Increased temperature-related newspaper coverage and more negative sentiment during hot and cold extremes

More frequent, intense, and prolonged temperature extremes due to climate change increase the risk of human morbidity and mortality. However, public perception of these risks is often low, while people's awareness is crucial to reducing the health impact of temperature extremes. News media plays a key role in raising awareness by providing essential information on heat waves and cold spells such as releasing warnings, sharing weather forecasts, and discussing impacts. Any sentiment conveyed within newspaper articles about temperature extremes, either through positive, negative, or neutral phrasing, can influence the audience's perception of risk and, potentially, their actions. Newspaper coverage of temperature extremes and the related sentiment may be influenced by multiple factors other than the actual temperature anomalies, such as political alignment or editorial decisions, but also potentially the countries’ vulnerability to climate change. In this study, we analyze and compare the sentiment of temperature-related newspaper articles in eight countries (Israel, Malaysia, New Zealand, Philippines, Singapour, Pakistan, South Africa, and the United Kingdom) with different climates and societal vulnerabilities to climate change (food, water, health, ecosystem services, human habitat, and infrastructure). We consider leading English language, national newspapers and use daily maximum temperature data from the day of each article’s publication from the ERA5 reanalysis. The sentiment is determined in an automated way based on the fraction of positive and negative words in text. In addition to the sentiment, we determine whether or not each article mentions climate change. We find clear differences during times of extreme temperatures versus times with near-normal temperatures in all countries. During days with comparatively cold or warm temperatures (i) more temperature-related articles are published, (ii) their sentiments are more negative, and (iii) climate change is mentioned less frequently. While the latter finding is surprising, it suggests that there are unobserved confounding factors that require further research, which might relate to other events and anomalies occurring simultaneously. A comparison of the results across countries shows more negative sentiment and fewer mentions of climate change in countries with higher climate change vulnerability. Being aware of these media reporting patterns of extreme temperatures may help media outlets reassess their role in aiding public health responses.