GNGTS 2018 - 37° Convegno Nazionale

136 GNGTS 2018 S essione 1.1 A CONSISTENT AND UNIFORM RESEARCH EARTHQUAKE CATALOG FOR THE ALPARRAY REGION I. Molinari 1 , M. Bagagli 1 , T. Diehl 2 , E. Kissling 1 , J. Clinton 2 , D. Giardini 1 , S. Wiemer 2 , the AlpArray Working Group 1 ETH Zurich, Department of Earth Sciences, Institute of Geophysics, Zurich, Switzerland 2 ETH Zurich, Swiss Seismological Service, Zurich, Switzerland The AlpArray seismic Network. To improve our understanding of the seismotectonics and the seismic hazard in the greater Alpine, a homogeneous earthquake catalog in terms of location and magnitude is needed. This requires merging of waveform data of many regional and national seismic networks. Moreover, different national and regional bulletins often report the same earthquake with different location and magnitude. Phase misidentification, non- uniform quality assessment, differences in picking procedures leads to inconsistencies between the different catalogs. The AlpArray initiative (www.alparray.ethz.ch ) is a large-scale European collaboration (~50 institutes involved) to study the entire Alpine orogen at high resolution with a variety of geoscientific methods. The AlpArray Seismic Network (AASN) provides unprecedentedly uniform station coverage for the region with more than 650 broadband seismic stations, 300 of which are temporary (Hetényi et al. , 2018). It is a joint effort of ~25 institutes from 10 nations, operates since January 2016 and is expected to continue until the end of March 2019. The data will be publicly available three years after the end of the project. We collect all the broadband, high-gain short period and low-gain strong motion stations within and around of the AlpArray region. The final data set consists of more than 1000 stations (~ 25 TB). Aims and challenges. Our main goals are: i) consistent and precise hypocenter locations and ii) provide preliminary but uniform magnitude calculations across the region. For this study, we collected the first two years of data (2016-2017) from more than 1000 stations (Fig. 1) and we systematically checked the data and metadata quality. The challenges posed by this work have both a scientific and technical nature: from the processing huge amount of data, to consistent phase picking and identification across station array, from network heterogeneities and the fact that many stations are not routinely picked to the establishment of velocity models for high-precision hypocenter locations. Workflow. The designed workflow is based on the iterative use of automatic P-wave pickers, detection and nonlinear location algorithms with a high-quality re-picking approach finally providing consistent phase arrival times in combination with a picking quality assessment. First, we detect events in the region in 2016/2017 using the STA/LTA based detector of the SeisComP3 monitoring system in off-line mode. We separate the picking ( scautopick ) from the detection to allow parallel picking. Picks from all stations are streamed to scautoloc and scevent in playback mode leading to a reduction of the playback time and processing of the ~1000 stations simultaneously of 300 times (one month of SeisComP3 automatic catalog in ~ 12 hours). To minimize the impact of erroneous automatic triggers on the location, the initial automatic hypocenters are derived by the Equal-Differential Time (EDT) algorithm implemented in the NonLinLoc location software using 1D alpine model (Diehl et al. , 2009). In addition, a quality-score is determined for each automatic origin and the one with the highest score is the preferred solution for the event. Among the detected events, we select 30 geographically homogeneously distributed events with magnitudes ≥2.5 representative for the entire catalog. We manually pick the selected events to establish a consistent phases reference data set, including arrival-time time uncertainties. The reference data are used to adjust the secondary, high-quality automatic post-pickers and to assess their performance, AutoPSy (Bagagli et al. , in prepartion). The post-picker is an iterative phase picker, searching for the most likely seismic phases arrival time with phase detection. The first-arrival P-onset and the S-onset of the seismogram are used for the earthquake location.