Abstract

In this paper, we develop a modern Fortran code to approximate network cross-correlation coefficients for seismic event detection. The code is available via an online repository. The algorithm exhibits efficient memory usage, small output size, and amenability to parallelization. Although exact NCC values are not calculated, the accuracy is deemed to be sufficient based on statistical investigation. A benchmark and practical application are also presented, suggesting the viability of performing matched filter analysis of years-long continuous multi-channel seismic data and thousands of template events on a workstation within a runtime extending from a few hours to a few days.