3:30 PM - 3:45 PM
[HDS11-11] Extraction of Tsunami Time-Series Data at Amaharashi Coast, Takaoka City, for the 2024 Noto Peninsula Earthquake from Live Camera Images
Keywords:Tsunami, Image Analysis, 2024 Noto Peninsula earthquake
Introduction:
Tsunami due to the Noto Peninsula earthquake (M7.6), which occurred at 16:10 on January 1, 2024, was observed along the coastal areas of the Sea of Japan. The early arrival of the tsunami at the Toyama tidal station implies the possibility of an additional tsunami source in the Toyama Bay other than crustal deformation associated with the earthquake (Tsushima et al., 2024, this meeting). In this study, we analyzed live camera images to obtain more information on tsunamis along the Toyama Bay coast other than the tidal stations.
Data and Methods:
We were provided with the video data from the live camera of the Takaoka Cable Network, which are used as weather information for Amaharashi Coast in Takaoka City, along Toyama Bay. The method used to exract the time-series tsunami waveform from the video data is as follows:
(1) Convert video data into still images at 1-minute intervals.
(2) The height per pixel of the image was calculated using the height of rocks on the beach measured in the field.
(3) The position of the coastline captured in each of the still images in (1) was read by eye. Using (2), the difference in water level at each time point from that before the tsunami arrival was calculated.
(4) The time-series tsunami waveform was summarized from (3).
Results:
The time-series tsunami waveforms for the Ameharashi Coast (Fig. 1) were obtained. Still images after sunset were not available to determine the sea level from the images due to darkness. Fig. 1 compares the extracted data of this study with the observation at the Fushiki Port tide station (Port and Harbor Bureau), which is located approximately 3 km in a straight line from the Amaharashi Coast. The waveforms were generally similar between the data extracted from the video data and Fushiki Port, but wave heights tended to be slightly larger on the Amaharashi Coast.
Discussions and Future Works:
The difference between the maximum tsunami heights at the Amaharashi Coast (1.0m, this study ) and those observed at Fushiki Port tide station (0.8m, Port and Harbor Bureau) may be related to the fact that the latter is located at the mouth of the Oyabe River, while the former is directly facing the bay. The tsunami arrival time at Toyama, Fushiki Port and Amaharashi Coast are three, four, seven minutes after the earthquake, respectively; the finding of these difference is valuable for analysing the location of additional tsunami source in Toyama Bay.
By extracting time-series water level waveforms from video data, the spatial density of tsunami observations has increased after this study. The method engaged in this study is applicable to many live cameras along the coast and other tsunami cases. Although nighttime data could not be obtained in this study, high-sensitivity camera or floodlighting can be the solution of this restriction; automatic recognition of images may enable to automate the data extraction procedure in this study. After these conditions are met, real-time tsunami observation by live cameras can be realized. Video cameras are simpler observation facilities than typical tsunami observations such as using tide gauges and water pressure gauges installed along the coast, and have potential to realize inexpensive multi-point real-time tsunami observation network.
Acknowledgments: The images used in this study were provided by Takaoka Cable Network. We would like to express our gratitude to them. This study was partly supported by JSPS KAKENHI Grant Number JP23K17482.
Tsunami due to the Noto Peninsula earthquake (M7.6), which occurred at 16:10 on January 1, 2024, was observed along the coastal areas of the Sea of Japan. The early arrival of the tsunami at the Toyama tidal station implies the possibility of an additional tsunami source in the Toyama Bay other than crustal deformation associated with the earthquake (Tsushima et al., 2024, this meeting). In this study, we analyzed live camera images to obtain more information on tsunamis along the Toyama Bay coast other than the tidal stations.
Data and Methods:
We were provided with the video data from the live camera of the Takaoka Cable Network, which are used as weather information for Amaharashi Coast in Takaoka City, along Toyama Bay. The method used to exract the time-series tsunami waveform from the video data is as follows:
(1) Convert video data into still images at 1-minute intervals.
(2) The height per pixel of the image was calculated using the height of rocks on the beach measured in the field.
(3) The position of the coastline captured in each of the still images in (1) was read by eye. Using (2), the difference in water level at each time point from that before the tsunami arrival was calculated.
(4) The time-series tsunami waveform was summarized from (3).
Results:
The time-series tsunami waveforms for the Ameharashi Coast (Fig. 1) were obtained. Still images after sunset were not available to determine the sea level from the images due to darkness. Fig. 1 compares the extracted data of this study with the observation at the Fushiki Port tide station (Port and Harbor Bureau), which is located approximately 3 km in a straight line from the Amaharashi Coast. The waveforms were generally similar between the data extracted from the video data and Fushiki Port, but wave heights tended to be slightly larger on the Amaharashi Coast.
Discussions and Future Works:
The difference between the maximum tsunami heights at the Amaharashi Coast (1.0m, this study ) and those observed at Fushiki Port tide station (0.8m, Port and Harbor Bureau) may be related to the fact that the latter is located at the mouth of the Oyabe River, while the former is directly facing the bay. The tsunami arrival time at Toyama, Fushiki Port and Amaharashi Coast are three, four, seven minutes after the earthquake, respectively; the finding of these difference is valuable for analysing the location of additional tsunami source in Toyama Bay.
By extracting time-series water level waveforms from video data, the spatial density of tsunami observations has increased after this study. The method engaged in this study is applicable to many live cameras along the coast and other tsunami cases. Although nighttime data could not be obtained in this study, high-sensitivity camera or floodlighting can be the solution of this restriction; automatic recognition of images may enable to automate the data extraction procedure in this study. After these conditions are met, real-time tsunami observation by live cameras can be realized. Video cameras are simpler observation facilities than typical tsunami observations such as using tide gauges and water pressure gauges installed along the coast, and have potential to realize inexpensive multi-point real-time tsunami observation network.
Acknowledgments: The images used in this study were provided by Takaoka Cable Network. We would like to express our gratitude to them. This study was partly supported by JSPS KAKENHI Grant Number JP23K17482.