Hurricane Dorian Gives Scientists a Chance to Test Local Data Devices

This 'spaghetti map' from the Weather Channel shows the different forecast tracks produced by the many computer models.
This ‘spaghetti map’ from the Weather Channel shows the different forecast tracks produced by the many computer models.

According to a local scientist, Hurricane Dorian was an unexpected gift. A gift, he said, because it provided an opportunity to test the reliability of local weather data devices.

Unexpected, he said, because Dorian shifted course and passed completely over St. Thomas. That was something retired University of the Virgin Islands Professor Roy Watlington said hasn’t happened in a long while.

Watlington is a physicist and oceanographer whose research over a quarter of a century has provided insights on climate change, currents, tsunamis and weather patterns. He is also the founder of Ocean Coastal Observing Virgin Islands, a registered non-profit focusing on ocean research.

Because of efforts made by members of OCOVI and their collaborations with scientists in Puerto Rico, weather data collected in the Virgin Islands Wednesday made its way into weather segments seen on national news networks.

By Wednesday morning – after forecasters with the National Hurricane Center forecast the storm crossing eastern Puerto Rico – there was a shift in direction that put St. Thomas in Dorian’s crosshairs.

On CNN, weatherman Chad Meyers said he tracked the storm throughout Tuesday night, searching for an explanation. “The American model, the European model, how one is better than the other. They both have junk,” Meyers said. “There’s something going on in the atmosphere that they don’t see; that those models don’t know about.”

Weather Underground blogger Dave McDermott posted his thoughts on the sudden shift on social media Thursday. McDermott suggested that when Tropical Storm Dorian made landfall on the island of St. Lucia, the storm veered off to the east of Puerto Rico, scoring a direct miss.

Watlington called that a plausible explanation, but not enough by itself. Technological developments in data collection over the past 15 years had improved weather forecasting, he said. At least for seasoned scientists. To the layman, Dorian’s unpredicted shift was mind boggling.

“I was speaking to a family member who was angry. He said these forecasts models are just B.S. But they are not B.S.,” he said. “Right now the Florida coast knows it’s going to get hit. Fifteen years ago it would have been a mystery.”

To properly track a storm, scientists rely on a formula, Watlington said. They start with boundary conditions, check satellite data and ground truths. Finally, he said, they check the observations of Hurricane Hunter aircraft and collect stories from persons going through the storm.

Computer models, like those seen on TV, cannot compare to following the formula, he said.

“Computer models start with equations. Equations are purely theoretical. They give sample results that may not have a lot to do with reality. Then you try to make that model real by setting boundary conditions,” Watlington said.

A closer approach to weather reality relies on ground truths, hard data supplied by weather stations. The one most cited as Tropical Storm Dorian took a Wednesday morning jog and rolled over Charlotte Amalie came from a weather station set up on Buck Island in 2010.

“We’re so proud of Buck Island. Two years ago it gave the maximum readings on Hurricane Irma. It was the first time the National Hurricane Center used data from the Virgin Islands in its weather reports,” he said.

But sometime after providing those readings, Irma became overbearing and the structure supporting the Buck Island instruments gave way. Since then the instruments have been remounted and were on the job late Tuesday night into Wednesday morning, Watlington said.

Ground truths were also supported by undersea gliders. Tube like devices that function like underwater drones, gliders have been deployed in the waters around the Virgin Islands for almost 10 years, the oceanographer said.

When a storm like Dorian passes over, the gliders dive deep, he said. As it passes, gliders rise to the surface and send their collected data to weather satellites.

Still, when added together, predicting the path of an ocean going storm is still inexact. Watlington said that’s because ground truths are being collected over water. Compared to measuring a storm crossing land where there may be weather stations at reliable intervals, there is no comparison.

What ground truth devices deployed across St. Thomas were able to supply this time was evidence that the eye of Dorian passed over the island in its entirety. Something the researcher said he could not recall seeing before.

Weather stations at Two Brothers, Buck Island, Rupert Rock, and a newly placed station at Mountain Top, captured the moment.

It was achievable because Dorian came in as a small tropical storm over a small island – St. Thomas – and morphed into a Category 1 hurricane while in transit.

“It is the first time in a long time that we had a direct crossing. This one entered some place around Buck Island and left somewhere around Santa Maria, Hull Bay. Both sides of the eye wall entered St. Thomas,” he said.