Fizzled flood forecast won't keep Fargo from relying on National Weather Service
FARGO -- It was a rough couple of weeks for the National Weather Service.
As spring took forever to bloom and snow kept falling, weather service forecasters warned in mid-April that there was a 40 percent chance of topping 2009's record 40.84-foot crest.
Just two weeks later, the flood fizzled into a relative nonevent with a preliminary crest of 33.32 feet Wednesday, still above major, the flood stage of 30 feet but well within nearly all of Fargo's permanent flood protection.
"We're kinda scratching our head," said Steve Buan, a service coordination hydrologist with the weather service's River Forecast Center in Chanhassen, Minn.
City and county officials have repeatedly defended the weather service in this odd flood year, but Fargo leaders also say when it comes to flood planning, they cull as much information as possible from various official sources.
"We rely on the local weather stations," City Administrator Pat Zavoral gave as an example. "I mean we have as much confidence in, like, John Wheeler and the other weather people in the area as anybody else. So we look at that."
Then there's Mayor Dennis Walaker, who every year takes multiple trips around the valley to look at snowpack. It's a process that he says is "a matter of comfort" for him -- to go out and look at the river and the snowpack and confirm the numbers the weather service is forecasting.
He also listens to officials from the Minnesota Department of Natural Resources, Wilkin County, the U.S. Army Corps of Engineers, and yes, local weather forecasters like WDAY's John Wheeler, who Walaker said has a "wealth of history" in the valley to contribute.
"None of this process is done without having all the information you can possibly get," the mayor said. "It's a lot easier to make decisions when you have that kind of a group around you."
Dry soil slurps up melt
So why was the National Weather Service forecast so far off this year? Buan said it has mostly to do with the historically late spring. As winter crept later and later into March and more snow fell, the weather service's models, based on historical data that only dates back about 100 years, didn't know how the soils would respond, Buan said.
"We had no way of checking how that would work because we had never seen it before," he said.
Those models in late March predicted that between 60 and 75 percent of the water in the heavy snowpack would run off into streams, which would have been similar to the flood of 2006, when the river hit 37.13 feet, Buan said. Expecting similar runoff, forecasters issued a severe flood outlook in mid-April.
But the runoff didn't happen. Spring took an extra month to melt, and the dry soils opened up like never before, Buan said.
Preliminary data from the southern Red River Basin shows that 70 percent of the snow's water content absorbed into the dirt, leaving only 30 percent to run off, an "alarmingly low rate" compared to past floods, Buan said.
"We thought 60 to 70 percent runoff was something more on the low side, and now we're seeing 30 (percent) when you've got big snowpack," he said.
To compare, 80 to 100 percent of the water content in the snow in 2009 ran off into streams, leading to the worst flood ever recorded in the Fargo area.
"We've seen now the two ends of the spectrum, so now we can hopefully establish these bookends," Buan said.
It wasn't just the weather service that was fooled by this year's flood.
Models run by the U.S. Army Corps of Engineers showed "similar results," said Alex Nelson, a hydrologic engineer for the corps who worked this spring as a weather service liaison at the River Forecast Center.
Nelson said the corps' model, which was set up to map the river channel for the proposed $1.8 billion Fargo-Moorhead diversion project, was usually within a foot of the predictions the weathers service released.
"For the most part, we were fairly close with the (weather service) forecast on a daily basis," he said.
During floods, the corps uses its model not to forecast, but to give the weather service forecasters another set of data that they can use to improve their outlooks.
Buan said once the U.S. Geological Survey confirms this year's flood levels, the weather service can go back, rerun and tweak the models.
"We're always learning," he said.
Too long to prepare?
Zavoral said he has only one criticism of the weather service -- its probabilistic flood forecasts released early in the year can create "a certain level of angst" in the community, especially when they're predicting a severe flood.
"It gets everybody in a heightened alert and then we have to start reacting," Zavoral said. "We can say 'Well, don't worry about it.' But on the other hand, we do use them as our forecaster, so how can we tell the public don't worry about it (when) they're forecasting something real high, real early?
"A month to prepare is great, but not three months," Zavoral added.
This year, based on the probability-based forecasts in early and mid-April, volunteers built about 1.5 million new sandbags, none of which got wet in Fargo and only 100,000 of which were even placed.
Walaker said he believes the weather service is still living down the flood of 1997, when it incorrectly predicted levels in Grand Forks and the city's levees failed.
But he says that's not why he seeks out other information and continues to make his own somewhat folksy predictions.
The National Weather Service remains the crucial forecaster for flood fighters in the area.
"I'm not going to bet the protection levels of the community based on my predictions. I mean it's just as simple as that," he said. "We have to go with the National Weather Service. Can you imagine the fallout there would be if we didn't protect to what the National Weather Service told us, and we failed?"