UPDATE, 08/26: Associated Press corrects itself, acknowledges USGS estimate represents reduction in Marcellus gas resources.
UPDATE 2, 08/26: Washington Post blog indicates that USGS estimate may address a subset of resources estimated by EIA. If so, it means that Marcellus Shale gas resources are approximately 50% of what they were a week ago (as opposed to 20%). I'm contacting the EIA study author to get to the bottom of this.
The new estimate:
84 trillion cubic feet of undiscovered, technically recoverable natural gas and 3.4 billion barrels of undiscovered, technically recoverable natural gas liquids according to a new assessment by the U. S. Geological Survey (USGS).
These gas estimates are significantly more than the last USGS assessment of the Marcellus Shale in the Appalachian Basin in 2002, which estimated a mean of about 2 trillion cubic feet of gas (TCF) and 0.01 billion barrels of natural gas liquids.
84 trillion is considerably more than 2 trillion, so this is a huge increase, right?
Only relative to the 2002 USGS estimate: which no one uses.
The most recent authoritative estimate was put out in July by the Energy Information Administration (part of the Department of Energy). That estimate: 410 trillion cubic feet. Other, earlier estimates are in that same ballpark.
Well then, it's a pissing match between EIA and the USGS, right?
Nope. In reaction to the new USGS estimate, the EIA is on record as saying that the USGS estimate is more authoritative:
“We consider the USGS to be the experts in this matter,” Budzik [of the EIA] said. “They're geologists, we're not. We're going to be taking this number and using it in our model.”
In a nutshell: because of the USGS estimate, the best information now indicates the Marcellus Shale contains approximately 20%* of the natural gas we thought it did a week ago.
*Correction from original post: math error overstated natural gas in new USGS estimate