Scientists made significant progress in 2015 measuring methane emissions from the natural gas industry, continuing a years-long quest to quantify the industry’s contribution to climate change. What they found adds to a growing body of evidence that methane leaks are sporadic, difficult to predict, and often far larger than existing government estimates.

Many of the studies came from the Environmental Defense Fund’s $18 million project. Launched in 2011, it aims to measure emissions from every sector of the industry, including production, storage, transmission and natural gas vehicles. The project has drawn praise for its scope, vision and scrupulous methods. It’s also been criticized for accepting industry funding and sometimes relying on collaboration with oil and gas operators to obtain measurements.

Over a 20-year period, methane is 86 times more powerful at warming the planet than carbon dioxide. Over 100 years, its potency dwindles to 34.

This means that even small methane leaks throughout the system can erase any climate benefit of burning natural gas instead of coal.

The most recent EDF paper, released in December, found methane emissions from Texas’ Barnett Shale were 90 percent higher than estimates from the U.S. EPA estimates. The study marked the end of a massive two-year campaign to gather data through “top-down” and “bottom-up” techniques (collecting data from the air and on the ground, respectively)—two methods that often yield conflicting numbers.

EDF’s study found greater agreement between the methods than previous studies, and the authors created a statistical analysis to more accurately predict the presence of “superemitters”—facilities that emit more than the expected volume of methane. In the Barnett, they found that half the emissions at any time came from just 2 percent of the facilities. The emissions varied over time and by location, which will complicate efforts to find and fix the largest emitters.

Superemitters were also important in a separate EDF study, which found that natural gas storage sites and compressor stations, which pressurize the gas for transport, leak $240 million worth of methane nationwide per year. In that case, more than 20 percent of the leaks came from 4 percent of the facilities. The total amount released was close to EPA estimates.

EDF’s work came under intense scrutiny this summer, when Touché Howard, a methane expert who has worked on several EDF-funded studies, published a peer-reviewed paper that described a flaw he had found in a commonly used methane detector. The flaw causes the detectors to underestimate methane emissions. Howard believes the problem affected the EDF paper, an allegation the authors deny. The implications go far beyond EDF: hundreds of technicians use the same type of instrument to report industry emissions to the EPA. Bacharach Inc., the manufacturer of the instrument, said the detector wasn’t intended for the type of methane measurements being taken today, and would revise its manual to reduce the possibility of future problems. The company has approached the EPA to discuss further testing.

Other methane studies emerged from efforts not related to EDF. A city-wide study in New York found more than 1,000 methane leaks from Manhattan’s pipelines, a leak rate far larger than those found in Cincinnati and Durham, N.C.—two other cities where similar studies had been done. Researchers attributed the difference to Manhattan’s aging infrastructure, which is full of older pipes that are prone to leaks.

In Cambridge, Mass., a nonprofit is mapping local pipeline leaks to highlight the problem. As of September, the group had mapped more than 20,000 leaks in the state using data provided by local utilities. The organization published maps showing the exact location of each leak. Because utilities only have to repair leaks that pose an explosion risk, many leaks remain unplugged; the oldest leaks on the maps date back to 1985.

Last January, a separate study published as part of the EDF series found that methane leaks in Boston were two to three times the EPA’s estimates. It was the first peer-reviewed study of leaks from urban areas.

On the policy front, the EPA proposed methane rules as a first step in the Obama administration’s goal to slash emissions from the oil and gas sector 40-45 percent by 2025 (compared to 2012 levels). Environmentalists criticized the proposal because they rely in part on voluntary action. The EPA recently concluded a public comment period, and the final rules will be released in 2016.