Hourglass in the snow

ITEM: NASA data indicate that 18 of the 19 hottest years on record have been observed since 2000.

ITEM: January 2019’s polar vortex broke or tied 340 Midwestern records for extreme cold in just two days – January 30 and 31 – as temperatures plunged to 30 below, 40 below, and even colder.

ITEM: Earth as of the start of April 2019 had experienced 411 consecutive months warmer than the global average, so no individual born after 1984 has experienced during his or her lifetime a month considered cooler than the historical norm.

All of which, and much more, raises some interesting questions:

  • If extremes keep recurring – as they seem to be doing – will people at some point simply stop paying attention?
  • As climate change makes these events more common, is there a point where people just stop talking about them? Pretty much accepting them (beware – cliché coming) as one more “new normal”?

Frances Moore, assistant professor of environmental science and policy at the University of California, Davis, and colleagues set out to answer these questions using what may seem like a fairly unconventional tool: Twitter. While this social media tool may seem ideal for keeping up with pop culture shenanigans, it’s become an increasingly important research tool for studying people’s perceptions about a wide array of issues and topics, including climate change.

Moore’s study, published in PNAS in February, found people base their perceptions of normal weather patterns on a two- to eight-year span of time – a short timeframe that can distort perceptions of the effects of climate change. Moore and her colleagues expressed concern about the “boiling frog” effect – accepting as normal a gradually more hostile environment rather than choosing to correct this environment – one they would have avoided if they knew what was truly happening.

“What’s worrying about this is that the constant rate of adjustment, this rate of normalization of two to eight years, that’s pretty quick compared to the rate at which climate change happens,” Moore says. “So the risk is that if you’re forgetting what happened before eight years ago, then you’re never really going to be able to put the weather that you experience into this longer-term context that really describes the overall effects of climate change.”

Moore and fellow researchers analyzed more than two-billion tweets from the continental U.S. posted between March 2014 and November 2016. They extracted weather-related tweets that contained certain words (such as rain, snow, windy, and chill, among others) from a Twitter data set, analyzing the “remarkability of temperature” by using the “bag-of-words” classification.

They found people would tweet about unusual weather conditions like extreme heat or cold events, but the more often these events occurred, the fewer the number of people who posted about them, suggesting they had begun to accept them as normal.

“What we show is that, if you have unusual temperatures and this is the first you’ve ever experienced it, that generates a big change on Twitter and people are talking about it a lot,” Moore says. “But if you have that same change … two years in a row, then people begin to stop talking about it. And if you have that same change eight years in a row, then people completely stop talking about it. So what that implies is that people’s idea of normal has shifted from what it used to be to this new state that’s defined by what happened two to eight years ago. And so we’re estimating this is kind of what people think of as normal just based on the rates at which they stop tweeting about unusual temperatures when they get them repeatedly year after year.”

The best-case scenario for the lack of conversation on Twitter about extreme weather could mean people are finding ways to adapt to extreme conditions, so the weather isn’t affecting them as much. However, researchers don’t believe people have stopped talking about the weather as a result of their personal adaptations. They say instead that people are still negatively impacted by extreme weather patterns even if they don’t mention them specifically in their tweets. Why do they believe this? Because people are grumpy when extreme weather events occur, and extreme temperatures – whether hot or cold – make people more likely to post negative tweets.

The research team used “sentiment analysis” techniques to examine each word and assign it a score based on positive or negative associations. They found extreme temperatures can cause negative sentiments, even if the extreme conditions are repeated over and over again.

“Even when you stop talking about them, it seems like these very unusual temperatures are still having a negative effect on your well-being,” Moore says. “It’s not the sentiment of tweets about weather, so this is not people saying ‘wow, I really hate this cold weather.’ This is the sentiment when they’re not talking about weather, so it’s kind of all the other tweets, they’re kind of talking about all these other different subjects but, because it’s hot or because it’s cold outside, it’s making people just a little bit more miserable, and that’s reflected in the kind of language they choose to use.”

Researchers use social media to learn about climate change beliefs.

Moore’s recent study builds upon previous research about social media and climate change. In 2015, a PLOS ONE study analyzed tweets from September 2008 to July 2014 that used the word “climate.” The researchers found Twitter to be a valuable tool for sharing climate change information, and they wrote in the paper that “We find that natural disasters, climate bills, and oil-drilling can contribute to a decrease in happiness while climate rallies, a book release, and a green ideas contest can contribute to an increase in happiness.” They found climate change advocates were more likely to use the words than deniers.

Value in ‘bipartisan communications networks’

Last September, sociologist Damon Centola, professor of communication at the University of Pennsylvania, and colleagues published a study in PNAS entitled “Social learning and partisan bias in the interpretation of climate trends.” The researchers found “bipartisan communications networks” created the best environment for people throughout the political spectrum to interpret data about climate change. When they added political logos to the same bipartisan information, people became polarized in their beliefs.

“What’s most interesting about Twitter is the phenomenal wealth, the volume of data that are produced, and that people are unselfconscious when they’re producing it,” Centola says. “It’s difficult to give people a questionnaire and ask them about their opinions on climate change, but if you can just sort of tap into their natural discussions about these kinds of topics, it gives you a lens into how people are thinking and talking normally.”

He points out Twitter can capture people’s conversations among peers – something researchers would have a very hard time observing otherwise. “We talk about the weather with our peers and we sort of remark on things and don’t remark on things. And as it becomes more and more common to talk about the weather a certain way, then we sort of accept it because everyone else accepts it, and Twitter can be a medium to see these social norms changing,” Centola says.

However, he also mentions that using Twitter as a research tool does come with limitations. “We only see the networks that are in Twitter, we don’t see all the other conversations that are happening outside of Twitter,” Centola says.

While some scientists continue to study how the planet is changing, others will analyze how people react to these changes, what they believe about them, and how they may adapt to them. And social media is just one tool that can help them learn more.

Kristen Pope

Kristen Pope

Kristen Pope is a freelance writer and editor who frequently writes about climate change, ecology, wildlife, conservation, and many other topics for a wide variety of publications. She has a masters degree...