A slightly weaker hurricane Irene moved north toward North Carolina Friday morning as hurricane warnings and watches were posted along the coast from the North Carolina-South Carolina border to the mouth of the Merrimack River north of Boston.
Packing maximum sustained winds of 110 miles per hour, Irene's hurricane-force winds now extend some 90 miles from its center, with tropical-storm winds extending 290 miles.
The storm's track has changed little in the past 24 hours. At 8 a.m. Eastern Daylight Time, forecasters said they expected Irene's eye to make its initial landfall Saturday as the storm strikes the Cape Hatteras region. Then it moves back over open water, hugging the coast. The eye is expected to make landfall and cross central Long Island in New York on Sunday, then move up through central New England.
Hurricane warnings were posted from the Little River Inlet, N.C., to Sandy Hook, N.J. The National Hurricane Center in Miami issues the warnings 36 hours before tropical-storm-force winds are expected to arrive to make preparations dangerous.
Forecasters also have posted a tropical-storm warning along a stretch of coastline from just north of South Carolina's Edisto Beach to Little River Inlet.
Hurricane watches stretch from Sandy Hook to the mouth of the Merrimack River north of Boston. Hurricane watches are issued some 48 hours ahead of the first tropical-storm-force winds.
In addition to the storm surge Irene is expected push into coastal areas, inland flooding could be a serious problem because of the storm's size and track, forecasters say. While the eastern half of a storm tends to have the strongest winds, the western half tends to dump the most rain.
Flood and flash-flood watches, issued by local National Weather Service forecast offices, have started to pop up along the eastern seaboard.
In eastern Pennsylvania, southeast New York, and much of southwestern New England, soil moisture is significantly above normal for this time of year, says David Vallee, lead hydrologist at the Northeast River Forecast Center in Taunton, Mass. Intense summer thunderstorms have dropped enough rain over southern New England to push monthly rainfall totals for August from 2 to 5 inches above normal.
As a result, rivers are running at unusually high levels for this time of year, he adds. "Thats not a good thing when you have the potential for a foot of rain to come down this weekend," he says.
Forecasters say the story is similar in the mid-Atlantic region, where some locations have seen storms dump more rain than New England has received.
The immediate concern is for flash-flooding, which carries a slightly looser meaning in the East than out West, where even seemingly distant cloud bursts can send sudden torrents of water surging down narrow canyons and ravines far from a storm.
In the east, the phrase covers urban flooding or inundation of culverts, depressed highway underpasses, and other low-lying areas that can occur quickly with short but intense downpours – "when you're going to be floating cars off the highway," Mr. Vallee says.
When a watch is posted, forecasters expect flash-flood conditions in the watch area within 12 to 18 hours. A warning "means within the next one to six hours, you're going to get wet," he adds.
With flood watches posted for regions such as New York's Adirondack Mountains, which cradle the headwaters of the Hudson River, river flooding also can remain a threat for days after a tropical cyclone vanishes from the radar screen.
To improve these storm-related flood forecasts, researchers with the National Severe Storms Laboratory (NSSL) in bone-dry Norman, Okla., are working with local National Weather Service forecast offices in North Carolina to put a new water-level forecasting model through its paces during Irene's encounter with the state. The tool is designed to be a one-stop source for forecasting floodwater depth and extent for an entire river basin – from mountains to mouth, according to Suzanne Van Cooten, a member of the team of federal, state, and university scientists working on the project.
The project began after back-to-back tropical cyclones that struck North Carolina in 1999. The two packed a double punch, with rivers cresting as high as 24 feet above flood stage and a storm surge that reached as high as 13 feet. By the time the second storm, Floyd, left the area, 52 people had been killed and some 7,000 homes destroyed in the coastal plain, which in essence lies east of Interstate 95.
In assessing forecasters' performance afterward, they had all the right tools in their toolkit. But key pieces weren't connected in ways that could have better represented the potential for the flooding that took place.
"When the research teams took it all apart, they said: Oh my gosh, we didn't account for the flood flows from Dennis," the previous storm, says Dr. Van Cooten, with the NSSL.
The experimental model the project has produced aims to remedy that situation. And it is designed to project flood levels all along a river, rather than at the relatively few points where the US Geological Survey has installed flood gauges.
Although the model is still experimental, forecasters in North Carolina will be using its output to help them with flood warnings, Van Cooten says. The model was put to the test last year with hurricane Earl.
It's unclear when the model will shed its training wheels and be declared "operational."
"As a former forecaster, I would have liked this sooner rather than later," she says.
Using it with two hurricanes helps ensure that the model is doing its job well, but two remains a small sample. And the team faces more mundane but important tasks, from writing the users' manual to ensuring that the model's links to the outside world have enough bandwidth to handle the data to and from forecast offices in the midst of a storm.
Interest in the "total water-level prediction" tool runs high, "but I could not tell you a time frame" for its shift to operational status, Van Cooten says.