5:15 PM - 6:30 PM
[SSS06-P02] A preliminary study on data quality characterization for the DONET seismic catalogue
Keywords:Seismicity, Statistical analysis, Data quality characterization, DONET, Nankai Trough
To understand the characteristics of seismicity occurring in the source area of the Nankai Trough earthquake, research based on past observation data is indispensable, and a seismic catalogue that includes earthquakes in and around this area, if available, becomes one of them. It is also necessary to examine whether or not the data in such catalogue can be used for this research, and the results of evaluating the data quality will be the basic information for establishing the research direction. In this presentation, we report preliminary results from a study to evaluate qualities of the seismic catalogue including earthquakes observed by the DONET (Dense Oceanfloor Network system for Earthquakes and Tsunamis).
In this study, quality evaluation was carried out, using three techniques that utilize statistical properties of earthquakes. The first is based on the frequency-magnitude distribution of earthquakes. Using this technique, it is confirmed that larger earthquakes tend to have fewer numbers and smaller ones tend to have more numbers, and that this tendency is well correlated with the Gutenberg-Richter's law. The second approach is to plot the frequency of earthquakes as a function of hour. Natural earthquakes should occur day and night, and it is suggested that if the frequency of earthquake occurrence changes depending on time, it originates from anthropogenic causes (e.g., human activities and artificial earthquakes). The frequency was found to be approximately constant, which is considered to be due to few anthropogenic effects because the DONET is installed offshore. The third approach is based on the assumptions that the smaller magnitude is, the harder earthquake detection is, and that earthquakes with magnitudes, above a certain threshold value, are observed without missing. The threshold is called completeness magnitude, Mc, above which all earthquakes consider to be detected. This is an indicator of the quality assessment of seismic catalogues. As a result of the completeness analysis, we observed spatiotemporal variability of Mc, as observed for other seismic networks such as the ones used to create the catalogue maintained by the Japan Meteorological Agency.
This study on data quality characterization for the DONET catalogue is currently underway. Our results in the presentation will include the most recent findings.
In this study, quality evaluation was carried out, using three techniques that utilize statistical properties of earthquakes. The first is based on the frequency-magnitude distribution of earthquakes. Using this technique, it is confirmed that larger earthquakes tend to have fewer numbers and smaller ones tend to have more numbers, and that this tendency is well correlated with the Gutenberg-Richter's law. The second approach is to plot the frequency of earthquakes as a function of hour. Natural earthquakes should occur day and night, and it is suggested that if the frequency of earthquake occurrence changes depending on time, it originates from anthropogenic causes (e.g., human activities and artificial earthquakes). The frequency was found to be approximately constant, which is considered to be due to few anthropogenic effects because the DONET is installed offshore. The third approach is based on the assumptions that the smaller magnitude is, the harder earthquake detection is, and that earthquakes with magnitudes, above a certain threshold value, are observed without missing. The threshold is called completeness magnitude, Mc, above which all earthquakes consider to be detected. This is an indicator of the quality assessment of seismic catalogues. As a result of the completeness analysis, we observed spatiotemporal variability of Mc, as observed for other seismic networks such as the ones used to create the catalogue maintained by the Japan Meteorological Agency.
This study on data quality characterization for the DONET catalogue is currently underway. Our results in the presentation will include the most recent findings.