Developed urban datasets for WRF-STILT: impacts on near-surface CO2 concentration in urban functional zones

Accurate urban carbon monitoring increasingly relies on atmospheric simulations to assess and refine emission inventories, but this “top-down” application is often limited by transport errors caused by oversimplification of urban representations in mesoscale models. This paper develops a high-resolution local urban dataset for Hefei, China, and explores its impact on near-surface CO 2 concentration in urban functional zones within the Stochastic Time-Inverted Lagrangian Transport Model (STILT) framework coupled with Weather Research and Forecasting (WRF). We innovatively integrated multi-source data, including the Global Impervious Surface Area (GISA) and Essential Urban Land Use Categories (EULUC), to reclassify the urban area into three distinct categories: low-intensity residential, high-intensity residential, and commercial/industrial/transportation. Key urban parameters (e.g., building height, road width, anthropogenic heat) were statistically derived from local geographic information data, replacing the model’s default values. The results demonstrate that the refined urban representation substantially improves the model’s performance, particularly in simulating near-surface wind fields, with the correlation coefficient (R) increasing from 0.82 to 0.89. The simulated near-surface CO 2 concentrations successfully capture the typical diurnal variation patterns and spatial heterogeneity among four representative functional zones, showing strong consistency with ground-based observations from a novel photoacoustic spectrometer (PAS). Improvements are most evident where urban morphology and activity patterns strongly modulate mixing and advection, while complex public areas remain challenging due to sub-grid heterogeneity and intermittency. These results demonstrate that integrating remotely sensed imperviousness, functional zoning, and localized urban canopy parameters can reduce transport-driven errors in city-scale CO 2 modeling, providing a more reliable meteorological basis for urban emission evaluation and inventory verification.