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%.     

变异函数在大厂锡矿田92号矿体中的应用

变异函数是地质统计学的核心内容和基本工具,它能够较好地研究一个矿床的区域化变量的基本特征,特别是能够反映变量的随机性和结构性的变化。本文在对大厂某矿体地质特征进行研究后,计算了该矿体的变异函数,并且对计算结果进行了分析。结果表明,利用变异函数建立矿床数学模型进行矿体储量估算是一种切合实际的方法。     

以地质统计学为基础的矿业软件在中国的历史和现状

本文回顾了以地质统计学为基础的矿业软件在中国的发展历史，评价了软件开发和应用的现状，指出了应用过程中存在的误区和解决途径，并以实例说明了软件应用带来的成效。总结应用经验，文章还指出了软件应用过程中几个应该引起重视的重要参数和问题，以供参考。     

基于地统计学定阶的松毛虫发生面积组合预测*

针对当前松毛虫滞后阶数确定方法存在局部最优、耗时长等问题,提出一种基于地统计学(GS)快速定阶的松毛虫发生面积组合预测模型(GS-ARIMA-SVM)。首先采用差分自回归移动平均(ARIMA)对松毛虫发生面积进行线性建模预测,然后采用GS对松毛虫发生面积非线性部分进行快速定阶和样本重构,最后采用支持向量机(SVM)对非线性部分进行建模预测,从而获得组合模型预测值。并对辽宁省朝阳市松毛虫发生面积数据进行了仿真实验。仿真结果表明,GS-ARIMA-SVM预测精度明显优于参比模型,更能反映松毛虫发生的复杂动态变化规律。     

基于三维软件资源储量估算对比研究

文中通过实际项目数据基于三维软件中的资源储量估算模块,结合传统方法和地质统计学方法进行估算对比,得到资源储量报告,结果说明不同的估算方法在合适的参数控制下,其估算结果基本一致,误差范围完全满足对资源量储量估算的要求。同时也发现不同类型、不同工程控制程度的矿床,估算适用方法也不同,需要注意对不同的情况进行分析,结合不同的估算方法进行验证对比,选择比较适合的估算方法,减少因估算方法的不同带来的误差。     

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.     

基于地统计学的土壤肥力空间变异分析——以天津市武清区为例

以天津市武清区为例,通过选取土壤pH值、有机质、全氮、速效磷、速效钾为土壤肥力指标,运用综合指数法对土壤肥力状况进行评价;并将地统计学与GIS相结合,进一步分析土壤各肥力因子及总体肥力质量的空间变异规律.结果表明：所选的5种肥力指标中,pH值具有弱空间相关性,有机质、全氮、速效磷、速效钾均表现中等强度的空间相关性;运用Kriging空间插值发现pH值较高的部分处于研究区中部,而较低的部分处于北部,有机质含量由北向南逐渐增大,全氮、速效钾均呈同心圆模式由内向外逐渐降低,其核心点为黄庄街道,速效磷含量由北向南逐渐降低;整个武清区土壤肥力质量呈现北高南低的趋势,以中等质量土壤为主.该评价结果能有效为规范农户种植、施肥行为和制定相关农业规划提供科学依据.     

Statistical analysis of the porous microstructure as a method for estimating reservoir permeability

Back-scatter scanning electron microscope images of cross-sections of several porous rocks were analyzed to determine the statistical properties of the porous microstructure. For statistically homogeneous media these properties are the porosity and autocorrelation function. A length scale (integral correlation scale), characteristic of the spatial distribution of porosity, was obtained as the integral of the autocorrelation function. The permeability of a wide variety of rock samples, including those investigated by Coskun and Wardlaw (1993), was adequately described by an empirical equation of the form k α φ^aI_S^b, where φ is the porosity and I_S is the integral correlation scale. The results obtained have useful application in the estimation of reservoir permeability from samples not amenable to experimental testing (e.g., drill cuttings) and provide support for the use of statistical methods for the generation of 3-D model porous media.     

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.