Abstract: Urban forests serve as a crucial ecological barrier to address climate change and carbon goals by optimizing human settlement quality, enhancing carbon sequestration, and improving urban climate. However, carbon fluxes within urban forests are largely unexplored. In this study, we used eddy covariance technology and a flux footprint model to analyze the spatial-temporal variations and key drivers of carbon fluxes in Beiling Park of Shenyang City from 2023 to 2024. The results showed that annual carbon fluxes ranged from -17 to 13 μmol·m^-2·s^-1, with daily averages fluctuating between -6.32 and 2.1 μmol·m^-2·s^-1. The forest in the park functioned as a carbon sink for only 44 days in each year. The carbon sink capacity varied seasonally, with the strongest carbon sink in summer and a carbon source in winter. Temperature was the dominant factor influencing carbon flux. We used an extended STIRPAT model to project carbon emissions under baseline scenario and multi-source data to estimate carbon emissions of urban functional zones. An analysis of the spatial correlation between carbon flux and carbon emissions revealed that high carbon emission areas concentrated in industrial and commercial zones, which was highly overlapped with carbon flux hotspots. This study provides a valuable dataset and methodological reference for monitoring carbon fluxes in urban forests, underscoring the importance of optimizing urban planning to enhance carbon sinks. Future research should integrate multi-source remote sensing and model simulations to elucidate the underlying mechanisms.

Key words: carbon flux, carbon emission, eddy covariance technology, flux footprint model, STIRPAT model