PS. Their methodology is absolutely standard for atmospheric science, especially when speaking of a specific basin which contains some 1% of the entire US wellheads.

And from the Salt Lake paper cited by oui, there's this:

The study based its findings on measurements recorded during a four-hour window Feb. 3, a calm, cloudless day that was perfect for measuring methane concentrations. There was almost no snow on the ground so the boundary layer over the basin was unusually high for winter, which allowed gas field emissions to mix evenly with the air, Sweeney said.

The basin's oil and gas infrastructure serves 6,000 wells that account for 1 percent of the nation's natural gas production.

The team found it leaked 60 tons of natural gas an hour during the Feb. 3 window.

"Most days we measured concentrations far greater than what we reported in the paper," Sweeney said. The new study was not designed to determine points of leakage.

To sum up, fracking should have to prove it's safety prior to implementation, rather than society has to prove its danger after the fact.

"Life shrinks or expands in proportion to one's courage." - Anaïs Nin