Hurricane Dorian highlights urgent need for improved forecasts

The catastrophic impacts of Hurricane Dorian are a grim reminder of the need to invest in research and technology to improve forecasts of hurricanes and other storms. Weather prediction has become significantly more reliable in recent years, but we still need far more precise forecasts of the track, intensity and timing of hurricanes to safeguard society.

The good news is that forecasts have improved dramatically just in the past couple of decades. Meteorologists can now predict the track of a hurricane three days in advance to within about 120 miles, which is twice as accurate as forecasts in the late 1990s. This has important real-world implications: Most Floridians could continue with their daily routines as Dorian churned offshore, whereas Hurricane Floyd in 1999, which followed a similar track as Dorian, triggered widespread and chaotic evacuations because no one could foretell where it would go.

Even the more precise track forecasts, however, still contain too much uncertainty. As Dorian neared the Caribbean, meteorologists struggled to predict whether it would hit Puerto Rico, Hispaniola or the Bahamas. When the storm moved toward the mainland United States, it tracked so closely to the coast that they could not be confident it would remain offshore. As a result, officials had to make costly preparations, including relocating military helicopters and ships from coastal bases in Florida and Georgia and dispatching the U.S. Navy fleet in Norfolk out to sea.


Weather models also failed to predict Dorian's rapid intensification or show that its top sustained winds would reach 185 mph, making it among the most powerful Atlantic hurricanes on record. They generally did not anticipate Dorian's fateful stall, which resulted in the storm pummeling Grand Bahama island for 40 hours with devastating impacts on lives and infrastructure. 

These shortfalls demonstrate that improving weather prediction must remain a national priority, especially as coastlines become more densely developed and increasingly powerful storms pose greater risks. Public safety officials need accurate and timely forecasts to evacuate those in harm's way, mobilize search-and-rescue teams, and avoid unnecessary evacuations that can cost millions of dollars and create widespread disruptions for residents, businesses, and the military. Reliable forecasts also enable companies and utilities to make preparations and recover more quickly.

Congress took an important step toward improved forecasting in 2017 with an initiative to leverage the unique breadth and depth of the atmospheric science research community at universities across the country. It authorized the National Oceanic and Atmospheric Administration (NOAA) to establish the Earth Prediction Innovation Center (EPIC), bringing together researchers and forecasters in an ongoing partnership to identify shortfalls in weather models and make the necessary steps to improve them. This initiative, under the leadership of NOAA Interim Administrator Neil Jacobs, will help the United States close the forecasting gap with Europe, where researchers and forecasters work with a common software to make continuous upgrades to their main forecasting model. 

The new U.S. approach will enable cutting-edge research to be moved much more quickly and effectively into real-time forecasting. It builds on the example of NOAA's successful creation of the National Water Model in 2015, a forecasting system that uses an advanced research hydrology model developed at the National Center for Atmospheric Research to make unprecedented predictions of streamflow and flooding across the contiguous United States. It was implemented by the National Water Center in Tuscaloosa, Alabama.

Congress must now take additional steps to advance hurricane research and forecasting. These include funding a new generation of specialized satellite instruments and radar, as well as unmanned aerial vehicles and marine gliders, that can provide more detailed observations of the atmospheric and oceanic conditions in and around the storm. Such observations, rapidly fed into high-resolution models and run on increasingly powerful supercomputers, can give forecasters critical information about subtle events, such as steering currents in the atmosphere or upper-ocean heat content, that can influence a hurricane’s power and influence whether it remains offshore or take a dangerous turn toward a populated coastline.


These improvements will require strategic investments. In addition to new observational instruments, Congress should prioritize research to assimilate the increased observational data into computer models more quickly and effectively, and funding for faster supercomputers to run the increasingly complex models and generate timely forecasts.

Significant advances in weather prediction will take time. For that reason, it is critical for the nation to move quickly in laying the groundwork for improved forecasts. The investments that we make today will pay substantial dividends every time that future storms threaten the United States. More accurate and timely forecasts will save lives and property, protect businesses, and help to safeguard our national security. 

Antonio J. Busalacchi is the president of the University Corporation for Atmospheric Research, a nonprofit consortium of 117 North American colleges and universities focused on research and training in the Earth system sciences.