U.S. Spanish speakers are “less likely than English speakers to receive, understand, and respond to tornado watches and warnings,” based on data from multiple years of surveys, finds a study in Bulletin of the American Meteorological Society.
The study incorporates data from 14,341 individuals who responded to the University of Oklahoma’s nationwide Severe Weather and Society Survey between 2018 and 2023. Participants were asked to distinguish between a tornado watch and a tornado warning, and to agree/disagree with statements such as, “Sometimes I miss tornado warnings that are issued for my area.”
The paper’s authors — América R. Gaviria Pabón, Makenzie J. Krocak, Joseph T. Ripberger, and Joseph E. Trujillo-Falcón — created overall metrics based on those responses to describe how well English- and Spanish-speaking populations were able to understand tornado warnings, their likelihood of receiving the warnings in the first place, and their likelihood of taking protective action in response. In all cases, there was a significant gap between Spanish and English speakers, with Spanish speakers overall falling slightly below average and English speakers falling slightly above average. Those differences remained significant even when controlling for other demographic factors like age and education, or for a person’s exposure to tornado warnings.
| Tornado Warning Reception | Tornado Warning Comprehension | Tornado Warning Response |
| English speakers (mean score): 0.01 | English speakers (mean score): 0.03 | English speakers (mean score): 0.02 |
| Spanish speakers (mean score): -0.24 | Spanish speakers (mean score): -0.29 | Spanish speakers (mean score): -0.16 |
For example, Spanish speakers were more likely than English speakers to report that they had missed tornado warnings in their area (37% of Spanish speakers vs. 29% of English speakers), had been unsure whether a tornado warning applied to their area (34% vs. 24%), had been too busy to respond to tornado warnings (21% vs. 13%), or had been unsure what to do in response (20% vs. 16%).
Comprehension of important tornado warning information was often low for both groups — for example, just over half of English speakers (55%), and only 38% of Spanish speakers, knew that a tornado warning indicates an imminent threat within 1 hour of issuance. 79% of English speakers and 65% of Spanish speakers correctly identified a description of a tornado warning.
“We often talk about access to information, but understanding that information is just as important. Cultural background, prior experience, and familiarity with severe weather all shape how people interpret and respond to risk messages.”
-América Gaviria Pabón
Author Q&A
We spoke with lead author América Gaviria Pabón to learn more about the paper and the implications of what the researchers found.
Are you seeing any trends in these data over the years?
Across all years, the biggest difference between English and Spanish speakers continues to be comprehension. Spanish speakers consistently showed lower comprehension scores, and that gap has remained about the same over time.
When it comes to receiving and responding to tornado warnings, the differences are smaller and fairly steady too. There’s a small sign that reception may have dipped slightly for Spanish speakers, but not enough to be certain. Overall, the results show that the language gap hasn’t really improved or worsened; it’s stayed about the same from year to year.
What are some factors that might be affecting Spanish speakers’ comprehension and response to U.S. tornado warnings?
Several factors make it harder for Spanish speakers to understand and respond to tornado warnings. Most weather risk information in the U.S. is still issued primarily in English. Yet, as of 2020, more than 40 million people in the U.S. reported speaking Spanish at home; about 61% said they spoke English “very well,” while 39% said they spoke it less than “very well” (Mast & Ruiz 2023).
We also know that risk perception isn’t universal. It’s shaped by local context, community history, cultural values, and past experiences (see, for example, Sutton et al. 2021; Morss et al. 2020; Mileti & Sorensen 1990; Burns & Slovic 2012). So when warnings aren’t linguistically or culturally resonant, Spanish speakers may have a harder time making quick, confident protective decisions (Aguirre 1988). Trujillo-Falcón et al. (2022, 2024) have also pointed out that inconsistent translations and a lack of bilingual resources remain major barriers.
We’ve seen this play out in real events. After the 2011 tornado outbreak in Alabama and Georgia, a survey of Spanish-speaking residents found that while many relied on Spanish-language TV and radio, they often didn’t know specific protective actions, like where to take shelter (Stewart, Capiello, & Ocampo 2014). That tells us the issue goes beyond simply having Spanish-language warnings:it’s about ensuring people truly understand, trust, and can act on them.
While resources have expanded considerably in the last couple of decades, it is important to note that a ton of rural communities have yet to have adopted multilingual stations. As previous studies have noted, especially in the Southeast United States, these also tend to be the most vulnerable to tornadoes.
Ultimately, language is more than just a tool for communication; it’s tied to trust and identity. Poorly translated or culturally disconnected warnings can easily be misunderstood or dismissed. Making warnings accessible in both Spanish and English isn’t just about inclusivity, it’s about building trust in institutions and supporting timely, life-saving action.
What are some limitations with the study that still need to be addressed?
One limitation is that we used an online survey. That made it possible to reach people across different regions, but it probably left out Spanish speakers with limited internet access or lower literacy (Brown 2015).
Survey participation also tends to be less common in many Spanish-speaking communities, so those who did respond may be more self-selective (De La Rosa et al. 2012). Because of that, our sample may underrepresent people who face the greatest challenges in receiving and understanding weather information.
Another limitation is that our data are self-reported. People’s responses about how they receive or react to tornado warnings might not always match what they’d actually do in a real situation. These responses are still valuable for understanding perceptions,confidence, and protective action intentions, but they don’t fully capture the complexities of real-time behavior.
Finally, this study focused only on Spanish speakers. The same approach could be extended to other language communities in the U.S. to see whether similar patterns emerge or if there are unique barriers tied to different linguistic or cultural contexts.
Is there anything you particularly want to highlight from this study?
What really stands out from this study is the gap in comprehension between English and Spanish speakers. We often talk about access to information, but understanding that information is just as important. Cultural background, prior experience, and familiarity with severe weather all shape how people interpret and respond to risk messages.
It’s also important to remember that many Spanish speakers in the U.S. come from places — both within and outside the country — where tornadoes and other severe weather events are not common. As people move into areas that experience these hazards more often, there’s a constant need to keep communicating and re-explaining what these warnings mean and what actions to take, even if it feels repetitive. That ongoing communication helps make sure everyone, regardless of background or experience, has the knowledge to stay safe.
