SAMPLING DESIGN FOR CLASSIFYING CONTAMINANT LEVEL USING ANNEALING SEARCH ALGORITHMS
SAMPLING DESIGN FOR CLASSIFYING CONTAMINANT LEVEL USING ANNEALING SEARCH ALGORITHMS
复制标题
DOI：
10.1029/93wr02301
复制
发表时间：
1993-12-01
影响因子：
5.4
通讯作者：
KILLAM, BR
中科院分区：
文献类型：
--
作者：
CHRISTAKOS, G;KILLAM, BR
提出了一种对空间分布的污染物水平进行随机采样的方法。采样的目的是将污染区域划分为污染物浓度高低区。特别是，给定一组对场地内污染物的初步观察，希望找到一组额外的采样地点，以一种考虑到场地的空间可变性特征并优化特定场地清理过程中从物理、监管和货币考虑中产生的某些目标功能的方式。由于兴趣在于将区域划分为污染物阈值水平以上和以下的区域，因此一个自然标准是错误分类的代价。得到的目标函数是与采样相关的空间损失函数的期望值。随机期望涉及污染物水平及其估计的联合概率分布，其中后者是通过空间估计技术计算的。实际计算需要对污染域进行离散化处理。因此，任何合理规模的问题都会导致组合学排除穷举搜索。使用退火算法，虽然不是最优的，但可以快速有效地找到一组良好的未来采样位置。为了深入了解该方法的参数和计算要求，对一个算例进行了详细的讨论。空间采样设计在实践中的实施将为废物场地修复、地下水管理和环境决策提供必要的模型输入。
A stochastic method for sampling spatially distributed contaminant level is presented. The purpose of sampling is to partition the contaminated region into zones of high and low pollutant concentration levels. In particular, given an initial set of observations of a contaminant within a site, it is desired to find a set of additional sampling locations in a way that takes into consideration the spatial variability characteristics of the site and optimizes certain objective functions emerging from the physical, regulatory and monetary considerations of the specific site cleanup process. Since the interest is in classifying the domain into zones above and below a pollutant threshold level, a natural criterion is the cost of misclassification. The resulting objective function is the expected value of a spatial loss function associated with sampling. Stochastic expectation involves the joint probability distribution of the pollutant level and its estimate, where the latter is calculated by means of spatial estimation techniques. Actual computation requires the discretization of the contaminated domain. As a consequence, any reasonably sized problem results in combinatorics precluding an exhaustive search. The use of an annealing algorithm, although suboptimal, can find a good set of future sampling locations quickly and efficiently. In order to obtain insight about the parameters and the computational requirements of the method, an example is discussed in detail. The implementation of spatial sampling design in practice will provide the model inputs necessary for waste site remediation, groundwater management, and environmental decision making.