Methodological Reporting in Humanitarian and Disaster Simulation Under Uncertainty

Simulation and simulation-based optimisation support humanitarian logistics and disaster-management decisions under uncertainty, but the evidential basis of their recommendations depends on how clearly publications connect model specification, computational implementation, experimentation, evaluation and uncertainty analysis. This secondary methodological review reanalyses a fixed corpus of 109 peer-reviewed journal studies spanning discrete-event simulation, agent-based modelling, system dynamics, hybrid simulation and simulation-optimisation. Indicators were organised across four methodological domains and analysed using indicator-specific denominators, publication-level co-occurrence configurations, descriptive comparisons and exploratory Fisher’s exact tests for the simulation-optimisation subgroup. Conceptual model structure and scenario design were commonly reported, whereas computational environments, statistical characterisation of stochastic outputs, joint reporting of calibration, verification and validation procedures, and input-model uncertainty were less consistently visible. Descriptively, simulation-optimisation studies reported replications and structured experimental designs more frequently than other simulation studies, but no subgroup association remained statistically significant after Holm adjustment. The principal reporting discontinuities lay between conceptual models, executable implementations, computational experiments and the decision claims derived from them. The review develops conditional reporting priorities that relate methodological evidence to model architecture, uncertainty conditions and intended humanitarian use. These priorities provide a structured basis for appraising simulation-supported decisions while recognising that reporting requirements vary across paradigms, experimental designs and operational contexts.