How and Why to Record Broad Band Seismic Signals

Abstract

MOST seismological observatories operate narrow band recording systems that record only that part of the seismic spectrum in which the signal/noise ratio is usually a maximum. These systems are ideal for detecting small seismic signals (say body wave magnitude m_b 4.0 to 5.0) such as those generated by underground explosions of a few kton but have the disadvantage that the recorded signals are distorted versions of the true ground motion. To study seismic source functions particularly from body waves requires a broad band (BB) system such as the Kirnos system operated in the USSR¹ which has a constant magnification over the range 0.1 Hz to 10 Hz. The drawback of such systems is that the recording band includes the strong microseismic peak at around 0.17 Hz and thus only relatively large signals, above ∼m_b = 5.5, are recorded. Here we describe how both broad band and narrow band recordings can be obtained from a single conventional long period (LP) seismometer. The chief cost of adding a broad band system to an established observatory is then the provision of a suitable recorder. We also try to demonstrate the advantages of BB recording.