When you buy through affiliate links in our content, we may earn a commission at no extra cost to you. Learn 
In the aftermath of an active 2023 hurricane season—the fourth highest since records began—all indicators point to another demanding year for hurricane forecasters in 2024.
Yet, there is hope on the horizon as artificial intelligence (AI) and machine learning (ML) algorithm applications are gaining traction when applied to hurricane forecasting and intensity modeling.
Hurricanes — tropical cyclones that occur in the North Atlantic and the central and eastern of the North Pacific — are notoriously tricky to forecast, as no two hurricanes are ever the same. Changes in atmospheric conditions and warmer sea temperatures can rapidly energize a hurricane’s intensity, bringing with it devastating wind speeds, increased rainfall, and, frequently, a change in its trajectory.
Traditionally classified based on wind speeds as measured against the Saffir-Simpson scale, hurricanes are classed from a Category 1, escalating upwards towards a Category 5. However, scientists are beginning to suggest that the increased number of mega-hurricanes (with wind speeds in excess of 309km per hour) could now be labeled Category 6.
Key Takeaways
- 2023 was the fourth most active hurricane season on record.
- According to NOAA, hurricanes that make landfall cost an average of US $22.8 billion per event.
- Innovations in AI demonstrate the opportunities to outperform traditional forecasting methods in predicting storm trajectories.
- The rise in mega-hurricanes and the proposal for a new Category 6 classification underscore the influence of climate change and the need for AI to help forecast storm intensity.
With five tropical cyclones surpassing the Category 6 threshold since 2013, it presents a strong case that climate change is beginning to take effect, and hurricane forecasters are turning to AI for help.
Aside from the increased risk to life, the average cost of a hurricane that makes landfall equates to US $22.8 billion per event, according to the National Oceanic and Atmospheric Administration (NOAA), which is why you can sense the urgency in scientists’ quests for improved hurricane forecasting.
While researchers agree that 100% accuracy in hurricane prediction may never be achieved, recent results from increasingly sophisticated AI tracking and modeling show promise in improving conventional forecasting techniques.
From real-time tracking and path prediction to anticipating hurricane wind speeds and storm surge levels to facilitate advanced early warning systems, we’ll explain how AI is making an impact on monitoring these monster storms and its potential to save lives ahead of future hurricanes.
AI’s Pivotal New Role in Hurricane Ground Tracking
Perhaps AI’s most valuable trait is its unparalleled ability to analyze and interpret extensive datasets far faster and more efficiently while also reducing the cost of hurricane forecasting.
Fortunately, scientists have an abundance of hurricane data gathered from as far back as 1851. Of course, the data variables have dramatically increased since hurricane monitoring first began, but AI has an extensive archive of backdated historical information on which to calculate future predictions.
On top of the decades of hurricane data, the National Hurricane Centre (NHC) in Florida and private firms, such as WindBorne Systems in California, are also utilizing real-time data collection to facilitate AI in early-stage detection further, as well as hurricane tracking and strength predictions.
A prime case study of this was WindBorne’s advanced AI global forecast model, WeatherMesh, which tracked the predicted path of Hurricane Ian, a Category 5 storm in 2022.
Utilizing advanced, AI-driven weather balloons and their deep learning numerical weather prediction (DLNWP) model, WeatherMesh claims to have outperformed the National Weather Service’s (NWS) in both short-term and long-term forecasting of up to 70 hours.
AI’s Ability to Monitor Hurricanes is Leading to Advanced Warning Notifications
The capacity of weather scientists to predict where a storm will make landfall is a crucial aspect of hurricane forecasting. After all, gaining a better understanding of a tropical cyclone’s projected path allows emergency services to accurately enforce early warning systems that help evacuate impacted areas and potentially save lives.
AI’s growing abilities in determining the storm’s trajectory, combined with its potential to forecast other weather-related phenomena associated with hurricanes, including wind speeds, rainfall, and storm surge projections, could soon provide even more advanced warning to the necessary agencies to escalate evacuation orders when needed.
A prime example of where this advanced technology would have been beneficial was during Hurricane Katrina, a Category 5 storm that impacted large swathes of Louisiana, Mississippi, and Alabama, but most notably the destructive flooding of New Orleans.
While Katrina devastated a vast area of the southern US coastline, its rapid intensification in the Gulf of Mexico, coupled with the resulting storm surge, completely overwhelmed New Orleans’ flood defenses, resulting in catastrophic damage and loss of life.
Incidents like Katrina have certainly fuelled research as to how AI might be able to predict potential escalations in future hurricane metrics, with a recent article published by the American Meteorological Society (AMS) detailing a study of a two-method approach utilizing ML convolutional neural networks with one-billion parameters to predict current and short-term tropical cyclone intensity changes.
These two AI models, known as ‘D-Print’ and ‘D-Mint’, demonstrated an ability to accurately predict hurricane wind speeds to within 10 knots, utilizing real-time infrared imagery, environmental scalar predictors, and microwave energy.
Examples of these key innovations in monitoring and predicting hurricane conditions demonstrate AI’s potential to identify future Katrina-esq intensifications, leading to advanced evacuation and preparedness in both rural and urban areas.
AI Hurricane Forecasting – Creating A Safer Tomorrow
Due to the heightened risk of climate change affecting tropical cyclones across the globe, it seems only logical for weather scientists to keep pursuing AI’s continued integration into hurricane tracking and prediction models.
Considering a warming planet is the perfect catalyst to alter the formation and intensity of every hurricane, applying AI to process swaths of real-time data instantly, like barometric pressures and sea surface temperatures, makes a lot of sense.
By integrating live readings from weather balloons, meteorological satellites, and radars alongside analyzing historical data, AI is set to become a primary tool for early detection and interpretation of hurricanes’ devasting attributes, even before they develop, easing the burden of modern-day meteorologists.
For now, traditional physics-based predictions and data interpretation still remain vital ahead of the 2024 hurricane season. Yet, ultimately, the continued adoption and integration of AI into the science of hurricanes can only improve our chances of reducing the impact of these monumental storm systems on those affected.
