P wave detection thresholds, Pn velocity estimates, and T wave location uncertainty from oceanic hydrophones

Citation
Pd. Slack et al., P wave detection thresholds, Pn velocity estimates, and T wave location uncertainty from oceanic hydrophones, J GEO R-SOL, 104(B6), 1999, pp. 13061-13072
Citations number
34
Categorie Soggetti
Earth Sciences
Journal title
JOURNAL OF GEOPHYSICAL RESEARCH-SOLID EARTH
ISSN journal
21699313 → ACNP
Volume
104
Issue
B6
Year of publication
1999
Pages
13061 - 13072
Database
ISI
SICI code
0148-0227(19990610)104:B6<13061:PWDTPV>2.0.ZU;2-U
Abstract
P wave arrivals recorded by the U.S. Navy's SOund SUrveillance System (SOSU S) hydrophone arrays were used to estimate earthquake detection thresholds and Pn velocities in the northeast Pacific Ocean. The Navy hydrophones have been used successfully to detect and locate oceanic earthquakes using thei r waterborne acoustic tertiary (T) waves; however, use of these hydrophones for seismic body wave detection allows regional seismic analyses to be ext ended to the oceanic environment. The P wave detection threshold of the SOS US hydrophones was quantified using the epicentral distance and magnitude o f 250 northeast Pacific Ocean earthquakes. Earthquakes with body wave magni tudes as low as 2 have detectable P wave arrivals at epicentral distances o f less than or equal to 500 km. Earthquakes with mb between 3.5 and 5 were detected similar to 50% of the time at distances of 100-1500 km, while even ts with m(b) > 5 were all detected, even out to distances of 1000-1500 km. Both P and T wave hydrophone arrival times were used to estimate the epicen ters of 100 earthquakes. The peak amplitude of the T wave coda and the onse t of the P wave were used as the earthquake arrival times to estimate event locations. T wave arrival time residuals have a Gaussian distribution with zero mean, which implies that using T wave peak amplitude is consistent wi th using the P wave onset as the arrival time. There are typically less tha n or equal to 6 stations used to derive a T wave based location, hence loca tion error ellipses are not well constrained, A Monte Carlo technique was e mployed to estimate T wave event location uncertainty. T wave locations hav e error bars of similar to 1 km in latitude and longitude when >3 hydrophon es are used for a location estimate. The detected P wave arrivals and earth quake locations were used to measure Pn velocities. Pn velocity values of 7 .9 +/- 0.1 and 8.0 +/- 0.1 km/s were found for the Pacific and Juan de Fuca plates, respectively. A Pn velocity of 7.5 +/- 0.1 km/s was measured for r ays traveling northward from the Mendocino Triple Junction along the Cascad ia subduction zone. A Pn velocity of 7.7 +/- 0.3 km/s was estimated for ray paths originating onshore western North America and traveling to the offsh ore hydrophones.