Introducing ISMDq—A Web Portal for Real-Time Quality Monitoring of Italian Strong-Motion Data

Abstract

Abstract We present the Istituto Nazionale di Geofisica e Vulcanologia Strong-Motion Data-quality (ISMDq)—a new automatic system designed to check both continuous data stream and event strong-motion waveforms before online publication. The main purpose of ISMDq is to ensure accurate ground-motion data and derived products to be rapidly shared with monitoring authorities and the scientific community. ISMDq provides data-quality reports within minutes of the occurrence of Italian earthquakes with magnitude ≥3.0 and includes a detailed daily picture describing the performance of the target strong-motion networks. In this article, we describe and discuss the automatic procedures used by ISMDq to perform its data-quality check. Before an earthquake, ISMDq evaluates the selected waveforms through the estimation of quality indexes employed to reject bad data and/or to group approved data into classes of quality that are useful to quantify the level of reliability. The quality indexes are estimated based on comparisons with the background ambient noise level performed both in the time and frequency domains. As a consequence, new high- and low-noise reference levels are derived for the overall Italian strong-motion network, for each station, and for groups of stations in the same soil categories of the Eurocode 8 (Eurocode 8 [EC8], 2003). In absence of earthquakes, 24 hr streaming of ambient noise recordings are analyzed at each station to set an empirical threshold on selected data metrics and data availability, with the goal to build a station quality archive, which is daily updated in a time span of six months. The ISMDq is accessible online (see Data and Resources) from August 2020, providing rapid open access to ∼10,000 high-quality checked automatically processed strong-motion waveforms and metadata, relative to more than 160 Italian earthquakes with magnitude in the 3.0–5.2 range. Comparisons between selected strong-motion data automatically processed and then manually revised corroborate the reliability of the proposed procedures.