Characterizing climate change sentiments in Alaska on social media

Area of Science:

• Environmental Social Science
• Computational Social Science
• Climate Change Communication

Background:

• Global concern over climate change impacts is high, yet public perception varies significantly based on local contexts.
• Understanding regional differences in climate change awareness is crucial for effective policy and communication.

Purpose of the Study:

• To investigate public perceptions of climate change in Alaska using Twitter data.
• To explore the relationship between local socioeconomic and environmental factors and climate change sentiments.

Main Methods:

• Utilized geo-located tweets from 2014-2017, identifying climate-related content with a deep learning framework.
• Applied lexicon-based sentiment analysis and extracted eight core emotions from tweets.
• Employed regression models to assess the influence of regional attributes on climate change sentiments at the census tract level.

Main Results:

• Observed an overall upward trend in climate change sentiments on Alaska's Twitter, with peaks during summer months.
• Found that census tracts with higher Native Alaskan or American Indian populations expressed more negative sentiments.
• Identified road density as a significant factor, indicating more climate change discussion in areas with denser infrastructure.
• Discovered associations between specific emotions and climate change sentiments, such as inverse relationships for disgust/surprise and positive for sadness/trust.

Conclusions:

• Public awareness of climate change is dynamic and influenced by local socioeconomic and environmental conditions.
• Twitter data provides valuable insights into regional climate change perceptions.
• Tailoring climate change communication strategies to local contexts is essential for effective engagement.