Lessons learned from the 2018 Season

[galaxy401]

As hurricane season is wrapping up, I think it's time to reflect on what has happened this year and what we can learn from it. What are some lessons that you took out of this season regarding forecasts, climatology, etc?

For me, and I'm sure for several people, it is to not judge a whole season with what it looks like in August. Even going deep into the season as the last week of August, conditions looked hostile out there with very cold SSTs. Yet that rapidly went away in September and the Atlantic saw an active year and an active MDR.

[CyclonicFury]

The biggest lesson I learned from the 2018 season was to not underestimate the Atlantic in an active era, even when the SST profile is not conducive.

Although the Atlantic MDR was extremely cold relative to average early in the season, the African monsoon was strong. That was a key indicator that the season could be more active than forecast. When we had Beryl in July, many did not take it as a sign that the season could be active, but it ultimately was one.

Finally, just because it is an El Niño does not mean the Caribbean and Gulf will be dead all season. We saw this with Michael in October which nearly reached Category 5 strength.

[AnnularCane]

Take pre-season forecasts with a grain of salt. Although that is something I've been learning over the past few years, not just this year. Sometimes I still forget though.

This isn't specific to any particular year, but something people seem to forget a lot: An average or below average season doesn't mean you're safe.

[Hurricaneman]

Take pre-season forecasts with a grain of salt. Although that is something I've been learning over the past few years, not just this year. Sometimes I still forget though.

This isn't specific to any particular year, but something people seem to forget a lot: An average or below average season doesn't mean you're safe.

1983 and 1992 prove that

[Cerlin]

Hurricanes never fail to surprise us. Just when we think we know all there is to know about the quality of a season by August, September fools us all. Just when we think a storm can’t do something or that all the analogs and models point to a general trend or path for a storm, it can be completely different and defying in days or even hours.

Hurricane Michael taught me never to write off a storm before it’s finished. I found myself continually saying “this won’t be more than a TS” or “it’ll peak at a Cat2” or “it can’t intensify while hitting land”, and yet every time I was wrong. Hurricanes will always be predictably unpredictable and each storm isn’t finished until it’s finished. As someone who is learning meteorology, this season has taught me valuable things about model reading and forecasting that I’ll be able to use for the future. Thank you to everyone for your valuable insight on a wacky 2018 season and for inspiring me to learn more and more about meteorology and hurricanes.

EDIT: Just wanted to add that I went through the Florida panhandle this morning and the destruction of the trees on I-10 was so astonishing—really shows how destructive Michael was, even after a month and a half since landfall. A gas station that I stopped at just barely had it’s pumps working, and this was well off the coast too.

[AnnularCane]

Hurricane Michael taught me never to write off a storm before it’s finished. I found myself continually saying “this won’t be more than a TS” or “it’ll peak at a Cat2” or “it can’t intensify while hitting land”, and yet every time I was wrong. Hurricanes will always be predictably unpredictable and each storm isn’t finished until it’s finished. As someone who is learning meteorology, this season has taught me valuable things about model reading and forecasting that I’ll be able to use for the future. Thank you to everyone for your valuable insight on a wacky 2018 season and for inspiring me to learn more and more about meteorology and hurricanes.

Oh yeah, I just remembered another one. Not all northern Gulf storms will weaken before landfall.

[cheezyWXguy]

Given that Michael significantly exceeded original landfall intensity expectations, and Florence fell significantly short of landfall intensity expectations, I think the 2018 Atlantic season perfectly exemplifies the following points:

1) Track forecasts have improved immensely over the years, but even though intensity forecasts have improved as well, there is a long way to go with them. What often makes it difficult, is that small-scale or storm-structural factors that models may not be able to pick up on can have significant effects on intensity. This level of nuance in forecasting is generally lost on the general public and those in the risk areas, and can result in a perceived "cry wolf" effect, like in Florence, or a "caught with your pants down" effect like in Michael.

2) There are a lot more factors to a storm's impact than its maximum sustained winds. Storm surge, duration, rainfall rates, and overall size are often more important and affect a far larger amount of people. Surge and flooding kill more people than winds, and a sole reliance on Saffir-Simpson ratings is dangerous. Major props to NOAA/NHC for amending the warning system to account for this through issuance of storm surge warnings, flood warnings, and severe weather warnings, as well as use of extreme wind warnings to identify the specific area expected to endure maximum sustained winds.

3) Even chasers aren't perfectly adept, and can still find themselves getting more than they bargained for, despite the increase in technology and knowledge available at their fingertips. Watching Jeff Piortrowski endure Harvey in the blue shed last year was one thing, but Michael really hit that point home when several chasers were missing for hours after the storm had passed. There were moments in hearing about all of this during the aftermath that reminded me of the El Reno tornado in 2013.

4) But most importantly, and all of the above factors into this, if you live in an area that has been issued a mandatory evacuation, just go. This is easy for me to say, because I live 300 miles from the coast. And in a lot of areas, road networks are not sufficient for mass movement of people in that scale -- urban or rural. Gridlock often occurs and when panic is incited, just as many people can die in the evacuation as the hurricane itself (see Hurricane Rita). But at the end of the day, things can be replaced, and lives cannot. Prepare as much as you can in advance, protect your home, take note of what you valuables you would bring when evacuating, what items you would leave. Know where you would go if you had to leave. Know your area's topography. If your beachfront home is 3 feet above sea level and is expecting an 8 foot surge, it doesn't matter what logistical challenges you face in evacuation. You will very likely die if you stay.

[Hurricanehink]

This year featured tropical threats to Hawaii, Mexico, Florida, North Carolina, Portugal, and Greece. Also one of these years, Europe is gonna get hit by a bonafide hurricane, if not from the Atlantic (almost like Leslie) than from the Mediterranean (like Zorbas). As the equator regions get warmer, expect more development and intense hurricanes in the mid-tropics. Florence had the potential to be a big storm in the Carolinas per computer models, and perhaps it got too strong too far east and entrained too much dry air. Future years like 2018 could steer an intense hurricane into the Mid-Atlantic or Southern New Jersey if they move faster and form closer. If Michael was on the Atlantic side and was a Category 4 moving northward, it would've had more latitude, perhaps a Hazel 54 equivalent. Interestingly, this year had little activity in the Caribbean, in contrast to 2017. The worst Caribbean impacts this year were in May from Alberto. That makes it the 7th of the last 12 years with a May tropical cyclone.

The main lesson from Michael IMO is that the Gulf Coast needs to be ready for intense hurricanes from June to November. Hopefully the US military learns that it shouldn't keep multi-million dollar aircraft in the tropics without being able to evacuate them all! Tyndall Air Force Base lost $6 billion in equipment. For perspective, that is 0.8% of the U.S. military's $716 billion spent in 2018.

Clearly the Atlantic is still capable of churning out some active seasons, but it was the Pacific that was more active. I fear that Mexico and Hawaii are due for some strong hurricanes in the near-future. In the WPAC, the Philippines continue to bear the brunt of powerful typhoons, as did the Northern Marianas Islands. We'll see the long-term effects of powerful storms in these areas. In the Northern Indian Ocean, Cyclone Sagar reached caused damaging rainfall as far west as Djibouti and far northwestern Somalia, the worst storm in many years. The Arabian Peninsula and India remain a common target of storms.

[TheStormExpert]

Do not put too much stock into pre/early season forecasts. As we saw even back in mid-late May or as late as early-mid summer the forecasters were expecting a bonified El Niño to develop and shut things down. That clearly failed to happen (once again) even though we only saw 2 major hurricanes...

...Also, DO NOT focus on the numbers! Florence and Michael both being the only major hurricanes, with Florence originally forecasted to be a major hurricane U.S. landfall would have made a good argument case like in 1992 with Andrew. Still Florence was a major impact in terms of flooding for the Carolinas and the SE U.S. Then came Michael the 3rd most intense U.S. landfalling hurricane behind Camille, and the 1935 FL Labor Day Hurricane, it's possible it was even a Cat.5 at landfall! Regardless it broke several notable records for the FL Panhandle being the only Cat.4 or stronger to ever strike the panhandle and in October.

All in all 15/8/2 is impressive knowing what was being forecasted early on and the overall negative state the Atlantic was in most of the season.

[hcane27]

I believe that I realized yet again that the forecasts are based on forecasts. If an earlier-in-the-chain of forecasts , chances are the rest will as well , leading to the inevitable question ... "How did they get it so wrong !?!?!?!?!"

[hcane27]

left out the "busts" in the fiorcast chain ....

[hcane27]

Man !!!! typing is Atrocious !!!! should be forecasts .....

[al78]

One lesson to be learned is when it comes to seasonal forecasts (or any forecasts), probabilistic forecasts are important. Although the deterministic seasonal forecasts went for below normal activity, the probabilistic forecast of a near or above average season was not zero. People can get too hung up on the deterministic numbers, which is only the most likely outcome, and ignore the range of possible outcomes and their probabilities based on the hindcast error distribution.

[Hammy]

I think the biggest lesson from this season is not to use climatology as an absolute--most people here (myself included on several occasions) and even many weather "pundits" so to speak were guilty at one point or another of rejecting possibilities on the grounds of "that's never happened before" when nearly every single storm this season (as well as August as a whole relative to what happened later) defied climatology in some way or another.

[cycloneye]

Dr Rick Knabb has some lessons learned points.

 https://twitter.com/DrRickKnabb/status/1068633925269053440