GeoPathSim, Python workflow for geostatistical simulation of mineral structures along transport pathways, Dresden, Germany; Open access in RODARE

GeoPathSim: A Python workflow for geostatistical simulation of mineral structures along potential transport pathways based on classified mineral images

Switch on selection | Login

Duckstein, A. ; Pospiech, S. ; Tolosana Delgado, R.

GeoPathSim is a Python-based research software workflow for deriving, analysing, and stochastically simulating mineral structures exposed along potential transport pathways in heterogeneous geological materials. It was developed within the SANGUR project (Systematic Sensitivity Analysis for Mechanistic Geochemical Models using Field Data from Crystalline Rock; ref. no. 02E12112A) to support the characterization of mineral assemblages that may be accessible along potential transport pathways and to provide structural input for subsequent geochemical and transport-related analyses.

GeoPathSim operates on spatially resolved, classified two-dimensional raster data, such as SEM-based automated mineralogy maps or comparable categorical images of mineral or material classes. The preprocessing workflow converts classified images into integer rasters, harmonizes mineral classes, calculates bulk and mask-area mineral compositions in volume and weight fractions, and can derive mineral-neighbour statistics from the original structure.

Two complementary stochastic simulation approaches are implemented. For visible fractures or other explicitly traceable pathways, the one-dimensional workflow extracts mineral sequences, estimates Markov transition probabilities, and generates stochastic pathway realizations. For samples without clearly traceable pathways or when complete spatial mineral fields are required, the two-dimensional workflow estimates empirical indicator variograms, fits Matérn models, and generates hierarchical multi-Gaussian mineral-field realizations.

Additional workflows support the aggregation of results from multiple 1D and 2D simulations, mineral-neighbour analysis, categorical smoothing, bulk-versus-pathway comparison, mineral highlighting, image cropping, and representative-volume-element size analysis. GeoPathSim records explicit random seeds, configurations, manifests, logs, and structured numerical outputs to facilitate reproducible analyses. The software is designed to run on conventional desktop or laptop computers and does not require dedicated high-performance computing infrastructure.

Keywords: Geostatistical simulation; Markov chains; Multi-Gaussian simulation; Transport pathways; Spatial data analysis; Python research software

Software in the HZDR data repository RODARE Publication date: 2026-08-13 Open access DOI: 10.14278

Veröffentlichung im HZDR-Daten-Repository RODARE - Id 4932

Bautzner Landstraße 400, 01328 Dresden - Germany Phone: +49 351 260 - 0 Email: kontakt@hzdr.de

Helmholtz-Zentrum Dresden-Rossendorf - Bautzner Landstraße 400, 01328 Dresden Germany

Telefon: +49 351 260-0 - https: