The construction of an emergency online public opinion event graph, centered on the public opinion object, aims to illustrate the correlations between the public opinion object and public opinion ontology. This supports the monitoring and governance of emergency online public opinion, focusing on risk prevention, mitigation, and emergency management. Firstly, considering the diversity and ambiguity of online public opinion information, we design the structure of an online public opinion event graph for emergencies. This structure is divided into the public opinion subject layer, public opinion ontology layer, and public opinion object layer, and we propose the construction process and method for the online public opinion event graph. Secondly, we explore the application scenarios of the online public opinion event graph and introduce a series of online public opinion evolution analysis methods based on the event graph. Finally, an empirical study is conducted on the Zhuozhou rainstorm of 2023, constructing the online public opinion event graph for the Zhuozhou rainstorm disaster and analyzing the evolution process of online public opinion using the event graph to delineate the development trends of key events. The study demonstrates that the online public opinion event graph effectively describes the subjects, ontology, and objects of public opinion and their interconnections, presents the evolution trend of online public opinion, identifies key nodes of risk construction, and reveals secondary and derivative risks of key events. This provides new perspectives and reference bases for the governance of online public opinion in the context of emergency management.