Abstract: Spatial analysis of carbon metabolism is the basis for optimizing urban land use and low-carbon development. In this study, we constructed a horizontal carbon flow model to calculate carbon metabolism density in Yulin City, Shaanxi Province. We employed the approach of ecological network utility to calculate and standardize carbon flux across different land-use types, and obtain the horizontal carbon flow matrix (A). Then, we acquired the effective utilization matrix (D) and overall utility matrix (U), to examine the impact of land use changes on the spatial pattern of unban carbon metabolism and to analyze the spatiotemporal variations of carbon flow in Yulin City. We used a geographic detector model to analyze the driving factors of carbon metabolism and identify the primary drivers for urban carbon metabolism. The results showed that land use change aggravated the disturbance of carbon metabolism in Yulin City from 2000 to 2020, and the net carbon flow was negative in both horizontal and vertical directions. Shenmu City and Yuyang District of Yulin City were the most important negative carbon metabolism compartments. Six counties in the south were positive carbon metabolism compartments. Annual average precipitation and NDVI were the primary drivers of variations of urban carbon metabolism, while annual average temperature and nighttime light exerted relatively minor influence. In conclusion, our results have reference significance for optimizing urban land use and facilitating low-carbon development in semi-arid areas, as well as realizing increasing carbon sink and reducing carbon emissions.

Key words: urban carbon metabolism, carbon flow, ecological network utility, land use change, geographic detector