Al-Ahdal Tareq, Barman Sandra, Dafka Stella, Alahmad Barrak, Bärnighausen Till, Gertz Michael, Rocklöv Joacim
Heidelberg Institute of Global Health (HIGH), Heidelberg University Hospital, Heidelberg University, Heidelberg, Germany.
Interdisciplinar Center for Scientific Computing (IWR), Heidelberg University, Heidelberg, Germany.
iScience. 2025 Feb 7;28(3):111966. doi: 10.1016/j.isci.2025.111966. eCollection 2025 Mar 21.
Expressions in social media can provide a rapid insight into people's reactions to events, such as periods of climatic stress. This study explored the link between climatic stressors and negative sentiment on the X platform in Germany to inform climate-related health policies and interventions. Natural language processing was used to standardize the text, and a comprehensive approach for sentiment analysis was utilized. We then conducted spatiotemporal modeling fitted using integrated nested laplace approximation (INLA). Our findings indicate that higher and lower level of temperature and precipitation is correlated with an increase and decrease in the relative risk of negative sentiments, respectively. The findings of this study illustrate that human sentiment of distress in social media varies with space and time about exposure to climate stressors. This emotional indicator of human exposure and responses to climate stress indicates potential physical and mental health impacts among the affected populations.
社交媒体中的表达能够快速洞察人们对诸如气候压力时期等事件的反应。本研究探讨了德国X平台上气候压力源与负面情绪之间的联系,以为与气候相关的健康政策和干预措施提供信息。使用自然语言处理对文本进行标准化,并采用了一种全面的情感分析方法。然后,我们进行了时空建模,采用集成嵌套拉普拉斯近似法(INLA)进行拟合。我们的研究结果表明,较高和较低水平的温度及降水量分别与负面情绪相对风险的增加和降低相关。本研究结果表明,社交媒体中人类的痛苦情绪会因接触气候压力源的时空变化而有所不同。这种人类接触和应对气候压力的情绪指标表明,受影响人群可能会出现身心健康问题。