四川旱涝震灾害的人工神经网络外推预测     

李祚泳  邓新民  黄志英 《成都信息工程学院学报》1997,(2)

选取旱、涝、震灾害之间的一步转移概率作为Ｂ－Ｐ人工神经网络训练样本的输入信息，建立了四川旱、涝、震的人工神经网络灾型预测模型。Ｂ－Ｐ网络模型应用于实例预测结果与用主分量分析法的趋势预测结果精度接近。  相似文献   

Thermal conductivity modelling of terrestrial soil media—A comparative study     

B. Usowicz  J. Lipiec  A. Ferrero 《Planetary and Space Science》2006,54(11):1086-1095

The paper describes prediction of thermal conductivity in terrestrial soil media. The model operates statistically by probability of occurrence for contacts between particular fractional compounds. It combines physical properties, specific to particular compounds, into one apparent conductance specific to the mixture. The concept of substituting grain compounds by hypothetical spheres is an essential tool to control porosity by the number of spheres, their radii and probability of contacts between them. The spheres are equal in radii. The spheres substitute compounds, regardless of the phase state. Control of particular phase states is possible by means of specific properties assigned to the spheres, at the input to the model. Performance of the model is successfully proved for many diverse terrestrial soil media in a wide range of bulk density, composition, water and water vapour content. Only the compounds of sand decline from the expected values and require introducing a correction to the thermal conductivity of sand grains. One possible explanation is that the thermal conductivity of sand is uncertain. Nevertheless, the model is useful and worth extending beyond terrestrial purposes.  相似文献   

�ֽ�ؿ��˶�����ѧ����״̬������Ԥ�����о�   总被引：2，自引：2，他引：0  

周硕愚  吴云  施顺英  姚运生 《大地测量与地球动力学》2007,27(4):92-99

??????????????????????????о????40????????????????????α估GPS????????????????????????????????????????????????????????????????????????????????????????￡????????????????????????????????????????????????????????????????????????????????????????????????????????????顢???塢???????????,????????????????????????????????????????????????????1??????????????????????????????????????????????????????????????????????鵽????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????SOC??????????????????????????????????????????????????????????????????????е?????????  相似文献   

地震发生的极限时间   总被引：7，自引：3，他引：4  

刘正荣 《地震研究》2002,25(2):95-99

指出了成组出现的地震一般所遵从的规律，并以实例形式计算了中国历史上几次巨大地震，1996－1976年间各大地震间所满足的关系，并以越南来州地震为例说明了以较小地震报大震的方法。  相似文献   

半Lagrange法在数值预报中的应用和近年来的发展(Ⅰ)     

廖洞贤 《成都信息工程学院学报》1989,(2)

本文就五十年代以来半Lagrange法在数值预报中的应用和近年来的发展进行评述,介绍我国在这方面的工作。还就此法和Euler法的优缺点作了比较;指出了存在的问题。  相似文献   

Ab initio prediction of 29Si-NMR chemical shifts     

Chu Shidong  Li Yingxia  Song Ni  Guan Huashi 《中国海洋大学学报(英文版)》2002,1(1):63-65

The ability of several ab initio models to predict experimental ²⁹Si-NMR chemical shift is examined. The shielding values of trimethylsilyl chloride (A), t-butyldimethylsilyl chloride (B) and allytrimethylsilane (C) are calculated by GIAO, CSGT and IGAIM methods, using HF/6–31G^*, B3LYP/6-31G^*, HF/6–311+G^**, B3LYP/6–311+G^** and MPW1PW91/6–311+G^** models respectively. The ²⁹Si chemical shifts calculated by GIAO method using HF/6–311+G^** model are highly in agreement with those obtained experimentally. All of the models above reproduce the trends of chemical shifts in all cases studied, suggesting that the models are of practical value.  相似文献   

用自激励门限自回归分析方法模拟和预报太阳黑子数的长期变化     

张桂清 《中国天文和天体物理学报》1988,(2)

本文用一种新方法——自激励门限自回归分析方法对太阳黑子相对数年平均值进行拟合和预报检验,并对未来第22周逐年年均值作出预报。 目激励门限自回归分析模型的形式如下: 在对1956至1985年逐年太阳黑子相对数年均值的预报检验中,最大拟会误差为40.6,最小拟合误差为0.3,平均拟合误差为±12.5。 对1986至1997年逐年太阳黑子相对数年均值的预报见表(4)。定出第21周极小在1986年或1987年,极大在1990或1991年,极大值R_M=81.2±16.2。  相似文献   

“00.7”北京特大暴雨模拟中气象资料同化作用的评估   总被引：18，自引：0，他引：18  

张朝林  Kuo Ying Hw  范水勇  仲跻芹 《气象学报》2005,63(6):922-932

针对2000年7月4～5日北京地区的一次特大暴雨过程(24 h降水量达240 mm),文中利用MM5/WRF三维变分系统和MM5非静力模式,对此次特大暴雨过程中的各种气象监测资料(地基GPS大气柱水汽含量、常规探空、高空测风、地面常规观测和地面自动气象站)的同化作用通过观测系统数值试验进行了评估.结果表明与传统的客观分析方案相比较,MM5/WRF三维变分同化系统可直接引入非常规地基GPS大气柱水汽含量监测资料,提供更好的大气初始分析场.在三维变分同化方案下,各种大气监测资料均对改进此次特大暴雨模拟有不同程度的贡献,其中,常规探空和高空测风监测资料对改进预报结果的影响最大,地面常规观测和地面自动气象站观测资料作用次之,地基GPS大气柱水汽含量资料在与其他大气监测资料相互优势互补后,可很好地改善模式大气的分析质量,通过三维变分同化技术在区域数值天气预报模式初始场中引入地基GPS大气水汽监测网资料,使此次强降水个例的6 h和24 h测站降水预报的TS评分值在1,5,10和20 mm预报检验阈值下分别提高了1%～8%.研究结果对利用三维变分数值系统,评估气象监测网资料在改进高影响天气事件预报中的作用有借鉴意义.  相似文献   

Hydrogeological and Gasgeochemical Earthquake Precursors – A Review for Application     

Jens?HartmannEmail author  Jason?K.?Levy 《Natural Hazards》2005,34(3):279-304

Even if earthquake precursory signals can be identified, how can they be useful? This paper investigates relationships among the attributes of 229 proposed earthquake related gasgeochemical and hydrogeological precursory signals, and applies these results to improve future earthquake prediction strategies. Sub-groups of these reported signals and relationships between sub-groups are established using parameters, including earthquake magnitude, signal duration, precursory time, and epicentral distance to the monitoring site (original studies are used wherever possible to improve data quality). A strong correlation (r=0.86) between signal duration and precursory time was identified. This suggests a relationship between the investigated precursory signals and tectonic processes related to the referenced earthquakes. Moreover, these signals are categorized into four groups, reflecting differences in monitoring station densities, measurement methods and physical processes related to signal occurrence: (a) radon exhalation from the earth’s crust, (b) exhalation of other gases (helium, argon and others), (c) temporal variation in water level or discharge of springs and (d) temporal variation in temperature and dissolved ions in the water of the monitoring sites. In addition, boundary functions are used to separate signal group subsets. Finally, it is shown how these boundary functions can be used in the context of an earthquake prediction strategy by identifying potential minimum magnitudes and maximum epicentral distances from the monitoring site.  相似文献   

青海五龙沟地区1:5万遥感信息找矿勘查与靶区预测研究   总被引：1，自引：0，他引：1  

高香香  邹林  高光明  杨自安  张普斌 《矿产与地质》2005,19(4):418-421

在青海五龙沟地区开展15万遥感信息找矿勘查与靶区预测研究,是以矿化蚀变的遥感异常提取为主线,结合成矿源场、成矿位场遥感信息的解译,通过遥感与地质、矿产、化探等多源信息的空间集成和综合分析,进行15万遥感岩石地层-构造-蚀变岩填编图与遥感找矿预测,优选出了13处新的找矿靶区.经有关地勘单位野外实地检查,取得了很好的应用效果,并证明了该方法的有效性.  相似文献