Disaster(natural and man-made)severity assessment plays a crucial role in effective disaster classification and management. This article introduces a methodology for assessing disaster severity using various indicators such as the relative value of direct economic losses,casualties,injuries(=serious+minor),and relocation indicators(=number of resettlement population×Average resettlement time). A unified natural disaster magnitude calculation model is proposed and demonstrated using specific examples of earthquakes,floods,and typhoons. Furthermore,the model is extended to assess the severity of man-made disasters,particularly production safety accidents. The article discusses the limitations of current severity classification standards for such accidents and presents a tailored severity classification scheme,magnitude calculation model,and workflow. The proposed model is rigorously derived mathematically and offers a unified framework controlled by the highest loss indicator while considering contributions from other indicators. Moreover,it allows for scalable and comprehensive disaster severity assessment across different disaster types. Suggestions are provided to enhance the statistics of injuries and resettlement populations for natural disasters,ensuring a more accurate reflection of the disaster's impact and scope.