« Continue Browsing

e-mail article Print     e-mail article E-mail

Erik Burgess, Published April 25 2013

Forecasts tougher due to mystery beneath snow

FARGO – The science of measuring how much of a spring snowmelt has sunk into soil – a crucial part of flood predictions – is far from perfect, and it’s one of the reasons forecasting the Red River’s rise has been difficult this year.

In fact, as far as Bob Dablow is concerned, some of it is far from science. He is one of many volunteers who help the weather service gauge soil infiltration.

Dablow sticks a big rebar pole into the ground once a week during flood season to determine how far down the frost extends and if the pole is wet, somewhat like using a toothpick to see if a pan of brownies is done.

“No rocket science here at all,” said Dablow, also the mayor of Sabin, Minn.

The data collected by volunteers, and organizations such as the Minnesota Department of Natural Resources, is important for a simple reason. Melted snow that seeps into the ground isn’t running into streams and rivers.

But methods of collecting the data are a bit archaic, said Steve Buan, a service coordination hydrologist with the National Weather Service office in Chanhassen, Minn.

Other techniques to gauge the moisture in soil include shoving PVC pipe into the dirt and filling it with a measured amount of water and waiting to see how much, if any, soaks in.

The weather service is forced to rely on these manual techniques to gather soil moisture data largely because of the climate in this area of the United States, Buan said.

“A lot of electronic stuff doesn’t like to freeze. It especially doesn’t like to freeze if there’s water around it,” he said.

Calculating temperature, wind speed and precipitation have all gotten more sophisticated, said Greg Gust, warning coordination meteorologist for the weather service in Grand Forks. Gust said that’s driven in part by the needs of the transportation and aviation industry.

“Trying to get the tools to help the hydrology out is the challenge now,” Gust said. “In a perfect world, there would be a more dignified, more formal, more well-articulated means of doing it.”

Each year, as winter rolls in, the weather service develops and relies on a frost index to help predict spring floods. The frost index accounts for temperatures, depth of snow and wetness of soil over the course of winter, which allows forecasters to predict whether the ground will be susceptible to infiltration in the spring or not.

But Buan said the frost index usually needs adjusting as melting starts.

“You have to trust that it’s accurate, but you can’t check it until you see: Is the water getting to the streams or is it not?” Buan said. “So always the first areas to melt, you have a chance you could be wrong.”

That is what happened this year. The first runoff appeared to show a hard frost index, a tough layer causing much of the melted snow to flow into the river system. Based in part on an expectation of a hard frost layer, forecasters predicted last week that the Red River this year had a 40 percent chance to top the record 40.84-foot flood of 2009.

As the melt continued, the ground opened up more in southern Wilkin and Richland counties, allowing forecasters to adapt the frost index and lower the crest range for Fargo to 38 to 40 feet.

“It’s still subjective,” Buan said, which is part of the reason why no crest prediction had been made yet for Fargo as of Thursday. “We still got a lot of snow to melt in Wilkin County.”

It’s not that the techniques are erroneous, Buan said. They’re helpful, but you can only dig a hole in so many spots. Volunteers are trained to take representative samples, and it helps that forecasters have time to adjust the index and provide better predictions, Buan said.

But better soil measurement techniques would be especially helpful when it comes to predicting a flood early in the season, like in February, when that year’s frost index isn’t fully formed and forecasters rely on historical data.

Of course, all of that goes out the window in a year like this year, Buan noted, when winter has lasted into April and the crest is the latest it’s ever been in recorded history.

“You’re beyond your historical period,” Buan said. “You have nothing to look back to.”

Dablow has been volunteering his time to provide information to the weather service for more than 20 years, and he says he’s glad he’s not the one charged with predicting a crest.

“I’m not a meteorologist. I’m just the guy that gets them the information,” Dablow said. “What they do with it from that, I think that’s where all the science comes in.”


Have a comment to share about a story? Letters to the editor should include author’s name, address and phone number. Generally, letters should be no longer than 250 words. All letters are subject to editing. Send a letter to the editor.


Readers can reach Forum reporter Erik Burgess at (701) 241-5518