RAPID ESTIMATION OF PARAMETERS FOR THE M6.8 EARTH-QUAKE ON JANUARY 7, 2025 IN DINGRI(XIZANG, CHINA) BASED ON DATA-DRIVEN METHODS

doi:10.3969/j.issn.0253-4967.2025.03.20250014

Abstract

Abstract:

On January 7, 2025, a magnitude 6.8 earthquake struck Dingri County, Shigatse City, Xizang, China, causing significant casualties and property damage. Rapid and accurate estimation of earthquake magnitude, instrumental intensity, and ground motion parameters is essential for seismic hazard assessment and emergency response. Magnitude, as a key indicator of the energy released by an earthquake, lays the foundation for preliminary disaster assessment. Instrumental intensity, calculated from the intensity of ground motion observed by instruments, can be used directly to determine the extent of damage and the severity of disasters. Ground motion parameters, such as PGA and PGV, are widely used in seismic design, disaster assessment, and seismic damage prediction and are important metrics for evaluating the impact of earthquakes on buildings and infrastructure.

In this study, a data-driven multi-task joint estimation framework is proposed that combines the SeismNet model for rapid magnitude and instrumental intensity estimation with the CRAQuake model for rapid estimation of ground shaking parameters. The framework is applied to the January 7, 2025, Dingri earthquake of magnitude 6.8, where the magnitude, instrumental intensity, and ground motion parameters are estimated and analyzed in parallel. The study starts by filtering and processing the strong motion data obtained, and then estimates the magnitude, instrumental intensity, and ground motion parameters by parallel computation. The results show that: 1)The estimated magnitude provided by SeismNet is 6.17 when the seismic data are input for 3s. With the increase in seismic wave duration, the estimated magnitude gradually approaches the catalog value, and the estimated magnitude is 6.71 at 7 seconds, with a significant reduction in the error. 2)For the instrumental intensity estimation, the results obtained by SeismNet are the same as those of the instrumental intensity flash report when the seismic data are input for 8 to 10 seconds. When data of 6 seconds or longer were used, there were no false alarms or omissions, showing a high degree of accuracy. 3)The estimates of ground-motion parameters provided by the CRAQuake model are in good agreement with observations, providing reliable results within a few seconds, especially for PGA, PGV, and other parameters, with minor and stable errors.

These results indicate that the data-driven estimation model exhibits strong generalization ability in the Dingri earthquake, particularly in the epicenter region and the early post-earthquake period, providing fast and reliable decision support. With the increase of seismic wave duration, the estimation results of SeismNet and CRAQuake are more stable, the errors are gradually reduced, and the estimation accuracy is significantly improved. Through parallel computing, these two models can estimate multiple seismic parameters at the same time, which not only enhances the estimation efficiency, but also provides efficient and comprehensive technical support for earthquake emergency response. Additionally, data-driven methods offer significant advantages in earthquake emergency response, particularly in large-scale earthquake disasters. These methods can quickly estimate magnitude, instrumental intensity, and ground motion parameters, providing more accurate decision support. The results offer new technical insights and methodological support for future large-scale earthquake emergency response and lay the foundation for the widespread application of data-driven methods in the earthquake field.

Key words: Dingri M6.8 earthquake, data-driven models, magnitude, instrumental intensity, ground motion parameters

摘要：

2025年1月7日, 西藏藏族自治区日喀则市定日县发生6.8级地震, 造成了人员伤亡和严重的财产损失, 快速准确地估算地震震级、仪器烈度及地震动参数对于地震应急与减灾具有重要意义。文中基于快速估算地震震级与仪器烈度模型SeismNet和快速估算地震动参数模型CRAQuake对定日6.8级地震的震级、仪器烈度和地震动参数进行了并行估算和分析, 并验证模型性能和积累震例经验。研究结果如下: 1)在输入3s地震波时, SeismNet即可估算震级的平均值为6.17级, 且随着地震波时长的增加, 估算震级的平均值与编目震级逐渐接近; 2)在仪器烈度估算方面, 当SeismNet输入8~10s的地震波时, 能够输出与烈度速报比较一致的结果, 且输入6s及以上地震波时, 无误报或漏报现象; 3)CRAQuake估算的地震动参数与观测值具有较好的吻合度, 能够在数秒内给出相对可靠的结果。这些结果表明, 基于数据驱动的估算模型在本次地震中表现出优异的性能, 具有较强的泛化能力, 显著提升了地震应急响应的效率与准确性。文中的研究成果为后续应急管理提供了重要的参考依据和技术支持, 同时也表明数据驱动方法能为大地震应急提供有价值的评价结果和重要的参考依据。

关键词: 定日6.8级地震, 数据驱动模型, 震级, 仪器烈度, 地震动参数

ZHAO Qing-xu, RONG Mian-shui, ZHANG Bin, WANG Ji-xin, KONG Xiao-shan, LI Xiao-jun. RAPID ESTIMATION OF PARAMETERS FOR THE M6.8 EARTH-QUAKE ON JANUARY 7, 2025 IN DINGRI(XIZANG, CHINA) BASED ON DATA-DRIVEN METHODS[J]. SEISMOLOGY AND GEOLOGY, 2025, 47(3): 969-983.

赵庆旭, 荣棉水, 张斌, 王继鑫, 孔小山, 李小军. 基于数据驱动模型的2025年1月7日西藏定日6.8级地震参数快速估算[J]. 地震地质, 2025, 47(3): 969-983.

Figures/Tables 12

Fig. 1 Relationship between PGA and hypocentral distance(a)and distribution of the mainshock epicenter and selected stations(b).

Table 1 Basic information of selected stations

台站名称	仪器类型	北纬/(°)	东经/(°)	震源距/km	P波到时/s
DB001	强震仪	29.10	89.20	183.60	57.59
DS002	强震仪	28.90	88.00	70.38	36.61
DX001	强震仪	29.44	88.20	127.66	46.45
RKZ	强震仪	29.26	88.80	156.65	52.65
D0001	烈度仪	29.30	87.20	92.50	40.80
D0003	烈度仪	30.07	86.90	182.31	58.26
D0004	烈度仪	28.37	87.70	30.12	31.52
D0005	烈度仪	28.24	88.40	97.99	42.14
D0007	烈度仪	29.09	87.60	67.75	35.78
D0009	烈度仪	29.68	89.00	199.94	58.61

Fig. 2 Acceleration waveforms and P-wave arrival times of stations with minimum hypocentral distance and maximum PGA.

Fig. 3 Acceleration response spectra of stations with minimum hypocentral distance and maximum PGA.

Table2 Instrumental intensity and ground motion parameters of selected stations

台站名称	仪器烈度 (速报结果)	仪器烈度 (本文计算结果)	PGA /cm·s^-2	PGV /cm·s^-1	PGD /cm	SA(0.3) /cm·s^-2	SA(1.0) /cm·s^-2	SA(3.0) /cm·s^-2
DB001	2.9	2.90	3.14	1.34	0.87	6.34	4.11	7.45
DS002	5.5	5.54	40.30	3.64	2.18	124.01	20.35	12.53
DX001	4.8	4.83	17.32	4.11	3.53	39.55	10.31	8.53
RKZ	3.6	3.61	4.30	2.51	1.82	7.97	8.35	9.73
D0001	6.1	6.10	62.59	6.79	5.81	125.47	51.87	16.06
D0003	4.2	4.24	7.39	4.55	3.71	17.51	12.31	9.77
D0004	7.3	7.28	114.76	14.86	5.21	285.08	170.17	28.06
D0005	4.6	4.62	20.72	2.24	1.45	59.53	7.81	7.01
D0007	8.0	8.02	396.02	28.29	8.29	1066.71	277.97	79.00
D0009	4.1	4.09	16.27	1.29	1.00	68.47	6.52	5.11

Fig. 4 Schematic of rapid estimation of magnitude, instrumental intensity, and ground motion parameters.

Fig. 5 Results of SeismNet estimation of magnitude for Dingri M6.8 earthquake.

Fig. 6 Distribution map of instrumental intensity for the Dingri M6.8 earthquake(Cited from the Strong Motion Observation Center, Institute of Engineering Mechanics, China Earthquake Administration).

Fig. 7 SeismNet estimated instrumental intensity results for the M6.8 Dingri earthquake(3s, 5s, 8s, 10s).

Fig. 8 Threshold determination results of instrumental Intensity for Dingri M6.8 earthquake estimated by SeismNet.

Fig. 9 Results of ground motion parameter estimation for Dingri M6.8 earthquake using CRAQuake with 3s seismic wave input.

Fig. 10 Results of ground motion parameter estimation for Dingri M6.8 earthquake using CRAQuake with different seismic wave durations.

References 37

[1]	金星. 2024. 地震预警与烈度速报-风险与控制[M]. 北京: 科学出版社.
	JIN Xing. 2024. Earthquake Early Warning and Intensity Rapid Reporting: Risks and Control[M]. Science Press, Beijing (in Chinese).
[2]	金星, 张红才, 李军, 等. 2012. 地震预警震级确定方法研究[J]. 地震学报, 34(5): 593-610.
	JIN Xing, ZHANG Hong-cai, LI Jun, et al. 2012. Research on earthquake early warning magnitude estimate[J]. Acta Seismologica Sinica, 34(5): 593-610 (in Chinese).
[3]	刘辰, 李小军, 景冰冰, 等. 2019. 地震预警PGV-Pd关系参数的距离分段特征[J]. 地球物理学报, 62(4): 1413-1426. DOI
	LIU Chen, LI Xiao-jun, JING Bin-bin, et al. 2019. The distance segmentation characters of PGV-Pd relationship parameters for earthquake early warning[J]. Chinese Journal of Geophysics, 62(4): 1413-1426 (in Chinese).
[4]	马美帅, 王延伟, 汪祚赚, 等. 2022. 基于累积绝对位移值的震级估算方法[J]. 自然灾害学报, 31(2): 93-101.
	MA Mei-shuai, WANG Yan-wei, WANG Zuo-zhuan, et al. 2022. Magnitude estimation based on the cumulative absolute displacement[J]. Journal of Natural Disasters, 31(2): 93-101 (in Chinese).
[5]	赵庆旭, 王延伟, 莫红艳, 等. 2024. 基于多输入高斯过程回归的震级快速估算方法[J]. 地震学报, 46(5): 806-824.
	ZHAO Qing-xu, WANG Yan-wei, MO Hong-yan, et al. 2024. Rapid magnitude estimation based on multi-input Gaussian process regression[J]. Acta Seismologica Sinica, 46(5): 806-824 (in Chinese).
[6]	中国地震局. 2020. 中国地震烈度表GB/T 17742-2020[S]. 北京: 中国标准出版社.
	China Earthquake Administration. 2020. The Chinese seismic intensity scale, GB/T 17742-2020[S]. Standards Press of China, Beijing (in Chinese).
[7]	Allen R, Kanamori H. 2003. The Potential for earthquake early warning in southern California[J]. Science, 300(5620): 786-789. doi: 10.1126/science.1080912. PMID
[8]	Bindi D, Pacor F, Puglia R, et al. 2011. Ground motion prediction equations derived from the Italian strong motion database[J]. Bulletin of Earthquake Engineering, 9(6): 1899-1920. doi: 10.1007/s10518-011-9313-z.
[9]	Campbell K, Bozorgnia Y. 2008. NGA ground motion model for the geometric mean horizontal component of PGA, PGV, PGD and 5% damped linear elastic response spectra for periods ranging from 0.01 to 10s[J]. Earthquake Spectra, 24(1): 139-171. doi: 10.1193/1.2857546.
[10]	Campbell K, Bozorgnia Y. 2011. A ground motion prediction equation for JMA instrumental seismic intensity for shallow crustal earthquakes in active tectonic regimes[J]. Earthquake Engineering & Structural Dynamics, 40(4): 413-427. doi: 10.1002/eqe.1027.
[11]	Campbell K, Bozorgnia Y. 2014. NGA-West 2 ground motion model for the average horizontal components of PGA, PGV, and 5% damped linear acceleration response spectra[J]. Earthquake Spectra, 30(3): 1087-1115. doi: 10.1193/062913EQS175M.
[12]	Dai H Z, Zhou Y Y, Liu H Y, et al. 2024. XGBoost-based prediction of on-site acceleration response spectra with multi-feature inputs from P-wave arrivals[J]. Soil Dynamics and Earthquake Engineering, 178:108503. doi: 10.1016/j.soildyn.2024.108503.
[13]	Festa G, Zollo A, Vesuviano I. 2008. Magnitude estimation from early radiated energy[J]. Geophysical Research Letters, 35(22): L22307. doi: 10.1029/2008GL035576.
[14]	Hu J J, Ding Y T, Zhang H, et al. 2023. A Real-time seismic intensity prediction framework based on interpretable ensemble learning[J]. Seismological Research Letters, 94(3): 1579-1602. doi: 10.1785/0220220167.
[15]	Ide S. 2019. Frequent observations of identical onsets of large and small earthquakes[J]. Nature, 573(7772): 112-116. doi: 10.1038/s41586-019-1508-5.
[16]	Kale Ö, Akkar S, Anasri A, et al. 2015. A ground-motion predictive model for Iran and Turkey for horizontal PGA, PGV, and 5% damped response spectrum: Investigation of possible regional effects[J]. Bulletin of the Seismological Society of America, 105(2A): 963-980. doi: 10.1785/0120140134.
[17]	Kanamori H. 2005. Real-time seismology and earthquake damage mitigation[J]. Annual Review of Earth and Planetary Sciences, 33(1): 195-214. doi: 10.1146/annurev.earth.33.092203.122626.
[18]	Karim K, Yamazaki F. 2002. Correlation of JMA instrumental seismic intensity with strong motion parameters[J]. Earthquake Engineering & Structural Dynamics, 31(5): 1191-1212. doi: 10.1002/eqe.158.
[19]	Olson E, Allen R. 2005. The deterministic nature of earthquake rupture[J]. Nature, 438(7065): 212-215. doi: 10.1038/nature04214.
[20]	Peng C Y, Yang J S, Xue B, et al. 2013. Exploring the feasibility of earthquake early warning using records of the 2008 Wenchuan earthquake and its aftershocks[J]. Soil Dynamics and Earthquake Engineering, 57:86-93. doi: 10.1016/j.soildyn.2013.11.005.
[21]	Peng C Y, Yang J S, Zheng Y, et al. 2017. New τ_c regression relationship derived from all P-wave time windows for rapid magnitude estimation[J]. Geophysical Research Letters, 44(4): 1724-1731. doi: 10.1002/2016GL071672.
[22]	Shieh J, Wu Y, Allen R. 2008. A comparison of τ_c and τ^p_max for magnitude estimation in earthquake early warning[J]. Geophysical Research Letters, 35(20): L20301. doi: 10.1029/2008GL035611.
[23]	Wald J, Quitoriano V, Heaton T H, et al. 1999. Relationships between peak ground acceleration, peak ground velocity, and modified mercalli intensity in California[J]. Earthquake Spectra, 15(3): 557-564. doi: 10.1193/1.1586058.
[24]	Wang Y W, Li X J, Li L, et al. 2021. New magnitude proxy for earthquake early warning based on initial time series and frequency[J]. Seismological Research Letters, 93(1): 216-225. doi: 10.1785/0220210106.
[25]	Wang Y W, Li X J, Wang Z F, et al. 2022. Deep learning for magnitude prediction in earthquake early warning[J]. Gondwana Research, 123:164-173. doi: 10.1016/j.gr.2022.06.009.
[26]	Wu Y, Kanamori H. 2005. Rapid assessment of damage potential of earthquakes in Taiwan from the beginning of P waves[J]. Bulletin of the Seismological Society of America, 95(3): 1181-1185. doi: 10.1785/0120040193.
[27]	Wu Y, Kanamori H. 2008. Exploring the feasibility of on-site earthquake early warning using close-in records of the 2007 Noto Hanto earthquake[J]. Earth Planets and Space, 60:155-160. doi: 10.1186/BF03352778.
[28]	Wu Y, Kanamori H, Allen R, et al. 2007. Determination of earthquake early warning parameters, τ_c and P_d, for southern California[J]. Geophysical Journal International, 170(2): 711-717. doi: 10.1111/j.1365-246X.2007.03430.x.
[29]	Wu Y, Zhao L. 2006. Magnitude estimation using the first three second P-wave amplitude in earthquake early warning[J]. Geophysical Research Letters, 331(16): L16312. doi: 10.1029/2006GL026871.
[30]	Zhang B, Yu Y X, Li X J, et al. 2021. Ground motion attenuation relationship of horizontal component of PGV and PGD in southwest China[J]. Chinese Journal of Geophysics, 64(8): 2733-2748.
[31]	Zhang B, Yu Y X, Li X J, et al. 2022. Ground motion prediction equation for the average horizontal component of PGA, PGV, and 5%damped acceleration response spectra at periods ranging from 0.033 to 8.0s in southwest China[J]. Soil Dynamics and Earthquake Engineering, 159:107297. doi: 10.1016/j.soildyn.2022.107297.
[32]	Zhang H C, Jin X, Wei Y X, et al. 2016. An earthquake early warning system in Fujian, China[J]. Bulletin of the Seismological Society of America, 106(2): 755-765. doi: 10.1785/0120150143.
[33]	Zhao Q X, Rong M S, Wang J X, et al. 2024a. An end-to-end multi-task network for early prediction of the instrumental intensity and magnitude in the north-south seismic belt of China[J]. Journal of Asian Earth Sciences, 276:106369. doi: 10.1016/j.jseaes.2024.106369.
[34]	Zhao Q X, Rong M S, Zhang B, et al. 2024b. An All-in-one rapid prediction of ground motion intensity measures hybrid network for multi-task in the North-South seismic belt of China[J]. Earth and Space Science, 11(10): e2024EA003664. doi: 10.1029/2024EA003664.
[35]	Zhu J B, Li S Y, Song J D. 2022. Magnitude estimation for earthquake early warning with multiple parameter inputs and a support vector machine[J]. Seismological Research Letters, 93(1): 126-136. doi: 10.1785/0220210144.
[36]	Ziv A. 2014. New frequency-based real-time magnitude proxy for earthquake early warning[J]. Geophysical Research Letters, 41(20): 7035-7040. doi: 10.1002/2014GL061564.
[37]	Zollo A, Amoroso O, Lancieri M, et al. 2010. A Threshold-based earthquake early warning using dense accelerometer networks[J]. Geophysical Journal International, 183(2): 963-974. doi: 10.1111/j.1365-246X.2010.04765.x.