This story was originally published by High Country News and is reproduced here as part of the Climate Desk Collaboration.

At a specialized cold lab in Boulder, Colorado, where indoor temperatures hover around minus 4 degrees Fahrenheit (minus 20 degrees Celsius), atmospheric researcher Anthony Bernal Ayala slices hailstones with a band saw, revealing layers of icy growth like the rings of a tree. These frozen rings — clear or cloudy, thick or thin — indicate the conditions in which the hail formed: A clear layer signals the water froze slowly, while a cloudy layer means it happened quickly, trapping air bubbles. Bernal Ayala then analyzes the crystalline structures, particles, and isotopes preserved in these layers.

“All that information can tell you a wonderful story of how the hailstone was growing, but nobody — not until now — was able to tap into that history,” said Bernal Ayala, who works for Atmospheric and Environmental Research, part of the JANUS Research Group.

Bernal Ayala’s research, conducted at the National Center for Atmospheric Research (NCAR), is part of a historic effort to study one of the costliest weather hazards. Researchers hope their findings will help protect lives and property from the stones that careen to Earth, battering roofs, pummeling crops, and pockmarking cars.

From 2023 to 2025, annual severe thunderstorm losses exceeded $50 billion nationwide, according to the Insurance Institute for Business and Home Safety — more than double the losses seen just a decade earlier. Hail accounted for 60 percent to 80 percent of the damage.

In Colorado, one of the country’s most hail-prone states, hail — not wildfire — is the main factor driving up insurance premiums. And researchers expect larger, more damaging hailstones to become more common with climate change.

“It’s going to just keep getting costlier and costlier,” Bernal Ayala said.

Even basic questions about hail have been notoriously hard to study. But a long-awaited resurgence in research, including the first major US field campaign to study hail in more than 40 years, could help change that. 

ICECHIP — the “In-situ Collaborative Experiment for the Collection of Hail In the Plains” — aims to improve forecasts, make warnings more precise and help communities better prepare. It received $11 million in funding from the National Science Foundation in 2024, just before deep cuts hit the foundation in 2025. 

For six weeks in spring 2025, about 100 scientists and university students chased hailstorms across the Central Plains and Front Range, from the Canadian to the Mexican border, using radar trucks to track storms and drones to take aerial footage. Researchers photographed hail with high-speed cameras, measured its impact with a tool called a disdrometer, recorded hail’s distribution and size based on its imprints on foam pads, captured it in specialized funnels, and crushed it to test its strength. 

Scientists from the Australian Bureau of Meteorology also used balloons to release golf ball-sized sensors into storms. Once aloft, each hail-like probe collected data on how fast it was traveling and whether ice was growing on it; a microphone even recorded water droplets and hailstones hitting it.

“This has really been the first time where we’ve been able to do a real comprehensive look at hail on the ground, hail aloft, the storm itself, the environment around the storm — just everything all at once,” said Becky Adams-Selin, ICECHIP’s lead investigator and a principal scientist at Atmospheric and Environmental Research, JANUS Research Group. 

ICECHIP’s samples were packed into 26 cooler-sized boxes and brought to NCAR. This internationally renowned hub for weather and climate research has been used to study hail for decades. In one of its last major campaigns, in the 1970s, the US investigated Soviet claims that it could suppress hail. 

Researchers say there’s no other US lab like it; the backup plan to store ICECHIP’s samples involved using a walk-in freezer from a fast-food restaurant.

Now NCAR’s own future is uncertain: In late 2025, the Trump administration said it would dissolve the center and transfer its work out of state. The University Corporation for Atmospheric Research (UCAR), which manages NCAR, sued the administration in March, alleging that it’s targeting NCAR to punish Colorado for retaining its mail-in voting system and because Gov. Jared Polis initially refused to release former Mesa County clerk Tina Peters, a 2020 election denier imprisoned for breaching election security and undermining election integrity.

In June, a federal judge temporarily blocked the White House from dismantling NCAR. The government has until Sept. 30 to respond to the lawsuit.

“I can’t even quantify the impact that it would have to our community if we lost NCAR,” Victor Gensini, professor of meteorology at Northern Illinois University, told HCN. “They are the beating heart.”

Previous research has predicted climate change’s effects on hail, noting that warmer temperatures could melt smaller hailstones before they land but also fuel stronger updrafts, keeping hailstones airborne longer and allowing them to grow bigger, enabling the most damaging hail to become more common. 

A record-setting stone recovered in Yuma County, Colorado, in 2023 measured more than 5 inches across. That year, the state logged 176 reports of hail 2 inches or greater, surpassing the previous record of 91 reports in 2018, according to the Colorado Climate Center.

Scientists say it’s difficult to know how much of the rise in damage is due to the increase in large hail versus the increased development in hail hotspots. The connection between a rapidly growing human footprint and higher insured losses has been dubbed the “expanding bull’s-eye effect.”

Gensini, a principal ICECHIP investigator, cites the development boom east of the Front Range, along Colorado’s Interstate 25 corridor, which has grown explosively in recent decades. The population of the Fort Collins metro area alone has doubled since 1990 to around 375,000 residents, according to the US Census Bureau.

This is happening across the Western US, as previously rural or open areas become densely populated, increasingly expensive targets for hail.

“Colorado is the poster child for what I expect the next 25 years to look like for hail,” Gensini said.

In 2019, Fort Collins became one of the first cities to adopt hail-specific roofing requirements for new builds and remodels. Marcus Coldiron, the city’s chief building official, said it was partly in response to all the shingles and roofing materials taken to the landfill after a catastrophic hailstorm in 2018.

It’s still unclear how well the requirements are working: Coldiron said fewer permits were being issued for roof replacements and fewer shingles were going to the dump. However, Fort Collins also hasn’t experienced a hailstorm as devastating as the one that hit in 2018.

A growing number of Front Range communities now require impact-resistant shingles in their building codes, and Colorado legislators recently passed a law offering grants to residents who install more resilient roofing.

ICECHIP’s researchers are also investigating how they can improve building materials as billion-dollar losses from single storms become more common.

“The dollars are simply piling up so fast,” said Ian Giammanco, a lead research meteorologist at the Insurance Institute for Business and Home Safety, one of ICECHIP’s key partners. “We’re still trying to understand how do we get the best materials onto roofs, onto people’s homes, onto businesses to actually start shaving off these losses.”   

Welcome to The Midterms, Actually. Every week I’ll be writing about two things: one big idea or trend I see forming across the 2026 primaries, and one thing I’m hearing from a voice on the ground. Let’s get into it. You can sign up here to get it in your inbox every week. Let’s get into it.

One big idea: Latinos didn’t come home. They’re just in motion.

In the last presidential election, President Donald Trump’s strong support with Latino voters ended up one of the biggest stories of the night. After years of Democrats telling themselves that the growing number of Hispanic Americans would help lead to a permanent liberal majority, their dreams were first dashed in 2016 (when many of those voters stayed home) and then buried in 2024 (when they, and other nonwhite voters, swung toward Trump).

But just as Democrats have at times been too confident that minority voters would fall in line, it now seems Republicans have replicated their mistake. Poll after poll shows a historic reversal among Latino voters in Trump’s second term, and these days even Republican members of Congress are beginning to call Trump out on his deportation efforts. 

Why? It comes back to the numbers.

As Nate Cohn observed in the New York Times this week, Democrats now lead Republicans among Hispanic likely voters by 35 percentage points on the generic congressional ballot, 64-29 — a 28-point swing from 2024, when Kamala Harris carried Hispanic voters by just 7 points. The shift is a big reason the polling has Democrats in “blue wave” territory this November.

Last week on America, Actually I explored this shift in detail — by talking to Stephanie Valencia, co-founder and president of Equis, the research firm that has spent a decade digging into the Latino electorate. And the first thing she wanted us all to know: Get rid of the idea that Latinos will ever be permanently Democratic or permanently Republican. Part of the electoral power is that a larger section of the electorate defines itself more independently.

“Latinos often get given the title, the sleeping giant, which I just want to rid that out of everybody’s vocabulary. We should be called the swinging giant based on how much shift we have seen election to election. A third of Latinos consider themselves kind of independent or apolitical nonpartisan, and that’s that swinging middle that really is deciding every single election.”

Here are three more trends Valencia pointed out, which help explain why this all-important voting group has done such a drastic 180 on Trump in his second term.

1. It’s the economy (again)

As with all things in Trump’s second term, it starts with the fundamental problem of his second term: He has not delivered on lower prices. Equis found that Trump’s net approval with Latinos on cost of living sits at -52 — and his number on the economy (-45) is worse than his overall approval with the group (-31). Both are dramatically worse than where he sits with non-Latinos, who put him at -39 on cost of living and -25 on the economy. Among Latino men 18 to 39, cost of living hits a stunning -63.

Here’s what Valencia said:

“Latinos make up more of the wage workforce than any other racial subgroup. And so what that means is when wages are down and inflation is high, Latinos feel it more. We have less savings. We have less rates of home ownership. So the kind of safety nets that other people might have, we just don’t have.”

She also pointed this out:

“In Trump 1.0, his approval on the economy and the cost of living was actually better than his everyday approval… And he has been underwater with Latinos almost this entire term so far on the cost of living and the economy. It’s almost been a complete and total reversal.”

One of the arguments Trump made to Hispanic communities (and other groups also) was that he’d economically empower them, and that this was more important than prioritizing social causes that didn’t have a material benefit. That’s also been his fundamental flaw. From tariffs to rising gas prices, he has not only not delivered — people also feel worse off than before.

2. The immigration backlash

Valencia’s data was undeniable: More and more Latinos turned on Trump as he broadened his deportation efforts, which were broadly unpopular. But what I didn’t understand is why it took until Trump’s second term for people to realize those threats were legitimate, considering they were so front and center during his campaign. (They held up “Mass Deportation Now!” signs at the Republican National Convention!)

Here’s what Valencia said:

“I think this is really important for us as avid news consumers to understand what it is that the average voter, and especially that voter who’s coming in and out of elections, the kind of news they’re consuming. What they were seeing on mainstream media, perhaps even Fox News at the time, was a focus on Trump surrogates who said he’s gonna focus on serious criminals, securing the border, and violent criminals. And so that’s what people bought.”

“Fast-forward to 2026, the people who feel most regretful who voted for Trump in 2024 are most regretful because of immigration. They feel like there was a broken promise there. Like, they were told he was gonna focus on the border, which everybody agreed was chaos, and he was gonna focus on violent criminals and serious threats, and at the end of the day, he has not complied with that.”

There’s also an element of disbelief. When Trump said he would deport “bad people,” a lot of people thought he was only talking about people in other communities — or immigrants with criminal legal histories — not the more far-reaching immigration sweeps we’ve seen since he’s been in office. A striking ad from Rep. María Elvira Salazar (R-FL) this week where she directly criticizes Trump’s immigration policy explicitly made this distinction. 

The Equis numbers track that shift almost perfectly. In December 2024, 65 percent of Latino Trump voters believed he would target only criminals and recent arrivals. Today, 62 percent of Latinos overall say he’s going after undocumented immigrants broadly — and among his own Latino voters, the share who say that has jumped from 26 percent to 37 percent. That perception is the whole ballgame: Latino Trump voters who still believe he’s targeting only criminals give him +84 net approval. Among those who’ve concluded he’s targeting everyone, it collapses to +31.

The backlash isn’t about the border. It’s about what’s happening in the interior — the raids, the family separations, the citizens and legal residents swept up. 

Equis finds 68 percent of Latinos say Republicans promised to deport criminals and deported hardworking people instead — and 67 percent say Democrats promise immigration reform and never deliver. Forty-two percent say both parties broke their promises.

“I think it is a huge mistake for Democrats to assume or to rest on their laurels and think that because of what people are feeling around Trump and Republicans and the sense of disappointment and broken promises, that that is automatically going to translate to them as a party and as a brand. Right now, the Democratic brand has suffered a repeated set of blows among Latino voters. 20 years of comprehensive immigration reform that hasn’t been delivered. It’s a thing that comes up in almost every focus group.”

3. Meeting voters where they are

The third trend isn’t about persuasion at all. It’s about the fact that a lot of campaigns are still bad at the mechanics of reaching this electorate — and Valencia’s examples were almost comically basic.

Their testing found that more than 70 percent of Latinos who see an ad in Spanish come away thinking that candidate or party cares more about Latinos. That’s not a persuasion effect. That’s a respect effect, and it’s enormous for the price of one ad buy.

So how do campaigns blow it? Some examples, according to Valencia: translating English ads directly into Spanish, rather than tailoring them to the language and culture; and assuming English speakers and Spanish speakers are separate categories when voters often shift between media in both languages. 

The deeper error is the monolith problem — the group is still treated as a voting bloc that moves in unison, rather than a diverse set of people from varying countries of origin and contexts. And the fastest-moving division isn’t the country of origin at all. It’s age — an electorate replenishing itself faster than either party’s model of it.

“There are a million Latinos who turn 18 every single year, and an 18-year-old who turns 18 this year and will vote in this year’s election was born the year Barack Obama was elected president. … The era of politics for an 18-year-old today is chaos, Covid, and corruption. They do not remember the hopium of Obama and the beautiful coalition that we built. That is all ancient history to them.”

A million new voters a year, completely up for grabs. Nobody is inheriting these voters. Neither party owns them. Because as Valencia put it: Demographics have proven not to be destiny — effort and intention matter more.

What I’m hearing

This week I called Matthew Eadie, a reporter at Evanston Now, who has been covering an ICE stop in Evanston, Illinois that has since attracted national attention. On the morning of Sunday, September 20, federal agents stopped a man walking on Howard Street. He turned out to be a US citizen, and not the person they were looking for. Agents reportedly left him on the ground with injuries to his head, neck, and teeth. DHS says he “refused to identify himself. Now the community is demanding answers.”

What’s the community reaction been?

People are really furious. If you remember the fall here — Operation Midway Blitz, the heightened Border Patrol enforcement across the Chicago area, Gregory Bovino parading around the streets with camera crews, the protests at the Broadview processing center — people here were very aware of how impactful it was. It was also happening around the elections, so it was prominent news everywhere and everybody was talking about it.

Over the summer it really died down, and I don’t think people have paid much attention since. ICE has been around all year, but they haven’t been as flashy about it, which has been talked about nationally too. This was a flare-up, and it’s a reminder that it’s still going on even after Bovino left. There have been protests at Broadview this week unrelated to the Evanston incident. So either people are starting to pay more attention as the election gets closer, or the incidents are actually picking up again.

What are the unanswered questions you’re still chasing?

Who were they looking for? Were they actually looking for somebody? That’s been their story — a specific individual — and a lot of people I’ve talked to are really skeptical that’s the case. One witness at the scene told me he felt like they just picked him and drove up on him. Essentially what’s been called a “Kavanaugh stop” — if they’re suspicious of you, they can pretty much find a reason to detain you. And then they dropped it once they realized either that he was a US citizen, which he says he is, or that there were too many people around taking videos for it to be worth their time.

So: What happened in the lead-up, and who specifically were they looking for? 

Afterward, the agents talked to Evanston police — that’s an Evanston police policy; they’re supposed to try to get information from ICE officers. I’ve FOIAed all the body cam footage. I want to see what those conversations were like, and what exactly the agents told police officers when they talked privately.

Part of what we’re trying to do is see the issues that play out from the bottom up rather than the way they get framed nationally. Beyond immigration, is there anything in your coverage that’s shaping local politics more than the national conversation reflects?

The national issues matter to everybody, but they hit communities in completely different ways. Evanston is affluent — college town, Northwestern, wealthier than most suburbs. Affordability is a buzzword that comes up everywhere, but people here talk about it differently.

This is an overwhelmingly progressive Democratic city. Some people in Evanston certainly need it to be more affordable for themselves. But you also have progressive voters saying we want to make Evanston more affordable and more inclusive for other people. Sometimes the affordability conversation gets framed as “life is too expensive for me and we need to make it cheaper for me,” which is true in a lot of places. In somewhere like Evanston, where there are a lot of affluent people who are also very progressive voters, it’s more: If we actually want to say we’re a progressive community, we have to show it. And showing it maybe means building affordable housing.

It’s also a city that passed reparations, which the federal government is challenging right now. That’s going to be a federal court case through the end of this year and probably into next year — the feds challenging the city’s reparations program. That’s something the city prides itself on. Those are big issues here that come out of the national conversation but sound different than how we hear them in national media.