孤家子—后五家户气田地震反演与储层建模综合研究

地震反演和储层建模是储层横向预测的两种方法。地质统计学地震反演是一种基于模型的地震反演，采用地质统计学空间插值技术建立模型，将测井曲线或测井解释成果曲线演化为“假阻抗”来合成地震记录，寻找子波，再将地震道直接反演成储层属性。这种方法避免了波阻抗转换为储层属性所造成的误差。储层建模横向预测中，应用协克里金算法联合地震反演与原始测井两种数据，综合建立地质模型。在孤家子—后五家户气田实际应用中，选取具有代表性的R2.5梯度曲线进行反演，反演结果反映了砂体分布。结合各井点测井解释数据，进行了储层建模，提高了横向预测精度。     

金属加工厂附近农田土壤锌污染的地统计学分析

目的为评价金属加工对农田土壤环境造成的影响.方法采用经典统计学与地统计学相结合的方法,研究了沈阳市南郊某有色金属加工厂附近农田土壤全锌含量的空间分布特征.结果在1000m距离内,土壤全锌含量与样点离加工厂烟囱的距离之间符合二阶多项式关系,并与距离呈显著负相关关系.0~10cm及10~20cm土层土壤全锌含量分别达226.92mg.kg-1和214.24mg.kg-1,远高于该地区土壤锌含量的背景值.高斯模型与实验半方差变异函数的拟合效果最好,其次是线性模型.0~10cm及10~20cm土层土壤全锌含量具有十分相似的空间分布格局,离加工厂烟囱距离越远,土壤全锌含量越低.结论克里格插值得出的土壤全锌含量的空间分布图很好地反映了农田土壤受到金属加工厂污染的情况.     

老挝波罗芬高原铝土矿床资源评价的地质统计学模型

文章目的是研究一种适合于表生层状铝土矿床的自动化建模和资源估算方法,实现大型矿床尺度上的自动化建模和资源估算。在对老挝波罗芬高原铝土矿钻孔数据进行地质统计学分析的基础上,计算和拟合矿区的变异函数模型;在变异函数变程范围内对见矿工程的顶、底板进行数字网络化处理,建立矿区的三维模型;在三维模型范围内部建立矿块模型;对矿块进行资源类别的自动划分和估值,完成矿区资源量的分类和估算。该模型方法自动化程度高,特别适合于波罗芬高原铝土矿这类地表风化壳矿床的建模和资源量的分类估算。     

基于区域地层压力描述的东坪区块井身结构优化

东坪区块是青海油田近年来开发的重点区块,由于地质条件复杂和前期对该区块地层压力认识不清,因此井身结构设计存在缺陷,钻井过程中频繁出现漏、喷、塌、卡、断等各种井下复杂情况和事故。针对目前东坪区块地层压力认识不清的问题,首先收集区块内已钻井的测井等资料,利用Eaton法建立已钻井的单井地层压力剖面;然后基于地质统计学和克里金插值法,将地层压力以地层层组为单元外延展开,从而实现地层压力的区域化描述;最后根据区域地层压力的预测结果,对区块内待钻井进行井身结构优化。研究结果为青海油田东坪区块井身结构优化及钻井液密度精细化设计提供了依据,可以有效降低钻井复杂事故的发生频率,对于安全钻井具有重要意义。     

河套灌区地下水埋深时空变异特征及其影响因素

选择内蒙河套灌区义长灌域作为研究区,利用研究区67眼监测井的地下水埋深资料,采用传统统计学、地统计学和灰色关联分析法,对灌域1990-2015年地下水埋深的时空变异性及其主控因素进行了分析,结果表明:(1)义长灌域26a来潜水位下降近0.4m,平均下降速率1.6cm/a;(2)义长灌域26a地下水埋深的最优变异函数有高斯、指数和球状模型,以高斯模型为主,区域地下水埋深具有较强的空间相关性,变程多在5～7km;(3)空间插值结果表明,不同时期地下水埋深的空间分布存在一定的差异性,但总体上以1.5～2.5m范围为主,平均占比60%;(4)根据灰色关联分析法和偏相关分析,发现蒸发量和地下水补给量是义长灌域地下水埋深的主要驱动因素。     

Spatial and temporal distribution of areas with drainage problems as estimated by different interpolation techniques

This study used spatial interpolation methods and a geographic information system that examines spatial and temporal variations in areas with drainage problems. Monthly groundwater depths were measured from 2005 to 2009, and the lowest depths were used to create groundwater maps for each year using five different interpolation. Cross‐validation using root mean squared error (RMSE) and the correlation coefficient (r) as indices were used to compare the accuracy of each method. A comparison of interpolated and observed values indicated radial basis function (RBF) to be the optimal method for interpolating lowest groundwater depth in 2005, whereas simple kriging (SK) was found to be the optimal method for 2006, 2007, 2008 and 2009. With the exception of 2007, from 2005 to 2009, approximately 8% of the study area was determined to have a groundwater depth ranging from 0 to 1 m. Between 2005 and 2009, the percentage of the area found to have drainage problems that negatively affected plant growth varied from 60.83% to 86.53%.     

地下水环境氯离子时空变异性模拟研究

地下水所在的多孔介质空隙几何条件复杂多变,水文地质参数往往呈空间随机分布,致使地下水水流系统中参数具有时间和空间上的变异性和不确定性。本文着重阐述浅层地下水环境评价中的时空变异性,通过实验模拟浅层地下水非均质运移环境,利用时空地质统计学计算了区域化变量氯离子的时空变异函数,构建了乘积—和时空理论模型,并对时空估计插值方法的应用作了进一步探讨。结果表明, 从时间和空间上的变程可以得出浅层地下水水质指标的演变在3.9 h左右,空间上为0.5 m左右;时间边界上的变异性相对空间边界上的要好一些;时空变异性能更精确地反映和估计环境变量在时间和空间的分布,对浅层地下水污染物的迁移运转及预测研究具有重要的现实意义。     

Obtaining soil and land quality indicators using research chains and geostatistical methods

Soil and land quality indicators play an important role in the assessment and evaluation of soil and land quality. In contrast with the general definitions of soil and land quality, working with indicators demands a better awareness of at which scale level measurements were made, at which scale calculations and models were developed and validated, and at which scale answers are needed. We propose that soil and land quality indicators may be classified by three characteristics: 1) scale level, 2) complexity, and 3) transferability. Each characteristic is represented by an axis in the Soil and Land Quality Indicator Diagram.Indicators with a high complexity can not be measured directly, but need to be calculated with one or more models, eg. pedotransfer functions and hydrological simulation models. For the application of the indicator it is then important to know how the indicator value was obtained, i.e. which models were used. A specific sequence of models used for obtaining an indicator value is called a research chain and is indicated in the Scale Hierarchy and Knowledge Type Diagram. The use of research chains allows the user to consider and evaluate alternative options for the assessment of a specific indicator.In this study values for three soil quality indicators were obtained through two alternative research chains. The research chains differed by the choice of used pedotransfer functions and soil hydrological models. The two research chains yielded for each of the three indicators two sets of thirty year averages for 166 locations in the study area. Per location the obtained indicator values were compared with a t-test. The research chains were found to yield significantly different values for all three indicators.The spatial and temporal variability of the data was analyzed for each step, i.e. per model, along both research chains. Alternative models yielded different spatial and temporal variability structures. Therefore, the choice of research chain not only affects the mean value of an indicator, but also the associated spatial and temporal variability structure. Knowledge of the spatial and temporal variability is important for upscaling purposes.Based on these results we conclude that the successful application of soil and land quality indicators depends on:1) the definition of suitable indicators based on scale level, complexity, and transferability; 2) the careful selection and definition of research chains; and 3) the combined presentation of indicator values and used research chains.     

储层建模稳健变差函数的求取

根据油气储层垂向钻井的物性分布特点,针对奇异值、混合分布和漂移三个主要影响油气储层物性参数稳健变差函数求取的地质因素,选取剔除奇异值、限制特高邻差值、正态变换等方法消除奇异值影响,利用细致分相或截尾处理降低混合分布的影响,通过计算去除趋势后剩余值变差函数的方法处理漂移作用,并提出了油气储层物性参数稳健变差函数求取的工作流程。