It Takes Just a Few Years For Extreme Weather to Start Seeming Normal, Study Finds

Just a normal day during heat wave.
Just a normal day during heat wave.
Photo: Getty

Our response to climate change is often compared to the hypothetical boiling frog scenario, where a frog sits in a pot of water as the heat is slowly turned up until it’s boiled to death. Unfortunately, while frogs really don’t like being boiled at all, a new study on responses to extreme weather suggests the metaphor may be just about perfect for humans.


Research published on Monday in the Proceedings of the National Academy of Sciences looked at people’s tweets during historically hot or cold weather to see how they responded. It took tweeters just five years to normalize once-extreme temperatures, indicating that the slow background burn of climate change could just become the same old, same old.

Climate change is making our weather increasingly violent, which is likely helping drive an uptick in the number of Americans concerned about global warming. The link between extreme weather and climate is perhaps clearest when it comes to heat waves, which is what led researchers at the University of California, Davis to look at how people reacted to hot weather. After all, if people are sufficiently freaked out by every heat wave, it stands to reason they would be more inclined to advocate for policies to address climate change.

“It’s a really interesting question of how climate change is experienced by people as the climate shifts,” Frances Moore, the study lead author who researches economic and social impacts of climate change, told Earther “We realized that the Twitter dataset could be a great way to measure this and how quickly ideas people’s ideas of weather change.”

Moore and her colleagues whittled down the 2.18 billion tweets sent in the U.S. between March 2014 and November 2016 (what are we doing with our lives, people?) to a set of 6,000 selectively sampled tweets about the weather aggregated to the county level. They also performed a “sentiment analysis” of all tweets that didn’t mention the weather, scoring them on scale that ran from positive to negative. They then compared the Twitter data with the temperature anomaly during the week the tweets were sent.

The findings show that people tweet more about the weather as the temperature veers into hot as balls or colder than Mars territory. The tweets reveal an asymmetrical breakdown, where people tweet more about the weather when cold weeks are colder than normal and hot weeks are hotter than normal. In other words, an unusually cool summer day doesn’t elicit as much of a reaction as an excessively hot one. But the results also show that the novelty of extremes can wear off as we become used to them. If a county had experienced an average of five years in a row of abnormally hot weather, then weather tweets began to fall away as people grew accustomed to the new normal (the same was true for cold years).

But here’s the weird part. Looking at the bigger dataset of sentiment analysis shows that even when people weren’t tweeting about the weather, their tweets got more negative as temperatures got more extreme.


“That [effect] exists even if people stop talking about it,” Moore said.

Basically, the results suggest extreme temperatures are affecting peoples’ moods even if they stop talking about the weather, which matters in a world where background temperatures are incrementally rising. It’s possible people could be forgetting what normal used to be, while still being perturbed by the uncomfortable new weather. Moore said that helping folks put their weather experience “in a longer term context is going be one of the challenges in climate communications.”


Yotam Ophir, a postdoctoral researcher in science communication at the Annenberg Public Policy Center at the University of Pennsylvania, told Earther this approach represented an “important step, but with limitations.” Even with a huge dataset, he said Twitter users still tend to be “young, highly educated, and urban. So what we can learn from a study such as this one is how people who use Twitter talk about weather on Twitter.”

Nevertheless, Ophir went on, “this study emphasizes that the question of how to communicate scientific information in ways that will break psychological barriers such as declining remarkability or politicization of science is now the most urgent one we face as science communicators.”


Moore said future research could focus on other climate impacts, like how people react to flooding from rising seas. Boiling is, after all, hardly the only challenge humanity faces.

Managing editor at Earther, writing about climate change, environmental justice, and, occasionally, my cat.


Yotam Ophir, PhD

Following some comments online, allow me to clarify what I (quoted in the article- Yotam Ophir) meant by saying we should “break psychological barriers”, as the topic is complex and my space in the article comment was very limited. Experts and the public have very different perceptions re what is considered risk and how risk should be evaluated. This is completely reasonable and should be seen as the result of communicating a complex and uncertain knowledge that is hard to comprehend, even if you’re an expert, but definitely without the proper education, expertise, and knowledge. The particular needs and limitations of audiences of different background should be respected when communicating science, and by saying “break barriers” I do not imply that one’s values and beliefs are not legitimate. On the contrary... My belief as a science communication researcher is that the solution is science education— giving people the tools they need to comprehend how science works and how it differs from other forms of knowledge. What I do mean is that we should accept that humans are limited in their capacity to process complex information, as well as often biased in doing so. For example, If one’s inclination is to evaluate trends in climate by thinking about the local weather in her neighborhood, then yes - this is a barrier- familiarity effect- we should attempt to break. If one tends to accept sensational news coverage (or depictions in fictional movies such as 2012) over less-exciting-though-more accurate- scientific evidence, thus remembering extremely cold and snowy days over trends in climate over time-As Donald Trump did in recent tweets and is known as the Local Weather Effect- then yes, let’s think of ways to break these psychological limitations. If people struggle with graphs and numbers but easily accept good vivid stories - we should find a way around it. Finally, and most worrisome these days, if one knows the facts but reject it since Alex Jones said it’s a hoax - then tear down that psychological wall as well. -Yotam Ophir, PhD, The Annenberg Public Policy Center at the University of Pennsylvania