The Source-Scanning Algorithm: mapping the distribution of seismic sources in time and space

TLDR

We introduce the Source‑Scanning Algorithm (SSA) to image seismic source distributions in both time and space. SSA computes a brightness function by summing absolute amplitudes at predicted arrival times for trial locations and origin times, then systematically searches model space and time for the maximum to identify the source distribution. SSA exploits waveform arrival times and amplitudes without synthetic seismograms or pre‑picked phases, and tests show it robustly recovers source configurations within one grid interval, as demonstrated by locating a tremor event during the recent ETS sequence in northern Cascadia.

Abstract

We introduce a new method, which we call the Source-Scanning Algorithm (SSA), for imaging the distribution of seismic sources in both time and space. Using trial locations and origin times, the method calculates the ‘brightness’ function by summing the absolute amplitudes observed at all stations at their respective predicted arrival times. The spatial and temporal distribution of sources is then identified by a systematic search throughout the model space and time for the maximum brightness. The greatest advantages of this method are that: (1) it exploits waveform information (both arrival times and relative amplitudes) without the need to calculate high-frequency synthetic seismograms; and (2) it requires neither pre-assembled phase-picking data nor any a priori assumptions about the source geometry. A series of tests using synthetic data have shown that this method is robust and can faithfully recover the input source configuration to within 1 grid interval. Finally, we demonstrate the value of the algorithm by locating a typical tremor event with emergent waveforms that occurred during the recent episodic tremor and slip (ETS) sequence in the northern Cascadia subduction zone.