Roanoke Times
Copyright (c) 1995, Landmark Communications, Inc.
DATE: SATURDAY, November 4, 1995 TAG: 9511050013
SECTION: VIRGINIA PAGE: A1 EDITION: METRO
SOURCE: S.D. HARRINGTON STAFF WRITER
DATELINE: LENGTH: Long
"It was just one of those things beyond the scope of experience any of us here had," he says now. "It was just a helpless-type feeling."
But 10 years later, the National Weather Service has a more advanced flood warning system, and McIntosh believes Western Virginia will be better served if a flood like the one in '85 hits again.
McIntosh, who retired from the weather service's Roanoke office in June, still has vivid memories of the flood of '85, which claimed 10 lives in the Roanoke Valley.
He was preparing to leave Roanoke early that Nov. 4 for a regional weather service managers' meeting in Charleston, S.C.
A low-pressure system would continue to drench the Roanoke Valley that day, after about 4 inches had fallen on the valley the week before. But the rain would not last long enough to do any serious harm, the meteorologist had predicted. The weather service headquarters in Washington, D.C., had called off a flood watch for the area that morning.
But when McIntosh left his house that morning, he noticed that the rain was coming down harder than expected, so he stopped by the office to check it out. He decided not to leave the valley that morning.
By 9:30 a.m., chaos erupted. Telephone lines at the weather service went dead, and the office couldn't get reports from observers of flood-prone areas.
People in the office weren't aware of the flooding already happening in the valley. Because of the downed lines, McIntosh also had problems contacting emergency services coordinators in the area.
Around 10:30 a.m., McIntosh issued a flash flood warning for Patrick, Floyd, Montgomery, Craig and Roanoke counties. But the warnings came too late, he says. Many places near the Roanoke River and area creeks were already flooded.
"The horse was out of the barn," McIntosh says.
McIntosh was no stranger to high-pressure situations; he had served as a weather forecaster in the Navy for fighter pilots on aircraft carriers. But this would turn out to be his worst day during 17 years as a civilian meteorologist.
Could the Roanoke Valley be caught by surprise like that again?
"If it repeated itself now, things are in place to do it better," McIntosh says, referring to a system of flood warning devices that keep the National Weather Service updated every 15 minutes on water levels in rivers and streams and how much rain has fallen in areas vulnerable to flooding
IFLOWS, or Integrated Flood Observing and Warning System, was in the experimental stages when the '85 flood hit, says Blacksburg hydrologist Mike Gillen, who was working in Richmond at the time.
McIntosh says the system was actually developed as a result of another dramatic flood in Virginia: Hurricane Camille in 1969.
Virginia was one of several states to implement the IFLOWS initially, but the Roanoke Valley was not included in that program. The nearest area being monitored by IFLOWS was in the extreme Southwest portion of the state.
The flood of '85 changed that.
In 1988, Congress approved $200,000 for flood control projects in Roanoke, part of which was used to fund an expansion of IFLOWS into Roanoke and surrounding counties.
About 15 rain gauges were placed in areas upstream from Roanoke, Gillen says. Data from the rain gauges is crunched into a central computer in Blacksburg, where most of the Roanoke office was transferred last year. From the data, the amount of rainfall that will flow into the Roanoke River can be determined.
Devices placed at the bottom of the river and streams measure the water's pressure and translate that data into water levels. Another way the weather service monitors water level is with floatation devices in monitoring wells near rivers.
Before this system was in place, Gillen says, there was no way to monitor activity in specific areas - only general estimates of rainfall.
In 1985, the weather service office in Roanoke, which was responsible for 14 counties, relied on a few on-site observers, scattered rain gauges and a radar that could give only rough estimates of how much rain would fall on isolated areas.
Now, with the combination of IFLOWS and Doppler radar, which uses rainfall data in isolated areas to generate an overall estimated rainfall, Gillen says warnings of up to six hours can be given for any significant amount of flooding.
"Unfortunately, it doesn't prevent the flood, but it gives emergency services coordinators enough time to anticipate what could happen," Gillen says.
To illustrate his point, Gillen describes a rainstorm in April 1992 when the Roanoke River was threatening to flood. Gillen, who worked in Roanoke then, says the weather service was able to give the city between four and six hours advance warning that the river would crest at 18 feet. The forecast was later revised to only 10 feet.
What if the phone lines go dead again?
IFLOWS has a radio communication system to let the weather service communicate with emergency services coordinators.
McIntosh isn't sure how much more notice meteorologists could give residents before floods hit.
"It's not infallible," he says. "It's not going to tell you this morning that tomorrow evening we're going to have a similar flood" to the one in 1985.
But he is confident that the few extra hours forecasters now can give could reduce the impact that flash floods have.
"It's probably enough to prevent a lot of deaths," McIntosh says.
by CNB