With rapid urbanization, urban public transport plays an increasingly critical role in improving service quality and travel experience. However, emergencies such as extreme weather and equipment failures often force metro and bus systems to adopt emergency measures including long/short routing adjustment and skip-stop operation. These measures disturb passenger travel and impair the structural and functional stability of the transit network. Therefore, scientifically evaluating the impacts of stations on network robustness under different failure scenarios and identifying critical stations are of great significance for scenario-based risk management. Based on network science theory, this study takes the urban bus-metro composite network as the research object considering research complexity. It analyzes node importance from the perspectives of topological structure and passenger flow characteristics, evaluates the robustness of the composite network, and constructs the Multimodal Station Impact (MSI) model to comprehensively quantify and classify node importance. The results indicate that single-indicator analysis based on either topology or passenger flow cannot fully assess node importance and network robustness. The MSI model quantifies the comprehensive influence of urban public transport nodes, and mitigates the bias caused by single-indicator identification to a certain extent.