A frightening weather forecast showing a hurricane that struck the Gulf coast in the second half of June turned to social networks this week – but do not panic.
It is the first “ghost hurricane” of the season.
A similar media threshing plays every season of hurricanes, especially at the beginning: a worse and worse picked scenario model race becomes viral, but more often than not, will never materialize.
Unofficially debusing “Ghost Storms” or “Ghost Hurricanes”, these tropical systems appear regularly in meteorological models – computer simulations that help meteorologists to predict future conditions – but never seem to manifest in real life.
The model responsible for this week was the global forecast system, also known as GFS or American model, managed by the National Oceanic and Atmospheric Administration. It is one of the many used by forecastists around the world.
All models have known biases or “quirks” where they tend to suddenly or under-predict certain things. The GFS is known to overexploit tropical storms and hurricanes in longer -term forecasts that look more than a week in the future, which leads to these false alarms. The GFS is not alone in this area – not all models have difficulty predicting tropical activity so far in advance – but it is known to do so.
For example, the GFS could spit a prediction for an American hurricane of about 10 In days, only for this chance to disappear completely as the forecast date approaches. This can happen at any time of the year, but is the most common during the hurricanes season – from June to November.
This is exactly what happened during last week when the forecasters keep an eye on the first storm of 2025 Atlantic hurricanes season.
No weather forecast model is designed exactly the same way as another, and that is why everyone can generate different results with similar data.
The reason why the GFS has more false alarms when you look for more than a week than similar models – such as European ECMWF, CMC of Canada or UKM from the United Kingdom – is because it is exactly what it is programmed to do, according to Alicia Bentley, the project of project of global verification of the environmental modeling center of the Noaa.
The GFS was built with a “convection diagram of cumulus configured low”, according to Bentley. In clear language, this means that when GFS thinks that there could be thunderstorms that develop in an area where tropical systems are possible – on the oceans – it is more likely to jump to the conclusion that something tropical will develop than to ignore it.
Other models are not designed to be as sensitive to this phenomenon, and they therefore do not show a tropical system until they are more confident than the right conditions are in place, which generally occurs when the forecasts are closer in time.
The Western Caribbean Sea is one of the favorite places of GFS to predict a ghost storm. It is because of the Central America gyre: a large disorganized area of ​​showers and thunderstorms that turn in the region and its surrounding water.
The combination of abundant humidity and rotation in the atmosphere makes it a first -rate reproductive ground for storms, especially at the start and during the top of the hurricanes season. Given the sensitivity of the model, it is quick to jump on these possible storms.
But this sensitivity has an advantage: by highlighting almost everything that could become tropical, the GFS lacks very little real storms.
Its tendency to cry Wolf is not ideal, but the GFS team found that it was worth giving the model a greater chance to catch each storm and better predict the intensity of each than to prioritize less false alarms, explained Bentley.
“It was essential to improve the probability of detecting tropical cyclone training and intensity forecasts of the tropical cyclone … and we have done it,” said Bentley.
During the 2024 Hurricane season, the GFs had the slightest error During the forecast of the intensity of tropical cyclones – tropical depressions, tropical storms and hurricanes – from any other world forecast model in its class: ECMWF, CMC and UKM. However, the ECMWF and the UKM have surpassed the GFS in the tropical track provide for more than five days in the future.
Despite the frequency to which the GFS evokes ghost systems in its longer term forecasts, it cannot be updated.
“The crucial role of the forecastist is to understand the known biases of a model and to use this knowledge to their advantage to produce a better forecast,” said Bentley.
Knowing the GFS clings to everything that could even become vaguely tropical well in advance helps the forecasters to keep an eye on the areas where the conditions can finally meet to create the next hurricane.
According to Bentley, the most reliable solution to predict tropical behavior is to take advantage of the forecast of the sets, according to Bentley.
“A deterministic model like the GFS produces a forecast at the same time; it gives an answer,” said Bentley. “An overall forecast can show you a variety of possible results, as well as forecasts that resemble possible aberrant value.”
Unlike click bait on social networks, no well -thought -out forecasts are made from a single model. The forecasters use everything available to them – deterministic and overall models, observations, climatology and more – to predict the weather as precisely as possible to Give people the time and the information they need to stay safe.
The National Hurricane Center, for example, generally uses a mixture of different types of models to make their forecasts. This strategy, combined with in -depth expertise, has led to their Most precise track forecasts Registered for the Atlantic last season.
