A man backpacks up a rocky trail to an air-sampling station
The first air sample analyzed for the new carbon-14 method of estimating fossil fuel emissions was collected from this site at Niwot Ridge, Colo. Credit: James Murnan, National Severe Storms Laboratory

Estimating how much carbon dioxide (CO2) is released by burning fossil fuels has traditionally resembled a large-scale mathematics test. Accurately summing up the number and types of vehicles on the road, evaluating current emissions standards, adding in emissions from power plants, and accounting for efficiencies require detailed calculations and data.

Since the 1990s, scientists at the Environmental Protection Agency (EPA) have been estimating emissions using these complex accounting procedures.

Now, a group of researchers has taken a different approach: measuring atmospheric CO2 directly. But first, they had to separate the fossil fuel–derived carbon from natural sources such as volcanic emissions. Using carbon-14 as a marker, the team parsed out CO2 sources, providing monthly to annual measurements of fossil fuel emissions.

They found CO2 fossil fuel emissions were about 5% higher than EPA estimates. They note their approach can give more frequent and focused measurements, which is useful in policy and climate modeling research.

Measuring Emissions

Greenhouse gas emissions are a direct driver of climate change. International treaties like the 2015 Paris Agreement and the 1992 United Nations Framework Convention on Climate Change (UNFCCC) treaty have specific targets for reducing greenhouse gas emissions.

Because of the UNFCCC treaty, the United States is obligated to report greenhouse gas emissions. “Our reporting obligations don’t stop, even if we pull out of the Paris climate accord,” said Sourish Basu, an atmospheric scientist at the University of Maryland and NASA Goddard Space Flight Center and lead author of the paper published in the Proceedings of the National Academy of Sciences of the United States of America.

Since the UNFCCC treaty, the EPA has used a bottom-up approach to estimate emissions. They look at activities that produce CO2: use of cars, trucks, planes, power plants, and heating. Then they look at the emission factor, or how much CO2 is produced by each activity. In the end, researchers multiply activities and emission factors and sum up everything to get an estimate of CO2 emissions for the country.

“These are very detailed statistics,” said Basu, adding that EPA researchers do a good job of trying to capture all ways CO2 is emitted.

“This is a very hard job, but at the end of the day, it’s an accounting process,” he said. “It’s entirely possible to miss something.” For example, using the wrong emission factor—such as an outdated fuel efficiency for a car—can skew estimates.

Basu and his colleagues decided to do a top-down approach and take direct measurements of CO2 in the atmosphere. “If you just look at the atmospheric data, because everything that’s emitted has to show up in the atmosphere, you can construct an independent estimate of [emissions],” explained Basu.

The team has been setting up a sampling network across the country for almost 2 decades. The geographic coverage and number of stations make a “robust” sampling of total U.S. emissions, said Basu.

Parsing Out Carbon Dioxide

“Now we have great carbon isotope data that we can use to figure out how much fossil fuel emissions [are] coming from the U.S.”

Atmospheric measurements of CO2 take all sources into account: those from fossil fuel combustion and carbon emissions from other sources. The trick is to figure out what part of the total emissions comes from fossil fuels.

Fossil fuels are made of carbon-based material (ancient plants and animals) that is millions of years old. The carbon in fossil fuels has long since stopped the process of radioactive decay.

“When the carbon gets to a fossil fuel state, it’s completely devoid of carbon-14,” said Basu.

Carbon-14 is the only known radioactive isotope of carbon. Separating it out means the remaining carbon in an air sample is fossil fuel derived.

“Now we have great carbon isotope data that we can use to figure out how much fossil fuel emissions [are] coming from the U.S. from the atmospheric data,” said Eri Saikawa, an atmospheric chemist at Emory University who was not involved with the study. “That is extremely fascinating.”

Man prepares air samples with a complex machine in a lab
University of Colorado Boulder researcher Chad Wolak prepares air samples for carbon-14 measurement by colleagues at the University of California, Irvine. Credit: Scott Lehman, University of Colorado Boulder

The team measured 1,000 air samples from 2010. Using a dual-tracer inverse modeling framework and measurements, the team found their top-down estimate was 5% larger than the EPA’s estimate: 1,653 teragrams of carbon per year from fossil fuels compared to about 1,581 teragrams of carbon per year.

Although there was a 5% increase using the top-down method, Saikawa said that statistically, it was not much different than the bottom-up approach. “Actually, the bottom-up and the top-down [methods] are overlapping,” she noted. “I thought that was a pretty good result.”

Saikawa noted both methods have uncertainties. “Nothing is perfect,” she said. “But I think having both sets to work with to see how we can improve is very important.”

“I think that’s the most exciting point,” said Saikawa. “We might actually be able to reduce the uncertainty that we currently have, in terms of the CO2 emissions estimates.”

Tracking Carbon Goals

The team wants to expand their work to include multiple years to note any trends in estimates or emissions over time. Accurately estimating the amount of CO2 emissions is important for long-term climate policy and planning, especially related to the Paris Agreement.

“The thing is, even though we as a country might be thinking of withdrawing from the Paris accord, there are several entities within this country who have decided, ‘No, we are going to have [our own trajectories] anyway,’” said Basu. “But in order to independently verify those trajectories using our atmospheric measurement–based method, we need more measurements.”

“I hope that we can use this type of great work as a way to increase the network [of sampling] and then make sure that we have a good set of measurements available,” said Saikawa. “I would love for them to be able to get data not at the national level, but then at the state level,” she said. “That could potentially show a much more difference in a specific region.”

—Sarah Derouin (@Sarah_Derouin), Science Writer


Derouin, S. (2020), Radioactive bookkeeping of carbon emissions, Eos, 101, https://doi.org/10.1029/2020EO145841. Published on 22 June 2020.

Text © 2020. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.