In-Situ Pumping Test for Multilayer Hydrogeological Site in Taiwan

Pingtung plain is located the in the southwestern Taiwan, and rainfall concentrated from May to October with the average annual precipitation of 2000 mm. However, topographic steepest rushing stream lead to most of the precipitation becomes runoff and drains to the ocean in short time. Due to the shortage of surface water, the groundwater is an important one of much water recourses. Additionally, the government will be set up artificial lakes in proximal-fan of pingtung plain, which increases recharge and supply to usage of the agricultural and aquaculture. However, the locations of pumping wells are decided not only affecting the developmental quantity of the groundwater but economic growth serious limited. Therefore, MODFLOW-96 was used to simulate the groundwater distribution and optimal the better recharge zone in the regional scale. Based on the model calibration and verification results, the tuku farm of the kaoping lake study is better recharged from Laonong Stream and the northeast, and the safe yield is much than other study zone. Additionally, directional variations in permeability anisotropic formations have important effects on velocities and storage of the groundwater recourses. We have further utilized modifying ANN (artificial neural networks) approach, as well as incorporating the Papadopoulos analytical solution [Lin et al, 2010], to estimate the directional and magnitude of the permeability parameters for pumping test at the tuku farm of the kaoping lake study, which the methodology will be improve accuracy of the estimation parameter. According to drawdown record data of six observation wells, results suggest that the locations of the pumping wells are set up in the northeast and northwest which since sedimentary formations are more permeable along the major direction from the northeast and northwest. Hence, the information can be helpful the groundwater management and supply in the Pingtung plain.