Social media posts reveal how communities cope with extreme heat

As extreme heat becomes more frequent and intense across California, understanding which communities are struggling and what kind of help they need is crucial for safeguarding public health. A recent study from UC Irvine suggests that social media may provide valuable insights into these challenges. Researchers analyzed 24,598 heat-related tweets from across the state between 2016 and 2022.

They discovered that factors such as location, mode of transportation, and available resources influenced how people talked about and coped with extreme heat. The findings, published in Environmental Research Communications, indicate that social media could serve as a supplementary tool for gauging community experiences and tailoring public health interventions.

Nicolas Rodriguez Garcia, a researcher involved in the study, notes that social media offers a unique perspective on how people experience heat in their daily lives. By understanding these experiences, officials can design more effective heat-safety messages and address gaps in support for vulnerable populations. Extreme heat poses a growing threat to public health, and factors like housing, income, occupation, and even type of transportation are known predictors of vulnerability.

Instead of focusing solely on temperature or heat-related illnesses, the UC Irvine team examined how people described the impact of heat on their daily lives and their coping mechanisms. They found that individuals living in sensitive housing—such as RVs, vans, or boats—were more likely to post about coping strategies like staying hydrated, using air conditioning, or altering exercise routines.

These residents may benefit from targeted interventions to reduce indoor heat exposure. The study categorized posts into two main themes: coping strategies and life impacts. Coping strategies included measures like staying hydrated, using air conditioning, or adjusting exercise plans. Life impacts encompassed health effects, work disruptions, canceled activities, concerns about pets or plants, changes in household expenses, and other consequences of high temperatures.

When comparing these patterns to county-level indicators related to income, housing, transportation, and tree canopy, the researchers found that housing was linked to posts about coping strategies, while active commuting was associated with posts about life impacts. Additionally, using a large language model, the team grouped thousands of social media posts based on how people discussed heat.

The model's accuracy in categorizing posts ranged from 74% to 90% for different categories. The study revealed notable differences based on housing. Counties with a high concentration of RV, van, and boat households were 15% less likely to post positive tweets about heat and 22% less likely to discuss coping strategies compared to counties with lower levels of such housing.

This finding is significant because these types of homes often have poor insulation, making them prone to overheating during extreme temperatures. Transportation patterns also revealed vulnerabilities. Counties with higher percentages of people using walking, biking, or public transportation showed a greater likelihood of negative heat-related posts and descriptions of heat's effects on daily life.

Active commuters are more directly exposed to heat while traveling, highlighting the importance of transportation infrastructure in protecting communities from rising temperatures. The researchers also noted disparities among counties with larger Hispanic populations. These areas showed a slightly higher likelihood of posts describing the impacts of heat on daily life but a lower likelihood of posts about coping strategies.

While social media patterns should not be considered direct measures of population-level vulnerability, the study suggests that social media can offer valuable insights to complement existing public health surveillance. By combining signals from social media with data on housing, transportation, and socioeconomic conditions, public health agencies could identify where specific types of heat communication are most needed.

For instance, communities with sensitive housing might benefit from messaging focused on keeping indoor spaces safe, while areas with many active commuters could benefit from heat-conscious transportation planning and interventions to reduce exposure during active travel. Ultimately, the researchers see social media as a complementary tool, not a replacement, for traditional public health surveillance.

By integrating social media data with other measures, officials can better understand how extreme heat affects people as events unfold and tailor interventions accordingly.