Experts say a combination of warm water in the Gulf of Mexico, weak winds and just enough storm surge all worked together to produce the record rainfall totals that made Hurricane Harvey one of the most devastating storms in U.S. history.
“It turned out to be one of the most prolific rainfall events in the history of the United States,” says Mike Schichtel, a senior branch forecaster at the National Weather Service’s Weather Prediction Center in Silver Springs, Md.
The National Weather Service reported rainfall totals of 51.88 inches at a rain gauge in Cedar Bayou, Texas, which set a record. Three other locations in the Lone Star State — Clear Creek at I-45, Dayton and Mary’s Creek at Winding Road — all recorded rainfall totals of more than four feet. Nineteen other spots in Texas got more than 40 inches.
When most people think of a hurricane, they envision a fast-moving storm that makes landfall and roars across an area, ripping off roofs, leveling trees and toppling road signs. Although maximum wind speeds for Hurricane Harvey were recorded at more than 130 mph, for the most part a lack of winds in the upper atmosphere was listed as one of the reasons the storm dropped so much rain on one area.
“When you’re looking for heavy rainfall potential, you have a weather pattern in advance where the steering winds are weak,” says Schichtel.
Steering winds or currents are the winds that keep a hurricane moving. Those around Harvey weren’t strong enough to force it out of the area, so the storm stayed in one spot and dropped epic amounts of rain.
Hal Needham is a storm surge expert and founder of the firm Marine Weather and Climate in Galveston, Texas. He says the storm surge in Galveston Bay and other parts of the Gulf of Mexico wasn’t very strong, maybe two or three feet, but the wind patterns held the surge in one position. This basically created a dam that blocked drainage of the floodwaters out to the bay and the Gulf. “What we did see was prolonged onshore winds, and the levels of Galveston Bay were elevated for days,” he says. Conversely, when Hurricane Ike hit Houston in 2008 it packed 143-mph winds and a 17-foot surge, but not as much rain.
Needham says the worst situation came when the winds weakened and Harvey was downgraded to a tropical storm. “My concern was that people were looking only at a category number, which only has to do with wind,” he explains. When the storm moved slightly inland and the winds died, most people assumed the worst was over. “The storms with weaker winds can actually have more rainfall,” he says.
The third ingredient in the recipe for the Houston disaster was that the waters of the Gulf of Mexico were warmer than usual. Needham says the Gulf was running two to three degrees warmer, which has been a trend for the past few years. “Hurricanes are basically a machine and their fuel is warm water,” he says.
“The warmer the water, the more vapor there is in the air above that water,” says Kenneth Kunkel, a researcher with the North Carolina Institute for Climate Studies. He estimates that the warmer weather in the Gulf created at least 10 percent more water vapor than if it were cooler.
When Harvey moved back out to sea, re-energized and then made landfall in Beaumont, Texas, on Aug. 30, the rains came down even harder, measured at as much as four inches an hour. Needham says it was actually worse for Houston that the storm stalled about 150 miles south because it could continue to draw moisture and keep producing rain.
The other unique characteristic of Harvey was how localized it was. Houston and Beaumont experienced record flooding, but 20 miles away in Lake Charles, La., rainfall totals were less than half and floodwaters only reached a foot. “The heavy bands of rain came in at the same places and you could move 20 miles and have very different amounts of rain,” says Kunkel.
He says the forecast models accurately predicted that the storm would stop in one area and drop a lot of rain, but it was difficult to pinpoint where it would happen. “The models were quite good at forecasting that it would slow down and stop and meander in one spot,” says Kunkel.
Although it may have been difficult to forecast the location, the National Weather Service’s Schichtel says the early warnings were accurate. “For this particular system, I was working about six days prior and there were good indications that it would form,” he says. “This is one storm that was very well forecast. We tried to get the word out well in advance, and we were very particular with wording we used.”
This article originally appeared in the October 2017 issue.