Select Page

When hurricanes form, meteorologists like me bring out beautiful maps with precise detail. The hurricanes forecast track even has a line down the middle. And yet, as has been proven this week with Hurricane Dorian, theres a whole lot about hurricane forecasting we still dont know.

The Windward Islands, Puerto Rico and the Dominican Republic were scared then spared.Were not sure yet, but thats probably going to be the case for Florida too.

This is not a question of whether forecasters can work harder or put in more hours. When it comes to predicting hurricanes, weve reached the current limits of science.

There is nothing on Earth more complex than the atmosphere. Thats the bad news. The good news is that the atmosphere has to follow the laws of physics.

It should be simple.Assign values to the atmosphere, crunch it with the laws of physics and voila.

Heres where the complexity comes in:In order to model the atmosphere, we have to approximate it.For the so-called American Model, our governments GFS or Global Forecast System computer model, thats every 18 miles. We make believe each of these grid squares is homogenous. Theyre not.

Eighteen miles is pretty rough when you look at the atmospheres intricacies.In practice it works well for synoptic systemslike the larger weather systems which travel across America during the winter leaving a trail of snow hundreds of miles long. But hurricaneseven monsters like Dorianare so small they often fall through the cracks.

Step one in modeling a hurricane is giving it a thorough exam.The models are initialized with surface observations like you might get at an airport, plus radar, satellites, buoy and ship reports and, during hurricane season, airplanes that purposely fly into the storm and drop instrument packages right in the eye.

Sometimes the planes have mechanical trouble and cant fly.In the ocean, where Dorian has spent most of his time, surface data is tough to come by. On Friday Dorian wasnt within 250 miles of any buoy or weather station. This makes initialization tough.

There are specialized computer models just for tropical systems with higher resolution but a shorter time spanand certainly they helpbut Floridians have already spent real money, effort and grief for what will mostly be a false alarm for them.

It sucks to be wrong.We tell people we can predict the future.Mostly we can. Its scary to worry some people might not trust us next time.

As of early Monday morning Hurricane Dorian was about 40 miles east of Freeport, Grand Bahama Island, 125 Miles from West Palm Beach (and a little closer to Mar-a-Lago, to answer some Facebook friends). Dorian is a Category Five storm with sustained winds at 175 mph.That slight drop in wind speed to 175 mph might represent the beginning of an eyewall replacement cycle. Dorian is moving to the west at 5 mph but will slow to walking speed or slower. Some guidance even shows it reversing course for a short time.

All this means the Northern Bahamas will continue to take a long and crushing beating with winds at tornado speedbut lasting a day or two. Storm surge tides of 18-23 feet might overwash some islands entirely.


On these flat islands, just a few feet above sea level, it will be as bad as any recent natural disaster weve seen.

Most likely (but not with certaintysee above) Florida will be spared a direct hit. There will be squally rain, gusty winds, excited surfers and the potential for a damaging storm surge if Dorian moves farther west.

A somewhat stronger upper wind flow will gradually turn Dorian north and up the coast.Winds will diminish but the rainfall will not. Its likely some areas in the Coastal Carolinas will get a foot or more of rain toward the end of the week.Some of these areas saw historic flooding last year.

After the Carolinas its off to sea, where Dorian will lose its tropical characteristics and just become another weather system.One model has Dorians remnants bringing snow to Greenland next week.

Read more: https://www.thedailybeast.com/why-hurricane-dorian-is-so-hard-to-predict