Abstract
No single emission model suits every heavy-duty vehicle (HDV) application, yet guidance on model selection is lacking and the context-dependent suitability of emission models has not yet been translated into practical guidance for model selection. Following PRISMA guidelines, we systematically reviewed 70 peer-reviewed HDV emission modelling studies (2016 to mid-2026) from Scopus and Web of Science. Studies were characterised by model category, input data, outputs, validation, localisation, application scale and modelling objective. Emission modelling systems such as COPERT and MOVES predominated (47%), although approaches are diversifying. Only 40% of studies validated their estimates, and 43% used localised parameters. Pollutant coverage was explained better by model family than by application scale. This evidence was encoded in a transparent web-based decision-support system that ranks candidate models against user-specified scale, objective, outputs and data, with adjustable weights and traceable recommendations. Rankings reflect correspondence with user requirements rather than predictive accuracy. The system offers researchers and policymakers a reproducible starting point for HDV emission model selection.