【Purpose/significance】 By exploring the distribution characteristics of social media topics and their differences in different

sentiment, this paper explores the characteristics and laws contained in the expression of opinions of social media users. The research helps formulate scientific content release and emotional guidance strategy, which can improve the effectiveness of risk communication management in emergency.【Method/process】 Based on the two dimensions of publication and users, this study designed topic distribu⁃tion indicators to describe the dispersion state of social media topics. Analysis of variance was used to test the significant difference of topic distribution characteristics in affective tendency, figure out the internal factors affecting the evolution of theme emotion.【Result/conclusion】 The experiment uses the daily posts of high-influence users on Sina Weibo as sample data, and the analysis found that:(1) By focused opinion output, users cultivate their influence in specific fields. (2) High-influence users tend to express clear emotion in their posts and the emotion evolutes over time. (3) The topic distribution popularity and breadth have significant differences in the positive and negative sentiment of posts, and are not affected by time factors.【Innovation/limitation】 This paper proposes a framework of topic distribution characteristics and discusses the differences between topic distribution and emotional tendencies. The research conclusions can provide references for formulating scientific risk communication strategies. In the future, prediction of opinions and emotion in a specific emergency can be realized by exploring the similarities and differences between users' regular publication and emergency-related publication. Thereby, assisting public opinion risk management in emergencies.