LAUSR

Abstract The observation of yield-drawdown response in the course of pumping test, as well as its derived application for aquifer characterization, is an essential practice in a geotechnical investigation. On the basis of that, for providing a reasonable estimate of aquifer transmissivity on site, a number of empirical models have therefore been proposed. Yet, the majority of previous studies are conducted based on investigations at homogeneous alluvial plain, very few, particularly in the case of Taiwan, are carried out focusing on mountainous fractured-rock aquifers. Aiming to address this problem, the observation data of 53 pumping tests obtained from a multi-years in-situ geological investigation program are analyzed, and the applicability of empirical log-log linear relationship between T and Q/s is examined. Our analysis reveals that the yield-drawdown responses derived from different geologic formations are quantitatively distinct, and those empirical rules derived from the alluvial landform should be used with care when applying in mountainous fractured-rock aquifers. Based on the analysis of statistical results in our study, with a systematic investigation and clarification of the Q-s relationship in advance, a satisfactory and cost-effective solution is obtained.