Dipartimento di Scienze Fisiche Università di Napoli Federico II

A local magnitude scale has been defined for southern Italy, in the area monitored by the recently installed Irpinia Seismic Network.Waveforms recorded from more than 100 events of small magnitude are processed to extract synthetic Wood– Anderson traces. Assuming a general description of peak-displacement scaling with the distance, by means of linear and logarithmic contributions, a global exploration of the parameter space is performed by a grid-search method with the aim of investigating the correlation between the two decay contributions and seeking for a physical solution of the problem. Assuming an L2 norm, we found M log A 1:79 log R 0:58; yielding an error on the single estimation smaller than 0.2, at least when the hypocenter location is accurate. Station corrections are investigated through the station residuals, referring to the average value of the magnitude. Using a z test, we found that some stations exhibit a correction term significantly different from 0. The use of the peak acceleration and peak velocity as indicators of the magnitude is also investigated.

The Campi Flegrei (southern Italy) is one of the most active calderas in the world. This caldera is characterized by episodes of slow vertical ground movement, called bradyseism. With several hundred thousand people living within its borders, this area is in a high-risk category should there be an eruption. The seismological monitoring system in the Campi Flegrei is based on nine seismic stations, eight of which are equipped with short-period seismometers (1 Hz), and one with a broadband seismometer (60 sec–50 Hz). While all of the seismic stations are located on land, part of the seismic activity occurs in the undersea area of the Pozzuoli Gulf (Campi Flegrei), where there are no seismic stations. This gap in the data coverage produces a biased and incomplete image of the volcanic area.We carried out an experiment in the Pozzuoli Gulf with the installation of two broadband seismic stations on the seafloor with remote and continuous data acquisition for a duration of 31 days between January and March 2005. Using the data acquired, we have computed the power spectral density (PSD) to characterize the background seismic noise, and to evaluate the true noise variation, we have generated the seismic noise probability density functions from the computed PSD curves. The results of our analysis show that the broadband seismic noise is high when compared with the Peterson noise model (land model), but for periods less than 0.3 sec, the seismic noise on the seafloor is lower than the recordings on land over the same period range. The last bradyseismic crisis (1982–1984) highlights the importance of this frequency range, where most of the spectral content of the recorded earthquakes was observed. Finally, we evaluate the detection threshold of a new seismic station located on the seafloor of the Campi Flegrei caldera considering the characteristics of the local seismicity. This analysis shows that the detection threshold for the sea-floor stations (Mw ∼ 0:2) is less than that for land stations (Mw ∼ 0:8).

An effective early-warning system must provide probabilistic estimates of the location and size of a potentially destructive earthquake within a few seconds after the event is first detected. In this work we present an evolutionary, real-time location technique based on an equal differential time (EDT) formulation and a probabilistic approach for describing the hypocenter estimation. The algorithm, at each timestep, relies on the information from triggered arrivals and not-yet-triggered stations. With just one recorded arrival, the hypocentral location is constrained by the Voronoi cell around the first triggering station constructed using the travel times to the not-yet-triggered stations.With two or more triggered arrivals, the location is constrained by the intersection of the volume defined by the Voronoi cells for the remaining, not-yet-triggered stations and the EDT surfaces between all pairs of triggered arrivals. As time passes and more triggers become available, the evolutionary location converges to a standard EDT location. Synthetic tests performed using the geometry of the Irpinia seismic network, southern Italy (ISNet), and the simulation of an evolutionary location for the 2000 Mw 6:6 Western Tottori, Japan, earthquake indicate that when a dense seismic network is available, reliable location estimates suitable for early-warning applications can be achieved after 1–3 sec from the first event detection. A further simulation with an Mw 6:7 southern Greece earthquake shows that at a regional scale, the real-time location can provide useful constraints on the earthquake position several seconds before a non-real-time algorithm. Finally, we show that the robustness of the algorithm in the presence of outliers can be effectively used to associate phase arrivals coming from events occurring close in time, and we present a preliminary algorithm for event detection.