甚高频时岩石介电常数的测试方法和特性

本文讨论20-200MHz频率范围内平行板电容法测量岩石介电常数和电导率的方法,指出常用的静电场公式的局限性,给出较精确的似稳场动态公式,并讨论20-50MHz频率范围内的岩石介电特性。     

加卸载响应比理论在中强地震（7〉M≥6）预测中的应用

本文运用加卸载响应比理论对中强地震前的加卸载响应比的变化进行了研究。结果表明，中强地震前加卸载响应比明显上升，表现为三种不同的形态，持续时间为半年到两年。     

滇西实验场两次中强地震前后尾波衰减的区域特征分析

本文在地方震尾波散理论模式的基础上，研究了１９９２年１２月发生在滇西地震预报实验场永胜期纳地区的两次Ｍｓ５．４，５．１地震前后的序列地震尾波振幅衰减率β和尾波Ｑ＾－１值随时间和空间变化的衰减特征。结果表明，尾波衰减率在空间分布上表现出明显的小区域范围的非均匀性，在主震震中附近的近场区内，主震期间的β值是０．００７６ｓ＾－１，相应的Ｑｃ＾－１值是０．００５６；余震期间的β值是０．０２０９ｓ＾－１，相     

复杂性理论与滑坡动态预测研究新思路

概述了复杂性理论的发展及其对滑坡动态研究的影响；在概括滑坡预测研究现状的基础上，分析了滑坡动态预测的研究进展，最后是关于预测的几点思考，亦即滑坡动力学预测的研究方向。     

临汾地震安全性评价的极值理论方法

本文根据极值理论建立了烈度超越给定值Ｉｊ的平均重现期公式，根据最大熵原理建立了未来Ｔ年烈度超越给定值Ｉｊ的概率和烈度超越给定值Ｉｊ的平均重现期之间的关系式。利用１０２２─１９９３年的地震资料，计算了临汾（３６．１０°Ｎ，１１１．５０°Ｅ）周围４级以上地震在临汾造成的烈度，绘制了未来１００年临汾的烈度超越概率曲线（地震危险性曲线）。结果表明，计算场地烈度的极值理论方法可作为地震危险性分析的综合概率法的补充和验证。     

自组织人工神经网络在地震综合预报中的应用

运用三维Ｔ·ｋｏｈｏｎｅｎ自组织人工神经网络进行地震综合预报，以云南多震区─滇西南地震区为例进行实际判别计算，预测成功率达１００％。结果表明．神经网络方法性能较好，可望成为进行地震实际预报的一种有效的辅助手段。     

Application of stress-pore pressure coupling theory for porous media to the Xinfengjiang reservoir earthquakes

Theory of the coupling of stress-pore pressure in the saturated, elastic porous media is used in the study of the formation mechanism of the Xinfengjiang reservoir-induced earthquakes. Based on the results, it is believed that compared with the mechanism of additional stress in the vicinity of the reservoir, the mechanism of the coupling of additional stress and pore pressure may be more well-founded for the occurrence of reservoir-induced earthquakes.     

地震波场反演的BG－逆散射方法

本文讨论利用三维反射地震数据进行波场反演的一种方法，旨在取得高分辨率的地球模型．这种方法用Ｂａｃｋｕｓ－Ｇｉｌｂｅｒｔ的理论构造波动方程非线性反问题的逐次线性化迭代格式，用逆散射原理导出泛函的Ｆｒｅｃｈｅｔ导数，并用最佳折衷准则求解线性化后的方程组．根据迭代过程中不断提高分辨率的思想和减少计算成本的原则，设计了可供实用的反演算法流程．     

Micro-scale modelling in the study of plume evolution in heterogeneous media

The migration of contaminants in heterogeneous aquifers involves dispersive processes that act at different scales. The interaction of these processes as a plume evolves can be studied by micro-scale modelling whereby two scales, a local- or micro-scale and an aquifer- or macro-scale, are covered simultaneously. Local-scale dispersive processes are represented through the local dispersion coefficient in the transport equation, while large-scale dispersion due to heterogeneities is represented through the resolution of the flow field and the diffusive exchange between streamtubes. The micro-scale model provides both the high degree of resolution compatible with local-scale processes, and the extent required for the approach to asymptotic conditions, using grids of up to a million nodal points. The model is based on the dual potential-streamfunction formulation for flow, and the transport problem is formulated in a natural coordinate system provided by the flownet. Simulations can be used to verify stochastic theories of dispersion, without the restrictive assumptions inherent in the theory. For the two-dimensional case, results indicate convergence of the effective dispersivity to the theoretical macrodispersivity value. Convergence takes place within a travel distance of about 50 correlation lengths of the hydraulic conductivity field. However, the approach taken to asymptotic conditions, as well as the macrodispersivity value, may differ for different realizations of the same medium. The influence of early-time events such as plume splitting on the asymptotic convergence remains to be investigated.