Abstract
It is very important to grasp the groundwater behavior not only on the design of underground structures but on their maintenance. This paper presents an application of multivariate regressive model to the groundwater system including structure, precipitation and other influential factors. First, we describe the modeling by selecting and newly adding manipulated variables with the use of simulated data. Second, we have applied the modeling method to the observed data of the Kikuma Test Plant for underground oil storage. We have also found this method is applicable for the groundwater system including natural factors.
