Variability in local weather patterns has long been suggested as a major barrier impeding laypeople from recognizing long-term climate trends. However, as humans are able to detect and interpret rapid signal fluctuations, it seems psychologically plausible to assume that they are able to integrate short-term variations of weather variables into their mental representations of climate change. Using a combined analysis of social media and weather station data, here we investigated the impact of the short-term volatility of local temperature on climate change-related tweets from 2014 to 2017. We found a nonlinear hockey stick relationship between weekly temperature volatility and climate change-related tweets, a volatility rise of 1 °C corresponds to an 82% increase in climate change tweets when volatility is above 3.5 °C. This volatility effect was observed from 2016 onwards, suggesting a recent change in people’s mental representations of climate change. This study provides empirical evidence illustrating that in the public mind, climate change may not be represented as a mere temperature increase any more, but as a disruption of the climate system in general.

长期以来，人们一直认为本地天气模式的变化是阻碍外行人认识长期气候趋势的主要障碍。但是，由于人类能够检测和解释快速的信号波动，因此从心理上看似合理的假设是，他们能够将天气变量的短期变化纳入他们对气候变化的心理表示中。使用社交媒体和气象站数据的组合分析，在这里我们调查了局部温度的短期波动对2014年至2017年与气候变化相关的推文的影响。我们发现每周温度波动与气候之间存在非线性曲棍球棒关系。与变化相关的推文，当波动性高于3.5°C时，波动性上升1°C对应于气候变化推文增加82％。从2016年开始观察到这种波动性影响，表明人们对气候变化的心理表述最近发生了变化。这项研究提供了经验证据，表明在公众心目中，气候变化可能不再仅仅表示温度升高，而是总体上破坏了气候系统。