One reason for the short shrift is that there hasn’t been a good way to reliably estimate the escape of carbon dioxide from waterways across large areas. Peter Raymond, professor of ecosystem ecology and Butman’s coauthor and Ph.D. advisor, is one of several researchers who has been studying the role fresh waterways play in carbon cycling for years, but the work on carbon dioxide evasion had been limited to relatively small scales. Then, recently, he and Butman realized there might be a way to do a comprehensive estimate for the entire country thanks to a relatively new database from the U.S. Geological Survey (USGS) which, with NASA and the National Science Foundation, is a project sponsor.
For decades the USGS and the Environmental Protection Agency have been compiling data on the water chemistry for tens of thousands of streams and rivers across the country, but there was no way to link this data with information about the physical characteristics of these waterways, such as their flow rates and topography. But in 2006, the agencies began creating the National Hydrography Dataset Plus, which put all the data in a form researchers can use. “Before that, there was really nothing like it,” said Butman. “There were just bits and pieces out there.”
Before coming to Yale, Butman had worked with large area calculations using satellite remote sensing and Geographic Information Systems (GIS) to study the amount of biomass and carbon in forests. Working with Raymond, he was able to apply similar techniques to developing a computer model to analyze the total surface area of all the streams and rivers in the database, which maps the locations of streams and rivers based mainly on topography.
The next step was to figure out an average for how much carbon dioxide the waterways are releasing. Carbon dioxide is a weak acid in water, so it affects pH, and water temperature affects how much carbon dioxide can stay dissolved in water. Raymond and Butman developed another model that exploited these and other aspects of aquatic chemistry to process waterway chemical information from the USGS dating back to the 1920s, as well as data from numerous other sources. This model generated estimates for the amount of carbon dioxide escaping from a given swath of water at a specific time. Ultimately they compiled this information to settle on an estimate of the average escape rate.