Testing earthquake prediction methods: «The West Pacific short-term forecast of earthquakes with magnitude MwHRV ≥ 5.8»

Analyzing the tables and probability maps posted by Yan Y. Kagan and David D. Jackson in April 2002–September 2004 at http://scec.ess.ucla.edu/~ykagan/predictions_index.html and the catalog of earthquakes for the same period, the conclusion is drawn that the underlying method could be used for prediction of aftershocks, while it does not outscore random guessing when main shocks are considered.     

Contemporary vertical crustal deformation in Cascadia

Data from 29 tide gauges and 113 pairs of first and second order leveling lines are analyzed to determine the pattern of vertical deformation in the Pacific Northwest of the United States. The data span nearly 100 years and represent the interseismic elastic deformation related to the great earthquake cycle. Uplift rates calculated from leveling surveys are adjusted to a model surface in the tidal reference frame using a robust, weighted, linear, least square technique. Rapid uplift occurs in two distinct broad regions along the coast separated by a narrow zone of slow subsidence. Vertical deformation rates range from > 4 mm/year of uplift on southern Vancouver Island to > 2 mm/year of subsidence in northern coastal Oregon. The deformation pattern is consistent with the results of previous studies and subduction models.     

Subsurface dams to harvest rainwater— a case study of the Swarnamukhi River basin, Southern India

Declining water level trends and yields of wells, deterioration of groundwater quality and drying up of shallow wells are common in many parts of India. This is mainly attributed to the recurrence of drought years, over exploitation of groundwater, increase in the number of groundwater structures and explosion of population. In this subcontinent, the saving of water has to be done on the days it rains. India receives much of its rainfall in just 100 h in a year mostly during the monsoon period. If this water is not captured or stored, the rest of the year experiences a precarious situation manifest in water scarcity. The main objective behind the construction of subsurface dams in the Swarnamukhi River basin was to harvest the base flow infiltrating into sandy alluvium as waste to the sea and thereby to increase groundwater potential for meeting future water demands. An analysis of hydrographs of piezometers of four subsurface dams, monitored during October 2001–December 2002, reveals that there is an average rise of 1.44 m in post-monsoon and 1.80 m in the pre-monsoon period after the subsurface dams were constructed. Further, during the pre-monsoon month of June, much before construction of subsurface dams in October 2001, the water level was found fluctuating in the range of 3.1–10 m, in contrast to the fluctuation ranging from 0.4 to 3.1 m during the period following the construction of dams. Hence, the planning of rainwater harvesting structures entails thorough scientific investigations for identifying the most suitable locations for subsurface dams.     

大规模地形实时绘制的算法研究

讨论了地形模型的多分辨率表示机制、节点评价函数、动态简化算法、裂缝的解决方法、三角网快速构建等地形可视化中的关键问题。实验结果表明，本文提出的算法实时、高效，可以满足大规模地形可视化要求。     

地震属性分析中水平切片的应用

利用地震数据体中含有的丰富地质信息,可以有效地进行储层预测。层拉平水平切片地层学解释方法是先将地震剖面目的层段做层拉平处理,然后对切片顺地层层面追踪,勾绘出各种地层现象,并以地层学的观点加以解释。该方法在吉林探区扶新南部地区的应用中,对于识别河道的展布趋势取得了较好的效果,而且发现了多期叠置的扇型沉积体。另外,还可指导反演数据体的解释,做到对岩性体从定性到定量的描述。     

北京市区春季燃烧源大气颗粒物的污染水平和影响因素

以大气中PM2.5和PM10为研究对象,于2005-03-13—25共7天的时间内,在中国地质大学(北京)测试楼顶、首钢焦化厂和首钢东门设立3个采样点进行采样监测。结果表明:PM2.5和PM10质量浓度的日变化呈现一定规律性,在不同时段PM2.5和PM10的质量浓度不尽相同,且变化较大,在特定时刻出现峰值,主要受污染源排放和气象因素的控制;PM2.5和PM10质量浓度随气温的升高而降低,这与高温有利于颗粒物扩散、低温容易形成逆温层有关;在一定的相对湿度范围内(以大气中水汽不发生重力沉降为界限),PM2.5和PM10质量浓度与相对湿度呈正相关关系;而当发生降水时,由于水滴的冲刷和附带作用,PM2.5和PM10质量浓度降低;PM2.5和PM10质量浓度与风级呈明显的负相关关系。通过北京市与国内8个省会城市的PM2.5和PM10质量浓度的对比,发现北京市PM2.5和PM10污染比较严重,PM2.5和PM10质量浓度分别超过了1996年中国制定的PM10排放标准和1997年美国EPA制定的PM2.5排放标准。     

绕坝渗流地下水位的时空分布模型研究

许多大坝的失事是由于高地下水位引起坝肩失稳所致。绕坝渗流是影响坝肩高地下水位的主要因素。为此通常将大坝基础防渗帷幕延伸到坝肩岸坡内一定距离,以减小绕坝渗流影响。而防渗帷幕运行性态随时间变化,为了评价坝肩防渗帷幕和地下水位的运行性态,首先分析了地下水位观测资料和水位、降水、温度、时效等时空影响因素及其表达式,随后基于岸坡地下水位观测资料,利用最小二乘法建立了大坝岸坡地下水位的时空分布模型。通过比较模型剩余标准差和测点的剩余标准差,可以确定坝肩地下水位的异常测点,分析岸坡防渗薄弱部位,掌握坝肩岸坡渗流场时空分布规律,监控绕坝渗流的性态。     

MRI分割的几何活动轮廓模型的半隐式AOS迭代方法

几何活动轮廓模型以其计算稳定等优点，被广泛用于图像分割。给出一种基于AOS(additive operator splitting)算法的快速活动轮廓模型，将AOS方法与几何活动轮廓模型相结合构造一种半隐式迭代模型，它对迭代步长没有限制，从而可以选择较大的迭代步长，提高了模型的时间效率，并使用区域信息构造系数矩阵，使模型具有全局性。对MR(magnetic resonance)图像分割的实验表明，该模型可以得到较好的分割结果，同时时间效率也有大幅提高，方便了实时应用。     

基于小波统计特性的遥感图像像素与特征联合最优融合方法

遥感影像的IHS融合方法由于匹配误差导致光谱畸变和退化,而小波变换在变换域具有良好的分频特性,小波系数的统计特性反映了遥感影像的边缘、线和区域等显著特征。提出了基于小波统计特性的遥感影像的像素和特征联合最优融合方法,在IHS空间,对强度分量I的高频部分利用多分辨率小波融合方法进行影像的高频细节特征融合,低频部分选取光谱信息和空间分辨率评价指标作为融合权系数求优指标,进行像素级最优融合,实验结果证明了该方法的有效性。