フィルター分離AR法とカルマン・フィルターによる洪水予測法に関する研究

抄録

A new method of flood prediction is presented which applies the physically-based hydrological system analysis (filter-separation AR method proposed by Hino and Hasebe) together with the Kalman filtering.
The principle of the method consists of separating through the numerical filters the total runoff time series into long-, intermediate and short-term components or groundwater flow, interflow and surface-flow components. As a total system, a hydrologic system is a non-linear one. However, once it is separated into two or three subsystems, each subsystem may be treated as a linear system.
The rainfall components into each subsystem a estimated inversely from the runoff component separated from the observed flood.
For the short-term runoff component, the Kalman filtering theory is applied to revise the first estimated rainfall through the errors between the observed and the predicted runoff.