土壤污染調查之採樣策略研究

A Study on Sampling Strategy for Contaminated Soil Investigation

陳志豪(Chih-Hao Chen)

指導教授 ： 張尊國

國立臺灣大學/生物資源暨農學院/生物環境系統工程學研究所/碩士(2005年)

https://doi.org/10.6342/NTU.2005.01389

摘要

本文以不同的角度切入三個在土壤污染調查時常遇見的問題(1) 可否找到高濃度分佈的地點，以及其命中率；(2) 場址受到污染的程度；(3) 高濃度區分佈之範圍。分別進行採樣策略合宜性的探討。
在搜尋高污染區方面，應用空間分析的方式，進行採樣配置的高污染區命中率分析，並比較空間模擬退火佈點採樣策略與網格式佈點採樣策略在搜尋污染場址中高污染區之能力。本研究的結果指出空間模擬退火佈點法，在搜尋污染場址之高污染區方面確實比網格式佈點法更優良，在相同的條件下，相同的採樣密度和同一污染場址，模擬退火佈點法比網格式佈點法有較低的錯失率。
在判斷場址受到污染程度的部份，統計理論提供對場址資訊量化的科學性證據，可作為進行採樣調查時控制成本與效益的實用工具，研究中利用控制可容忍誤差範圍方式進行階段性採樣合宜性之探討，結果顯示，在規劃合理之信賴度及容許的誤差範圍下，評估調查土壤污染調查所需之樣本數量，能讓調查結果明確的代表母體的特性。
在探討高濃度區分佈範圍方面，於研究中分別使用克利金法與序列指標模擬進行污染範圍的推測，並透過實例分析兩種方法的特性。結果顯示，利用克利金法推測污染範圍能有效的輔助判斷性採樣之設計，而模擬法則能提供使決策者更能掌握空間變數在未採樣位置的變化情形，也提供設計採樣時更多元的資訊。

關鍵字

土壤污染 ； 採樣策略 ； 空間模擬退火法 ； 地理統計

並列摘要

Three typical objectives of a sampling design for contaminated soil investigation are: (1) to identify the location of “hot spots”, (2) to estimate contamination levels, (3) to delineate the pollution patterns and range of contaminated sites. This research addresses the issue of suitability about the sampling strategy for above objectives.
In first case, objective of sampling is to determine where "hot spots" are present. Spatial analysis method is presented to assess the capability of searching for hot spots of sampling strategy, includes Spatial Simulated Annealing (SSA) and grid sampling strategy. The results indicate the Spatial Simulated Annealing is better than grid sampling design when searching for hot spots. Grid sampling design has the higher miss rate than Spatial Simulated Annealing when deal with the same pollution site with same sampling density.
When sampling strategy were aimed at estimating average concentrations or total amounts of pollutants in environmental media, Statistics theory provides a basis for balancing decision uncertainty with available resources. This section discusses optimize the design for obtaining data by specify limits on decision errors stage by stage. The results reflect that assessing sample size needed for investigation by well-planned confidence could make the outcome achieve sampling objective with required performance.
To delineate the pollution patterns and range of contaminated sites, two approaches are presented: ordinary kriging (OK) and sequential indicator simulation (SIS) approach, and illustrated the characteristics by case studies. Mapping the values of continuous soil attributes by kriging perform well to assist judgmental sampling, while the simulation-based approach provide assessing the uncertainty about the value of soil properties at unsampled locations, and thus incorporate this assessment in subsequent decision-making processes.

並列關鍵字

Soil contamination ； Sampling Strategy ； Spatial Simulated Annealing ； Geostatistics

參考文獻

1. 行政院環境保護署, 2002。「農地土壤重金屬調查與場址列管計畫」。

45. Singer, D. A., 1975, “Relative efficiencies of square and triangular grids in the search for elliptically shaped resource targets,” Journal of Research of the U.S. Geological Survey, Vol. 3, No. 2, pp. 163- 167.

13. Aarts, E., Korst, J., 1990, Simulated Annealing and Boltzmann Machines — A Stochastic Approach to Combinatorial Optimization and Neural Computing, Wiley, New York.

15. Brus, D. J., De Gruijter, J. J., 1997, “Random sampling or geostatistical modeling? Choosing between design-based and model-based sampling strategies for soil (with discussion),” Geoderma, Vol. 80, pp. 1-44.

16. Burgess, T. M., Webster, R., 1980, “Optimal interpolation and isarithmic mapping of soil properties,” Journal of Soil Science, Vol. 31, pp.315-331.