About this platform

The Seasonal Hurricane Predictions platform provides free access to seasonal forecasts issued by a number of universities, private sector entities and government agencies around the world for the upcoming hurricane season in the North Atlantic.

The platform allows forecasters to make these predictions freely available to the wider public, compare the predictions with real-time hurricane activity, and obtain the best estimate of upcoming hurricane activity.

This platform has been co-developed by the Barcelona Supercomputing Center (BSC) and the Colorado State University (CSU), with support from the global (re)insurer AXA XL. Learn more about us by visiting our Team page.

What can you find on the Seasonal Hurricane Predictions platform?

The home page shows the mean hurricane activity predicted for the season, as well as the number of hurricanes that have occurred to date. It also provides a visualization and overview of the predictions for the current hurricane season in a plot that displays the latest forecasts submitted individually by the main forecasting centers. The information on the main page provides useful material to the wider public, journalists, and other audiences interested in the evolution of the hurricane season. The plots and raw data can be downloaded for use, ensuring appropriate attribution.

The predicted hurricane activity for the season is shown in three categories: low, normal and high activity.

More detailed information is available in the predictions page, which provides predictions on the number of hurricanes, named storms, major hurricanes and accumulated cyclone energy. 

Other available information on the platform includes a post-mortem analysis of previous hurricane seasons, information on past annual activity dating back to 1966, and background information related to hurricanes and factors influencing their variability.

What was the motivation behind this platform?

Hurricanes typically occur in the Atlantic Ocean region between June and November each year. These violent storms are formed over the ocean and can move over land, causing destruction to homes and other infrastructure and posing a serious threat to the affected populations.

The first attempt at forecasting the Atlantic tropical cyclones dates back to 1984. Performed at CSU, this first forecast utilized the state of El Niño, the Quasi-Biennial Oscillation in the stratosphere, as well as the surface pressures in the Caribbean and Gulf of Mexico to predict the Atlantic hurricane activity.

More than three decades later, seasonal hurricane predictions are produced ahead of each hurricane season by a range of different research groups, from meteorological services to academic groups and catastrophe modelers. These predictions cover almost all major ocean basins that support tropical cyclone formation. For the Atlantic basin alone, a large number of groups are currently producing seasonal predictions in the months leading up to the climatological start of the peak of the hurricane season (August 1st).

These predictions are generally made freely available by the groups who produce them. However, gathering this large amount of information in one place and making comparisons between different predictions to obtain the best scientific estimate for the upcoming season is a challenging task, in particular considering that the number of such predictions is likely to keep increasing in the future.

To respond to this need, the Seasonal Hurricane Predictions platform was created to centralize predictions on the upcoming hurricane season and provide forecasters a platform to submit their latest seasonal predictions, and at the same time facilitating access to these predictions for experts, as well as the wider public.