Extreme precipitation has increased most rapidly in the Northeast over the past 25 years. According to previous studies, the amount of extreme precipitation—rain or snow that leaves behind one to two inches of water in a single day—has increased by almost 50 percent since 1901 to 1995.
A new Dartmouth study sheds light on how changes in temperature and precipitation brought on by global warming affect flooding and streamflow in the Northeast. The results have been published in the American Water Resources Association Journal.
The scientists analyzed how precipitation, including snowfall, winter downpours on snow occasions, springtime snowmelt, and soil conditions influence streamflow. They concentrated on four northeastern watersheds: the Mattawamkeag Waterway in northeastern Maine; the northern New Hampshire river known as the Dead Diamond; the White Waterway in eastern Vermont; and West Virginia’s Shenandoah River.
“The standard and antecedent precipitation indices are essentially measures of how wet the land surface is already, which impacts runoff and streamflow.”
first author Charlotte Cockburn, Guarini ’21,
Streamflow in the three northern watersheds is unequivocally impacted by snowmelt, while the Shenandoah Waterway watershed is impacted more by precipitation. The four watersheds were chosen because they are unregulated rivers, which means that there is no dam to control the streamflow, and they cover a wide range of latitudes.
The team used historical relationships between streamflow and factors to create a machine learning model for the first part of the study that included: temperature; precipitation (whether snow or rain); the “antecedent precipitation index,” or the amount of moisture held in the soil prior to a storm; the “standardized precipitation index,” which is used to distinguish between periods of rain and periods of dryness; and furthermore, streamflow.
They used observations of snow depth from the Northeast Regional Climate Center and streamflow data from the U.S. Geological Survey, as well as more than 95 years of historical climate data from 1915 to 2011.
“Both the predecessor precipitation record and the normalized precipitation file are essentially proportions of how wet the land surface is as of now, which influences overflow and streamflow,” says first creator Charlotte Cockburn, Guarini ’21, who was an expert’s understudy in studies of the planet at Dartmouth at the hour of the examination.
“A lot of that water can be absorbed by the soil if you have a really big rainstorm on a relatively dry surface, but if you have multiple rainstorms leading up to the really big rainstorm, there is no room for the water in the soil, which causes higher streamflow.”
That occurred in August 2011, when Typhoon Irene, known as Hurricane Irene in quite a bit of New England, caused devastating flooding, various passings, and billions of dollars in harm, Cockburn notes.
To anticipate streamflow during the very long virus season of November to May, the group utilized normal temperature, three-day and 30-day precipitation, and three-day and 30-day snowfall as factors in their model. To imitate snowmelt, they developed a submodel. Based on the model variables, the model would then predict streamflow for a specific date, such as April 1, 2009.
“For setting, the most elevated streamflow in Upper East watersheds will in general happen in the spring, close to now, when there is snowmelt, more precipitation occasions than in the colder time of year, no vegetation to haul water out of the dirt, and when the dirt is either immersed or frozen,” says senior creator Jonathan Winter, an academic administrator of geology at Dartmouth.
One of the problems with the model, as the researchers explain in the study, is that it is trained to rely on snowpack as an important driver for projecting cold-season streamflow.
Therefore, streamflow decreases are predicted by the model when it encounters future dates when there will be less snowpack as a result of global warming. However, as Cockburn points out, “the models don’t exactly capture the dynamics of winter changes in streamflow because they are trained on the past,” and “we expect rain to be a much more important driver of winter streamflow in a world that is warmer due to climate change.”
The team used a climate projection from 2070 to 2099 to force the machine learning model in the second part of the study to see how streamflow would change under that climate.
The key discoveries are:
Across watersheds and seasons, three-day precipitation and starting soil dampness are the main factors that decide streamflow in the Upper East.
Because the Mattawamkeag River’s watershed is both the largest and most northern, it is less sensitive to short-term, extreme precipitation events and more sensitive to snow. Thirty-day snowmelt and rainfall are important to the river’s streamflow.
How New England watersheds react to the transition from winters dominated by snow to winters dominated by rain will determine the flow of streams during the colder months to come.
Changes in rainfall are almost exclusively responsible for future warm-season streamflow.
According to Winter, “the Northeast will have increased streamflow and higher flood risk” if, as climate models predict, “the Northeast gets wetter soils and more heavy rainfall events.”
The Northeast’s snowpack was lower than usual this winter as a result of temperatures that were more than 4 degrees Fahrenheit above average.
“We are going to experience winters like the one we just had more frequently in the future. Winter says, “It’s a glimpse of what’s to come.” However, our analysis reveals surprisingly that streamflow sensitivity to precipitation is more important than snow in the Northeast.”
According to Winter, “understanding how streamflow may change in a warmer and wetter climate is important as these dynamics have implications for flooding, ecosystems, water resources, and hydropower” in light of the effects of climate change.
More information: Charlotte Cockburn et al, Drivers of future streamflow changes in watersheds across the Northeastern United States, JAWRA Journal of the American Water Resources Association (2023). DOI: 10.1111/1752-1688.13120
