Smarter hurricane warnings save billions

Hurricanes are the biggest storms on the planet and the costliest natural disasters in the U.S., often decimating property and infrastructure as they crash onto the southeastern coast.

But better weather forecasting can reduce that damage by nearly a quarter, offering enormous savings, according to a new Cornell study.

Improved hurricane forecasts from 2007 to 2022 led to a 23% reduction in the cost of damages, or $2 billion saved per hurricane - more than the entire federal budget for weather forecasting, the researchers found.

"These forecasts are much more valuable than the costs to improve them. If you have better forecast data, you can do a better job of making decisions on how to deal with the storm," said study co-author Ivan Rudik, the Kenneth L. Robinson Professor of Applied Economics and Public Policy in the Charles H. Dyson School of Applied Economics and Management at the Cornell SC Johnson College of Business.

The research, "The Social Value of Hurricane Forecasts," was published April 3 by the National Bureau of Economic Research. Rudik's co-author is Renato Molina, assistant professor at the University of Miami.

The research has implications for policymakers and federal spending, especially as climate change increases the strength of these storms, the researchers said.

"Our results suggest investment in forecasting is going to pass a cost-benefit test for the U.S.," Rudik said. "It's a huge capital investment, and our results are finding it still makes sense."

Better forecasting enables counties and states to better prepare before hurricanes make landfall, Rudik said. "If the forecast says a hurricane could hit anywhere from Miami to Charleston, it's hard for you to make a decision in that case," he said. "But if the forecast says it's definitely going to hit South Florida and only South Florida, you're going to be able to make much better decisions about what you want to do, whether you're the governor of those states or just a citizen trying to make a decision about evacuating."

These decisions could include protective measures such as handing out sandbags and building temporary levees to mitigate flooding, authorizing overtime pay for police officers to conduct evacuations, and flying in generators to keep hospitals running. And despite hurricanes' massive destructive force, little was known about the role of information in their management. "The evidence out there was pretty limited," Rudik said.

To better understand the value and economic impact of U.S. hurricane forecasts, they analyzed data from the National Oceanic and Atmospheric Administration (NOAA) and before and after a significant increase in federal spending starting in 2007 that bolstered storm tracking, wind intensity forecasts and computation capabilities.

This spending was a response to the devastating effects of Hurricane Katrina and 12 other hurricanes that hit the U.S. from 2002 to 2005. Ten of those storms in 2004 and 2005 alone were responsible for at least 5,200 deaths and $229 billion in damages.

As part of the study, the team developed a new county-level dataset of forecasts, wind speed, precipitation and forecast uncertainty from 31 Category 1 hurricanes and greater that hit the continental U.S. from 2005 to 2022. The datasets account for more than 70% of direct property damage and about 40% of deaths for all environmental hazards in the U.S. during that period.

The researchers used this data to estimate how emergency federal funding for damage prevention responded to forecast information, the costs of forecast errors in terms of damages and increased spending for post-hurricane recovery, and the marginal value of a forecast improvement. Then they used those estimates to value the dramatic improvements in wind speed forecast accuracy since the 2000s.

The researchers found counties that expected to get hurricane-force winds received $36 million more in federal funding for hurricane preparation than counties forecast to have lower, sub-hurricane-force winds. That suggests forecasts play a significant role in driving protective actions, Rudik said.

Underestimating wind speed and misclassifying by up to two categories increased county damages by $177 million and federal emergency recovery spending by $23 million. That's about 10% of county GDP or more than $3,800 per person.

The study attributes these benefits to targeted federal investments post-Katrina, particularly in computational capabilities and data collection methods.

"We're doing things like flying planes into hurricanes and dropping hundreds of sensors to collect better data that helps inform the models and makes better predictions," Rudik said.

"The cone of uncertainty has been getting smaller," he added, referring to the geographical area where forecasters are uncertain whether a hurricane will hit. "The better forecasts get over time, the smaller these forecast errors are over time and so the tighter the cone ends up being."