Editor’s note: This story is by Audrey Clark, a curatorial assistant at the Pringle Herbarium at the University of Vermont.
The floods caused by Hurricane Irene and the spring rain and snowmelt of 2011 were some of the worst Vermont has ever seen, but we may expect more and worse floods in the future, according to Mathias Collins, a hydrologist with the National Oceanic and Atmospheric Administration.
Collins and his research team studied stream gauge records for the past 100 years in the Northeast. They found a strong upward trend in the frequency and magnitude of floods, in spite of an increase in tree cover over the same time period. Flooding generally decreases after reforestation because trees pull moisture out of the soil, allowing it to absorb more precipitation.
“I consider these trends especially compelling given rural Northeast U.S. land-cover trends over the last 100 years. Much of this region has been reforesting dramatically over the same period of record. The fact that we are seeing upward trends in flooding at all is remarkable given the fact we should be seeing downward trends given the land-cover change,” said Collins.
Collins spoke at a conference on flood resilience in the Lake Champlain Basin at the University of Vermont on Monday. Conference participants included scientists and state agency personnel who convened to discuss the lessons learned from recent research and the last year of extreme weather events.
Even though speakers consistently expressed awe at the destruction caused by last year’s flooding, they are concerned that the worst is yet to come.
“We got lucky, believe it or not,” said Paul Sisson, a science and operations officer for the National Weather Service’s Burlington Weather Forecast Office. He points out that the state could get wetter soil conditions prior to major storm events, storms that stay over Vermont for longer, or even two storms back-to-back in one week, as happened with hurricanes Connie and Danny in 1955. “We have had some crazy, unusual things going on recently. If you’re a meteorologist, it’s an interesting time.”
Collins’ research found that, starting around 1970, flood frequency and magnitude underwent a significant stepped change. There is now about one more flood per year than there was prior to 1970. Collins attributes this to the North Atlantic Oscillation (NAO).
The North Atlantic Oscillation refers to the circulation of air currents along the surface of the North Atlantic Ocean. When the NAO is in its positive mode, temperature and precipitation on the East Coast are greater, and storms are pulled up the coast rather than being forced into the Gulf of Mexico.
“The NAO is one of the oldest-known, prominent, recurrent upper atmospheric circulation patterns,” said Collins. “Its alternate modes strongly influence climate variability along the East Coast. One notable feature of the NAO is that it’s been in a dominantly positive, or what in New England is described as a ‘wet mode,’ since about 1970.”
This “wet mode” is clearly related to the increase in flood frequency and magnitude, though the connection to climate change is not as clear.
“There are some implications that [the NAO] is being or will be affected by global warming or anthropogenic climate change,” said Collins. “I think it is important to remember that features like the NAO and other large-scale circulation patterns are part of the climate system. So if anthropogenic climate change is affecting the whole climate system I think it’s just reasonable to expect it is probably affecting these large-scale patterns as well in ways we don’t yet know. There are some modeling strategies that suggest that the NAO, in a global warming context, should remain in a positive mode.”
In other words, climate models predict that we’ll continue to see the post-1970 flood patterns.
Climate models also predict that over the next 100 years, the northeastern United States will experience a wetter spring and fall and a drier summer. Rain in the spring causes rapid snowmelt, which can lead to flash floods. If the soil is wet prior to a storm event, rainwater doesn’t soak into the already saturated soil. Instead it runs off into streams and rivers, causing flash floods.
A dry summer may dry the soil out enough to absorb fall floods. However, even after a droughty summer, a few rainy days can restore soil moisture enough to cause major flooding.
According to state climatologist Leslie-Ann Dupigny-Giroux, this is what happened in 2011. The summer was so dry that the state of Vermont was considering requesting a disaster declaration from President Obama. But several inches of rain in the weeks prior to Aug. 28 saturated the soil to 90 precent of normal. This was followed by Tropical Storm Irene’s long and heavy rains and catastrophic floods.
Hurricane Floyd in 1999 was remarkably similar to Irene in terms of storm track, precipitation, and duration. Why did Irene wreak such havoc, while Floyd did not?
Hurricane Floyd in 1999 was remarkably similar to Irene in terms of storm track, precipitation, and duration. Why did Irene wreak such havoc, while Floyd did not?
According to Sisson, the main factor was a difference in soil moisture. Floyd followed the driest summer on record for Vermont and was not preceded by any rainy days.
“Given two storms of similar magnitudes, one produced devastating flash flooding, the other didn’t. A big factor, we think, was due to the soil moisture conditions,” Sisson said. Soil moisture conditions are challenging to track and predict.
There’s no question the state is more prepared for catastrophic flooding than it was a year ago. Johnathan Croft, a GIS database administrator with the Vermont Agency of Transportation, explains that the state now has greater technical capacity, as well as experience, to map and respond to flood damage. They have, for example, made maps accessible on iPads, which can be used in the field to record road conditions.
But Bill Saunders, a senior hydrologist with the National Weather Service’s Northeast River Forecast Center, says that his group is still struggling to find models they can use to accurately predict wind and soil moisture conditions quickly enough for frequent forecasts.
More than technical limitations, perhaps the greatest obstacle to proper preparation is what we believe.
Saunders notes that forecasters under-predicted the spring 2011 lake levels because, “No one here had seen this lake do this before. People generally had a lot of disbelief associated with — this rise in this lake cannot continue to be sustained. We learned that the lake can rise a lot faster than we had expected before.”
On top of that, Saunders points out that there may have been some hesitance to predict record-breaking conditions.
“When you’re in a record-setting scenario, you have to stop and think, do I really want to put myself out there and say, ‘Yes, this is going to break the record by a foot.’”