New UMass research estimates carbon emissions from 22 million stream reaches across the US

AMHERST, Mass. — Using a sophisticated new modeling approach, researchers at the University of Massachusetts Amherst have estimated carbon dioxide emissions from inland waters to 22 million U.S. lakes, rivers and reservoirs. It marks the first time this approach has been applied at a continental scale, and reveals previous methods may have overestimated CO2 emissions by as much as 25%. 

In rivers, lakes and streams, CO2 is generated mostly through the breakdown of organic matter. Any time there is more CO2 in the water than the air, that water will “breathe” it out, emitting carbon as a gas — but how much CO2 , and how much is coming from each specific source, is still a big question mark. 

“We need to know how much CO2 is being generated so we can predict how it will respond to climate change,” says Matthew Winnick, assistant professor of Earth, Geographic and Climate Sciences and corresponding author on the paper published in AGU Advances. “As temperature rises, we tend to think that a lot of the natural carbon cycle processes will respond to that and potentially amplify climate change.” 

To date, these estimates have been generated by taking the average stream CO2 concentration over large areas and applying it across the waters in a given region. However, this fails to capture the range of CO2 emissions across different environments.  

“When you have a really turbulent stream, it’s going to de-gas a lot faster,” explains Winnick. Also, small, headwater streams receive more CO2 from groundwater compared to water further downstream, which also impacts how much CO2 that part of a river will emit. “Aggregating these really steep mountain reaches in with low slope areas just misses a lot of information,” he adds. 

In their alternative model, the researchers simulate carbon movement in each stream reach individually to get more realistic estimates of how much CO2 comes out of these bodies of water.  

Winnick and Brian Saccardi, a former graduate student in the Earth, Geographic and Climate Sciences Department and the co-lead author on this paper, previously tested their method using the East River watershed in Colorado’s Rocky Mountains. By partnering with Colin Gleason, Armstrong Professor of civil and environmental engineering, and Craig Brinkerhoff, then a UMass doctoral student and co-lead author of the study, they were able to apply their model to 22 million different stream reaches across the country. 

“We tested it out in the mountains in Colorado and we saw that Colin and Craig had been doing this large-scale river network modeling. This was a natural fit,” Winnick adds. 

They found that their estimates modeled emissions of 120 million metric tons of carbon, while the standard aggregate modeled estimated emissions of 159 million metric tons — a difference of 25%. “As a result, the whole budget shifts because these mountain areas play a really strong role in CO2 emissions at the continental scale,” Winnick says. 

One area where having more accurate estimations of carbon emissions may have an impact is in carbon sequestration efforts. Winnick nods to projects that add calcium carbonate minerals into streams in an attempt to convert the CO2 to a more stable form.  

“If we want to know if these methods can work, we need to know how much CO2 is in these river systems,” he says, as well as how CO2 changes dynamically over the length of a stream. “We know that CO2 varies wildly — it can change by orders of magnitude within tens of meters along a stream. So having ways to predict these dynamics will really help in evaluating whether and where carbon sequestration might be effective.” 

Another debate in this field is: where does the CO2 actually originate from — groundwater or stream corridors? If researchers are to accurately predict how carbon emissions may change in response to climate change, they need an answer to this question because different environments have unique responses to climate change. 

“If it’s coming from this near stream sediment, where water is actively being exchanged back and forth between the stream itself and the underground, it’s going to have a different response to changes in temperature or precipitation than it would if it was happening in the hillslope groundwater system,” Winnick explains.  

Winnick’s research has landed on the side of the stream corridor, which includes the water in streams and near-stream sediments, though he acknowledges that this is still very much an open question. “We hope this study will spur more efforts to get a more precise budget for where this CO2 is coming from,” he says. 

This research was supported by the U.S. National Science Foundation.