ニュートラルネットワークモデルを用いた地下水モニタリング解析に関する研究

A Stydy of Groundwater Monitoring Data Analysis Using Artifical Neural Network Model

渡辺 邦夫; Gautam, M. R.*; 三枝 博光

Watanabe, Kunio; Gautam, M. R.*; Saegusa, Hiromitsu

日本語要旨なし

The results of groundwater flow modeling are to be justified using groundwater monitoring data in the hydrogeological characterization. On the other hand, hydraulic continuities of the geological structures, all of which are considered to have great effect on groundwater flow and/or groundwater quality, are to be estimated using the groundwater flow monitoring data with hydraulic response to some impacts such as borehole drilling, pumping test and so on. Therefore, the groundwater monitoring is important for characterizing the geological and hydrogeological environments. In order to characterize of hydrogeological environment using the monitoring data, it is important to evaluate the influence of artifical and natural impact on the monitoring data. In this study, the following three research works were carried out based on the groundwater monitoring data collected at the Tono area. Artifical Neural Network (ANN) was adopted as the tool for monitoring data analysis.Runoff analysis for assessment of importance of soil moisture on runoff estimation in a catchment Analysis of water level fluctuation for determination influence factors in the water level fluctuation and for filtering out the influence factors from the water level data Analysis of hydraulic pressure fluctuation in deep geological formations for hydrogeological characterization and assessment of human influence on the pore pressure in deep formation Through this study, applicability of ANN for analysis and interpretation of the groundwater monitoring data could be confirmed and methodology for utilization the monitoring data for understanding and characterization of hydrogeological environment could be developed.