ヒ テイジョウ マルコフ モデル ニヨル カセン リュウリョウ シミュレーション
Stochastic Simulation of River Flow by Nonstationary Markov-Chain Model
d0030008n015.pdf 530 KB
The time series of the average value of 5-days flow can be represented by a nonstationary Markov-chain model with randam components, month by month. The randam components ε_t may be transformed into χ_t (χ_t = ε_t + B + b) and it's distribution can be fitted most satisfactorily to a lognormal distribution. Here,
B ; a gradually increasing value with some step. The greater it is, the more normalized distribution of χ_t is gained.
b ; a parameter which approximates the shape of lognormal distribution to that of the observed one, by Iwai method.
Having sequentially generated the river flow by that model with Monte Carlo simulation technique for 100 years at two points, the model has been proved to be available for practical purposes.
Bulletin of the Faculty of Agriculture, Shimane University
Shimane University, Faculty of Agriculture
Departmental Bulletin Paper
Faculty of Life and Environmental Science