JRODOS Multi-Model Atmospheric Dispersion Pipeline for Nuclear Emergency Planning

Multi-ADM-Analyzer is a modular Python framework for post-processing and visualizing atmospheric dispersion model outputs. The tool supports ensemble-based analysis of cell-wise simulation results by transforming binary model outputs into geospatial products suitable for inter-model comparison and publication-quality reporting. Its main workflows include probabilistic exceedance mapping, signed relative-frequency comparison between two Atmospheric Dispersion Models (ADMs), categorical overlay analysis among three ADMs, and threshold-based overlay analysis for identifying robust regions of probabilistic agreement. The framework reads scenario-specific binary result files, extracts one scalar value per computational grid cell and ensemble run, computes statistical indicators such as exceedance probability or relative activation frequency, joins the derived quantities to a geospatial computational grid, and renders classified map outputs.

The architecture separates configuration handling, path resolution, data loading, geospatial joining, statistical computation, plotting, and pipeline orchestration into dedicated modules. The design keeps the scientific workflow configuration-driven. ... mehrBy supporting harmonized map panels, fixed spatial extents, local or online basemap overlays, statistical summaries, and debug outputs, Multi-ADM-Analyzer provides a reproducible workflow for comparing ADM spatial footprints, identifying agreement and divergence among models, and supporting probabilistic interpretation of atmospheric dispersion simulations in emergency-planning and sensitivity-analysis contexts.