This study aims to develop a framework for identifying local circulation-dominated pollution hotspots using large eddy simulations (LES). The top-down nitrogen oxides (NOx) emissions based on satellite data provide improved accuracy, enabling clearer identification of pollution hotspots. A comparison between the top-down NOx emissions and government emission inventories is conducted to highlight the significance of the new emission method in shaping pollution hotspots. The framework utilized the Taiwan Vector Vorticity equation cloud-resolving Model (TaiwanVVM), which better resolves turbulence structures in local circulations and boundary layer processes. Coupled with a chemical module, the formation and evolution of pollution hotspots at the local scale can be closely examined in the high-resolution LES of TaiwanVVM. The composite results show similar daytime pollution pattern even under different emission inventories. However, there is a significant difference in the nighttime pollution hotspots. Detailed analysis revealed that the lee vortices can influence hotspot locations and facilitate offshore pollution transport, effectively reducing onshore concentrations even when boundary layer heights are low at midnight. Through the case study, the differences in emission source inventories directly contribute to nighttime concentration differences, while variations in chemical processes also influenced the formation of pollution hotspots. This framework emphasizes the importance of satellite-based emission inventories in identifying new pollution hotspots and highlights the role of lee vortices in shaping these hotspots, providing valuable insights for emission regulation through semi-realistic simulations and potential adaptation to future climate scenarios.