地球油气资源潜力是自然科学的基本问题

层序地层学自１９８７年问世以来，在国外如火如荼地发展起来，石油地质界对它的发展尤为关注。然而，在我国却显得举步维艰，进展缓慢。本文讨论了出现这种状况的起因：①不认识地震反射界面基本上是等时面；②认为地震／层序地层学是研究被动大陆边缘海相地层产生的，不适于非海相地层；③常规剖面中的地震反射频率和分辨率太低，不足以进行层序地层学研究；④层序地层学属纯理论科学，脱离油气勘探实际，采纳它的概念可能打气生产秩序。最后文章从正面阐述了开展层序地层学研究的必要性。     

8807号台风局地特大暴雨过程中尺度分析

分析了８８０７号台风所引发的局地特大暴雨成因：（１）台风提供了足够的水汽；（２）台风螺旋状云雨区和太阳辐射不均匀，形成了低层中尺度锋区，为暴雨发生提供极不稳定能量和触发条件；（３）局地地形和摩擦因子使湖北省宜昌地区成为这次对流云发展较有利的地点；（４）高空的中尺度降压系统和涡度因子加强了对流云团的发展。     

宁夏及青海地震活动性统计指标及未来地震趋势的综合概率预报

本文叙述了地震活动的综合概率预报方法，并对宁夏及青海地区（３５°００’－４１°００’Ｎ，１００°００’－１０７°４０’Ｅ）１９７０年以来的地震活动资料进行了统计计算，对１９９２年－１９９３年该区地震活动进行了综合预报效能检验。并据此对该区１９９４年－１９９５年的地震趋势做了估计。     

新疆钻孔应变异常特征及其与地震三要素关系研究

将新疆高精度钻孔应变仪１０年来观测到的应变异常划分为趋势异常、短期异常、临震异常三类，并对这三种异常的特征进行了分析总结．认为，地震三要素与应变异常参量关系密切，震级与异常量级、震中距有着确定性关系。     

地下流体前兆模糊识别

在地震的孕育、发生到发展的过程中，可能会出现多项地下流体观测量不同程度的前兆异常变化，这些异常变化是复杂多样、模糊不清的。引入模糊识别法对一定区域和一定时间段的地下流体观测数据进行了分析和处理，表明模糊识别法可较好地识别和分辨出地下流体前兆的地震异常。     

求震源机制P波初动解的格点尝试概率法

考虑到Ｐ波初动数据在震源球面上分布的不均匀性对节面解的可能影响，不用矛盾符号所占比例的大小作为选择可取节面解的标准，而是将震源球面分成许多小块，再确定有Ｐ波初动数字的小块为正号区或负号区的概率，最后以全部有数据小块的平均概率的高低作为选择解答的标准。用此改进方法重新测定了徐纪仁等（Ｘｕ ｅｔ．ａｌ，，１９８８）发表的青藏高原地区９９个地震的震源机制解，从中筛选出了２０个可信的解答。     

Study on electric variations of media in epicentral area by geomagnetic transfer functions

ＳｔｕｄｙｏｎｅｌｅｃｔｒｉｃｖａｒｉａｔｉｏｎｓｏｆｍｅｄｉａｉｎｅｐｉｃｅｎｔｒａｌａｒｅａｂｙｇｅｏｍａｇｎｅｔｉｃｒａｎｓｆｅｒｆｕｎｃｔｉｏｎｓＸｉａｏ－ＰｉｎｇＺＥＮＧ；（曾小苹）Ｙｕｎ－ＦａｎｇＬＩＮＩ；（林云芳）Ｚｈｏｎｇ－ＪｉｅＺＨ...     

应变能释放随时间变化的聚类方法研究

由于线性回归模型应用的广泛性，本文设定样品集的趋势是线性的，在用最优分割法对地震活动性类研究的基础上引进线性模型，扩大了最优分割法在地震活动性分类中的应用，将降类问题与建模问题有机地结合起来，去揭示地震活动性中的某种内在规律，这对于探索地震预报，研究地震发生的规律性有一定的实际意义。     

山地边界层中场变量的动力学研究

采用与实测较接近的二次函数来表达Ｅｋｍａｎ层中的湍流粘性系数Ｋ，在圆形气压场条件下，求得了山地上空边界层中的风速，进而求得散度、涡度和垂直速度等场变量随高度的分布。并作图分析了这些场变量的一些动力学特征。改进了以往在求解析解时，略去运动方程中湍流粘性力项中的关于高度的一阶导数项，以及取山坡面上风速为零作下边界条件等欠合理欠精确的做法。所求得的风速、散度、涡度和垂直速度均用简单的初等函数表示出来，有助于边界层参数化和深化对边界层动力学的认识。     

MONITORING OF 1991 EASTERN CHINA FLOODING WITH POLAR ORBITING METEOROLOGICAL SATELLITES FY-1 AND NOAA

With polar orbiting meteorological satellites FY-1 and NOAA,flooding was monitored in the areas of the HuaiheRiver basin and the Taihu Lake region during June and July 1991.All satellite images from FY-1 and NOAA for concerned areas before and during flooding were examined.Thoseof cloud-free,with small amount of cumulus or thin cirrus were selected to exam the situation.Navigation and projec-tion were carefully performed,to ensure the projected images at different time overlap accurately with each other in 1—2pixels.Channel 1 (CH1) and Channel 2 (CH2) data of FY-1 and NOAA satellites with wavelength of 0.58—0.68μm and0.725—1.1μm were used to monitor the flooding.Albedo of Channel 2 and normalized vegetation index (NDVI) wereadopted as indicators to identify water body from land.With histogram and man-machine interactive methods,analysiswas done.In cloud-free condition,the two indicators identified the same area and scope of the water body.Totally cloud-free image in a large area is quite rare.To understand flood process,it is necessary to use more fre-quent images.It was investigated to distinguish water from land in partly cloudy condition.The result showed that whenthere is small amount of cumulus or thin cirrus,satellite images are still valuable in monitoring water body.In case ofmonitoring area covered with cirrus,vegetation index is useful,and while there is small amount of cumulus on land,albedo of Channel 2 can be used.Ten images from May 16 to August 18 of 1991 were examined.The results show that in the Lixiahe area,JiangsuProvince,the area submerged in total was the largest;along main stream of the Huaihe River,the Chuhe River,andaround the Chaohu Lake,a large percentage of area submerged;while in the Taihu Lake area,less field submerged.Flood monitoring was performed for 87 counties in the region concerned.These counties were put in order accord-ing to the percentage of submerged area in total.This order showed the extent of disaster at one view point.