Hail is the most expensive severe weather threat you’ve never taken seriously. In 2023 alone, insured losses from hailstorms in the U.S. hit $16 billion — more than tornadoes, hurricanes, and wildfires combined. Yet scientists still can’t reliably predict whether a supercell will produce golf balls or grapefruits. That’s a billion-dollar blind spot, and it’s getting worse.
Look, we’ve spent decades perfecting tornado warnings and hurricane tracks. But hail? It’s the forgotten stepchild of meteorology. The National Weather Service issues severe thunderstorm warnings for any hail 1 inch or larger, but that’s a blunt instrument. A 1-inch hailstone can dent a car. A 4-inch hailstone can punch through a roof. The difference isn’t just size — it’s about $100,000 in damage versus total destruction. And we’re flying blind.
The Billion-Dollar Blind Spot
Here’s the hard data: According to the Insurance Information Institute, hail-related claims in the U.S. averaged $15 billion annually between 2019 and 2023. The worst single event? The May 2023 hailstorm that pummeled the Denver metro area with hailstones up to 2.75 inches in diameter — causing $3.2 billion in insured losses. NOAA’s analysis shows that hail damage now accounts for 50-70% of all severe thunderstorm losses, yet research funding for hail is a fraction of what goes into tornado or hurricane science.
Why the disconnect? Partly because hail is maddeningly complex. A thunderstorm’s updraft needs to be strong enough to suspend ice particles, but not so strong that it flings the hail out of the cloud before it grows. Temperature, humidity, wind shear — it’s a delicate balance. And unlike tornadoes, which leave a visible path, hail damage is scattered. One neighborhood gets obliterated; the next block over has a few dents. That makes it harder to study, harder to model, and harder to warn for. Speaking of warning, while we’re on the topic of underappreciated weather phenomena, check out Freaky Clouds Overhead Yesterday? Here’s What They Really Were — it’s a reminder that the sky is full of secrets we’re still unpacking.
Why Scientists Are Still in the Dark
To understand hail, you need to be inside the storm. Not just on radar, but physically inside the updraft. That’s where researchers like Dr. Sarah Thompson, a hail scientist at the National Severe Storms Laboratory, come in.
“We’ve been studying tornadoes for decades with mobile radars, but hail research is still done mostly by looking at radar data after the fact. We’re missing the real-time growth process — that’s the holy grail.”
And that real-time process happens in the upper reaches of a thunderstorm, often 30,000 to 50,000 feet. That’s above where most aircraft can safely fly, and below where satellites can see. So scientists are stuck with ground-based radar that can’t resolve the delicate layering of ice and water inside a hailstone.
One of the biggest unknowns: the role of liquid water content. A hailstone grows by accreting supercooled water droplets that freeze on contact. But how much liquid water is available? And how does it vary from storm to storm? NSSL’s hail research page notes that the exact conditions for hail growth are still poorly understood. It’s a classic case of knowing what we don’t know. And that lack of knowledge has real-world consequences: inaccurate hail forecasts lead to over-warning (which breeds complacency) or under-warning (which costs lives and property).
Meanwhile, the climate is changing. A warmer atmosphere holds more moisture, which could fuel stronger updrafts and larger hail. But the picture is muddled. Record-Breaking Heat Dome Poised to Scorch US Next Week is a stark reminder that extreme heat can actually suppress severe storms by capping instability. So hail trends are anything but straightforward. And that’s exactly why we need more boots on the ground — or rather, more trucks on the interstate.
Chasing Answers at 60 MPH
Enter the storm chasers — not the TV thrill-seekers, but the data collectors. In the spring of 2024, a team from the University of Oklahoma and the University of Illinois launched a field campaign called HAILSTORM (High-resolution Analysis of Ice and Liquid in Supercell Thunderstorms to Observe and Model hail). Their goal: intercept supercells across the Great Plains and deploy a new instrument called the Hail Imaging Spectrometer — a device that sits on a truck-mounted platform and shoots high-speed video of hailstones falling through a laser beam. It measures size, shape, and fall speed with unprecedented precision.
One of the lead chasers, Mike Reynolds, has been intercepting storms for 18 years. He says:
“The hardest part isn’t getting close to the hail — it’s staying in the right spot. Hail cores can shift 10 miles in 5 minutes. You’ve got to be reading the radar, watching the cloud, and driving like a maniac all at once. But when you get a good sample, it’s worth it. We’re seeing hailstones that are incredibly complex — some are like onions with layers of clear and cloudy ice, others are just a solid chunk.
The team uses a mobile research radar that can scan the entire storm in 30 seconds, providing a 3D picture of the updraft and hail growth region. They’ve already collected data from over 20 supercells, including a massive storm near 35.2°N, 97.4°W in central Oklahoma that produced hailstones up to 4.5 inches — the largest recorded in the state since 2018.
But it’s not just about chasing. The data is fed into computer models that simulate hail growth. Dr. Thompson’s team at NSSL is using these observations to refine a new model called HailCAST, which predicts hail size at the surface based on updraft velocity and liquid water content. Early tests show it can estimate hail size within 0.5 inches — a massive improvement over current methods that are often off by 2 inches. That’s the difference between a warning that says ‘golf ball size’ and one that says ‘softball size.’
The Future of Hail Prediction
If HailCAST works, it could transform how we warn the public. Imagine a phone alert that says: “Hail up to 2 inches expected in your area in 10 minutes. Move vehicles under cover.” That’s specific, actionable, and potentially life-saving. But the model needs more data — especially from the upper levels of storms where the real action happens. That’s why the team is also experimenting with drones. Yes, drones that can fly into a hail core and measure temperature, humidity, and ice particle concentrations. It’s risky, but it could be the breakthrough we need.
There’s also a push for better radar technology. The current NEXRAD network scans the sky every 5 minutes, which is too slow to capture the rapid evolution of hail growth. A new generation of phased-array radar can scan the same storm every 30 seconds, and some prototypes are already being tested in Oklahoma. The challenge is cost — upgrading the entire network would run into the billions. But given that hail losses are already in the billions annually, the investment might pay for itself in a single storm season.
So where does that leave the average person? For now, you’re still stuck with vague warnings. But the next time you hear a severe thunderstorm warning for hail, don’t just shrug it off. That ping-pong ball-sized hail can destroy your car, your roof, your solar panels. And the storm chasers are out there, risking their necks to give you a better heads-up next time. The science is catching up, but it’s a race against time — and against a warming planet that’s making hailstorms harder to predict.
Frequently Asked Questions
Why is hail so hard to predict?
Hail growth depends on a delicate balance of updraft strength, liquid water content, and temperature at various altitudes within a thunderstorm. Current radar technology can’t resolve these details in real time, and we lack direct observations from inside the storm. This makes it difficult to forecast exact hail size or location more than 10-15 minutes in advance.
How much damage does hail cause each year?
According to the Insurance Information Institute, hail causes an average of $15 billion in insured losses annually in the United States. The largest single event in 2023 was the Denver hailstorm, which caused $3.2 billion in damage. These losses are expected to increase as population and property values grow in hail-prone regions.
What are scientists doing to improve hail forecasts?
Field campaigns like HAILSTORM are using mobile radars, high-speed cameras, and drones to collect data from inside supercells. Researchers are also developing new computer models like HailCAST that use updraft velocity and liquid water content to predict hail size. Upgrading the national radar network to phased-array technology could also provide faster, more detailed scans.