Rhett Butler

The global seismographic network surpasses its design goal

This year, the Global Seismographic Network (GSN) surpassed its 128-station design goal for uniform worldwide coverage of the Earth. A total of 136 GSN stations are now sited from the South Pole to Siberia, and from the Amazon Basin to the sea floor of the northeast Pacific Ocean—in cooperation with over 100 host organizations and seismic networks in 59 countries worldwide (Figure 1). Established in 1986 by the Incorporated Research Institutions for Seismology (IRIS) to replace the obsolete, analog Worldwide Standardized Seismograph Network (WWSSN),the GSN continues a tradition in global seismology that dates back more than a century to the network of Milne seismographs that initially spanned the globe. The GSN is a permanent network of state-of-the-art seismological and geophysical sensors connected by available telecommunications to serve as a multi-use scientific facility and societal resource for scientific research, environmental monitoring, and education for our national and international community.

Global seismographic network records the Great Sumatra-Andaman earthquake

On 26 December 2004 the Indonesian sub-duction zone near the northern end of Sumatrabegan to rupture at 58 minutes,47 secondspast midnight Greenwich Mean Time.The rup-ture continued for approximately seven min-utes,extending northwestward along the SundaTrench for roughly 1200 km to the AndamanIslands.The seafloor displacement generateda massive tsunami that swept ashore with 10-mamplitude in northern Sumatra and expandedacross the Indian Ocean and Andaman Sea,striking Sri Lanka and Thailand within twohours of the rupture.Confirmed deaths alongthe coastlines of 11 Indian Ocean nationsexceed 220,000,marking this as one of themost lethal natural disasters in human history.The 2004 Sumatra-Andaman earthquake isthe largest event since the 1964 Good FridayAlaskan earthquake (

Regional seismic observations of the Ontong Java plateau and East Mariana Basin

P-waves recorded on Ponape Island at the northern end of the Ontong Java plateau have been investigated. Different modes of propagation in the distance range 12° to 17° Δ between paths from Melanesian earthquakes across the Ontong Java plateau and paths from Mariana earthquakes across the eastern Mariana basin suggest that the mantle underlying the regions is not homogeneous. Travel-times of P-waves beneath the Ontong Java plateau are slower than for paths beneath the east Mariana basin. The frequency content of the first-arriving P-waves is lower for paths across the Ontong Java plateau than for the east Mariana basin. The disparate crustal thicknesses of the Ontong Java plateau and east Mariana basin may influence the relative amplitude of oceanic Pn, with smaller amplitudes corresponding to a thicker crust. Pn-P differential times for the Ontong Java plateau show a general decrease with the depth of earthquake focus but no comparable trend is seen in the Mariana data, possibly because of the scatter. The observation of high-frequency oceanic Pn propagation across the Ontong Java Plateau is suggestive that the plateau is not of continental origin.

Seismic noise on Rarotonga: Surface versus downhole

Seismic noise data are presented from the new Global Seismographic Network station, RAR, on the Island of Rarotonga in the South Pacific. Data from the first new borehole site in the GSN are compared with a surface vault installation. Initial indications from the data show that borehole siting on a small island significantly reduces long-period (>20 s) horizontal seismic noise levels during the daytime, but little or no improvement is evident at periods shorter than 20 s or on the vertical component. The goal of the Incorporated Research Institutions for Seismology (IRIS) GSN program is broad, uniform coverage of the Earth with a 128-station network. To achieve this goal and provide coverage in oceanic areas, many stations will be sited on islands. A major siting consideration for these new stations is whether to build a surface vault or drill a borehole. Neither option is inexpensive. The costs for drilling a cased hole and a borehole sensor are large, but the benefit of a borehole site is that seismic noise is reduced during certain periods when a surface installation may be subject to wind, weather, and thermal effects. This benefit translates into recording greater numbers of smaller earthquakes and higher signal-to-noise ratio.

2003-2004 upgrades and additions to the Hawaii-2 Observatory

A permanent deep ocean scientific research facility the Hawaii-2 Observatory (H2O) was installed on the retired HAW-2 commercial submarine telephone cable in September 1998. H2O consists of a seafloor submarine cable termination and junction box in 5000 m of water located halfway between Hawaii and California. In 2003, a major upgrade to the communications architecture was completed, and the buried broadband seismometer the previously installed at H2O was augmented by two seafloor geomagnetic observatories, the small experiment module and the benthic biology experiment. During 2004, a borehole seismometer will be installed at ODP Hole 1224D located /spl sim/1.5 km from H2O. All of these experiments will be linked to shore in real-time.

High‐Frequency (>100 Hz) Earthquakes North of Moloka‘i Detected on the Seafloor at the Aloha Cabled ObservatoryHigh‐Frequency Earthquakes North of Moloka‘i Detected on the Seafloor

I describe the analysis of 10 earthquakes (ML 2.7–4.7) that occurred between 2012 and 2017, at depths of 10–33 km, north of Moloka‘i, Hawaiian Islands. Observed on the seafloor Aloha Cabled Observatory (ACO) 24 kHz hydrophone, the events show extraordinary high frequencies up to 165 Hz at distances between 130 and 213 km. The data show a low-frequency spectral decay rate of ω−2 that steepens to ω−4 beyond 50 Hz. I interpret this change in slope to be the small wavelength manifestation of a smoothly accelerating and decelerating dynamic crack source with larger-scale, variable-rupture velocity. Nearby KIP data from O‘ahu are also considered. Although limited to 45 Hz in bandwidth, comparable spectral slopes are observed. Applying corner frequency analysis to the KIP data, the apparent moments and moment magnitudes of the earthquakes compare well with Hawaiian Volcanos Observatory (HVO) ML. Characteristic fault dimensions are 0.4–0.9 km, and stress drops Δσ are 1–9 MPa. Electronic Supplement: Table of the crustal model used in calculations. Figures comparing the velocity response for the KIP STS-2 and the ACO hydrophone, illustrating the poor Pand S-wave fits of a single spectral slope from low to high frequency, and detailing Pand S-wave signal spectra with respect to pre-event background noise.