The interplay of news media, social media, and public search behavior during the 2022–2023 South Korea drought

Drought is a creeping phenomenon that aggravates adverse impacts over time, eventually altering emotion and social behavior. Web search engine, news media, and social media trace changes of public search behavior and online communication during the emergence of a drought. Understanding the interplay of drought characteristics with news media, social media, and public search behavior helps develop efficient drought preparedness and response plans, which are lacking. Here we assess changes of attention in web search engine and sentiment in mass/social media during the 2022–23 South Korea drought, harnessing Natural Language Processing and multiple digital tracing data. Results show that internet search activity volumes, daily news articles, tweets are highest under the nationwide drought conditions in June 2022. When the drought is intensified in March 2023, particularly in the Southwestern region, the local news outlets published news articles accompanied by increased internet search activities, but few Twitter/X posts. Our results indicate diverse social behavior patterns in web search engine, news outlets, and social media under the national and local drought conditions. Furthermore, news headlines show consistent emotion types (Expectancy, Anxiety, and Disappointment) during the 2022–23 drought. This study offers insights into how big data and AI can help design efficient drought preparedness and mitigation plans by analyzing interactions with news media, social media, and public information-seeking behavior.