pyKasso: An open-source three-dimensional discrete karst network generator

• Published in: Environmental modelling & software : with environment data news Vol. 186; p. 106362
• Main Authors: Miville, François, Renard, Philippe, Fandel, Chloé, Filipponi, Marco
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.03.2025
• Subjects: algorithms; anisotropy; computer software; Geometry; Groundwater; groundwater flow; Karst; karsts; Network; Stochastic; Geometry; Karst; Stochastic; Groundwater; Network
• ISSN: 1364-8152; 1873-6726
• DOI: 10.1016/j.envsoft.2025.106362

Modeling groundwater flow using physically based models requires knowing the geometry of the karst conduit network. Often, this geometry is not accessible and unknown. It is therefore crucial to be able to model it. This paper presents pyKasso, an open-source Python package that generates those geometry based on a pseudo-genetic approach. The model accounts for multiple data sources: a 3D geologic model, the position of known inlets and outlets, the statistical distribution of fractures or inception features, and known base levels. This approach simplifies previously published work by considering a 3D anisotropic fast marching algorithm. The paper presents the structure of the code and explains in detail how it can be used from the most simple 2D situation to a complex 3D case. •pyKasso is an open-source package for generating stochastic karst conduit networks.•It uses a 3D anisotropic fast marching technique that simplifies the algorithm.•pyKasso provides input geometries for groundwater flow simulations in karst systems.•pyKasso can be used to quantify karst network geometry uncertainty.