There are many groundwater pollution events were reported and investigated in Taiwan within recent ten years. Yet, a serious problem involved in some of those pollution events is that the sources of pollution were not cleary identified. The purpose of this project is to analyze those groundwater sampling and analyzed data using multivariate statistical analysis, including principal components analysis, factor analysis, and cluster analysis and to explore the possible pollution sources and to assess the feasibility of these methods in goundwater source identification. The achievements of this project include: 1. To review the related literatures. 2. To assess the feasibility of principal components analysis, factor analysis, and cluster analysis for exploring the possible pollution sources, and compare the results of these methods. 3. To provide the necessary information including the items, the quantity, the accuracy and the precision for detection data required in those statistical analyses, and to provide a guideline for the preprocess of the data. 4. To explain their strengths and the liminations of those statistical methods in analyzing three groundwater pollution cases happened in Taiwan. 5. To provide the necessary sampling items for the pollution source identification and the conditions for the use of these statistical methods. 6. To give trainning courses and provide related data and documentation regading those statistical analyses in case studies.

contaminant source identification,multivariate statistical analysis,groundwater contamination