利用中国地震台网中心提供的1970—2019年8月全国地震目录，在分析青藏块体最小完备震级的基础上，应用PI算法，选取空间网格尺度0.2°×0.2°，截止震级为ML 3.0，目标震级≥MS 6.0，预测时间窗设定为10 a、5 a、3 a不同时间尺度进行回溯性研究，并对预测结果进行ROC检验。结果表明，预测时间窗内的强震基本发生在PI算法所得异常区域附近，且预测结果明显优于随机概率法；PI算法对青藏块体中长时间尺度的强震具有一定的预测效能，而且将预测时间窗设定为3 a时，预测效能较好。此外，定性分析了活动块体边界带“目标”地震与地震热点丛集间的关系，发现青藏块体MS≥6.0地震与活动块体边界带地震热点间的对应关系较好。在实际应用中，将PI算法计算结果与活动块体边界带结合起来考虑强震相对危险区域时，该研究结果具有一定的参考价值。
Pattern Informatics is a medium- and long-term earthquake prediction method based on statistical physics and seismology. Some studies have shown that the method has a high prediction efficiency for seismically active areas. Based on the earthquake catalogue since 2017 provided by the China Earthquake Networks Center, this paper analyzes the minimum completeness magnitude of the Qinghai-Tibet black and adopts the PI method to carry out a retrospective study in the research area of (30.0°—41.0°N，88.0°—105.0°E). The spatial grid size is 0.2°×0.2°, the cutoff magnitude is ML 3.0, the target magnitude is ≥MS 6.0, and the learning windows and forecasting windows are 10, 5, and 3 years. The results showed that earthquakes have a good corresponding relationship with the seismic hotspots of the PI image in the prediction period, and the prediction results are obviously better than the random probability. The PI method has a certain prediction efficiency for the occurrence of earthquakes in the middle and long-timescale of the Qinghai-Tibet block, and the prediction efficiency is better when set learning window and forecasting window 3 years. In addition, this paper qualitatively analyzes the relationship between the “target” earthquakes in the active block boundary zone and the seismic hotspots. It is found that the corresponding relationship is very good between the MS≥6.0 earthquakes and the seismic hotspots in the active block boundary zone. Therefore, it is valuable to combine the results with the boundary zone of the active block in the study of earthquake prediction.
2021,42(1): 35-41 收稿日期：2020-04-29
DOI：10.3969/j.issn.1003-3246.2021.01.006
基金项目：青海省地震科学基金（项目编号：2020A04）
作者简介：余娜（1984—），女，工程师，硕士，从事数字地震资料的应用研究。E-mail：yuna2003@126.com
参考文献：
陈建志，Rundle J B，Turcotte D L，等. 从临界转变的角度理解地震预测技术[J]. 物理，2013，42（5）：329-333.
蒋长胜，吴忠良. 对地震预测的一个统计物理算法在川滇地区的回溯性预测检验[J]. 中国科学 D辑：地球科学，2008，38（7）：852-861.
吴哲. 中国大陆第5地震轮回青藏块体强震活动特征[J]. 地震地磁观测与研究，2018，39（6）：1-6.
袁伏全，孙世瑞，王小玲. PI算法用于青海地区中强震危险性预测的回溯性检验研究[J]. 地震研究，2016，39（Z1）：76-82.
张浪平，李志雄，邵志刚，等. 中国大陆活动地块边界带强震趋势研究[J]. 地震，2010，30（3）：53-63.
张小涛，张永仙，夏彩韵，等. 利用图像信息方法研究芦山MS 7.0地震前川滇及附近地区的图像异常[J]. 地震学报，2014，36（5）：780-789.
Chen C C, Rundle J B, Holliday J R, et al. The 1999 Chi-Chi, Taiwan, earthquake as a typical example of seismic activation and quiescence[J]. Geophys Res Lett, 2005, 32(22): L22315.
Holliday J R, Nanjo K Z, Tiampo K F, et al. Earthquake forecasting and its verification[J]. Nonlin Processes Geophys, 2005, 12(6): 965-977.
Holliday J R, Rundle J B, Tiampo K F, et al. Modification of the pattern informatics method for forecasting large earthquake events using complex eigenfactors[J]. Tectonophysics, 2006, 413(1/2): 87-91.
Jiang C S, Wu Z L. PI forecast for the Sichuan-Yunnan region: retrospective test after the May 12, 2008, Wenchuan earthquake[J]. Pure Appl Geophys, 2010, 167(6/7): 751-761.
Jiang C S, Wu Z L. PI forecast with or without de-clustering: an experiment for the Sichuan-Yunnan region[J]. Nat Hazards Earth Syst Sci, 2011, 11(3): 697-706.
Jiang C S, Wu Z L. Intermediate-term medium-range precursory accelerating seismicity prior to the 12 May 2008, Wenchuan earthquake[J]. Pure Appl Geophys, 2013, 170(1/2): 209-219.
Moore E F. Machine models of self-reproduction[J]. Proc Symp Appl Math, 1962, 14: 17-33.
Nanjo K Z, Rundle J B, Holliday J R, et al. Pattern informatics and its application for optimal forecasting of large earthquakes in Japan[J]. Pure Appl Geophys, 2006, 163(11/12): 2 417-2 432.
Rundle J B, Tiampo K F, Klein W, et al. Self-organization in leaky threshold systems: The influence of near-mean field dynamics and its implications for earthquakes, neurobiology, and forecasting[J]. Proc Natl Acad Sci USA, 2002, 99(Z1): 2 514-2 521.
Tiampo K F, Rundle J B, McGinnis S J, et al. Mean-field threshold systems and phase dynamics: An application to earthquake fault systems[J]. EPL, 2002, 60(3): 481-488.
Woessner J, Wiemer S. Assessing the quality of earthquake catalogs: Estimating the magnitude of completeness and its uncertainty[J]. Bull Seismol Soc Am, 2005, 95(2): 684-698.
Zhang Y X, Zhang X T, Yin X C, et al. Study on the forecast effects of PI method to the North and Southwest China[J]. Concurrency and Computation: Practice and Experience, 2010, 22(12): 1 559-1 568.
Zhang Y X, Zhang X T, Wu Y J, et al. Retrospective study on the predictability of pattern informatics to the Wenchuan M 8.0 and Yutian M 7.3 earthquakes[J]. Pure Appl Geophys, 2013, 170(1/2): 197-208.
