The Tristate Hailstorm: The Most Costly on Record (2024)

1. Introduction

On 10 April 2001, a strong long-lasting thunderstorm produced numerous hailstreaks that caused insured property losses easily exceeding those of any prior hail event since insurance records began in 1949. This record storm began in Kansas, and ended its west-to-east trek 8 h later in Illinois with large hailstones all along a swath 585 km long. The insured property losses of this “tristate hailstorm” amounted to $1.5 billion (Property Claims Service 2002). The major hail-producing storm was part of a 2-day period when other nearby damaging hailstorms occurred in Nebraska, Oklahoma, Ohio, Indiana, and Iowa, and the insurance industry labeled this as a hail catastrophe totaling $1.9 billion in losses. This ranks as the ninth most damaging weather catastrophe in the United States for the entire 1949–2001 period (Kerney 2002). Table 1 lists the nation's top 15 weather catastrophes, revealing that the April 2001 hailstorms were the only hail event to make the list.

Prior hail research has provided measures of losses for past damaging storms. Hailstorms damage both crops and property, and the average annual crop and property insured losses for the United States created by hail is $445 million in 2001 dollars (Changnon and Hewings 2001). In 1949 the property insurance industry established the Property Claims Service (PCS), a group of insurance loss experts, and the PCS began keeping records of all natural hazards that caused losses to insured property of $1 million or more in the United States, labeled as catastrophes. For each catastrophe, the loss assessors of the PCS identified the weather condition(s) causing the losses, creating an extensive loss database listing each event, its causes, losses, and area of occurrence. In 1983 the loss level used to identify an event as a catastrophe was shifted from $1 million to $5 million, as an adjustment for inflation. In 1997 the PCS made a shift from $5 to $25 million as a further adjustment for the effects of ever-growing inflation on the selection of catastrophes. A major insurance firm has analyzed the historical catastrophe data to assess risk, and each year this firm updated the past catastrophe values to match the latest year's conditions. This normalization resulted in a database allowing an unbiased comparison of current catastrophe losses with those in past years. This sizable annual loss adjustment effort required assessing each past event including its location, and three adjustment calculations were made to the original loss value for each catastrophe. One adjustment was designed to correct for time changes in property values and the cost of repairs, and hence, this also adjusted for inflation. The second adjustment addressed the relative growth in the size of the property market in the areas affected by the catastrophe using census data, property records, and insurance records. This adjusted losses for shifts in the insured property (wealth) between the year of a given storm's occurrence and the updated year (2001 in this study). The third adjustment factor was based on estimates of the relative changes in the share of the total property market that was insured against weather perils, using insurance sales records. These three adjustments were used to calculate a revised monetary loss value for each catastrophe so as to make it comparable to current-year values. Thus, adjustments made in 2001 for all past hail catastrophes dating back to 1949 allow a comparative temporal assessment of their losses. An extensive description of adjustment methods is available (Changnon and Changnon 1998). The insurance loss data provide reliable measures of loss but do not capture all losses. Past assessments indicate insured losses are usually 90% of the total losses from weather catastrophes (Changnon et al. 2000).

The highest previous hail catastrophe insured loss was $1.1 billion (adjusted to 2001 dollars) from a hail, wind, and heavy rain system that struck New Mexico and Texas on 5–6 May 1995. Some $200 million of the losses were due to flooding from the heavy rains, and the hail–wind losses were $0.9 billion. Some speculated that the Texas storm caused $2 billion in losses (Hill 1996), but the source of such an estimate is not documented, does not agree with the insured losses, and is not considered reliable. The tristate hailstorm losses were 65% larger than those in 1995, a difference that helps reveal its enormous magnitude. The most costly crop–hail loss came from a storm in June 1964 in Iowa–Illinois, causing $96.8 million in crop losses (Changnon 1966). The five states with the highest average hail losses are Texas, Oklahoma, Kansas, Missouri, and Illinois (Changnon 2001), and the tristate hailstorm affected three of the five top loss states.

2. Storm characteristics

The hail-generating thunderstorm on 10 April developed and remained for its duration at the south end of an east-moving mesoscale convective complex, becoming a long-lived supercell. The first large hail fell at 1450 LST at a small town 65 km southwest of Kansas City. In the ensuing 8 h, the supercell storm produced hundreds of hailstreaks before the hailfalls ended near Effingham, Illinois, at 2315 LST (Fig. 1). The result was a swath of nearly continuous hail that fell along an area that ranged from 10 to 25 km wide and was 585 km long. Past studies of hailswaths in the Midwest defined their lengths as ranging from 80 to 330 km with none as enormous as the long-lasting hailswath in April 2001 (Changnon 1977). Studies of hailswaths in the northern high plains defined average widths of 15–25 km, and the longest on record in a 3-yr study was 490 km (Frisby 1963). Thus, the length of the April 2001 hailswath exceeded anything previously documented from studies of hail in the high plains and Midwest.

The numerous locations where hailstones of 2.5 cm or larger size fell are shown in Fig. 1. Many locales reported high winds with the large hail, adding to the damage potential of the hail. Smaller hail fell at many locations between the large hail locales. Four small, short-lived (three with paths of 1 km or less, and one 5 km long) tornadoes occurred with the storm. One tornado was in central Missouri, two near St. Louis, and one in Illinois. Importantly, the production of large hail continued during the entire lifetime of this giant storm, another condition highly unusual for hailswaths (Changnon 1999).

The evolution of this long-lived storm shows changes in size and shape over time. A mesoscale convective system began developing around a low center along a stationary front in northeastern Kansas and southeastern Nebraska at midday on 10 April. At 1400 LST a cell developed at the southern extremes of the disorganized complex of showers and storms, an optimum location for a supercell to develop. At 1500 LST, the developing storm was a well-defined echo with a core of 50 dBZ (Fig. 2). (Note, all the echoes on Figs. 2 and 3 are at 0.5° antenna elevations, but at times, the considerable distance to the echo meant that the echo portrayed was aloft.) At 1600 LST the echo had enlarged to 50 km in diameter, had a maximum reflectivity of 74 dBZ, and hailstones with diameters >2.5 cm were being produced. The echo at 1700 LST was seen aloft at 75 km from the Springfield radar and had a notable vault indicative of a rotating updraft.

As the storm passed over Columbia at 1800 LST, the echo had enlarged and evolved into a bow shape, as seen by the St. Louis radar. Radar operators at St. Louis, using vertical scans of the storm, noted a well-defined vault in the storm, a condition typical of supercell storms (R. Przybylinski 2002, personal communication). At 1900 LST (Fig. 3) the bow-shaped echo also indicated the beginning of the vault circulation along its right flank, and at 2000 LST the storm was depositing hailstones ranging from 2.5 to 7.5 cm over St. Louis. By 2100 LST the echo shape had altered and large hail was still falling, but now in Illinois. At 2200 LST the storm was still producing large hail and had resumed a bow shape. Peak reflectivities had remained between 62 and 74 dBZ since 1600 LST. Movement speed, based on hail initiation times along the hailswath, revealed that the forward speed ranged from 76 to 95 km h⁻¹, reflecting the rapid movement of the storm due to strong winds aloft. The storm deposited heavy rainfall ranging from 1 to 8 cm along its path across Missouri and Illinois.

Conditions necessary for supercell development had begun to evolve on 9 April. A west–east-oriented warm front moved north of St. Louis and became stationary during the night of 9–10 April about 100 km north of the city. Very warm and moist gulf air flowed into the Missouri–Illinois–Kansas area. Surface maximum temperatures in the three-state area on 9 and 10 April were 29°–32°C, 10°C above average, and dewpoints were >17°C. A low-level jet was established from western Kansas eastward over the future storm area. Thus, by noon on 10 April, the ingredients for supercell development were present—an unlimited supply of warm, moist air, and strong vertical wind shear from eastern Kansas to Indiana. By noon of 10 April, the quasi-stationary front had moved south and became oriented in the afternoon from Kansas City to St. Louis. This provided a zone of strong instability and became the ultimate track of the supercell. The strong northward transport of warm, moist air by southerly winds further helped destabilize the storm's environment.

3. Hail damages and losses

Three factors were the cause of the enormous property damage produced by the tristate hailstorm. First was the enormous size of the hailswath, increasing the likelihood of damages. Second was the near-continuous production of hailstones ranging in size from 2.5 to 7.5 cm, and often associated with high winds. Third was the fact that hail fell over the south suburbs of Kansas City; over portions of Columbia, Missouri; and across the northern portions of the St. Louis metropolitan area in both Missouri and Illinois.

The insured losses in Missouri amounted to $1.4 billion, those in Illinois were $50 million, and those in Kansas were $45 million (Property Claims Service 2002). The $1.4 billion loss in Missouri was the state's single greatest insured loss since records began in 1949. Principal damages were to residences, buildings, and to vehicles. Primary targets for structural damage were windows, roofs, siding (west sides), and skylights. Missouri had 193 000 insurance claims, Illinois 15 500, and Kansas 14 500, a storm total of 223 000 claims. The claims in Missouri included 120 000 for personal fixed property, 8000 for commercial losses, and 65 000 for vehicles. The claims of automobile damages in the suburbs of north St. Louis totaled 38 000. Twenty-four jet aircraft at Lambert Field in St. Louis were badly damaged by the hail, and 67 flights were canceled for up to 24 h.

4. Summary

If the April 2001 storm had occurred 25 or more years ago, the losses would have been vastly less. Much of badly damaged suburban areas of St. Louis and Kansas City were then rural farm fields. Thus, there is no great surprise that the most damaging hailstorm of the past 53 years occurred recently. The 1990s had a series of quite damaging urban hailstorms that produced extensive damages in Denver, Fort Worth, Orlando, Wichita, and Oklahoma City (Changnon 1999). These have given rise to the recognition that the ongoing demographic shifts to urban areas, causing rapidly growing metropolitan areas, have increased the potential for costly losses from hailstorms and other forms of severe weather (Changnon et al. 2000). During the 1951–70 period, property hail losses represented 46% of the total hail losses, but in the 1981–97 period, property losses exceeded crop–hail losses and had become 61% of the national total from crops and property (Changnon 1999).