Abstract: Accurate monitoring of carbon sinks and carbon emissions at the urban scale in China is of great significance for assessing the progress towards carbon neutrality. To precisely characterize carbon source-sink features at high resolution, an integrated model for urban net carbon emissions was developed and applied to the Yangtze River Delta （YRD） urban agglomeration in China as a case study. Firstly, the CASA model was employed to calculate the spatiotemporal distribution of monthly net primary productivity （NPP） for 2021—2022. The results reveal significant seasonal variations in NPP, with peak carbon sequestration during summer reaching 70–100 g/（m²·month） in major cities, while winter values generally remained below 15 g/（m²·month）. Spatially, NPP exhibited a pattern of lower values in urban built-up areas and higher values in the vegetated southern regions. Additionally, an improved high-resolution spatial disaggregation method for carbon emissions, based on nighttime light data and population weighting, was proposed to downscale provincial-level emissions to a grid level. Results indicate that high carbon emission values are concentrated in urban centers, with per-grid emissions exceeding 5 000 t. Finally, the spatiotemporal distribution of net carbon emissions （NCE） was calculated at a 250 m resolution. The findings show that NCE in most cities exceeds 700 g/（m²·a）, with only a few cities, such as Chizhou and Xuancheng in Anhui Province, acting as carbon sinks. The overall carbon sink proportion relative to total carbon emissions in the YRD region is 13.1%, reflecting a currently high level of net carbon emissions in this urban agglomeration. Significant disparities in NCE exist among cities, with Shanghai exhibiting markedly higher annual NCE compared to cities rich in vegetation resources, such as Hangzhou and Huzhou. The model demonstrates a strong capability to accurately capture the characteristics of urban carbon sinks and emissions, offering finer spatiotemporal resolution and more detailed spatial variability than existing data products. Owing to the comprehensive data coverage and methodological generalizability, the model can be applied to estimate NCE distributions in other urban areas across China, thereby aiding in understanding disparities in urban carbon neutrality levels and supporting decision-making for urban development and the construction of low-carbon cities.