Understanding reforestation’s role is essential for forest ecosystem health, maintaining carbon cycle balance, and addressing climate change. However, the contribution of reforestation to aboveground carbon (AGC) accumulation and its contribution to climate change mitigation in urban landscapes remains poorly understood. This gap limits the development of effective strategies to maximise their carbon accumulation potential. Therefore, this study aims to understand the past, present and future aboveground carbon stocks in reforested urban landscapes using forest inventory observation data, combined with land use, topographic and global climate circulation model data (CMIP6). An artificial neural network regression model was selected to construct the spatial relationship between AGC and the ancillary datasets. Analyzing the spatio-temporal dynamics of aboveground carbon (AGC) from 2012 to 2040, this study employed a resolution of 1 km and a 20-year interval, focusing on three distinct climate scenarios outlined in CMIP6. The cross-validation outcomes substantiate the reliability of AGC predictions in reforested urban landscapes, albeit acknowledging uncertainties in specific study site sections. The mean AGC values recorded over the past three decades were observed at 55.94 and 84.02 t. ha−1. In contrast to historical AGC (2012–2022), future AGC exhibited notable fluctuations under diverse climate scenarios. Specifically, AGC showcased a gradual increasing trend under SSP1-2.6 low emissions, a diminishing trajectory under SSP2-4.5, and a markedly rapid decline under SSP5-8.5. This work enhances the management of urban forest carbon dynamics to support informed climate change mitigation policies and sustainability strategies.