Temperature Influences Mood: Evidence from 11 Years of Baidu Index Data in Chinese Provincial Capitals

Mengjiao Yin^*Mengmeng Zhu

• Taihu University of Wuxi, Wuxi, China

This study innovatively integrates big data from climate metrics and search engine user behavior to quantify the dynamic relationship between temperature changes and public negative emotions (depression, anxiety, loneliness), while constructing a "climate zone × season" framework to identify regional sensitivity scenarios. It also uses search index as an indicator instead of subjective reports to avoid social desirability bias. Utilizing panel fixed-effect models and heterogeneity analysis based on Baidu Index and meteorological data from 31 provincial capital cities in China from 2013 to 2023, it was found that for every 1°C increase in daily mean temperature (DMT), the search indices for depression, anxiety, and loneliness significantly increased by 22.71%, 18.76%, and 19.59% respectively (p<0.01). Conversely, for every 1°C increase in daily temperature range (DTR), these indices decreased by 30.35%, 31.19%, and 15.41% respectively (p<0.05). Threshold regression analysis indicated that the adverse effect of high temperatures on loneliness became insignificant when DTR exceeded 14°C. Heterogeneity analysis across regions revealed significant variations in the impact of temperature on negative emotions during specific scenarios such as cold seasons in severe cold zones and hot seasons in warm summer-cold winter zones, with the moderating effect of climate on mental health being highly dependent on regional climatic characteristics, seasonal patterns, and interactions with temperature metrics. This study provides empirical evidence for climate policy-making and mental health interventions based on objective data.