A comprehensive new study of atmospheric levels of methane, an important greenhouse gas released by leaky oil and gas operations and livestock, has found much higher levels over the United States than those estimated by the Environmental Protection Agency and an international greenhouse gas monitoring effort. [The paper, "Anthropogenic emissions of methane in the United States," is being published in the Proceedings of the National Academy of Sciences and should be available at this link shortly.]
The study, combining ground and aerial sampling of the gas with computer modeling, is the most comprehensive “top down” look so far at methane levels over the United States, providing a vital check on “bottom up” approaches, which have tallied estimates for releases from a host of sources — ranging from livestock operations to gas wells.
Read on for an excerpt from the news release issued by Harvard University, the home of two of the lead authors, Scot M. Miller and Steven Wofsy, which summarizes the main points well.
Later today, I’ll add answers to questions I posed to the authors as well as reactions from Robert Howarth, the Cornell environmental scientist who’s been studying methane levels and pushing for an end to hydraulic fracturing for natural gas, and Steven Hamburg of the Environmental Defense Fund, which is working with a host of academic and industry partners on a big emissions study, part of which was described here recently.
E.P.A. has not yet had the opportunity to fully review the PNAS study on methane emissions; however we are encouraged that more methane emissions measurement data are now available to the public.
E.P.A. is committed to using the best available data for our Inventory and continually seeks opportunities to update and improve our estimates. Research studies like these will add to our knowledge base of GHG emissions and will help us refine our estimates going forward.
E.P.A. is currently evaluating data received through the Greenhouse Gas Reporting Program , which provides valuable information on the location and magnitude of certain emissions sources from large facilities.
An E.P.A. review of methane emissions from gas wells in the United States strongly implies that all of these figures may be too low. In its analysis, the E.P.A. concluded that the amount emitted by routine operations at gas wells … is 12 times the agency’s longtime estimate of nine billion cubic feet. In heat-trapping potential, that new estimate equals the carbon dioxide emitted annually by eight million cars.
It’s important to note that the new study is a snapshot of conditions in 2007 and 2008, before concerns increased about the need for tighter standards for gas and oil drilling operations. The authors say a similar analysis is under for more recent years. A comparison will be very helpful in clarifying the impact of changed practices by energy companies so far, and what remains to be done.
Overall, according to the new study, total methane emissions in the United States appear to be 1.5 times and 1.7 times higher than the amounts previously estimated by the U.S. Environmental Protection Agency (E.P.A.) and the international Emissions Database for Global Atmospheric Research (EDGAR), respectively.
The difference lies in the methodology. The EPA and EDGAR use a bottom-up approach, calculating total emissions based on “emissions factors” — the amount of methane typically released per cow or per unit of coal or natural gas sold, for example. The new study takes a top-down approach, measuring what is actually present in the atmosphere and then using meteorological data and statistical analysis to trace it back to regional sources.
Generated by a large, multi-institutional team of researchers, the latest findings offer a robust and comprehensive baseline for assessing policies designed to reduce greenhouse gases. They also point to a few areas where the assumptions built into recent emissions factors and estimated totals may be flawed.
“The bottom-up and top-down approaches give us very different answers about the level of methane gas emissions,” notes lead author Scot M. Miller, a doctoral student in Earth and Planetary Sciences through the Harvard Graduate School of Arts and Sciences. “Most strikingly, our results are higher by a factor of 2.7 over the south-central United States, which we know is a key region for fossil-fuel extraction and refining. It will be important to resolve that discrepancy in order to fully understand the impact of these industries on methane emissions.”
Miller is a 2007 graduate of Harvard College and earned a master’s degree in engineering sciences at the Harvard School of Engineering and Applied Sciences (SEAS) in 2013. He studies in the lab of Steven C. Wofsy, Abbott Lawrence Rotch Professor of Atmospheric and Environmental Science at SEAS.
“When we measure methane gas at the atmospheric level, we’re seeing the cumulative effect of emissions that are happening at the surface across a very large region,” says Wofsy, a coauthor of the PNAS study. “That includes the sources that were part of the bottom-up inventories, but maybe also things they didn’t think to measure.”
Miller and Wofsy, along with colleagues at the Carnegie Institution for Science, the National Oceanic and Atmospheric Administration (NOAA), and five other institutions, used a combination of observation and modeling to conduct their analysis.
NOAA and the U.S. Department of Energy collect observations of methane and other gases from the tops of telecommunications towers, typically about as tall as the Empire State Building, and during research flights. The team combined this data with meteorological models of the temperatures, winds, and movement of air masses from the same time period, and then used a statistical method known as geostatistical inverse modeling to essentially run the model backward and determine the methane’s origin.
The team also compared these results with regional economic and demographic data, as well as other information that provided clues to the sources—for example, data on human populations, livestock populations, electricity production from power plants, oil and natural gas production, production from oil refineries, rice production, and coal production. In addition, they drew correlations between methane levels and other gases that were observed at the time. For example, a high correlation between levels of methane and propane in the south-central region suggests a significant role for fossil fuels there.