High-resolution net ecosystem productivity modeling reveals spatiotemporal heterogeneity of urban carbon metabolism

Urban ecosystems exhibit complex metabolic dynamics in regional carbon cycles, yet high-resolution quantification of urban carbon metabolism remains challenging in rapidly urbanizing regions. This study addresses this limitation by developing an unprecedented 250 m resolution modeling framework that combines the ecological CASA model with a soil respiration model, applied to the Pearl River Delta region of China. Using remote sensing data and ERA5-Land meteorological inputs, we quantified monthly net ecosystem productivity (NEP) for 2021-2022 across all cities, revealing significant spatiotemporal heterogeneity (-40 to 638 gC·m-2·a-1). Our findings indicate that forest ecosystems contribute 55.1% of regional net carbon absorption, while core cities like Dongguan and Zhongshan become net carbon sources when soil respiration exceeds vegetation productivity. The model outperforms traditional MODIS products in terms of resolution and spatial variability details (R2 = 0.79) and provides valuable insights for urban carbon neutrality strategies. The findings highlight the importance of maintaining green infrastructure and implementing targeted carbon management policies in megacity regions.