【正文】
sions in treating specific sites. Decision support systems (DSS) can assist in solving this problem. Many DSSs have been proposed for managing coal mining operations and groundwater remediation. However, there is a lack of study that bines the functions of mine characterization, numerical modeling, risk assessment and remediation technique selection within a DSS. The objective of this study is to address this gap in previous researches and develop an integrated decision support system (GCDSS) that supports all these functions for groundwater pollution control at coalmining contaminated sites. 2 Background: AMD and its treatment AMD from coal mining is a difficult and costly problem. It can seriously affect groundwater quality and cause metals to leach from mine wastes. AMD results from the oxidation of metal sulfides, particularly pyrite (FeS2). Under the acidic conditions, oxidation of pyrite occurs in the following reaction [1]: This reaction demonstrates the polluting capability of the oxidation of pyrite that every mole of pyrite can be converted to 16 moles of hydrogen and 2 moles of sulfate. Much acid is generated through this reaction. There are two methods for treating AMD: active treatment and passive treatment. Active treatment involves neutralizing acidpolluted water with alkaline chemicals which include limestone, hydrated lime, caustic soda, soda ash, and ammonia [2]. Active treatment is expensive and requires much time and manpower to maintain. Passive treatment employs naturally occurring chemical and biological reactions and requires little or no maintenance. Passive methods include anoxic drains, limestone rock channels, alkaline recharge of groundwater, and diversion of drainage through manmade wetlands or other settling structures. 3 Development of Decision Support System Knowledge Acquisition Knowledge acquisition is a bottleneck in DSS development and involves the processes of knowledge elicitation, analysis and representation. It is crucial 中國(guó)礦業(yè) 大學(xué) 12 屆 本科生 畢業(yè)設(shè)計(jì)(論文) 第 141 頁(yè) because output of the system is only as good as the input. The main sources of knowledge in this study are the domain experts, the statistical data about coal mining, and documents. GCDSS GCDSS consists of the modules for mine characterization, numerical modeling, risk assessment, and remediation technique selection. It also consists of a graphical user interface which allows the user to input and query the site related data, and shows the remendations and suggestions for the user. Details on the numerical modeling, risk assessment, and remediation technique selection modules are discussed as follows. The architecture of GCDSS is shown in Figure 1. Architecture of GCDSS Mine Characterization Module Mine characterization is crucial for the following numerical modeling, risk assessment, and the selection of remediation technologies in GCDSS. This 中國(guó)礦業(yè) 大學(xué) 12 屆 本科生 畢業(yè)設(shè)計(jì)(論文) 第 142 頁(yè) m