Bivariate probabilistic modelling of hydro-mechanical properties of vegetated soils

Vegetation retains soil suction, which significantly affects the shear strength characteristics of soil and slope stability. However, because of uncertainties (heterogeneous root distribution, spatial variability) associated with vegetated soil, probabilistic analysis on the hydro-mechanical properties of vegetated soil is a necessity. Moreover, mechanical (cohesion, c, angle of internal friction, φ) as well as hydrological properties (suction, ψ, water content, θ_w) are correlated in nature, and this has a significant effect on the reliability of geotechnical structures. However, these studies investigated the dependence structure of mechanical parameters only. The main objective of this study is to evaluate the dependence structure of both mechanical (c, φ) as well as hydrological (ψ, θ_w) parameters of homogeneously compacted vegetated soil on the basis of measured field data. This is done by using copula theory to map the copula density functions (copula probability distribution functions [PDFs]) of (c, φ) and dψ, θ_w) to their respective marginal distributions (PDFs) in order to simulate their bivariate distributions. The novelty of this work further lies in analysis of time-dependent behavior of θ_w by generating its PDFs, utilizing the concept of average mutual information (AMI). Thus, the information attained in this study about the inherent behavior of hydromechanical parameters of soil can be further utilized to estimate the reliability of vegetated slopes under varying climatic conditions.