[Purposes]Aiming at the limitation that existing assessment methods rely on static data and can hardly reflect the "tidal effect" of population flow, this paper proposes a dynamic ground risk map construction method based on time-varying coefficients of functional zones. [Methods] Firstly, a multi-dimensional risk quantification model covering crash casualties, dynamic noise and property damage is constructed. Secondly, POI data is adopted to identify urban functional zones, and a spatiotemporal mapping function is established combined with typical crowd travel characteristics, so as to realize low-cost and high-precision characterization of urban dynamic population distribution. On this basis, integrating the building sheltering effect and the secondary injury risk in special areas, a dynamic ground risk map with high temporal resolution is generated. Finally, simulation verification is carried out taking the typical urban area of Jiangning District, Nanjing as the case study.[Findings] This method can accurately capture the spatiotemporal evolution characteristics of risks in different functional zones, and effectively identify high-risk periods and areas. [Conclusions] This method effectively overcomes the limitations of static data, improves the accuracy of risk assessment, and can provide scientific decision support for refined urban UAV route planning and operation management in urban areas.