跳到主要內容 :::

年度	104	專案性質	非實驗性質
專案類別	研究專案	研究主題	評估
申請機構	國立台灣大學	申請系所	生物環境系統工程學系
專案主持人	林裕彬	職等/職稱	教授
專案中文名稱	土壤污染潛勢篩選情境之決策工具研擬
中文關鍵字	土壤污染潛勢、重金屬、情境模擬、篩選決策、GIS
專案英文名稱	Development of decision support tools for screenin
英文關鍵字	GIS
執行金額	750,000元
執行期間	2014/12/1 至 2015/11/30
計畫中文摘要	土壤重金屬污染可經由多種途徑影響人體健康，為保護國家社會經濟安全，必須掌握土壤重金屬分布情形，土壤重金屬污染潛勢(Pollution Potential)圖成為污染潛勢區篩選，以及污染潛勢區篩選決策作業之重要參考資訊，然而潛勢圖大多以現有重金屬調查資料為基礎，進行潛勢之空間分布推估。但由於經費限制因素，調查資料無法涵蓋所有的區域，使未調查點之污染潛勢推估具有某程度之不確定性(Uncertainty)，進而提高推估污染潛勢區域篩選誤判之風險，並對污染區篩選及篩選決策具不同程度影響。此外，污染物之間常具有某程度之空間相關特性，若忽略此特性即各別推估污染物潛勢，此分析結果有可能遺失此污染物之空間特性。因此，研擬具科學性及可靠性(Reliable)潛勢推估工具，以提供穩健(Robust)污染潛勢範圍篩選之決策工具，為多年來土壤重金屬污染研究與實務之重要工作。 本計畫以全臺地區為研究範圍，並以農試所調查之130,772筆資料為基礎，有別於傳統克利金法及條件模擬法，本計畫結合地理統計模擬方法－高斯連續模擬與U-WEDGE法，共同模擬(Jointly simulate)研究區域內的各種土壤重金屬空間分布之可能情境(Scenarios or Realizations)，並將模擬之土壤重金屬空間分布間的空間相關性將與原始量測調資料進行比較，以確定模擬之土壤重金屬空間分布間的空間相關性與原始量測資料空間是否特性一致。依據模擬結果及情境，進行空間未確定分析(Uncertainty analysis)，且依據不確定性分析及模擬情境，應用系統性的分析工具－Information-Gap Decision Theory (IGDT)進行污染潛勢區篩選不確定性之決策分析，確保因資訊不足所產生的不確定情況下，亦能系統性地進行穩健決策分析(Robust decision analysis)，提供穩健決策，亦即當最劣的情況(worst scenario)發生時(現實情況土壤重金屬之空間分布與模擬之結果產生差異時)，所做的決策能夠恰能滿足管理者的需求；換言之，本分析方法將使上述之不確定性對染潛勢區篩選的影響較小，同時提供染潛勢區篩選不同情境。本計畫同時結合內梅羅指標與模擬情境，進行污染潛勢篩選，並將上述方法與系統性決策分析工具建置於地理資訊系統，並發展此系統之軟體及使用者友善操作介面，完成操作介面之使用說明，並舉辨一場方法及軟體說明會。
計畫英文摘要	Potential pollution maps of heavy metals in soil play important role in screening and making specific decisions on polluted area. The distributions of the heavy metals in those maps are generated based on the existing survey data. However, without comprehensive survey data, the uncertainty caused by the estimation of distributions of the heavy metals in unsampling sites that will lead to the increasing the risk of misjudgment on contamination area, and result in the errors in screening or decision making of pollution area. Besides, spatial correlations between different heavy metals should be considered. Estimating the distribution of those heavy metals indepenedtly and without considering the spatial correlations is therefore inappropriate and may not retain the spatial correlation between each variate. Thus, to find out scientific and reliable tools for estimating pollution potential area and to provide the robust decision-making tools for screening pollution potential area are important works in heavy metals in soil. The geostatistical approaches, such as sequential Gaussian simulation and Uniformly Weighted Exhaustive Diagonalisation with Gauss iterations (U-WEDGE), are applied in this project to jointly simulate the possible scenarios or realizations of distributions of heavy metals in soil over entire Taiwan based on the 130,772 survey data provided by the Taiwan Agricultural Research Institute, Council of Agriculture, Executive Yuan. Additionally, the estimating spatial distributions of soil heavy metals must be consistent with the spatial characteristic of original survey data. The spatial uncertainty analysis will be made according to the simulation results and scenarios. In accordance with the uncertainty analysis and the resultant simulation, a structured tool, Information-Gap Decision Theory (IGDT), is applied to make decisions on pollution potential area screening. The IGDT provides a general framework for robust decision analysis under the uncertainty caused by the insufficient information. That is this decision may still meet the needs of decision makers even in occurrence of the worst scenario, such as the significant difference between simulated and actual distributions of heavy metals in soil. In other words, this method will result in the lower impacts of uncertainties on the screening of pollution potential area and will provide the different screening scenarios. The Nemerow index and the simulation scenarios are also integrated into screening the possible polluted area. Additionally, a user-friendly GIS based tool to co-simulate the spatial distribution of heavy metals in soil, and make robust decision under sever uncertainty will be delivered. Finally, the orientation for those methods and the GIS based tool will be held.