You might also like...
Check out some other posts from University of Florida
As Floridians brace for hurricanes amid the wild weather of 2025, some University of Florida researchers have their eyes on 2017’s Hurricane Maria, the deadly Category 4 storm that pummeled Puerto Rico.
Engineering professor and natural hazards researcher Brian Phillips, Ph.D., is leading UF’s efforts in a Hurricane Maria investigation conducted by the National Institute of Standards and Technology, known as NIST.
The goal is increased safety and resilience amid deadly conditions.
Maria killed nearly 3,000 people and caused more than $90 billion in damage. Most of the island’s wind sensors and weather stations failed as the storm raged, leaving responders and investigators with few reliable weather measurements.
What went wrong?
Phillips and UF storm researchers are helping answer that question — and provide safety and structural recommendations — as part of NIST’s Hurricane Maria investigation. The full report will be released in 2026, but NIST recently published preliminary findings; some of the hazard and structural load data was derived from wind tunnel tests at UF's NHERI Experimental Facility in the Powell Family Structure and Materials Laboratory on UF’s East Campus in Gainesville.
“Our wind tunnel has a strong reputation in the wind-engineering community for its unique flow control and measurement capabilities We worked with NIST to develop a test campaign to study the wind conditions Puerto Rico’s mountainous terrain and the resulting loads of critical infrastructure,” said Phillips, a civil and coastal engineering professor with UF’s Engineering School of Sustainable Infrastructure & Environment. “UF,” he added, “has one of the premier research wind tunnels in the country and it enables us to pursue impactful research like this.” As part of the NIST investigation, Phillips and his team created 1-to-3100 scale topographic models of regions in Puerto Rico — about 12 kilometers shrunk down to four meters, Phillips said. They set up those models in the wind tunnel and replicated wind flow over the topography.
“These initial tests were designed to understand the influence of the complex topography had on the wind,” Phillips said. Flow was measured using velocity probes and particle image velocimetry (PIV). These topographic model tests were followed by 1-to-100 scale tests on models of two hospitals in Puerto Rico. In addition to surface pressure measurements, the team conducted qualitative flow visualization tests using smoke, lasers, and high-speed cameras.
“The capabilities of the UF wind tunnel enabled us to investigate the hurricane winds at two different scales,” said NIST’s lead Hurricane Maria investigator, Joseph Main, “so we could measure how the winds were accelerated by Puerto Rico’s mountainous topography and then how those winds translated into loads on critical buildings.” Maria’s flooding blocked roads to hospitals and shelters. The hospitals themselves were heavily damaged by the storm, NIST reported. Reduced access to healthcare was a major factor in the death toll.
“It's good to take a step back,” Phillips said about the overall investigation. “Researchers are approaching the disaster from multiple angles, including the better understanding of the hazard, the performance of critical infrastructure, public response and recovery. “This holistic approach is needed to capture the complete picture and maximize what we can learn from the event. UF's primary contribution was understanding the hurricane wind field and the resulting structural loads, which is a critical piece of that puzzle.” In finding infrastructure vulnerabilities, researchers contend the goal is integrating their findings into design standards for Puerto Rico’s unique topography and building codes. The findings also could be valuable to other storm-prone regions with complex topography.
NIST launched the investigation in 2018, noting Hurricane Maria “set off a cascade of building and infrastructure failures across Puerto Rico that had lasting impacts on society, including health care, business and education.”
“Our goal is to learn from that event to recommend improvements to building codes, standards and practices that will make communities more resilient to hurricanes and other hazards, not just in Puerto Rico but across the United States,” Main said. The complete report is scheduled to be released in 2026, and NIST noted some findings may change before its release. But in July, NIST released some preliminary findings. They include:
Peak wind speeds over flat terrain reached 140 mph. They accelerated to over 200 mph in some areas due to the steep hills and mountains. The mountains also intensified the rainfall, which reached 30 inches in some areas. Only three out of 22 weather stations were fully functional during the hurricane. 95.3% of schools on the island lost power for an average of over 100 days. “One important preliminary finding from the study is that emergency preparations work,” NIST reported. “Businesses, schools and hospitals that took specific measures to prepare before Hurricane Maria were able to resume operations more quickly” said Maria Dillard, NIST’s associate lead Hurricane Maria investigator. Preparations included pre-established emergency plans, designated risk mitigation funds and backup power sources.
A sweeping new study from the University of Florida and University College London has found that daily reading for pleasure in the United States has declined by more than 40% over the last 20 years — raising urgent questions about the cultural, educational and health consequences of a nation reading less.
Published today in the journal iScience, the study analyzed data from over 236,000 Americans who participated in the American Time Use Survey between 2003 and 2023. The findings suggest a fundamental cultural shift: fewer people are carving out time in their day to read for enjoyment.
“This is not just a small dip — it’s a sustained, steady decline of about 3% per year,” said Jill Sonke, Ph.D., director of research initiatives at the UF Center for Arts in Medicine and co-director of the EpiArts Lab, a National Endowment for the Arts research lab at UF in partnership with University College London. “It’s significant, and it’s deeply concerning.” Who’s reading and who isn’t The decline wasn’t evenly spread across the population. Researchers found steeper drops among Black Americans than white Americans, people with lower income or educational attainment, and those in rural (versus metropolitan) areas — highlighting deepening disparities in reading access and habits.
“While people with higher education levels and women are still more likely to read, even among these groups, we’re seeing shifts,” said Jessica Bone, Ph.D., senior research fellow in statistics and epidemiology at University College London. “And among those who do read, the time spent reading has increased slightly, which may suggest a polarization, where some people are reading more while many have stopped reading altogether.” The researchers also noted some more promising findings, including that reading with children did not change over the last 20 years. However, reading with children was a lot less common than reading for pleasure, which is concerning given that this activity is tied to early literacy development, academic success and family bonding, Bone said.
Why it matters Reading for pleasure has long been recognized not just as a tool for education, but as a means of supporting mental health, empathy, creativity and lifelong learning. The EpiArts Lab, which uses large data sets to examine links between the arts and health, has previously identified clear associations between creative engagement and well-being.
“Reading has historically been a low-barrier, high-impact way to engage creatively and improve quality of life,” Sonke said. “When we lose one of the simplest tools in our public health toolkit, it’s a serious loss.” The American Time Use Survey offers a unique window into these trends. “We’re working with incredibly detailed data about how people spend their days,” Bone said. “And because it’s a representative sample of U.S. residents in private households, we can look not just at the national trend, but at how it plays out across different communities.” Why are Americans reading less? While causes were not part of the study, the researchers point to multiple potential factors, including the rise of digital media, growing economic pressures, shrinking leisure time and uneven access to books and libraries.
“Our digital culture is certainly part of the story,” Sonke said. “But there are also structural issues — limited access to reading materials, economic insecurity and a national decline in leisure time. If you’re working multiple jobs or dealing with transportation barriers in a rural area, a trip to the library may just not be feasible.” What can be done? The study’s authors say that interventions could help slow or reverse the trend, but they need to be strategic.
“Reading with children is one of the most promising avenues,” said Daisy Fancourt, Ph.D., a professor of psychology and epidemiology at University College London and co-director of the EpiArts Lab. “It supports not only language and literacy, but empathy, social bonding, emotional development and school readiness.” Bone added that creating more community-centered reading opportunities could also help: “Ideally, we’d make local libraries more accessible and attractive, encourage book groups, and make reading a more social and supported activity — not just something done in isolation.” The study underscores the importance of valuing and protecting access to the arts — not only as a matter of culture, but as a matter of public health.
“Reading has always been one of the more accessible ways to support well-being,” Fancourt said. “To see this kind of decline is concerning because the research is clear: reading is a vital health-enhancing behavior for every group within society, with benefits across the life-course.”
As students return to classrooms, Maya Israel, professor of educational technology and computer science education at the University of Florida, shares insights on best practices for AI use for students in K-12. She also serves as the director of CSEveryone Center for Computer Science Education at UF, a program created to boost teachers’ capabilities around computer science and AI in education. Israel also leads the Florida K-12 Education Task Force, a group committed to empowering educators, students, families and administrators by harnessing the transformative potential of AI in K-12 classrooms, prioritizing safety, privacy, access and fairness. How are K–12 students using AI in classrooms? There is a wide range of approaches that students are using AI in classrooms. It depends on several factors including district policies, student age and the teacher’s instructional goals. Some districts restrict AI to only teacher use, such as creating custom reading passages for younger students. Others allow older students to use tools to check grammar, create visuals or run science simulations. Even then, skilled teachers frame AI as one tool, not a replacement for student thinking and effort.
What are examples of age-appropriate tools that enhance learning? AI tools can be used to either enhance or erode learner agency and critical thinking. It is up to the educators to consider how these tools can be used appropriately. It is critical to use AI tools in a manner that supports learning, creativity and problem solving rather than bypass critical thinking. For example, Canva lets students create infographics, posters and videos to show understanding. Google’s Teachable Machine helps students learn AI concepts by training their own image-recognition models. These types of AI-augmented tools work best when they are embedded into activities such as project-based learning, where AI supports learning and critical thinking.
How do teachers ensure AI supports core skills? While AI can be incredibly helpful in supporting learning, it should not be a shortcut that allows students to bypass learning. Teachers should design learning opportunities that integrate AI in a manner that encourages critical thinking. For example, if students are using AI to support their mathematical understanding, teachers should ask them to explain their reasoning, engage in discussions and attempt to solve problems in different ways. Teachers can ask students questions like, “Does that answer make sense based on what you know?” or “Why do you think [said AI tool] made that suggestion?” This type of reflection reinforces the message that learning does not happen through getting fast answers. Learning happens through exploration, productive struggle and collaboration.
Many parents worry that using AI might make students too dependent on technology. How do educators address that concern? This is a very valid concern. Over-reliance on AI can erode independence and critical thinking, that’s why teachers should be intentional in how they use AI for teaching and learning. Educators can address this concern by communicating with parents their policies and approaches to using AI with students. This approach can include providing clear expectations of when AI is used, designing assignments that require critical thinking, personal reflection and reasoning and teaching students the metacognitive skills to self-assess how and when to use AI so that it is used to support learning rather than as a crutch.
How do schools ensure that students still develop original thinking and creativity when using AI for assignments or projects? In the age of AI, there is the need to be even more intentional designing learning experiences where students engage in creative and critical thinking. One of the best practices that have shown to support this is the use of project-based learning, where students must create, iterate and evaluate ideas based on feedback from their peers and teachers. AI can help students gather ideas or organize research, but the students must ask the questions, synthesize information and produce original ideas. Assessment and rubrics should emphasize skills such as reasoning, process and creativity rather than just focusing on the final product. That way, although AI can play a role in instruction, the goal is to design instructional activities that move beyond what the AI can do.
How do educators help students understand when it’s appropriate to use AI in their schoolwork? In the age of AI, educators should help students develop the skills to be original thinkers who can use AI thoughtfully and responsibly. Educators can help students understand when to use AI in their school work by directly embedding AI literacy into their instruction. AI literacy includes having discussions about the capabilities and limitations of AI, ethical considerations and the importance of students’ agency and original thoughts. Additionally, clear guidelines and policies help students navigate some of the gray areas of AI usage.
What guidance should parents give at home? There are several key messages that parents should give their children about the use of AI. The most important message is that even though AI is powerful, it does not replace their judgement, creativity or empathy. Even though AI can provide fast answers, it is important for students to learn the skills themselves. Another key message is to know the rules about AI in the classroom. Parents should speak with their students about the mental health implications of over-reliance on AI. When students turn to AI-augmented tools for every answer or idea, they can gradually lose confidence in their own problem-solving abilities. Instead, students should learn how to use AI in ways that strengthen their skills and build independence.
