NEW HURRICANE DORIAN REPORT DETAILS FORECAST CHALLENGES & SUCCESSES
Updated: May 1, 2020
The National Hurricane Center’s final report on Hurricane Dorian is out. Information gathered from across the Caribbean, the Bahamas, Florida, and up the coast to Canada is the final best estimate of exactly what happened over the three weeks from the time Dorian developed in the tropics until it dissipated in the North Atlantic.
The assessment of the top wind-speed in Dorian came in at 185 mph – the same estimate the NHC forecasters made as the Category 5 hurricane was hitting Abaco Island. That makes Dorian the strongest storm to hit the northern Bahamas in modern records.
But... the top-wind number is tentative, pending ongoing research. More on that later.
The report highlights the forecasting successes during Dorian, and the challenges. Before the storm got the Bahamas, the system’s track and intensity were not well forecast. Essentially, Dorian refused to do what the computer forecast models said it would.
To the good, it was very unusual to have a giant and powerful hurricane moving west across the northern Bahamas, and have no watches or warnings posted for Broward or Miami-Dade County. This forecasting success saved the residents of Dade and Broward a lot of money and aggravation.
Dorian’s extremely slow movement helped. There was confidence that the storm wasn’t going anywhere fast in the atmospheric environment of weak to nonexistent steering currents that was present over the Bahamas at that time. There would have been time to adjust the warnings if Dorian had significantly deviated from the forecast track.
Warnings were issued for Palm Beach County and points north on the Florida coast. In the end, they were not all needed, but the best computer models forecast the center to track close to the coast, so there was no choice. In the plot below, the blue lines are the official NHC forecasts, which were based on the best available computer models, and the white line is Dorian’s final track. You can see that they lined up well off Georgia and the Carolinas, but off Florida, the forecast tracks indicated that significant impacts along the coast were possible.
And there was more. A week before Dorian got anywhere near Florida, a weak version of the storm was approaching the Caribbean. At that point, the forecast models performed exceptionally poorly. The computer forecasts insisted that the center of an intensifying hurricane was going to track near or just west of Puerto Rico and over the Dominican Republic. In the end, Dorian hung a right and surprised everybody by tracking over the Virgin Islands. It was a very unusual miss. The forecast was so far off that hurricane warnings were not issued in advance.
I don’t remember that happening in modern times.
The graphic below shows the computer-model forecasts and the track Dorian ended up taking. The difference is shockingly large.
We always say, forecasts for weak systems generally have much larger errors. In this case, not only was Dorian a weak tropical storm, it also tracked right over the tall mountains of the island of St. Lucia. That combination scrambled the forecast. That rule about weak systems is important to remember for the future.
As far as how strong Dorian was going to be, during the time the hurricane was between the Caribbean and the Bahamas, the intensity forecasts were terrible as well. None of the most reliable computer models came close to estimating the extreme increase in strength. As a result, the official National Hurricane Center forecast didn’t pick up on Dorian’s incredible intensity jump until it was already happening.
The graphic below shows that most of the computer models forecast Dorian to strengthen, but nothing close to what actually happened. The wind speeds on the left are in knots, so to get miles per hour, you multiply by 1.15. The 185-mph peak came right when the Category 5 hurricane was approaching Abaco in the northern Bahamas.
We know that modern science delivers much better forecasts of where a hurricane is going than how strong it is going to be when it gets there. And Dorian was no exception. Even with the slight forecast miss offshore of Florida and the dramatic deviation in the Caribbean, the overall average track forecast errors for Dorian were quite low. No doubt, scientists will be researching what went wrong with both the track and intensity models for Dorian for some time.
An important research project is underway that could change Dorian’s peak-intensity number. In the super-intense hurricanes we’ve seen over the past few years – Irma as it was approaching the Caribbean islands, Michael as it was approaching Panama City, and now Dorian near the Bahamas – the instruments on the Hurricane Hunters that estimate the wind speed at the ocean’s surface have been showing some extraordinarily high values.
The calibration system we use today to convert the data measured by the plane into ground-level wind speeds cranked out numbers over 200 mph in Dorian, which gave everybody pause.
The question has been, are the super-high numbers representative of what’s actually going on at the earth’s surface, or are they artifacts of the measuring system or the instrument itself? The National Hurricane Center has been ignoring these extreme readings while research is ongoing to figure out what’s happening. The hope is that that research will be completed by early next year.
It’s possible that the estimated peak wind speeds in Irma, Michael, Dorian, and other strong Category 4 and 5 hurricanes will be adjusted based on new science.
This kind of revision has happened before. Hurricane Andrew was upgraded to a Category 5 based on new measurement techniques deployed in the late 90s that lead to a new understanding of the structure of strong hurricanes. It is possible that this new research will lead to another leap in understanding, and a revision to the wind-speed estimates in past super-strong storms.
Dorian reminded us that even with the best computer forecast models, and analysis by world-class forecasters at the National Hurricane Center, we still have to take into account the possibility of a significant forecast miss. It happens less often as science progresses, but Dorian was a great reminder of the uncertainty intrinsic in hurricane forecasting.
Sources in the Bahamas put the death toll from Dorian at between 200 and 300. It’s clear that we will never know exactly how many people died during the hellacious hurricane. And now, the Herculean effort to rebuild there is only beginning amid the COVID-19 pandemic.
For perspective, think about our friends on Abaco and Grand Bahama Islands and what they’re facing, when what we’re going through feels like too much.