Optimizing a seismic network for locating earthquakes is a crucial issue in seismology. Precise earthquake location is the key factor for most sophisticated seismological analyses and for the monitoring of tectonic and volcanic areas. Network testing and optimization should therefore be a basic procedure to plan, install, and improve a monitoring seismic network. Here, we evaluate the most appropriate methods for network testing, using two approaches based on linearized and Bayesian methods. We then propose a new procedure for seismic network optimization, based on the direct comparison of all the possible networks resulting from the permutation of M stations in N sites (with N>M). The performance of a network can be defined using different criteria. In addition to the covariance matrix as conventionally applied, we have introduced the condition number of the coefficient matrix. Our new procedure is very efficient for network optimization with respect to multiple criteria and overcomes several problems of current procedures. Applied to the Campi Flegrei volcanic area, the method has identified a key configuration for optimizing and improving the seismic‐monitoring network.

You do not currently have access to this article.