We have investigated flood runoff processes on hillslope through variation in time-varying parameters of the nonlinear relationship between basin storage and discharge. This paper described a method of identification of the parameters using stochastic system theory. Parameter argumentation was utilized in state space formulation, and a nonlinear filter and a nonlinear smoother were applied to data obtained at a small mountainous basin (catchment area 0.095 km^2)， It was found, by comparing the two methods, that filtering realized a speedy estimation, whereas smoothing had an accurate algorithm. It is concluded, from the simulations, that Single Stage Iteration Filter and Fixed Interval Smoother when used together produced a marked performance in state estimation.
Mendeley出力
