Kristina Gutchess takes samples in a stream. (Credit: Syracuse University)
For some time now, scientists have been concerned about the effects of using road salt on icy winter roads. Although it’s a naturally occurring substance, like most things salt can become toxic when levels of it are too high. It is a special problem for surface-waterways in urban areas with cold winters, which take on large amounts of road salt each year.
Groundwater is slow to refresh; this means it takes in salt slowly, but it also flushes salt away at a slow rate. Thanks to the use of road salt, some wells and aquifers have salt concentrations that are too high for people with high blood pressure, not to mention various amphibians and plants. This is where those proposed alternatives to road salt come in.
However, researchers from Syracuse University recently investigated and answered a different yet related question: how will climate change affect surface-water salinity that is already impacted by road salt? It’s a good question, given the reduction in snowfall some areas can expect based on various climate models.“This was an ideal location for the study because it contains two branches with contrasting land use: one branch dominated by urban lands with a major interstate highway that runs along the entire reach, and another with a primarily forested/agricultural land cover,” Gutchess explains. “This gave us an opportunity to simultaneously compare the response of two different systems.”
Use of road salt to de-ice roads and other paved surfaces has caused chloride concentrations in affected waters to rise. Historically, areas with more paved surfaces or urban land have experienced greater increases than their more rural counterparts, and this pattern existed in the Tioughnioga River Watershed. However, although experts have used other research models to assess the potential effects of de-icing practices in the future, different climate scenarios had not ever been used in any of these assessments.
“We calibrated the model to concentrations of chloride that were measured in weekly grab samples collected at the mouth of two branches of the Tioughnioga River,” details Gutchess. “After collection, the 100 or so grab samples were filtered in the lab and then refrigerated until analysis via Ion Chromatography.”
The team also installed Solinst Dataloggers, continuous water quality monitors, to record daily specific conductance so they could determine the linear relationship between the daily specific conductance and the grab samples. They then used that equation to calculate daily chloride throughout the time that the loggers were in the river. The team’s model predicted that the road salt usage of urban areas would continue to cause a greater increase in chloride than in rural areas—but not forever.
“Our model predicts that this will continue to be true for near future periods—pre-2050 or so,” remarks Gutchess. “Into the latter half of the next century, however, based on climate model predictions of increased temperatures during winter months, reduced winter snowfall totals will require that less salt be applied to roads and paved surfaces. This will reduce the annual salt contamination to previously affected waters and lead to a net decline in salinity over time.”
This change will take place more rapidly in areas that are less affected by increases in salinity in the first place, such as areas that are primarily forested. Urban or densely-paved areas that have experienced more road salt usage and, as a result, rapid and intense salinization, will see a slower change.
“Our simulations represent a range of plausible future outcomes specific to this study area, so they are not necessarily representative of future response in every location,” cautions Gutchess. “For example, in a region that has undergone a higher degree of salinization, the system could take longer to respond.”
The team used the INCA model framework for this study, modified to simulate chloride, rather than its default, nitrogen.
“The INCA model framework was ideal for these types of investigations because it is semi-distributed, meaning that the area is broken up into smaller areas that may or may not have different properties, and allows for variable time steps and applications rates,” describes Gutchess. “So by using the INCA model, we could simulate both changes in land use and changes in management practices, such as higher or lower daily salt application rates.”
This research is intriguing, in part because it teases out some of the complexity we must expect from the Gordian knot that climate change is. As the climate warms, a plethora of changes will affect each region, and inevitably some of these changes will be helpful—at least in the short-term. However, this does not diminish the need to work to mitigate the effects of climate change.
“I want to avoid any possible misperceptions or misinterpretations that these model results could mean that climate change is ‘good,’ as that is not at all what our work has revealed,” Gutchess warns. “Our model results do, however, highlight some pretty alarming changes to winter climate regimes in the northeastern US that have a significant impact on water quality—that by the end of the current century these regions could experience declines in annual snowfall totals significant enough to reduce approximately half a century’s worth of chloride contamination.”
What’s next for this research? Next, the team plans to model future changes in nitrogen in the same watershed, and Gutchess will begin working on the Pennsylvania and Ohio Water Energy Resources Study (POWERS) at Yale.