Informing the Boiling Frog: Binary climate data amplifies climate change perception

Climate change is causing real harm to our communities, but its impacts often appear as slow, gradual shifts in daily weather patterns. This subtlety creates a psychological barrier to action—as we adapt to incremental changes, each new "normal" dulls our perception of the crisis [1,2].

Our research group has been investigating the cognitive underpinnings of climate inaction for many years, exploring how people process climate information and why knowledge often fails to translate into action. This project emerged from two parallel explorations.

In the summer of 2020, we came across David Roberts' thought-provoking piece on shifting baseline syndrome, which articulated what may be the most insidious aspect of climate change: not that we'll wake up too late, but that we might never truly wake up at all. If humans continuously reset their perception of “normal” every few years, climate change might proceed as mere background noise despite mounting damages.

Simultaneously, Grace—then an undergraduate student at Princeton University—was undertaking an ambitious research project to understand climate impacts in our local community, Princeton, New Jersey. Although Princeton doesn't directly experience dramatic climate impacts such as floods, wildfires, or droughts, she realized that the town had been indeed impacted: the local lake (Lake Carnegie) hadn’t frozen over for a series of years in a row, preventing community members from ice skating or playing hockey on the lake. Through conversations with long-term residents and digging into research in Princeton newspaper archives, Grace discovered that there had been a stark decline in ice skating on the lake over the past century, and a sentiment of loss surrounding this decline. The loss of this annual community tradition provided a clear and personally relevant climate signal for residents who might otherwise overlook the gradual impacts of climate change. 

When our research paths crossed, we wondered: Could there be something special about binary climate data? Specifically, could binary climate data like lake freezing patterns be uniquely effective at countering the "boiling frog'' effect of climate apathy? 

We conducted a large-scale experiment with over 700 participants in which we presented the participant with either a plot of winter temperature history or a plot of lake freeze history for a fictional town. The data for both scenarios were generated to have matched correlation values. We then asked the participants questions to rate how much they think the town has been impacted by climate change, how much the temperature has changed, and how much the lake freeze frequency has changed.

The results of the experiment showed that participants who viewed the lake freeze climate visual tended to perceive greater climate change impact compared to participants who viewed the temperature climate visual. They also perceived a greater change in freeze frequency and, counterintuitively, a greater change in temperature. We found that these results remained consistent in a similar experimental setup with real-world lake freeze and temperature data.

We then asked the question: what might be causing people to perceive a greater climate impact in binary data? While multiple reasons, such as emotional connection and decreased mental effort, could play a role, we posited that binary data may be further heightening perceptions of climate change by creating an “illusion” of sudden shifts in the data. 

We conducted an experiment with ~400 participants in which the participant was again shown either a binary climate visual or a continuous climate visual, and we asked if they perceived a noticeable shift in the data, referred to as a changepoint. We also asked them to choose the year that they notice the biggest shift in the data pattern. We found that participants in the binary condition were more likely to perceive a changepoint compared to those in the continuous condition. Additionally, the participants who viewed the binary data showed more agreement on the location of the changepoint on the timeline. These findings suggest that the “illusion” of sudden shifts in binary data could be an underlying reason for elevated change perception.

Our findings indicate that presenting climate data in binary terms (yes/no, present/absent) makes climate change impacts more noticeable compared to continuous data (gradual increases or decreases). Gradual shifts in binary data creates an illusion of sudden changes, increasing the saliency of the climate trend. These results have immediate practical uses, especially for designing and visualizing climate data. For example, policymakers and journalists could use binary visuals to simplify and concretize complex climate data. Instead of showing gradual temperature increases, they could highlight clear shifts, such as the disappearance of white Christmases, the inability to ice-skate in winter, or summer activities disrupted by wildfire smoke. This approach could also help communities understand local climate changes, such as the decrease of lake freeze on Lake Carnegie in Princeton. In areas experiencing shifting rainfall patterns or more frequent droughts, binary data would highlight the worsening conditions by showing the shift from occasional to recurring severe droughts, helping communities recognize how once-rare events are becoming common.

This widespread inability to perceive gradual climate trends is often referred to as the “boiling frog” effect. Like the proverbial boiling frog who fails to notice gradually heating water, we fail to notice the danger of climate change until it's too late. While our study shows how we can make climate change more noticeable – a crucial first step toward meaningful engagement – future work should investigate how this work can be extended to drive concrete action.

References

1. Moore, Frances C., et al. "Rapidly declining remarkability of temperature anomalies may obscure public perception of climate change." Proceedings of the National Academy of Sciences 116.11 (2019): 4905-4910.
2. Marlon, Jennifer R., et al. "Detecting local environmental change: the role of experience in shaping risk judgments about global warming." Journal of Risk Research 22.7 (2019): 936-950.