10:45 AM - 12:15 PM
[SSS10-P19] A demonstration experiment to evaluate online seismicity, using the Hierarchical Space-Time ETAS (HIST-ETAS) model: A preliminary report
Keywords:Seismicity, ETAS, Automatic processing
Seismicity patterns that vary from one place to another show various clustering features. Kanamori (1981) pointed out that heterogeneity and complexity of the fault zone are responsible for observed variations. Stochastic point-process models for earthquake occurrence are typically used to understand such complex features. Stochastic models have to be sufficiently accurate in the sense that they are spatiotemporally well adapted to and are able to predict various local patterns of normal seismic activity. The epidemic type aftershock sequence (ETAS) and its space-time extension were introduced, precisely to address these purposes (Ogata, 1985, 1988).
Among candidate space-time ETAS models, the hierarchical version of the model (HIST-ETAS model, or the Hierarchical Space-Time ETAS model) has been intensively studied (Ogata, 2011, 2022). One of the features of this model is that its parameters depend on the earthquakes’ locations. Another feature of the HIST-ETAS model is its ability to implement space-time forecasting of history-dependent seismicity rate based on a previously defined time frame. A FORTRAN version of this computer program is publicly available and open source (Ogata et al., 2021).
Using this package, a demonstration experiment to evaluate online seismic activity is currently underway. Two experiment regions were employed: the first included all of Japan while the other only included the Nankai trough. The following three steps are being performed daily by automatic processing. (1) The creation of a catalog that consists of earthquakes that occurred until the day prior to the iteration, by extracting necessary information from the web page of the Preliminary Catalog using the Hi-net (High Sensitivity Seismograph Network) Automatic System, a hypocenter list that shows earthquake information determined by the NIED (National Research Institute for Earth Science and Disaster Resilience) Hi-net automatic hypocenter determination system. (2) Processing the output of various calculation results based on the input of earthquake information in the catalog created in (1) by the software package programs. (3) Automatically posting, on Twitter, figures that show the resulting output from the process in (2). Given that the experiment in ongoing, the Twitter account has been set to private mode and followers, including the author, follow this Twitter account.
In this presentation, a preliminary report on the overview of the ongoing demonstration experiment will be provided.
Acknowledgments: The author thanks Y. Ogata and J. Zhuang for discussion and Y. Noda for help with automatic processing within the experiment. This study was partially supported by the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan, under the STAR-E (Seismology TowArd Research innovation with data of Earthquake) Program (Grant Number JPJ010217), and Chubu Electric Power's research based on selected proposals. The Preliminary Catalog by the Hi-net Automatic System operated by NIED was used (https://www.hinet.bosai.go.jp/?LANG=en). The software package of the HIST-ETAS model was obtained from The Institute of Statistical Mathematics, Computer Science Monographs, No.35 (February 2021), Hierarchical Space-Time Point-Process Models (HIST-PPM): Software Documentation by Y. Ogata et al. (https://www.ism.ac.jp/editsec/csm/index.html).
Among candidate space-time ETAS models, the hierarchical version of the model (HIST-ETAS model, or the Hierarchical Space-Time ETAS model) has been intensively studied (Ogata, 2011, 2022). One of the features of this model is that its parameters depend on the earthquakes’ locations. Another feature of the HIST-ETAS model is its ability to implement space-time forecasting of history-dependent seismicity rate based on a previously defined time frame. A FORTRAN version of this computer program is publicly available and open source (Ogata et al., 2021).
Using this package, a demonstration experiment to evaluate online seismic activity is currently underway. Two experiment regions were employed: the first included all of Japan while the other only included the Nankai trough. The following three steps are being performed daily by automatic processing. (1) The creation of a catalog that consists of earthquakes that occurred until the day prior to the iteration, by extracting necessary information from the web page of the Preliminary Catalog using the Hi-net (High Sensitivity Seismograph Network) Automatic System, a hypocenter list that shows earthquake information determined by the NIED (National Research Institute for Earth Science and Disaster Resilience) Hi-net automatic hypocenter determination system. (2) Processing the output of various calculation results based on the input of earthquake information in the catalog created in (1) by the software package programs. (3) Automatically posting, on Twitter, figures that show the resulting output from the process in (2). Given that the experiment in ongoing, the Twitter account has been set to private mode and followers, including the author, follow this Twitter account.
In this presentation, a preliminary report on the overview of the ongoing demonstration experiment will be provided.
Acknowledgments: The author thanks Y. Ogata and J. Zhuang for discussion and Y. Noda for help with automatic processing within the experiment. This study was partially supported by the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan, under the STAR-E (Seismology TowArd Research innovation with data of Earthquake) Program (Grant Number JPJ010217), and Chubu Electric Power's research based on selected proposals. The Preliminary Catalog by the Hi-net Automatic System operated by NIED was used (https://www.hinet.bosai.go.jp/?LANG=en). The software package of the HIST-ETAS model was obtained from The Institute of Statistical Mathematics, Computer Science Monographs, No.35 (February 2021), Hierarchical Space-Time Point-Process Models (HIST-PPM): Software Documentation by Y. Ogata et al. (https://www.ism.ac.jp/editsec/csm/index.html).