The impacts of climate change and global warming have become more visible globally because of the increasing frequency of extreme weather events, abnormal heatwaves, and other climate crises. Besides the traditional survey method, it is benecial to automatically distillate climate change opinions from social platforms to measure public reactions quickly. We investigate how to organize climate change opinions on Twitter into meaningful categories to support perspective summarizing tasks. We nd that merely using the available taxonomy for this task is ineffective; hence we must consider the entire text content. We recommend ve high-level categories (Root cause, Impact, Mitigation, Politics or Policy, Others) and assemble ClimateTweets, a dataset with category and polarity labels. In addition, we construct category classication and polarity detection tasks with a range of opinion mining baselines. The experimental results show that both tasks are challenging for existing models. We release the ClimateTweets dataset to facilitate investigation in public opinion mining using text content and articial intelligent methods. We hope this study could pave the way for future studies in the climate change domain.