Abstract
Soil moisture is a critical variable for evaluating water movement in hillslope hydrology. To understand its temporal response patterns via statistical distribution, time series of soil moisture spatial distributions were collected for nine months along a steep montane hillside. In this study, a novel synthesized probability density function was introduced to accurately describe the soil moisture variability in conjunction with controls. The optimal number of componential distribution model was determined using hidden Markov model (HMM). A univariate HMM captured the profile development of latent soil moisture states as well as the transitional probabilities between regression and switch patterns that were useful for categorizing soil moisture variability corresponding to hydrological processes. The deeper and lower the locations in the hillslope, the greater the complexity in soil moisture latent states, indicating a higher number of flow paths. Both seasonal- and event-based soil moisture variations were adequately characterized through the identification of latent soil moisture states. The impact of environmental factors on soil moisture was evaluated via a comparative analysis of multivariate HMM and principle component analysis, revealing that the controlling factors were related to the context of seasonal and spatial variability, as well as wetting and drying periods during rainfall event.
Accepted Version (
Free)
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call