Reading the Earth's Signals
A seismogram is a time-series recording of ground motion caused by seismic waves. Modern broadband seismometers capture three components (vertical, north-south, east-west) of ground velocity over a frequency range from 0.001 to 50 Hz. The ability to read and interpret seismograms — identifying wave types, measuring arrival times, and estimating magnitudes — is the foundational skill of observational seismology.
Body Wave Arrivals
The first signal to arrive is the P-wave, a compressional pulse traveling at 5.8–8.1 km/s through the crust and upper mantle. The S-wave arrives second, traveling at roughly 60% of the P-wave speed. The time difference between P and S arrivals increases linearly with distance at about 1 second per 8 km, providing a quick distance estimate. For crustal earthquakes, P is typically a sharp onset followed by coda, while S has a more emergent character on horizontal components.
Surface Waves
At regional to teleseismic distances, surface waves dominate the seismogram. Love waves (SH motion) and Rayleigh waves (retrograde elliptical motion) travel slower than body waves but with less geometric spreading, so they carry the majority of seismic energy. Their dispersive character — long-period components travel faster because they sample deeper (faster) structure — produces a characteristic waveform train that sweeps from low to high frequency, enabling measurement of crustal and mantle velocity profiles.
Modern Analysis
Today's seismogram analysis uses digital signal processing: bandpass filtering isolates specific wave types, spectral analysis reveals source characteristics, and cross-correlation with synthetic seismograms (computed from Earth models) enables precise moment tensor inversion. Automated picking algorithms process thousands of earthquakes per day at global monitoring centers, while machine learning approaches are increasingly used for phase identification and event detection in noisy data.