灰色理论在干旱预测中的应用

介绍灰色系统理论及其建模原理,利用沁河流域上游沁源气象站42年的实测降水量资料建立灰色预测GM(1,1)灾变模型,对干旱灾害进行预测,经残差、后验差检验等分析,模型精度较高平均达98.8%.并对实测资料进行检验,效果较理想,为沁河流域上游抗旱及供水提供必要的预测信息.     

河流集合预报方法在洮河流域径流预报中的应用研究

为研究洮河流域径流预报,根据河流集合预报方法,利用红旗水文站实时监测资料,基于中国洪水预报系统开发的三水源蓄满产流、滞后演算和马斯京根模块构建全流域洪水预报模型,并应用该模型分析预测红旗水文站径流及其概率分布统计。结果表明:径流拟合过程拟合较好,取得了较好的预报效果,模型满足河流集合预报应用的条件,结论可为洮河流域水资源开发利用、防汛保安提供参考依据。     

大汶河流域地下水位动态预测

对大汶河流域地下水位动态变化进行概述,利用77眼测井实测资料,分析了大汶河流域六年的水位变化情况。对观测井1978～2003年的地下水位年变差系列值和相应的年降水量进行一元回归分析,建立一元线性回归方程式,进行了年降水量影响预测;对东平县大清河以南的地区进行了典型区回归预测;对典型区建立了数学模型,利用三维地下水有限差分计算软件MODFLOW进行地下水数值模拟,对实测与模拟水位过程线进行对比,进行了典型区模型预测。结果表明：率定参数正确、所建模型可靠,能基本反映该地区的地下水运动规律。     

基于主成分分析和贝叶斯正则化方法的神经网络年最大洪峰流量预测模型探讨

针时水文预测建模中输入因子过多而导致神经网络结构规模过大,泛化能力差的问题,利用主成分分析和贝叶斯正则化方法对神经网络进行改进,优化网络结构,从而提高泛化能力。以洮儿河流域镇西站年最大洪峰流量预测为例,研究结果表明,改进的神经网络预测方法与传统的神经网络方法相比,泛化能力有显著提高,而且网络的收敛也比较稳定,实际预测中效果良好。     

基于SRTM的SNOW—TOPMODEL模型在玛纳斯河流域的应用

基于分辨率90m网格的玛河流域SRTM数据,提取了玛纳斯河流域的河网和地形指数,生成的河网与实际河网基本接近。将加入融雪模块TOPMODE应用于玛河流域,对模型计算结果进行误差评价并分析误差产生的原因。由玛纳斯河流域应用加入融雪模块的TOPMODEL模型可以看出,该模型在玛纳斯河流域有较好的适应性,可以将此模型应用于玛纳斯河流域的洪水预报。     

VIC陆面水文模型在白莲河流域径流模拟中的应用

本文从气象与水文水资源学科交叉的角度引进VIC陆面水文模型,将其应用于湖北省白莲河流域,探讨模型在中小流域降雨径流模拟中的适用性。本研究利用白莲河流域DEM、植被、土壤等数字化资料,经过模型预处理,建立了白莲河流域VIC模型框架。并进行了模型参数敏感性分析。模拟结果表明,模型基本能够反映该流域的日径流水文过程,具有一定适用性。     

基于神经网络的采空塌陷预测

依据某煤炭开采区的勘察资料, 综合考虑影响采空塌陷的主要因素, 建立了预测采空塌陷的 BP神经网络模型。该模型结构为 7-10-2型。优化学习参数后, 用该模型对采空区塌陷进行了预测分析, 结果与实际情况完全吻合, 表明 BP神经网络模型应用于采空塌陷预测领域是行之有效的。      

流域产流产沙BP网络预报模型的初步研究

运用BP神经网络模型的基本原理,以流域降水条件为基本因子,建立了流域产流产沙BP网络预报模型.该模型能用于定量分析流域人类活动因素对流域产流产沙的影响.西汉水、大通江、香溪河流域资料验证表明,模型基本合理、可靠.     

济南市岩溶水动态变化的神经网络模拟及泉水喷涌趋势预测

本文根据影响岩溶水位的因素,利用改进的BP神经网络建立一个能够反映济南市泉域岩溶地下水动态变化的随机模型,并进行检验,与运用多元回归模型预测的结果相比较,结果表明:BP神经网络模型进行岩溶地下水动态变化预测是可行的,该模型具有较强的学习、容错和联想功能,对岩溶地下水动态变化的预测精度大大的提高。最后在模拟的基础上,又对泉水喷涌的宏观趋势作了进一步的预测和分析。     

人工神经网络在岩溶塌陷预测中的应用研究

本文利用已有资料建立了岩溶塌陷预测ＢＰ神经网络模型,并对桂林岩区进行了岩溶塌陷预测。结果表明,在岩溶塌陷预测中应用ＢＰ神经网络模型是行之有效的。     

回归分析在确定多金属硫化物矿石密度中的应用

在研究多金属硫化物矿床的特点,分析以往矿床勘查中确定矿石密度值难点的基础上,通过建立矿石密度值与矿石各组分的多元线性回归数学模型,预测矿床的矿石密度值。     

Summer monsoon onset over Kerala: New definition and prediction

The summer monsoon onset over Kerala (MOK) marks the beginning of the rainy season for the country. Associated with the MOK, significant transitions of large scale atmospheric and oceanic circulation patterns are observed over the Asia-Pacific region. In this study, a new method for the objective identification of MOK, based on large scale circulation features and rainfall over Kerala, is discussed. Further, a set of empirical models based on the principal component regression (PCR) technique was developed for the prediction of the date of MOK by keeping in mind the IMD’s operational forecasting service requirements. Predictors for the models were derived using correlation analysis from the thermal, convective and circulation patterns. Only five predictors pertaining to the second half of April were used in the first model (Model-1) so that the prediction of MOK can be prepared by the end of April itself. The second model (Model-2) used four additional predictors pertaining up to the first half of May along with two predictors used in the Model-1 for update prediction at the end of the first half of May. To develop each of the PCR models, Principal Components Analysis (PCA) of the respective predictor data was carried out followed by regression analysis of first two principal components (PCs) with the date of MOK. Both these models showed good skill in predicting the date of MOK during the independent test period of 1997–2007. The root mean square error (RMSE) of the predictions from both the models during the independent test period was about four days which was nearly half the RMSE of the predictions based on climatology.     

西北区冰雹日气候分析及预测方法研究

利用1970 —2001 年1 ～12 月中国西北区冰雹日数的气象实测资料，将年冰雹日数发生的变化趋势看作每年的一个气候现象，分析其发生的气候变化趋势、年代际变化以及与大气环流的相关关系，得出多雹年北半球500 h Pa 极涡强、位置偏东，副热带高压面积小、强度弱、位置偏南，乌拉尔山至巴尔喀什湖高度场为正距平，雅库次克至贝加尔湖高度场为负距平。少雹年北半球500 h Pa 极涡弱、位置偏西，副热带高压面积大、强度强、位置偏北，乌拉尔山至巴尔喀什湖高度场为负距平，雅库次克至贝加尔湖高度场为正距平区。从而找出冰雹气候趋势预测的着眼点，建立冰雹短期气候预测概念模型；此模型对冰雹发生变化趋势的预测具有一定的实用价值。     

水文相关中的最小二乘回归问题探讨

在研究了水文相关中利用最小二乘原理进行回归分析方法的基础上，指出了传统水文回归分析方法的使用条件、局限性和存在的问题。提出了不同性质的变量在回归分析时应区别对待的概念。结合水文观测的实际情况，研究了相关变量观测值均存在误差情况下，以变量的观测误差作权进行回归分析的概念，提出了建立基于最小二乘原理基础上的双加权回归模型，并研究了该模型的解算方法。同时推导了估计参数和预测值的误差估计公式，通过估计参数和模型误差的研究，有助于提高水文变量回归的精度和误差估计水平。     

灰色理论在新乡百泉泉水流量动态分析中的应用

新乡百泉具有供水、农灌、人文、旅游及生态等多种功能,研究其泉水流量动态,建立泉水流量动态预测模型,对泉域水资源评价和泉水资源保护具有重要意义。为了进一步研究新乡百泉天然状态下泉水流量动态特征、评价泉域岩溶水资源,基于1964—1978年泉水年均实测流量和泉域年均降水量资料,通过逐步回归分析,确定泉水流量的主要影响因素为前1年降水量,并建立了逐步回归模型,其回归效果显著;在逐步回归分析的基础上建立了泉水流量动态预测的GM(1,2)模型、NSGM(1, 2)模型和GM(0, 2)模型。结果表明：1964—1978年百泉泉水流量动态主要受泉域降水控制,且泉水流量滞后降水1年,反映了天然状态下泉水动态特征。3种灰色模型的精度等级均为最高级（优）。1964—1978年百泉泉水实测流量为2.347～6.448 m3/s,平均为3.904 m3/s;逐步回归模型预测值为1.882～6.383 m3/s,平均为3.904 m3/s;GM(1,2)模型预测值为2.327～6.448 m3/s,平均为3.939 m3/s;NSGM(1,2)模型预测值为2.133～6.448 m3/s,平均为3.927 m...     

随机森林模型预测岩溶区酸性煤矿井水锰污染

酸性煤矿井水严重威胁地下水的水质。如何更有效对受影响区域的地下水源进行动态监测是当前的一个关键问题。采用随机森林中的回归模型,利用自变量(采空区水位、岩溶水位、pH值、泉水流量、电导率)和因变量(污染离子浓度)的相关性,建立回归模型;使用测试数据进行误差分析,结果证明模型准度较高,所得预测值具有参考价值;得出各自变量对因变量影响的重要程度,分析结果与实际情况相符合。试验表明,随机森林回归模型在酸性煤矿井水污染预测方面具有适用性,可作为辅助手段监测水质污染情况,对今后工作有一定的指导意义和经济价值。     

Yearly variations in runoff as a function of precipitation variations for the East-European plain

Variations in climatic conditions for the East-European Plain (EEP) are analysed. The observations show that the time periods of high-water and low-water correspond to certain circulation conditions, which points to a conceptual possibility to construct a model correlating streamflow with atmospheric precipitation. Described is a linear model based on a regression equation for the principal components of runoff and precipitation. The model presents the best results for a territory where a strong reaction of the river water to a certain circulation type has been established beforehand.     

Prediction by regression and intrarange data scatter in surface-process studies

Modeling is a major component of contemporary earth science, and regression analysis occupies a central position in the parameterization, calibration, and validation of geomorphic and hydrologic models. Although this methodology can be used in many ways, we are primarily concerned with the prediction of values for one variable from another variable. Examination of the literature reveals considerable inconsistency in the presentation of the results of regression analysis and the occurrence of patterns in the scatter of data points about the regression line. Both circumstances confound utilization and evaluation of the models. Statisticians are well aware of various problems associated with the use of regression analysis and offer improved practices; often, however, their guidelines are not followed. After a review of the aforementioned circumstances and until standard criteria for model evaluation become established, we recommend, as a minimum, inclusion of scatter diagrams, the standard error of the estimate, and sample size in reporting the results of regression analyses for most surface-process studies.     

The impact of neighborhood services on land values: an estimation through the hedonic pricing model

A strong need exists to increase the knowledge and recognize the values of neighborhood services. This paper makes an attempt to examine the impact of neighborhood services on land values, through analysis based on results obtained from multi linear regression analysis. This case study was done in area of dense urban settlement to check the consequence of neighborhood services through the hedonic pricing model base variables which include the structural, locational, community and neighborhood services. The regression coefficient was checked at (p?<?0.05) level of significance for each variable. The primary data was collect through the questionnaire filling by random sampling and Punjab Urban Gazette was used for the verification of land values in study area. The results found a valuable potential on land and property values of neighborhood services through the regression analysis results express through R (0.926), R² (0.856) and coefficient tables to represent the effect of each individual variable on property and land values. Every individual variable play role in defining the values of land and property based upon its utilization such as larger the structure of a property, more the covered area, larger size of a property more the land area higher price. Similarly, for locational factor study in geography, a piece of land adjacent to the road have high land values 50–70% compare to those located at distance. Very less attention has been paid on such issues in the world due to lack of effective urban planning and research on such crucial issues. If empirical studies on such issue has been done, effective way can be obtained for urban planning.

     

地质数据回归分析的自变量迭代变换

回归分析是地质数据处理中应用最广的一种统计分析方法,尤其是其中的线性回归模型应用最广泛。模型中的变量,包括因变量和自变量,均被看作是线性的,但实际计算时,这些变量却可以看成是原始观测值的一种非线性的变换。