|Above: The back window of a car is broken in the Broadmoor area of Colorado Springs, CO, after a hailstorm with stones the size of softballs hit the area on Monday, Aug. 6, 2018. Image credit: Jerilee Bennett/The Gazette via AP.|
What had looked like an overdue, relatively quiet year for U.S. hail damage is falling more in line with the last decade of unprecedented losses, said experts at the North American Hail Workshop. This week’s meeting, hosted by the National Center for Atmospheric Research in Boulder, Colorado, took a multidimensional look at the hail problem, with some 200 specialists on hand ranging from cloud physicists to climate and catastrophe modelers to insurance analysts and roofing experts. Sponsors included the National Science Foundation, the Insurance Institute for Business & Home Safety (IBHS), and Building Envelope Consultants.
Last week’s hailstorm in Colorado Springs is the most recent of the summer’s big ones. The quick-hitting storm dropped softball-sized stones on the Cheyenne Mountain Zoo on August 6, killing two animals and injuring at least 14 people. It ranks #21 among all U.S. hailstorms to date in the number of people reported hurt, according to John Allen (Central Michigan University).
“The atmosphere decided to tell us why we were here,” said Ian Giammanco, lead research meteorologist at the IBHS Research Center, at the opening of the August 14-16 workshop. Giammanco co-chaired the meeting with NCAR senior scientist Andrew Heymsfield.
This year has left its mark on the nation’s two metropolitan areas that are most likely to be hit by high-end hailstorms. On June 6, an early-morning hailstorm in the Dallas–Fort Worth metroplex caused at least $825 million in damage, according to insurance broker Aon Benfield. Days later, on June 18, the Denver-Boulder metro area endured its own barrage of hail, part of a four-day onslaught of severe weather that left at least $1.4 billion in damage.
|Figure 1. Insured losses from severe convective storms (thunderstorms) went far beyond $10 million (in 2018 US dollars) in every year from 2007 through 2017. Hail makes up the majority of these insured losses, typically between 50% and 80%. Image credit: Steve Bowen, Aon Benfield.|
All told, each year since 2008 has produced well over $10 billion in U.S. insured losses from severe convective storms (thunderstorms) in 2018 dollars, as shown in Figure 1. That’s more than twice the inflation-adjusted damage rate that was typical in the early 2000s, and more than four times the rate seen in the 1980s. Most of these insured losses are hail-related, according to Steve Bowen, director of meteorology at Aon Benfield, who estimates that each of these years has likely seen at least $10 billion in insured damage from hail alone. Total damages from hail (including uninsured losses) most likely varied between $12 and $20 billion per year, he added.
Experts agree on a number of factors that are all pushing up the tab for hail over time. Yet the sharp rise of the late 2000s, and the newly sustained level of damage, defy simple explanation.
“Have we entered a new normal? Maybe we have,” said Bowen.
More people, more roofs
Perhaps the biggest factor at work is the growth in the number of U.S. households and in the amount of real estate in hail-vulnerable areas. Severe thunderstorms that once struck unnoticed beyond city limits are now inflicting huge tolls on suburban and exurban areas filled with big homes and big cars.
Even fast-growing Phoenix, not a usual hotbed for hail, hasn’t been out of the woods. The city experienced the single most damaging hailstorm in U.S. history on October 5, 2010, with a tab of $3.2 billion (adjusted to 2018 dollars). “Some homeowners had to wait more than a year before a licensed contractor was available to repair their roofs,” said meteorologist and operations analyst Bryan Wood (Assurant) in a Capital Weather Gang essay published August 8.
Bowen pointed out that the number of single-family homes in Texas spiked from about 2.4 million in 1950 to 10.9 million in 2017. Colorado now has 2.4 million single-family homes, compared to just 350,000 in 1950. On top of this, the average new-home size rose from about 1700 square feet in the mid-1970s to about 2500 feet today. All that square footage tends to increase the amount of roof area exposed to hail.
|Figure 2. A woman repairs her hail-damaged roof in Stavropol, Russia, on Sept. 10, 2007. An intense hailstorm seriously damaged some 150 houses in the southern Russian city. Image credit: Danil Semyonov/AFP/Getty Images.|
How roofs are built makes a difference too, according to Mark Bove, senior research meteorologist at the reinsurance firm Munich Re. Roof covers and sidings not rated for hail are still in common use, Bove said. Other stress points are gutters (which are easy to damage), non-impact-resistant windows and skylights, and roof decking that’s not sealed with bituminous tape or a felt covering. Once water gets in through a roof, it can cause a great deal of havoc in short order, especially with the ever-increasing amount of electronics in modern homes.
Greed is another factor that can’t be ignored. Aggressive roof contractors often swoop in following a big storm in hail country, and people on the fringes of a hard-hit area may work the system to try and get themselves a new roof like their ones their neighbors are getting. “We also see faked hail damage on occasion,” said Bove. One telltale sign: “The damage stops about a foot from the eave edge, because nobody wants to fall off the roof.”
Hail- and fire-prone Colorado experienced the nation's third-highest jump in home insurance rates from 2007 to 2015, according to data from the National Association of Insurance Commissioners analyzed by the Denver Business Journal. Insurers also have other options for addressing their sharp rise in hail risk, according to Bove. These can range from the use of hail deductibles (on the order of 1-2% of property value) to the exclusion of cosmetic damage that doesn’t affect roof function, or even to excluding hail from standard coverage altogether.
“Obviously the insurance industry is quite concerned about these loss trends,” said Bove. “Basically, they’re getting death by a thousand paper cuts.”
|Figure 3. A smorgasbord of objects is commonly used to help people estimate hail size—but do they help? Image credit: weather.gov.|
Counting the hailstones (lots of them)
The vast database of severe-hail reports maintained by the NOAA/NWS Storm Prediction Center—roughly 346,000 in all, going back to 1955—is a gold mine for insurers and scientists trying to get their arms around the problem of hail risk. It’s a mine riddled with hazards, though. For example, some volunteer observers are more eager than others, which can distort the regional picture of hail distribution over time.
Even the simple step of the National Weather Service raising the minimum diameter for severe hailstones from 0.75” to 1.00” in 2010 has had a surprisingly large impact on the database, according to CMU’s John Allen, who analyzed the 1955-2014 dataset with Michael Tippett (Columbia University) in the Electronic Journal of Severe Storms Meteorology. The higher threshold led to a drop in 0.75” reports and a simultaneous rise in 1.00” reports. “We have put large influences on our dataset that we probably shouldn’t have,” said Allen.
The longstanding practice of referring to everyday items like hen’s eggs or golf balls to estimate hail size may also be introducing unneeded error. According to Allen, who’s carried out some simple tests with undergraduate students, “it turns out that people are better at [directly] estimating hail size than using reference objects.” (Few people outside the weather or golf communities could tell you the diameter of a golf ball.)
|Figure 4. The widest and heaviest hailstone on record for the United States (8” in diameter and 31 ounces) fell in Vivian, South Dakota, on July 23, 2010. The stone's circumference of 18.625" was topped only by the 18.75" from a hailstone that fell in Aurora, Nebraska, on June 22, 2003. Image credit: NWS/Aberdeen, SD, via Wikimedia Commons.|
We know even less about how other characteristics of hailstorms are distributed, in part because such variables as hail depth or the wind-driven quality of a given hailstorm aren’t routinely reported. Researchers have started to analyze hail depth through a project based at Colorado State University called CHAT (Colorado Hail Accumulation from Thunderstorms). Since 2012, it’s gathered 150 hail-depth reports from volunteers, largely from Colorado but also in other parts of the U.S. and beyond. One perception that’s already been exploded by CHAT: the idea that hail accumulates to great depths only within the “Hail Alley” of the U.S. High Plains. The CHAT website includes photos of impressive haildrifts from Germany and Mexico City.
Researchers are using a variety of inventive approaches to better predict hailstorms and to assess whether and how the hail threat might change in a warming climate. We’ll have more on these topics in a subsequent post.
We are continuing to watch Tropical Storm Ernesto in the Atlantic, Hurricane Lane in the East Pacific, and a disturbance called 99L in the Atlantic tropics. Stay tuned for more on these and other tropical developments in our next post.