FUME: processing emission inputs for chemical transport models from continental to street scale

Abstrakt

Credible modeling of air quality largely depends on the availability of consistent, model-ready emission inputs, yet the preparation of emissions remains one of the most complex and error-prone steps in the modeling chain. We present FUME 2 (Flexible Universal processor for Modeling Emissions), an open-source emission processor designed to bridge heterogeneous emission inventories and chemical transport models at various spatial scales from continental to urban microscale. FUME provides a highly configurable, modular framework that unifies diverse input data formats, geometries, chemical species, and temporal resolutions into a common internal structure, enabling transparent spatial, temporal, chemical, and vertical disaggregation of emissions. Built as a Python application on top of a PostgreSQL/PostGIS backend, FUME implements user-defined transformation chains that allow flexible combination of inventories, application of spatial surrogates and emission scenarios, and integration of external models such as biogenic emission modules. The processor supports output of emissions in formats suitable for ingestion by multiple regional chemical transport models (e.g., CAMx, WRF-Chem), and the meter-scale model PALM, providing fully three-dimensional emissions accounting for urban structures. FUME can handle large, multi-inventory setups with a wide range of processing configurations. By emphasizing flexibility, transparency, and extensibility, FUME provides a robust and transferable solution for emission processing in research, operational forecasting, and policy-relevant air quality applications.