Last month it was Typhoon Kalmaegi hitting Vietnam. Before that it was Hurricane Melissa tearing through the Caribbean. Many of us watch these natural disasters from our living rooms, scrolling through news alerts and satellite images before moving on to the next headline. We unconsciously tell ourselves this is a tragedy happening over there, to other, unfortunate people in places with weak infrastructure. But that distance is an illusion, and a dangerous one.
What’s unfolding in Vietnam and Jamaica is not a Vietnamese or Jamaican problem. It is our collective present and near future.
Tomorrow it could be cyclones in the Indian Ocean, wildfires consuming towns in minutes, or floods swallowing cities that thought they were safe. The last time I was in Hawaii there were massive floods. In Madagascar, a tropical cyclone. We watched BMWs washed away in German streets that were never supposed to flood. These headline-grabbing events are a tiny fraction of reality. Nine million people were displaced by such disasters last year alone. Their stories rarely make the news, and when they do, coverage lasts only through the height of the crisis. When the storm passes and real human suffering begins, the news cycle moves on.
In many places, there’s barely recovery time between disasters. What does recovery look like when your temporary shelter from the last hurricane gets hit again?
The specific disaster may change with geography, floods, fires, drought, and typhoons, but the fundamental threat is universal. Climate extremes know no borders. In 2024, researchers proposed adding “Category 6” for storms with sustained winds over 192 mph, far beyond conventional Category 5 destruction. Catastrophic weather events are no longer anomalies. They’re becoming the norm.
When disasters strike, I immediately think about whether frontline communities received warnings, whether they had shelter, and how quickly we can capture the impact. But as a Remote Sensing Scientist and Professor leading NASA projects with access to powerful computers, datasets, and the skills to predict and analyze the impact types of events, I also think about documentation, mapping, and assessment. Sometimes I receive requests to help with analysis, but factoring in time, the most valuable resource in emergency response, emails and phone calls shared during crisis phases are often too late.
Then the numbers come: the scale, the devastation. But we never really grasp the human impact. The destroyed livelihoods. Families sheltering in place. The kids who won’t be in school for months. Waterborne diseases quietly spreading.
Here’s what makes these events ever damning: we’ve never had better tools to see catastrophe coming. Before Melissa made landfall, AI predicted its path days in advance with “impressive” accuracy. High-resolution satellite imagery, real-time data feeds, and computer simulations showed exactly where it would hit and how devastating it would be. I checked the Jamaican weather service website right after landfall, but it was down. On the weather system’s Instagram page, people wrote that they felt dismay and frustration. Meanwhile, the prediction systems continued to get praise even as folks affected by the storm suffered.
So, the multi-billion-dollar infrastructure to watch disasters unfold in real-time, in high definition, with unprecedented foreknowledge falls short, clearly—no translation, no delivery.
The technology works brilliantly. It can deliver insight after insight. But those insights die in the gap between prediction and prevention, between warning and action, between data and political will. If protecting lives is the goal of better forecasting, it needs to be paired with actions, workflow and infrastructure that protects people, systems that support recovery, not in silos and for scientific curiosity only.
In the information age, we’ve convinced ourselves that better information will solve the problem. But the people of Jamaica didn’t need AI to tell them hurricanes are dangerous. They needed resources, infrastructure, and systems capable of protecting them against what was coming.
Don’t misunderstand, we need better prediction. After all, events keep getting worse. But prediction without protection is just expensive, anxiety-inducing theater. It’s like we are building a global nervous system that can feel every disaster but a body that won’t move to stop the loss and damage they live in their wake.
Next time, it might be your city. Your neighborhood. Your home. Are you prepared? Is your community? Because those satellites and AI systems don’t deliver blankets, food, or water.
We cannot keep treating these events as isolated incidents, as someone else’s tragedy, as acceptable losses in places we deem peripheral. We cannot keep investing in the ability to watch while neglecting the capacity to act. Every flood, fire, and storm that levels a community despite our perfect predictions is an indictment of our priorities.
The question isn’t whether catastrophe will reach us. It’s whether we’ll finally invest in protection with the same urgency we’ve invested in prediction, in infrastructure, in evacuation systems, in resilient communities, in the unglamorous work of keeping people safe.
The storms will continue to come. We are getting better at predicting them. The only question is: will we finally invest in saving lives with the same drive and resources we’ve invested in watching them end?