Abstract
It is important to simulate accurately a rainwater infiltration process into slopes, in order to predict landslides caused by heavy rains. The accurate simulations of rainwater infiltration process into slopes require elucidating the soil moisture characteristics of soils. The authors have tried to estimate soil moisture parameters based on the field measurements of volumetric water contents by using the Particle Filter method, which is one of the sequential data assimilation methods. In this study, the discussion on determining the soil moisture parameters was carried out from the viewpoint of practical applications of the Particle Filter method. Firstly, discussion about appropriate value of standard deviation of observation noise, which is inherent to the Particle Filter method, was carried out. 60% of variation of measured volume water contents was recommended to an appropriate value of standard deviation of observation noise, in order to determine the adequate parameters of the soil moisture characteristics. Then, another discussion about adequate algorithm of the Particle Filter method was carried out, in comparison between SIR and MPF, which are representative algorithms, respectively, in the Particle Filter method. The appropriate soil moisture parameters could be determined even if either algorithm is applied. However, it was elucidated that the MPF is more effective algorithm than SIR.
