["Transactions in GIS, Volume 30, Issue 5, August 2026. ", "\nABSTRACT\nUnderstanding urban residential spatial patterns is vital for resource allocation and planning. Current approaches often overlook demographic shifts, reducing projection accuracy. To address this, we propose a novel framework that integrates an improved cohort‐component model with the PLUS model for multi‐scenario population simulation. This integration allows the dynamic coupling of demographic forecasts with land‐use change. Our method incorporates policy‐sensitive parameters, simulates land‐use transitions via random forest under three development scenarios—natural growth, cropland protection, and urban expansion—and downscales population using multi‐source POI and geospatial data. Applied to Zhengzhou City at a 30‐m resolution, our model demonstrates higher predictive accuracy, evidenced by a lower mean squared error compared to the WorldPop and LandScan datasets, and achieves a substantially better fit (R2 = 0.76) with the Seventh National Population Census of China data. These results validate the robustness of this policy‐sensitive approach for fine‐scale spatial population modeling.\n"]