This study aims to estimate the fatality risk of Unmanned Aerial System (UAS) operations from a population perspective using high-resolution de facto population data. In doing so, it provides more practical risk values compared to the risk values derived from the residential population data. We then set restricted airspace using the risk values and the acceptable level of safety. We regard the restricted airspace as airspace being blocked by a dynamic obstacle. Scenario analysis on the study area in Seoul, South Korea presents a richer set of results for time-dependent differences in restricted airspace. Especially during the daytime, the restricted airspace is clustered around commercial and business areas. The difference between restricting airspace based on residential population-derived risk and restricting airspace based on de facto population-derived risk is also observed. The findings confirm the importance of accurately taking into account population density when assessing and mitigating the risk of UAS operations. Sensitivity analysis also reveals the importance of an accurate estimate of population density when estimating the risk.