Abstract: Using Baidu index and meteorological data, we analyze the network attention of Shanghai meteorological information with “weather in Shanghai”, “high-temperature in Shanghai” and “typhoon in Shanghai” as search keywords. The results show that the network attention of “weather in Shanghai” shows an overall upward trend from 2014 to 2016, with no significant variation from 2017 to 2021; The attention is mostly high in spring and summer, while low in early autumn and late winter; The weekly variation of the network attention is not obvious, with Monday experiencing relatively higher attention; In case of legal holidays, the attention peaks 2 to 4 days prior to the holiday, the change characteristics differ from each other within the holiday, and the high attention remains within 3 days after the holiday. The network attention of “high-temperature in Shanghai” shows the consistence of the variation trends between that of the public attention and of the daily maximum temperature; The high attention goes to the first 40 ℃ and the continuous high-temperature weather, with the start of the attention as early as the late June and the early July. The network attention of “typhoon in Shanghai” shows the consistence of the variation trends between that of the public attention and of the daily maximum wind speed and the daily precipitation; The start time of the attention matches with the start time of the typhoon; The attention of the typhoon increases gradually and rises significantly around 2 days prior to the strike of the typhoon itself. Finally, suggestions on how meteorological authorities can optimize public weather services, establish a meteorological public opinion database, and utilize new media for data dissemination are propose in hopes of providing a reference for meteorological services.