Introduction
The seismograph is one of the most sensitive instruments ever built. Modern broadband seismometers can detect ground motions smaller than a hydrogen atom's diameter — vibrations from distant earthquakes, ocean waves, industrial machinery, and even atmospheric pressure changes all register on their records. The fundamental principle has not changed since John Milne built the first practical horizontal-pendulum seismograph in Japan in 1880: an inertial mass resists motion while the ground beneath it shakes, so a pen attached to the mass traces the relative motion of mass versus ground on a rotating drum of paper (or, today, produces a voltage signal proportional to ground velocity). The record produced is a seismogram — a time-series of ground motion that seismologists read as precisely as doctors read an electrocardiogram.
Modern seismometers are broadband instruments capable of recording ground motion over a frequency range from ~0.001 Hz (periods of 1,000 seconds — the Earth's free oscillations after great earthquakes) to ~50 Hz (local small earthquakes and explosions). They measure ground velocity in three orthogonal directions: vertical (Z component) and two horizontal components typically oriented North-South (N) and East-West (E). This three-component recording is essential: P-waves arrive most clearly on the vertical component, S-waves on the horizontal components, and surface waves on all three. Seismometers are installed in vaults to shield them from temperature fluctuations, wind, and human noise; ocean-bottom seismometers extend coverage to the seafloor; and borehole seismometers several hundred metres underground achieve the quietest recordings possible on land.
Reading a seismogram requires recognising the characteristic signatures of each wave type. The P-wave arrival appears as a small abrupt onset in the vertical component — the first sign that an earthquake has occurred. Several seconds to minutes later (depending on distance), S-waves arrive with larger amplitude on horizontal components. Then the surface waves roll in — Love waves on the horizontal components, Rayleigh waves on all three — with the longest periods and highest amplitudes. The time between the P and S arrivals (the S-P time) directly gives the distance to the earthquake: for a rough estimate, multiply the S-P time in seconds by ~8 km (5.0 mi). But a single station only gives distance, not direction. Locating the earthquake precisely requires at least three stations: each station defines a sphere (or circle on the surface) of possible earthquake locations at the calculated distance; the intersection of three spheres gives the hypocenter in three dimensions.
The Global Seismographic Network (GSN), operated by USGS and EarthScope (formerly IRIS), consists of 150 high-quality broadband seismograph stations distributed across all continents and ocean islands. Together with regional networks (SCSN in southern California, TexNet in Texas, the Japanese Hi-net with 800 stations), these networks provide near-real-time earthquake locations worldwide. After a large earthquake, USGS can issue a preliminary hypocenter location within minutes and a more refined solution within hours as more waveform data arrive. The Comprehensive Nuclear-Test-Ban Treaty Organization (CTBTO) operates the International Monitoring System (IMS) — 170 seismic stations globally — specifically to detect covert nuclear tests by their distinctive seismic signatures; this network also provides valuable scientific data.
Key Terms
An instrument that measures ground motion by suspending an inertial mass relative to the moving ground. Modern broadband seismometers record ground velocity over a frequency range of ~0.001–50 Hz and measure motion in three orthogonal directions (Z, N, E), outputting a voltage signal proportional to ground velocity. Installed in vaults, boreholes, or on the ocean floor, they can detect ground motions smaller than 10⁻¹² m — far below the threshold of human perception.
The time-series record of ground motion produced by a seismometer. A seismogram shows the characteristic P-wave arrival (small amplitude, sharp onset, dominant on the vertical component), the S-wave arrival (larger amplitude, dominant on horizontal components), and surface waves (long-period, large-amplitude wavepackets arriving last). Seismologists use seismograms to determine epicentral distance (from S-P time), earthquake magnitude (from wave amplitude), and source mechanism (from P-wave first-motion polarities).
The point on Earth's surface directly above the hypocenter (focus), where fault rupture initiates. The distance from an epicenter to a seismograph station is the epicentral distance. Determining the epicenter requires observations from at least three seismograph stations; each provides a circle of possible locations based on epicentral distance, and the three circles intersect at the epicenter. Depth of the hypocenter is also determined from waveform analysis.
The method of locating an earthquake epicenter using S-P times measured at three or more seismograph stations. Each station's S-P time gives the epicentral distance, defining a circle of possible epicenter locations on the map. The intersection of three such circles yields the epicenter. Modern computers extend this principle by solving an over-determined system with dozens of stations, using least-squares fitting of P-wave arrival times to find the best three-dimensional hypocenter.
A network of 150 broadband, three-component seismograph stations worldwide, operated by USGS and EarthScope (formerly IRIS). Stations transmit real-time data via satellite to the IRIS Data Management Center (now EarthScope). Data are freely available for scientific research and support earthquake location, seismic tomography, nuclear test monitoring, and studies of Earth's internal structure.