The number of hurricanes, major hurricanes, and U.S. landfalling hurricanes in the 2019 Atlantic hurricane season was predicted to be close to the 30-year average. In the event, the numbers matched that average exactly. The number of named storms, however, was anticipated to be slightly above average—and proved to be a little higher still.
Five storms directly impacted the mainland United States, but only 2 of them made landfall. The season will be remembered chiefly for the devastation of the Bahamas by Hurricane Dorian, the fifth Atlantic Basin Category 5 storm in four years and the slowest-moving on record. For the United States, this year’s most costly named tropical cyclone, the soaker Imelda, never achieved hurricane status but was extremely slow and wet.
With the season now well and truly over, we can announce the winner of AIR’s fourth client hurricane contest. We had a grand total of 121 entrants and, for once, no tie for first place. Congratulations to Jared Bailey, a Catastrophe Actuary at Aon! Jared has won an Ambient Weather WS-2902A Smart WiFi Weather Station with Remote Monitoring and Alerts. This is the second year in a row that the winner of this contest has come from Aon.
What Was It All About?
The object of AIR’s contest is to guess:
- How many named tropical storms will form in the Atlantic during the year?
- How many of those tropical storms will become hurricanes?
- How many will become major hurricanes (Saffir-Simpson Category 3 or higher)?
- How many hurricanes will make landfall in the mainland U.S.?
Jared predicted 18 named storms, 7 hurricanes, 3 major hurricanes, and 2 U.S. landfalls. The actual totals for the year were 17 named storms, 7 hurricanes, 3 major hurricanes, and 2 U.S. landfalls. So, Jared’s numbers for tropical storm count, major hurricane count, and landfall count were spot on and he was only out by one on the named storm count. Like last year’s winner, he had only 1 error point. Congratulations Jared!
Most entrants plumped for an above average number of named storms and many were close to the actual figure, but some went as low as 3 or as high as 48. Predictions for the other categories exhibited a similarly wide range.
How Did Staff at AIR Do?
The short answer is, “not any better.” AIR runs an almost identical hurricane contest for our employees, and this year 125 entries were submitted. We have no way of knowing what weird and wonderful techniques people used to come up with their predictions, but while many of them were surprisingly close to the actual figures some were wide of the mark. Assistant Vice President and Principal Engineer Karthik Ramanathan, Ph.D., won the contest with a perfect score; second and third places both had just one error point each and were once again decided using our infamous tie-breaker. The mean error this year was 7.2 (in 2018 it was 7.1—lower is better), and the maximum error was 20.
AIR’s hurricane contest is a fun reminder of just how difficult it is to predict seasonal Atlantic hurricane activity. We hope you enjoyed this year’s contest and look forward to launching the 2020 edition in May!