| 摘要: | 自工業化時代以來，各式工廠林立於城鎮周遭，雖然帶動當代經濟發展與創造地區就業機會，但許多工業在生產過程中，同時也耗盡環境資源而形成大量廢水、廢氣與固體排放物，不僅嚴重危害人體健康，更造成環境不可逆的現象。然而，隨著科技蓬勃發 展，各項產業結構正逐漸面臨轉型，因此許多工廠受大環境淘汰或搬遷，而對於遺留下來的高污染土地，即稱之為棕地（Brown field）。台灣推動工業化政策，由農業社會轉為工業社會，亦嚴重存在著棕地現象。而為解決此議題，本研究從近年國外諸多研究中，發現巨量資料（Big Data）之概念已被廣泛運用於各項領域，其資料重複性使用可由大量看似雜亂的數據中推測出潛在相關性，不僅有助於瞭解這個世界，更能以此改善決策方式達到重塑社會的走向。因此，為有效推動污染整治工作，若能透過巨量資料的應用，使之在整治前即充分運用污染場址過去的工業活動進行評估，且借以預測出不同場址未來各項污染趨勢，再以此污染趨勢為評估基準，進行一系列污染整治優先順序評定，勢必能使基金運用更有效率，且有助於加速與擴大整治工作的進行，並能有效解決土壤及地下水污染影響議題。 據此，為處理上述巨量資料（Big Data）所帶來的龐大數據，本研究將整合前兩年度研究成果，作為輸入巨量資料部分基礎數據，再加以彙整污染場址過去的工業活動，以分析未來污染趨勢，並以國外優先補助區域(Priority Funding Area, PFA)之概念來進行策略模式的擬定，續以模糊德爾菲法（Fuzzy Delphi Method, FDM）為研擬各項評估面向之操作基礎，經由檢定與整合專家問卷結果，並探討不同面向之間的層級關係，進而瞭解其相互影響作用，並進行初步篩選以研擬出評估準則影響要素，再進一步以分析網絡程序法（Analytic Network Process, ANP），來探討不同構面指標之重要程度，再以運用Super Decisions 分析軟體進行權重計算與排序，進而從不同專家意見中與巨量資料分析所呈現之未來污染趨勢，進行交叉比對與綜合分析，以此評選出污染整治各項關鍵影響因素，亦能由此進行重點整治與優先補助順序之考量，最後以期透過科學性之工具加以進行污染場址之改善與兼顧社會發展之需求，並綜合考量污染場址各項特性，以符合國際潮流之方式管理污染場址，以達國土資源永續經營。
All kinds of factory buildings in the surrounding towns, although driving contemporary economic development and create employment opportunities for the region, but many industries in the production process, but also the depletion of environmental resources and the formation of a large number of waste water, waste gas and solid emissions, not only serious harm to human health, More cause environmental irreversible phenomenon. However, the industrial structure is gradually facing a transition, many factories eliminated or relocated by the environment, and for the legacy of high pollution of land, which is called Brown field. Taiwan to promote the industrialization policy, from an agricultural society into an industrial society, there is also a serious phenomenon of brownfields. The study from abroad in recent years, many studies found the concept of Big Data has been widely used in various fields, and its use by the large number of seemingly repetitive data clutter data inferred potential relevance, not only helps to understand the world , better way to achieve this improved decision-making to reshape society. Therefore, in order to effectively promote pollution remediation work, if a huge amount of information through the application and expansion will help accelerate remediation work, and can effectively solve the soil and groundwater pollution issues. Accordingly, in order to handle huge data of the Big Data arising from this research will integrate the first two years of research results, a huge amount of information as part of the basic input data, coupled with aggregated contaminated sites of industrial activity in the past, in order to analyze the future trend of pollution , and Area, the concept of Priority Funding PFA to develop policy mode, continued to Fuzzy Delphi Method, FDM is elaboration of the basis for the assessment of the operation, through verification and integration experts questionnaire results, and to explore different aspects between hierarchical relationships, and to understand their mutual influence, and preliminary screening criteria in order to assess the impact of the elements of elaboration and further to the Analytic Network Process, ANP, to discuss the importance of the different facets of the index, then performed using Super Decisions analysis software weight calculation and sorting, and then from a different expert opinions and massive data analysis presented in the future trend of pollution, cross-comparison and comprehensive analysis, as key factors of the selected remediation, also thus focus on remediation subsidies and priority order considerations, the final improvements to the contaminated sites and taking into account the needs of society through the development of scientific tools to, and comprehensive consideration of the characteristics of contaminated sites, in line with international trends of the management of contaminated sites, in order to sustainable management of land and resources. |
