現場計測 (1)

[III-527] Application of the deep learning to estimating groundwater level within landslide body in a heavy snowfall area

*峯 啓一郎¹、柳浦 良行¹、三木 茂¹、武田 茂典¹ (1. 基礎地盤コンサルタンツ（株）)

Keywords:deep learning, groundwater level, snowfall area, time history, estimation

The tank model is also applied to the estimation of groundwater level in landslide. However, the tank model has difficulties that the estimation is in a heavy snowfall area, because water from deposited snow is continuously recharged into the ground without regard to the rainfall in a snowmelt season. Therefore, the deep learning, which is one of AI technologies, is applied to the estimation of groundwater level. It is resulted that the annual change of groundwater level is successfully simulated and estimated with good accuracy by using the deep learning. We will present the method to estimate the groundwater level and the results in the heavy snowfall area.