5:15 PM - 7:15 PM
[HDS10-P03] Probabilistic Tsunami Hazard Assessment for Active Faults in the Southwestern Sea of Japan
Keywords:Tsunami, Probabilistic Tsunami Hazard Assessment, the Sea of Japan, Active Fault in the Sea Area
1. Introduction
Along the coast of the Sea of Japan, earthquakes from active faults, such as the 1983 Japan Sea Mid-Central EQ (M7.7), 1993 Southwest Off Hokkaido EQ (M7.8), and 2024 Noto Peninsula EQ (M7.6), have caused significant tsunami damage. This study conducted a probabilistic tsunami hazard assessment (PTHA) targeting the southwestern part of the Sea of Japan to evaluate tsunami hazards generated by these earthquakes. The presentation covers the establishment of earthquake groups, construction of earthquake fault models, tsunami propagation calculations, setting of earthquake occurrence probabilities, calculation of hazard curves, and probabilistic integration of the curves.
2. Setting of Earthquake Groups
This PTHA targeted active fault zones evaluated in the " Long-term evaluations of offshore active faults in the Southwestern Sea of Japan," (HERP, 2022) short active faults indicated in the assessment, and background earthquakes (Mw6.9) in the southwestern region. For fault zones with multiple segments, individual earthquakes and interactions with adjacent segments were considered.
3. Construction of earthquake fault models
Each fault was modeled as a single rectangular fault for each segment, with the following parameters:
1) The endpoints were set using positions from the "Long-term Evaluation," with background earthquakes arranged at 0.2-degree intervals for both latitude and longitude.
2)The upper end depth was set to 0 km below the seabed, based on the "Long-term Evaluation."
3) The dip angles were based on descriptions from the “Long-term Evaluation”: "nearly vertical" was set to 90 degrees, "steep" to 70 degrees, and "moderate" to 45 degrees. For background earthquakes, a uniform dip angle of 90 degrees was applied.
4) The strike of background earthquakes was based on the average of the active faults (including short faults) present in the corresponding region.
5) The slip angle was set to 12.5 degrees, corresponding to the median of the strike-slip/vertical-slip ratio indicated in the "Long-term Evaluation."
6) The fault width was initially set as a "base value" derived from the thickness of the seismic layer and the dip of the fault, and the fault width for the earthquake fault models was set to "base value" + 5 km.
7) Fault length was set to values from the "Long-term Evaluation," with unspecified earthquakes calculated from Irikura and Miyake (2001).
8) Average slip amounts were based on magnitudes derived from the fault area calculated using Irikura and Miyake (2001).
9)A large slip area was established with slip amounts twice the average, covering 30% of the fault area (assumed similar to interplate earthquakes) and positioned at the shallow side contacting the upper end of the fault.
4. Tsunami Calculations
A computational area with a minimum mesh size of 50 m was established, and tsunami calculations indicated that maximum tsunami height were higher west of Hyogo Prefecture and in the Wakasa Bay area.
5. Setting of Earthquake Occurrence Probabilities
Earthquake occurrence probabilities for the active faults from the "Long-term Evaluation" were calculated as the sum of probabilities based on median average recurrence intervals and those for “earthquakes that are difficult to identify.” For background earthquakes, probabilities were derived from the G-R law combined with a Poisson process, using values from the National Seismic Hazard Maps for Japan.
6. Calculation of Hazard Curves
In previous PTHA studies on the Pacific side, maximum tsunami height for hazard curve calculations were set over 1.0 m. However, due to lower maximum tsunami height on the Sea of Japan side, this value was tentatively set to over 0.2 m.
7. Probabilistic Integration
Tsunami hazard curves from individual earthquake fault models were probabilistically integrated, considering uncertainties due to computational errors, following methodologies similar to those of Murata et al. (2024).
8. Results of the PTHA
Results indicated relatively high probabilities of tsunami occurrences exceeding 30 years along the coast from Tottori Prefecture to Fukuoka Prefecture and in the northern regions of Iki and Tsushima. Since the tsunami heights were relatively low to the east of Hyogo Prefecture, the probability of occurrence exceeding 30 years was low.
This study presents a PTHA for only earthquakes in the southwestern part of the Sea of Japan.
This study was conducted as part of a research project on "Hazard and Risk Assessment for Natural Disasters" by the National Research Institute for Earth Science and Disaster Resilience.
Along the coast of the Sea of Japan, earthquakes from active faults, such as the 1983 Japan Sea Mid-Central EQ (M7.7), 1993 Southwest Off Hokkaido EQ (M7.8), and 2024 Noto Peninsula EQ (M7.6), have caused significant tsunami damage. This study conducted a probabilistic tsunami hazard assessment (PTHA) targeting the southwestern part of the Sea of Japan to evaluate tsunami hazards generated by these earthquakes. The presentation covers the establishment of earthquake groups, construction of earthquake fault models, tsunami propagation calculations, setting of earthquake occurrence probabilities, calculation of hazard curves, and probabilistic integration of the curves.
2. Setting of Earthquake Groups
This PTHA targeted active fault zones evaluated in the " Long-term evaluations of offshore active faults in the Southwestern Sea of Japan," (HERP, 2022) short active faults indicated in the assessment, and background earthquakes (Mw6.9) in the southwestern region. For fault zones with multiple segments, individual earthquakes and interactions with adjacent segments were considered.
3. Construction of earthquake fault models
Each fault was modeled as a single rectangular fault for each segment, with the following parameters:
1) The endpoints were set using positions from the "Long-term Evaluation," with background earthquakes arranged at 0.2-degree intervals for both latitude and longitude.
2)The upper end depth was set to 0 km below the seabed, based on the "Long-term Evaluation."
3) The dip angles were based on descriptions from the “Long-term Evaluation”: "nearly vertical" was set to 90 degrees, "steep" to 70 degrees, and "moderate" to 45 degrees. For background earthquakes, a uniform dip angle of 90 degrees was applied.
4) The strike of background earthquakes was based on the average of the active faults (including short faults) present in the corresponding region.
5) The slip angle was set to 12.5 degrees, corresponding to the median of the strike-slip/vertical-slip ratio indicated in the "Long-term Evaluation."
6) The fault width was initially set as a "base value" derived from the thickness of the seismic layer and the dip of the fault, and the fault width for the earthquake fault models was set to "base value" + 5 km.
7) Fault length was set to values from the "Long-term Evaluation," with unspecified earthquakes calculated from Irikura and Miyake (2001).
8) Average slip amounts were based on magnitudes derived from the fault area calculated using Irikura and Miyake (2001).
9)A large slip area was established with slip amounts twice the average, covering 30% of the fault area (assumed similar to interplate earthquakes) and positioned at the shallow side contacting the upper end of the fault.
4. Tsunami Calculations
A computational area with a minimum mesh size of 50 m was established, and tsunami calculations indicated that maximum tsunami height were higher west of Hyogo Prefecture and in the Wakasa Bay area.
5. Setting of Earthquake Occurrence Probabilities
Earthquake occurrence probabilities for the active faults from the "Long-term Evaluation" were calculated as the sum of probabilities based on median average recurrence intervals and those for “earthquakes that are difficult to identify.” For background earthquakes, probabilities were derived from the G-R law combined with a Poisson process, using values from the National Seismic Hazard Maps for Japan.
6. Calculation of Hazard Curves
In previous PTHA studies on the Pacific side, maximum tsunami height for hazard curve calculations were set over 1.0 m. However, due to lower maximum tsunami height on the Sea of Japan side, this value was tentatively set to over 0.2 m.
7. Probabilistic Integration
Tsunami hazard curves from individual earthquake fault models were probabilistically integrated, considering uncertainties due to computational errors, following methodologies similar to those of Murata et al. (2024).
8. Results of the PTHA
Results indicated relatively high probabilities of tsunami occurrences exceeding 30 years along the coast from Tottori Prefecture to Fukuoka Prefecture and in the northern regions of Iki and Tsushima. Since the tsunami heights were relatively low to the east of Hyogo Prefecture, the probability of occurrence exceeding 30 years was low.
This study presents a PTHA for only earthquakes in the southwestern part of the Sea of Japan.
This study was conducted as part of a research project on "Hazard and Risk Assessment for Natural Disasters" by the National Research Institute for Earth Science and Disaster Resilience.