Stochastic simulation and uncertainty analysis of karst conduit network using pyKasso

Abstract

Owing to the intrinsic heterogeneity and anisotropy of karst systems, traditional hydrological exploration methods face significant challenges when investigating karst conduit networks. This study employs pyKasso for the stochastic simulation of karst conduit networks in Panzhou City, focusing on uncertainty analysis through local sensitivity analysis and Monte Carlo methods. The simulation process incorporates geological, topographic, and fracture data to create a realistic representation of the karst network. We found that the spatial configuration and characteristics of the karst network are significantly influenced by various input parameters such as fracture parameters, inlets, outlets, and cost ratios. We highlight the minimal influence of fracture densities and the substantial impact of the count of inlets and outlets on crucial network metrics. The results demonstrate the critical role of parameter sensitivity and variability in modeling the intricate karst systems, providing valuable insights for hydrogeological studies and the management of karst water.