In　times　of　crisis,　social　media　emerges　as　a　pivotal　channel　for　government-public　communication.　This　study　delves　into　the　dynamics　of　a　temporal　weighted　network　on　Weibo,　analyzing　39,567,227　reposted　data　points　related　to　the　COVID-19　pandemic,　specifically　during　its　critical　early　stage.　We　examine　the　information　dissemination　roles　of　eight　distinct　user　types:　state-run　news　media,　non-state-run　news　media,　influential　we　media,　normal　we　media,　celebrities,　organizations,　governmental　departments,　and　normal　users.　This　is　achieved　through　the　application　of　three　centrality　metrics:　temporal　weighted　degree　centrality　(TWDC),　temporal　weighted　closeness　centrality　(TWCC),　and　temporal　weighted　betweenness　centrality　(TWBC),　focusing　on　three　cases　(Top　10,　Top　20,　and　Top　50)　based　on　ranking　results.　Our　findings　reveal　that　while　state-run　news　media　were　prolific　in　information　publishing,　their　roles　as　central　and　mediator　agents　in　information　spreading　were　comparatively　weaker　than　those　of　non-state-run　news　media　and　we　media,　especially　influential　we　media.　This　pattern　underscores　potential　risks　in　false　information　propagation　and　highlights　gaps　in　the　government＇s　information　dissemination　strategy　during　the　early　stages　of　the　pandemic.　The　study　offers　targeted　theoretical　and　practical　implications,　suggesting　how　governments　might　more　effectively　leverage　social　media　for　crisis　communication,　thereby　enhancing　their　role　in　combating　misinformation　and　guiding　public　perception.　These　insights　are　crucial　for　framing　more　responsive　and　adaptive　communication　strategies　in　the　initial　phases　of　a　crisis,　providing　a　global　blueprint　for　effective　information　management　in　public　health　emergencies.　©　2024　Elsevier　Ltd