GPR在地下污染檢測之方面,是一門極有發展潛力與價值的實用性大地工程檢測技術。但地下油污染圖像的判讀分析解釋,一直無法精確的掌握其圖像顯示的特徵與實際上污染分佈的準確度,本研究利用砂箱模型模擬現地污染現象和特徵,包括乾土層、濕土層及部分飽和土層,其中部分飽和土層模擬地下水位以上、地下水位面、地下水面以下之滲漏點,以透地雷達搭配高頻天線(900MHz)實際施測,進行觀察砂土層中比水輕之輕非水相液體(Light Non-Aqueous Phase Liquid,LNAPL)污染,分析測線斷面中實際污染分佈情形與訊號處理後之雷達圖像,研究結果顯示利用希爾伯特轉換中之能階(Magnitude)所處理後之污染圖像,其在乾土層、濕土層及部分飽和土層中皆有不錯的辨識度,並配合實際案例與地電阻影像比對,進一步的驗證透地雷達在地下地層環境污染檢測之精確性與實用性,且提供實驗中後處理方式、污染圖像比與修正係數,期以用於後續處理此類案例中之參考依據,達到發揮此法之功能。
For detection of underground pollution, the use of GPR is a geotechnical detection technique that is promising and has a good deal of potentials for further developments. For the identification of pollution patterns and their interpretation, however, there has not been a way that is accurate enough to determine the characteristics of pattern displays and the precision of the actual distribution of pollution. The study intended to investigate the contamination and characteristics in geologic formations, such as dry soil layers, wet soil layers and partially saturated layers, using simulation of molding box. In the partially saturated layers, seeping points above, at and below water table were simulated. The combination of ground-penetrating radar and high frequency antennas (900MHz) was deployed for field survey, while the contamination of Light Non-Aqueous Phase Liquid (LNAPL) was explored in sand layers. The actual distribution of contamination in the surveyed profiles and the radar images after signal processing were analyzed. The result showed that the contamination images after processed with magnitude in the Hilbert transform provided acceptable identification for dry, wet and partially saturated soil layers. With the help of comparison between practical cases and electrical resistivity imaging, the accuracy and practicality of ground-penetrating radar were further verified for detection of contamination in underground environments. Also, a post-processing method for experiments, contamination image ratio and correction factors were provided for subsequent processing of similar cases, achieving the optimal use of this method.