| PaperNO | Paper / Abstract |
|
C4-001
16:10
|
16:30
|
NIED OBSERVATION NETWORK FOR EARTHQUAKE, TSUNAMI AND VOLCANO: MOWLAS
Lessons from the 1995 Kobe earthquake, earthquake observation system in Japan has dramatically changed since then. The Headquarters of Earthquake Research Promotion was established, and NIED has constructed and operated nationwide land observation networks. These are composed of High Sensitivity Seismograph Network Japan (Hi-net: around 800 stations), Full Range Seismograph Network of Japan (F-net: 73 stations), Kyoshin Network (K-NET: around 1050 stations), and Kiban Kyoshin Network (KiK-net: around 700 stations) that started to operate since 1996 a year after the 1995 earthquake, and the networks was completed around in 2000. On the other hand, seafloor observations were largely delayed than land observations. Considering the severe disaster of the 2011 Tohoku-Oki earthquake and tsunami, DONET with 51 stations along the Nankai Trough was installed by JAMSTEC and transferred to NIED. In addition, NIED established Seafloor realtime observation network for earthquakes and tsunamis along the Japan Trench (S-net: 150 stations) to cover Eastern Japan for earthquake and tsunami early warning as well as information delivery. The above six observation networks and Fundamental Volcano Observation Network (V-net), NIED has started integrated operation of MOWLAS (Monitoring of Waves on Land and Seafloor) as the NIED Observation Network for Earthquake, Tsunami and Volcano. Direct observation above the hypocenter has an advantage from both research purpose and earthquake and tsunami early warning. By using S-net and DONET observation data, the ground motion and tsunami detection can be earlier around 30 s and 20 min, respectively. This will contribute to increase lead time and accuracy for JMA operation and control systems by private companies. NIED preliminary analyses show to increase the detection capability of offshore seismicity that may contribute the discovery of new offshore phenomena. Seafloor observation needs the cost rather than land observations. Due to severe observation condition, number of stations and data quality is limited. It is important to consider the balance between land and seafloor observations, as well as to promote the methods for analyses of integrated land and seafloor observation data.
Shin Aoi
land observation network, MOWLAS, NIED Observation Network for Earthquake, seafloor observation network, Tsunami and Volcano
|
|
C4-002
16:30
|
16:50
|
DEVELOPMENT OF THE LOW COST EARTHQUAKE EARLY WARNING AND SHAKEMAP SYSTEMS IN TAIWAN
The National Taiwan University (NTU) developed an earthquake early warning (EEW) system for research purposes using low-cost accelerometers (P-Alert) since 2010. As of 2019, a total of 692 stations have been deployed and configured. The NTU system can provide earthquake information within 15 s of an earthquake occurrence. Thus, this system may provide early warnings for cities located more than 50 km from the epicenter. Additionally, the NTU system also has an onsite alert function that triggers a warning for incoming P-waves greater than a certain magnitude threshold, thus providing a 2–3 s lead time before peak ground acceleration (PGA) for regions close to an epicenter. Detailed shaking maps are produced by the NTU system within one or two minutes after an earthquake. Recently, a new module named ShakeAlarm has been developed. Equipped with real-time acceleration signals and the time-dependent anisotropic attenuation relationship of the PGA, ShakingAlarm can provide an accurate PGA estimation immediately before the arrival of the observed PGA. This unique advantage produces sufficient lead time for hazard assessment and emergency response, which is unavailable for traditional shakemap, which are based on only the PGA observed in real time. The performance of ShakingAlarm was tested with more than ten M > 5.5 inland earthquakes from 2013 to 2019. Taking the 2016 M6.4 Meinong earthquake simulation as an example, the predicted PGA converges to a stable value and produces a predicted shake map and an isocontour map of the predicted PGA within 16 seconds of earthquake occurrence. Compared with traditional regional EEW system, ShakingAlarm can effectively identify possible damage regions and provide valuable early warning information (magnitude and PGA) for risk mitigation.
Yih-Min Wu
earthquake early warning, seismic hazard mitigation, Seismology, shakemap
|
|
C4-012
16:50
|
17:05
|
STRONG MOTION OBSERVATIONS AND CHARACTERISTICS IN MEINONG AND HUALIEN, TAIWAN EARTHQUAKES
The Mw 6.5 Meinong earthquake occurred in 2016 was caused by a blind fault; however significant rupture directivity effect was observed and lead to severe building damages and fatalities in Tainan City. During the 2018 Mw 6.4 Hualien earthquake, Milun fault ruptured from north (near the epicenter) to south (Hualien City), several damages and fatalities were located in the region suffered from obvious rupture forward directivity effect. The co-seismic deformation during Meinong earthquake was therefore smaller than that during Hualien earthquake according to strong motion and GPS data. Except for strong motion records, 50-Hz high-rate GPS records at near-fault region were used as strong motion time histories to supplement spatial distribution of stations. Strong motions with pulse-like velocities were observed by both strong motion and high-rate GPS stations during both the Meinong and Hualien earthquakes and the observed maximum PGV are 86 cm/s and 146 cm/s, respectively. Additionally, strong soil nonlinearity was observed in Hualien City and therefore caused special V/H behavior of strong motions. Meinong earthquake reveal the importance of rupture directivity effect of unknown blind faults, which may induce extremely large strong motions for an area has no active fault.
Cheng-Ying Ho, Chun-Hsiang Kuo, Che-Min Lin, Ruey-Juin Rau, Kuo-Liang Wen
High-rate GPS, Rupture directivity effect, Strong motion
|
|
C4-011
17:05
|
17:20
|
GROUND MOTION CHARACTERISTICS IN HUALIEN, TAIWAN BASED ON MICROTREMOR OBSERVATIONS
On February 6, 2018, an earthquake of MW6.4 struck the Hualien area of eastern Taiwan. Large ground deformations caused by fault displacement were recorded by several seismometers. Seventeen people were killed and 280 people were injured as a result of building and infrastructure damage. According to reports of the Central Weather Bureau, in some places in the city the ground shaking reached level VII on the Taiwanese seismic intensity scale. In earthquake disaster mitigation it is important to understand the characteristics of ground surface vibration such as the dominant period and amplification. In this study we carried out microtremor observations of single instruments and arrays in Hualien. Earthquake ground motion data were also collected. Using these data, we estimated the dominant periods at 81 points in the city of Hualien and produced a dominant period contour map. Array observations at points across the fault were obtained and ground profiles were estimated. We discuss the ground vibration characteristics in the city based on the dominant periods, earthquake records and estimated ground profiles.
Junji Kiyono, Takuto Miki, Kenji Fukunaga, Jian-Hong Wu
2018 Hualien earthquake, array observation, ground vibration characteristics, H/V spectrum ratio, microtremor observation
|
|
C4-013
17:20
|
17:35
|
A GRAPHICS PROCESSING UNIT (GPU) BASED MICROSEISMIC MONITORING PLATFORM
Currently, the high-quality continuous seismic recordings in Taiwan bring both new research possibilities and challenges in seismology. However, little attention has been given to the wealth of seismic signal related to microseismic events in continuous recordings. A detailed and precise analysis of microseismicity may play a key role in understanding the physical mechanisms of various earthquake processes. We have developed a platform for monitoring microseismicities using template matching algorithm (TMA). A large amount of normalized correlation coefficient (NCC) calculations are required for earthquake detections and therefore a Graphic Processing Unit (GPU) based code has been developed for the massive computational task. The current GPU implementation has shown about 800 times speedup as compared to a sequential CPU code. We have applied TMA to detect foreshocks and aftershocks of 2018 Mw 6.4 Hualien earthquake, the template matching algorithm has detected significant among of foreshocks and more than seven times more aftershocks than those in Central Weather Bureau (CWB) catalog. Our algorithm is also able for different types of microseismic monitoring, such as non-volcanic tremors.
En-Jui Lee, Po Chen, Dawei Mu, Ruey-Juin Rau
Aftershocks, Foreshocks, Graphics Processing Unit (GPU), Microseismic
|
|
C3-016
17:35
|
17:50
|
RADON MONITORING IN TATUN VOLCANIC GROUP, MAI-TAO-SAN AREAS AND WAN-DAN AREAS OF TAIWAN FOR SEISMIC AND VOLCANIC STUDY
Radon monitoring has been carried out at Tatun volcanic group of northern Taiwan using solid state nuclear track detectors (SSNTDs) technique. The pre-calibrated radon-thoron discriminators with LR films has been installed in Hsiaoyoukeng (SYK), Dayoukeng (DYK), Bayen (BY) and Gungtzeping (GTP) of Tatun Volcanic area in a hole (about 50 cm depths) having different temperatures for a defined period (bi-weekly to monthly). The preliminary results have shown that our monitoring stations in TVG area are sensitive to the events with in distance 60 kms. We have kept integrated monitoring at TVG area to test and verify the hypothesis mentioned above. In Mai-Tao-San areas and Wan-dan areas we are carrying out integrated radon monitoring with passive (SSNTDs) and active (RAD7) detector in soil whereas we are also carrying out radon monitoring in water with active (RAD7) detector. The results of this integrated radon monitoring in water and soil at Mai-tao-san areas and Wan-dan areas of Taiwan for seismic and volcanic study will also be discuss in detail.
Arvind Kumar, Shih-Jung Lin, Cheng-Horng Lin, Vivek Walia
Mai-tao-san, RAD 7, Radon, Solid state nuclear track detectors, Taiwan, Tatun Volcanic Group, Wan-dan
|