods used, comparisons with the measurements and implications, as well as the impact on sound speed accuracy required, will be discussed.

The time, frequency, and angular spreads imposed on acoustic transmissions by inhomogeneities in the ocean volume have been measured over 1.5- and 18-mile purely refracted paths. The dependence of these spreads on acoustic frequency has been investigated and results over the range of frequencies from 400 to 5000 Hz are presented. Fluctuations in received intensity have also been quantified and results are presented in the form of log variance as a function of range and frequency. The transition from the weak to strong (saturated) regime is clearly visible as range and frequency are increased. The large quantity of available data, the long averaging times used, and the consistency of results over many experiments, together with the availability of detailed simultaneous measurements of oceanographic parameters, make the results especially valuable in the evaluation of acoustic propagation models. [Work supported by NAVSEA, ONR, and ARPA.]

An important component of the CTBT monitoring system is the 11-station hydroacoustic network designed to monitor the oceans for unannounced underwater explosions. Explosions that do not breach the sea surface consist of a series of pulses due to gas bubble oscillations. The period of oscillation is a well-known function of the yield and detonation depth for explosives up to several thousand kilograms of TNT equivalent. The presence of a bubble pulse is a strong indicator of an explosive source and, therefore, is useful in event characterization. Examples of underwater explosions, which can be used to calibrate signal processing algorithms and propagation models, are rare. A set of explosive and other impulsive events recorded on hydrophones at Wake Island and Point Sur, CA is presented. These include the French nuclear tests, volcanic events, and chemical explosions. The set also includes pressure time-series simulated using a normal-mode propagation model. The automatic signal processor has been developed to search the real cepstrum for peaks using a noise-spectral equalization algorithm similar to those used in passive sonar. Examples illustrate complications due to range, depth, ambient noise, and waveguide distortion.