You claim that the most recent estimates of future output for shale gas in the United States have become more conservative, but in our view this is a red herring (Nature 516, 28–30; 2014).

State-of-the-art projections for the world's future shale-gas supplies hinge on improved quantification of the uncertainty range and reducing its spread (called 'de-risking' a shale play) as experience and technology advance. Several of the studies you quote include uncertainty ranges that explain the current spread in forward projections of future US gas supplies. That crucial nuance was missing from your graphic, however, which shows only a simplified, discrete forward-production prognosis.

Comparing just one scenario from the study by the team at the University of Texas at Austin with another from the US Energy Information Administration's shale-gas outlook, omitting uncertainty ranges, creates an apparent mismatch where one may not in fact exist.

As a result of technology innovation (see also Nature 516, 7; 2014), the United States is today drilling 3-kilometre-long horizontal wells and conducting 30-stage fracture treatments at depths of 3.7 km. Further technological gains will increase global oil and gas output (see, for example, S. Neff and M. Coleman Energy Strategy Rev. 5, 6–13; 2014). Oil and gas prices also drive global shale development.

No one can accurately predict both the technology improvement rate and future wellhead prices, so we have to rely on a range of forecasts based on a variety of assumptions.