Zhang, Yi, Gomes, Antonio Topa, Beer, Michael 
ORCID: 0000-0002-0611-0345, Neumann, Ingo, Nackenhorst, Udo and Kim, Chul-Woo
(2019)
Modeling asymmetric dependences among multivariate soil data for the geotechnical analysis - The asymmetric copula approach.
SOILS AND FOUNDATIONS, 59 (6).
pp. 1960-1979.
|
Text
manuscript.pdf - Author Accepted Manuscript Download (1MB) | Preview |
Abstract
Multivariate information of soil parameters is quite important for the design and risk assessment of geotechnical engineering problems. It is necessary to have an accurate and realistic statistical multivariate model for representing the soil properties and thus evaluating the soil conditions. Thus, advanced multivariate modeling of soil parameters could help to improve the geotechnical engineering practice. In this paper, the asymmetric copulas are introduced to model the geotechnical soil data. Compared to extensive previous research on the use of symmetric copulas on the modeling of engineering data, this study is focusing on capturing asymmetric dependencies among the natural soil parameters, which are critical for engineering design. A copula-based multivariate probabilistic model is built based on a set of collected samples from a granite residual soil from Portugal. Several asymmetric copula functions, capable of capturing nonlinear asymmetric dependence structures, are tested and analyzed. The fundamental information on tail dependencies and measures of asymmetric dependencies are also exploited. To demonstrate the advantages of asymmetric copulas, its concept is compared with the traditional copula approaches for modeling site soil data. The performance of these asymmetric copulas is discussed and compared based on data fitting and extreme value characterizations.
| Item Type: | Article |
|---|---|
| Uncontrolled Keywords: | Geotechnical analysis, Asymmetric copula, Soil properties, Joint distribution, Multivariate analysis |
| Depositing User: | Symplectic Admin |
| Date Deposited: | 20 Nov 2019 09:48 |
| Last Modified: | 19 Jan 2023 00:19 |
| DOI: | 10.1016/j.sandf.2019.09.001 |
| Related URLs: | |
| URI: | https://livrepository.liverpool.ac.uk/id/eprint/3062473 |
Dimensions
Dimensions
