童慶斌臺灣大學:生物環境系統工程學研究所陳佳正Chen, Chia-JengChia-JengChen2007-11-272018-06-292007-11-272018-06-292004http://ntur.lib.ntu.edu.tw//handle/246246/55955摘要 對於控制及復育地下水污染之先決條件為必須探求出污染源之所在位址及其排放歷程,本研究發展地下水污染源位址之最佳化推估流程。地下水反向分析問題一般針對於水文地質參數之推估以利於有效模擬地下水流與傳輸情況,而確立地下水污染源與推求其排放歷程可謂為在能夠模擬區域地下水流與傳輸情況後反向分析之最終目的。因此本研究在探求地下水污染源位址之同時對於水文地質參數之延散度加以最佳化推估,再進行推求重現其污染源排放歷程。 本研究設計簡例應用於此類問題,希求以簡例先行驗證反向分析模式之可行性。在最佳化推估延散度部份,假設區域內建立一隨機延散度分佈視為真值,在一持續性點源污染排放之情形下利用遺傳演算法配合佛洛諾尹分區方法,以假設下僅有之水質觀測井利用遺傳演算法加以迭代優選出污染源之位址與最佳延散度值及其空間分佈;而研究結果更對於參數之維度與區域尺度內縮問題等加以深入探討對於結果之變化與影響。 在推求重現污染源排放歷程部分,本研究提出推估污染源排放起始時間與結束時間之觀點,以提升反向分析模式觀測值與模式輸出值比對之合理性為目的,利用擴增起始與結束時間變數組之統整性方法希求能推求重現出複雜之污染源排放歷程。在決定起始與結束時間變數之最佳組數時,亦利用在決定最佳延散度分區數之相同停止判別準則,分析其參數維度與結構誤差等表徵來決定最佳組數。 簡言之,本研究有效地確立出地下水污染源位址與最佳化延散度參數分區,以及重現污染源排放歷程,以設計簡例證實反向分析模式之可行性,可作為日後控制與復育地下水之應用。Abstract The initial step in groundwater contamination control and remediation process is to identify the location of the source and its release history. The general groundwater inverse problems focus on estimating hydrogeological parameters in order to simulate groundwater flow and transport process effectively. To identify the location and recover the release history of groundwater contaminant source may be an important goal of the inverse model after effective simulation. Therefore, this study identifies groundwater contaminant source with optimizing hydrogeological parameter dispersivity simultaneously. Thereafter, recovering release history of contaminant source is the other part of this study. This study designs cases to analog to these problems. It is wished to demonstrate the practicability of the inverse analysis model in advance. In optimizing dispersivity zoning patterns, true dispersivity in the study area are designed to randomly distribute. The true contaminant source is designed to emit steadily. With few observation wells, Genetic Algorithms and Voronoi Zonation method are used to calibrate the location of contaminant source and the optimized spatial distribution of dispersivity. In recovering release history of contaminant source, this study proposes a method to estimate release beginning, ending time and associated emission quantity. Furthermore, the unified block-out method is proposed to expand release beginning and ending time group variables to recover complicate release history. The same stopping criteria used in determining optimal numbers of dispersivity zone are adopted to determine the optimal number of variable groups. In summary, this study proposes methodologies which can identify groundwater contaminant source with optimizing dispersivity zoning patterns and recovering release history effectively. The results of designed cases prove the practicability. The methods can be references to future studies.Table of Content Chinese Abstract Abstract Table of Content………………………………………………………………Ⅰ List of Tables…………………………………………………………………Ⅲ List of Figures…………………………………………………………………Ⅴ Chapter 1 Introduction 1.1 Purpose of This Study…………………………………………………1 1.2 Spatial Analysis of Identifying Groundwater Contaminant Source…2 1.3 Release History of Identifying Groundwater Contaminant Source…3 1.4 Process of This Study…………………………………………………5 1.5 Structure of Thesis……………………………………………………6 Chapter 2 Literature Review 2.1 Groundwater Inverse Analysis…………………………………………7 2.2 Identification of Groundwater Contaminant Source Problem………9 2.3 Genetic Algorithms…………………………………………………11 Chapter 3 Development of Methodology 3.1 Voronoi Zonation Method………………………………………………13 3.2 Zonation Stopping Criteria……………………………………………15 3.3 Genetic Algorithms………………………………………………………24 Chapter 4 Spatial Analysis of Groundwater Contaminant 4.1 Introduction to Problems………………………………………………27 4.2 Spatial Analysis Process………………………………………………28 4.3 Design of Study Cases………………………………………………34 4.4 Results and Discussions………………………………………………39 Chapter 5 Identification of Groundwater Contaminant Release History 5.1 Introduction to Problems………………………………………………53 5.2 Methodology of Estimating Release History………………………54 5.3 Application to Simple Case…………………………………………59 5.4 Application to Complicated Case………………………………………66 Chapter 6 Conclusions and Suggestions 6.1 Conclusions……………………………………………………………81 6.2 Suggestions……………………………………………………………83 References………………………………………………………………………87 Appendix: Summary of Inverse Problems in Literatures………………………93643575 bytesapplication/pdfen-US地下水污染排放歷程反向分析遺傳演算法Release historyInverse analysisGenetic algorithmsGroundwater contaminant延散度空間型態及地下水污染源與其排放歷程反算分析之研究Inverse Analysis of Identifying Groundwater Contaminant Source and its Release History Along with Optimizing Dispersivity Spatial Patternsthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/55955/1/ntu-93-R91622010-1.pdf