Fuzzy neural network models for liquefaction prediction

Integrated fuzzy neural network models are developed for the assessment of liquefaction potential of a site. The models are trained with large databases of liquefaction case histories. A two-stage training algorithm is used to develop a fuzzy neural network model. In the preliminary training stage, the training case histories are used to determine initial network parameters. In the final training stage, the training case histories are processed one by one to develop membership functions for the network parameters. During the testing phase, input variables are described in linguistic terms such as ‘high’ and ‘low’. The prediction is made in terms of a liquefaction index representing the degree of liquefaction described in fuzzy terms such as ‘highly likely’, ‘likely’, or ‘unlikely’. The results from the model are compared with actual field observations and misclassified cases are identified. The models are found to have good predictive ability and are expected to be very useful for a preliminary evaluation of liquefaction potential of a site for which the input parameters are not well defined.     

Most likely locations of large earthquakes in the Kanto and Tokai areas, Japan, based on the local recurrence times

We examined the hypothesis that minima in local recurrence time, T_L, or equivalently maxima in local probability, P_L, may map asperities in the Kanto and Tokai areas of Japan, where the earthquake catalog of the National Research Institute for Earth Science and Disaster Prevention (NIED) is complete at the M=1.5 (M1.5) level. We mapped T_L (P_L) based on the a- and b-values of the nearest earthquakes within 20 km of every node of a grid spaced 0.01° for M7 target events. Only earthquakes within the top 33 km were used. The b-values increase strongly with depth, in several areas. Therefore, some of the T_L (P_L) anomalies are not revealed if data from the entire crustal seismogenic zone are mixed. Thus, we mapped T_L (P_L) separately for the top 15 km and the rest of the depth range, as well as for the entire seismogenic crust. The resulting T_L- and P_L-maps show that approximately 12% of the total area shows anomalously short recurrence times. Out of six shallow target events with M≥6.5 and which occurred since 1890, five are located within the anomalous areas with T_L <450 years. We interpret this to mean that areas with anomalously short T_L map asperities, which are more likely than other areas to generate future target events. The probability that this result is due to chance is vanishingly small. The great Kanto rupture of 1923 appears to have initiated in the most significant asperity we mapped in the study area. One anomaly is located in the northeastern part of the area of the proposed future rupture of the Tokai earthquake, and another one at its southwestern corner. The absolute values of T_L calculated are uncertain because they depend on the size of the volume used for the calculation.     

浅谈地震保险的必要性及其作用

以防震减灾为视角，针对地震灾害对社会发展及人民生活造成的重大损失，从强化地震保险意识的角度，论述地震保险的必要性和重大意义。同时对地震保险在心理、补偿、救灾、缓冲、防灾及附带等方面的作用加以阐述，从而证明要减少地震灾害的损失，充分认识地震保险的必要性及其作用，建立和完善有效的保险机制，对于减轻地震灾害，促进社会发展和人民安居乐业是一项必要的基础性工作。     

辽宁省地震前兆台网技术改造概述

“九五”期间实施的《辽宁省地震前兆台网技术改造项目》使辽宁省地震前兆观测技术系统由模拟到数字实现了质的飞跃，本文对项目的技术思路，实施情况，取得的成果及建设特色进行了介绍。     

地震前兆时空非均匀性指标Cv值的实验检验

在双向加载条件下，对两种构造和介质不同的岩石标本进行变形实验，测量了标本表面不同部位的应变．在此基础上，利用王晓青和陈学忠等提出的描述地震前兆非均匀分布的参量————Cv值，分析了标本变形失稳过程中应变异常的分布，以期从实验角度检验Cv值方法并探讨其物理意义．研究表明，Cv值的变化与岩石变形特征的变化有关，是描述前兆分布非均匀性的一种有效指标；Cv值在失稳发生前先异常上升，然后下降，并在Cv值恢复过程中或恢复后伴随着失稳事件的发生．因此，Cv值是一种有效的预报指标．      

1999年台湾7.6级大震与江苏-南黄海地区中强震预测

分析讨论了台湾地区7级大震与本区中强震之间相关关系，指出1999年台湾7.6级大震后2－3年内，本区将有5－6级中强震发生。同时应用可公度模型和“带头地震”的异年倍九法联合对本区中强震发震时间进行分析预测。结论表明，该方法可以较好地应用于本区短临地震预报实践。     

云南中强震的水氡前兆异常特征

2001年云南澜沧5.0、施甸5.2,5.9,5.3、宁蒗5.8、楚雄5.3和江川5.1级地震前，云南49项水氡前兆中的一些台站出现突出的异常。研究发现，根据这些异常项的异常特征和空间分布特征及历史震例，可对这几次地震预报的三要素，尤其是地点和时间做出较好的判断。     

地震预报思路的再讨论

文中回顾了1974年提出的“块、带、源、兆、触、报”六字预报思路，并基于近20年来的研究对此进行了补充和深化。     

川滇地区地震(Ms≥5.0)破裂类型与前兆异常分布的关系

对川滇地区 31次 (组 ) 5 .0级以上地震与余震次数的关系进行了统计 ,结合“强地震分为走滑型地震和断错型地震”的理论 ,确定了各次 (组 )地震的破裂类型。从物理机制上对地震破裂类型与震前前兆异常分布关系进行了讨论 ,由此进一步探讨了 5 .0级以上地震震后趋势预测     

大姚地热动态特征分析

通过大姚高精度水温观测井水地质、井孔条件的分析，震例观测资料、实验观测数据的处理，以及井水水质分析结果，表明大姚水温周期性变化的突出位置是井深75m左右，根本原因是该地层石膏脉稳定释热，含水层热水与井内冷水混合过程中，在通道内形成钙盐类沉淀物（如CaSO4CaCO3），堵塞对流通道，含水层内外压差突破被阻塞通道，热水反复侵入井内，便形成水温周期性变化。地震的孕育、发生和调整过程，由于应力场作用，产生附加地热场，水温基值发生变化，周期畸变或消失，这种附加地热动态是可以恢复的。所以认为大姚地热动态是井区特殊地层地热信息的反映。