This kind of explosive population growth, and the associated trillion-dollar development along our coasts, make it critical for the United States to further improve forecasts of hurricanes and their deadly combination of storm surges, flash floods, and tornadoes.
As we saw with Hurricanes Harvey and Irma, our forecasting abilities — while remarkably good — need to be sharpened, especially given the challenges of evacuating large metropolitan areas. With Harvey gaining power, Texas residents were alerted that a dangerous hurricane would strike somewhere along the state’s coastline and then pour down devastating rains for days. However, a few days before landfall, computer models could not predict its rapid intensification to a Category 4 behemoth or pinpoint locations for the historic rains that submerged Houston neighborhoods.
In the case of Irma, the computer models accurately foretold widespread impacts on Florida. Yet they could not settle well in advance on whether it would make landfall along the east or west coasts of the state’s peninsula, leading to havoc as numerous south Florida residents fled to the Tampa area and into harm’s way.
To be sure, the forecasts were far more accurate than would have been the case just a decade or so ago. A recent study found that, since 2000, the National Oceanic and Atmospheric Administration’s National Hurricane Center has cut the average error in its 48-hour track forecasts by more than half, from 150 miles down to just 70 miles. Some 10 days before Irma made landfall, while it was still forming off Africa, computer models were already indicating that it would reach the Caribbean or southeastern United States as a powerful hurricane — a long-range prediction that was unheard of two decades ago.
The forecasts are better because of farsighted investments that the U.S. government has made in satellite and other observing technologies that monitor potentially dangerous storms, computer models that generate increasingly realistic simulations of their future track and intensity, and supercomputers that run the models multiple times a day to update the forecasts.
Gone are the days when a storm could sneak up and kill thousands of people with little notice, such as what happened with the Galveston hurricane in 1900. While the loss of any life is a tragedy, the fact that dozens of people, rather than thousands, died in each of the two recent hurricanes is a testament to how far our science has come in providing value to society.
However, as more people and businesses flock to our coasts, hurricanes are menacing our economic health. Moody’s Analytics estimates that Harvey and Irma combined will cost about $170 billion, rattling our entire economy. In the wake of Harvey’s floods, Goldman Sachs warned that third-quarter economic growth would fall by as much as a full percentage point and cost 20,000-100,000 jobs in September. Inflation is also expected to tick up because of Harvey’s devastating blow to Texas oil refineries and the resulting jump in prices at the pump.
To protect our communities and our nation’s economy, we must make smart investments that will advance our forecasts. This would provide actionable intelligence for officials making evacuation decisions, utility crews positioning in advance to restore power to affected areas, and businesses safeguarding valuable inventory.
Although the scientific community has been mobilized to confront these challenges, the government in recent years has cut annual spending on its flagship hurricane forecast improvement program from $13 million to less than $5 million. Compared with the costs of Harvey and Irma, this is as though we own a $30,000 car and will not spend more than about 90 cents a year to protect it.
Other countries are outpacing us in weather prediction. In recent years, we have largely ceded our competitive edge to Europe, whose premier forecasting model often generates better predictions of hurricanes off our own coasts.
Increased research funding would enable us to gain new insights into fluctuations in hurricane intensity and rainfall — a major challenge in better anticipating the destructive potential of a storm at landfall. Scientists are also working on faster ways to take the immense stream of observations and feed that data into the forecast models in near real time, thereby improving predictions of a hurricane’s track. More research is needed as well into the best ways to communicate to the public about where particular threats, such as high winds, storm surge, and flash floods, are most likely to occur.
As a part-time resident of New Orleans and having graduated from high school and college in Florida, I have seen firsthand the devastation that a powerful hurricane can cause. I am well aware of the frustration that comes with watching a potentially dangerous storm gather strength far offshore and not know where it will land or how intense it will be.
With millions of lives and trillions of dollars of coastal assets at risk, hurricanes are a threat to our economic development and national security. It is time that we treat them as such.