ALBERT

Description: Probabilistic earthquake locations provide confidence intervals for the hypocentre solutions such as errors encountered in the position, the origin time, and in magnitude. If the relationship of the parameters relative to the local arrangement of the seismic network is considered, such as the node distance, the number of stations, the seismic gap, and the quality of phase readings), the uncertainties can then provide insights on the location capability of the network. In this paper, we collect the earthquake data recorded from the Italian Seismic Network for a time span of 5 years. The data pertain to three different catalogues according to the progressive refinement phases of the location procedure: automatic location, revised location, and published location. By means of spatial analysis,we assess the distribution of the location-related and network-related estimators across the study area. These estimators are subsequently combined to assess the existence of spatial correlations at a local scale. The results indicate that the Italian network is generally able to provide robust locations at the national scale and for smaller earthquakes, and the elongated shape of Italy (and of its network) does not cause systematic bias in the locations. However, we highlight the existence of subregions in which the performance of the network is weaker. At present, a unique 2D, 3-layer velocity model is used for the earthquake location procedure, and this could represent the main limitation for the improvement of the locations. Therefore, the assessment of locally optimized velocity models is the priority for the homogenization and the improvement of the Italian Seismic Network performance.

Description: Published

Description: 1061–1076

Description: 1IT. Reti di monitoraggio e sorveglianza

Description: JCR Journal

Description: Analyzing seismic data to get information about earthquakes has always been a major task for seismologists and, more in general, for geophysicists. Recently, thanks to the technological development of observation systems, more and more data are available to perform such tasks. However, this data “grow up” makes “human possibility” of data processing more complex in terms of required efforts and time demanding. That is why new technological approaches such as artificial intelligence are becoming very popular and more and more exploited. In this paper, we explore the possibility of interpreting seismic waveform segments by means of pre-trained deep learning. More specifically, we apply convolutional networks to seismological waveforms recorded at local or regional distances without any pre-elaboration or filtering. We show that such an approach can be very successful in determining if an earthquake is “included” in the seismic wave image and in estimating the distance between the earthquake epicenter and the recording station.

Description: Published

Description: 1347–1359

Description: 1T. Struttura della Terra

Description: 3T. Fisica dei terremoti e Sorgente Sismica

Description: JCR Journal

Description: The Arctic water cycle is changing dramatically as evidenced by marked shifts in Arctic sea ice conditions, atmospheric processes, and hydrological regimes. Evaporation from the increasingly ice-free Arctic ocean causes moistening of the atmosphere and serves as an unprecedented water source for the Northern Hemisphere. Stable water isotopes (δ18O, δ2H, d-excess) can be used to trace exchange processes between ocean and atmosphere including their potential to feedback into the global climate system. The MOSAiC expedition provided a unique opportunity to collect, analyze, and synthesize discrete samples of the different hydrological compartments in the central Arctic, comprising sea ice, seawater, snow, and melt ponds. Moreover, we present observations of atmospheric humidity, δ18O, δ2H, and d-excess, obtained from a cavity-ring-down spectrometer installed on RV Polarstern and operated continuously during the MOSAiC expedition. By analyzing discrete samples, we found that average seawater δ18O of -1.7±1.95‰ (n=126) conforms to observed and modelled isotopic traits of the Arctic Ocean. Second year ice is relatively depleted compared to first year ice with average δ18O values of -3.1±2.81‰ (n=397) and -0.7±2.28‰ (n=409), respectively. Snow on top of the sea ice (n=303) has the most depleted isotopic signature among all compartments shaping the Arctic water cycle (mean δ18O=-15.3±7.12‰) The atmospheric water vapour dataset reveals a clear seasonal cycle; significant positive correlations are found both with local specific humidity and air temperature. The comparison of synoptic events, characterized by abrupt isotopic fluctuations, with simultaneous observations from land-based Arctic stations indicates a strong influence of sea ice coverage on the isotopic signal. For an in-depth understanding of the isotopic changes, the observations are compared to an isotope-enhanced ECHAM6 atmosphere simulation. The model-data comparison assesses the capability of this state-of-the-art AGCM to capture the first-order evaporation/condensation processes and their seasonal evolution. Our dataset provides a comprehensive description of the present-day isotopic composition of the Arctic water covering a complete seasonal cycle. This will ultimately contribute to resolve the linkages between sea ice, ocean, and atmosphere during critical transitions from frozen ocean to open water conditions.