In the New Yorker last week, @billmckibben argued that methane leakage from natural gas has erased most of the declines in US emissions over the last few years. This is not the case, as I discuss in a new @TheBTI article: thebreakthrough.org/issues/energy/… 1/14 
The New Yorker analysis is based on a figure from Dr. Robert Howarth (@howarth_cornell). To maximize the calculated impact of methane emissions, it relies on applying a very-short 20-year timeframe. newyorker.com/news/daily-com… 2/14
Methane is a very potent greenhouse gas, packing a warming punch more than 100x stronger than a similar amount of CO2 while in the atmosphere. However, unlike CO2, methane has a very short atmospheric lifetime. Nearly all the methane emitted today will be gone in 10 years. 3/14
Given their different atmospheric lifetimes and warming effects, comparing CO2 and methane emissions can be quite difficult. The standard used in GHG inventories is to compare their average warming over 100 years, over which methane has a GWP around 34 times larger than CO2. 4/14
Not everyone takes this approach, however. Some have argued that a shorter 20 year period should be used instead. Over 20 years methane has a GWP 86 times larger than CO2 (rather than 34), as more remains in the atmosphere over the shorter time period. 5/14
The choice of GWP period comes down to the perceived importance of near-term vs. longer-term warming. And while there is a case to be made for the shorter time frame, it is a weak one. For example, a 20-year GWP can perversely incentivize reducing methane over CO2. 6/14
If we stop emitting methane, it disappears from the atmosphere quickly and its warming effects dissipate. CO2, on the other hand, sticks around for centuries. This suggests that lowering CO2 should be prioritized as it cannot be easily removed later on. 7/14
Just how much methane leaks from US natural gas systems have long been a source of controversy. While official EPA inventory estimates suggest a system-wide leakage rate of around 1.5%, dozens of studies have found much higher leakage rates. 8/14
Brandt et al 2014 suggested that actual leakage from natural gas systems was between 1.25 and 1.75 times larger than official inventories, implying a leakage rate of between 1.9% and 2.6%. A recent review by Alvarez et al 2018 found a similar leakage range of 2% to 2.7%. 9/14
The figure below shows how emissions from US natural gas compare to coal under these different leakage rate assumptions, as well as under Howarth's 3.5% leakage/GWP20 case. 10/14
Another way to look at coal vs gas is through modeling the radiative forcing (and warming effects) of both in the atmosphere, as in my 2015 paper. The figure below shows the climate impact of 30-years of coal and gas under leakage rates of 1% to 6%: sciencedirect.com/science/articl… 11/14
To summarize, under standard accounting methods (GWP100), natural gas leakage makes US GHG reductions modestly smaller (from 13% to 10%-12% depending on the leakage rate assumed), but contrary to the claims of @billmckibben does not undo the progress we have experienced. 12/14
At the same time, its important to emphasize that reducing methane emissions is important for mitigating climate change. Deep decarbonization will ultimately involve either replacing natural gas plants with zero-carbon alternatives or equipping them with CCS technology. 13/14
While gas has provided a bridge away from coal in the US, it is a bridge that is quickly reaching its end, as the number of coal plants rapidly shrinks and the falling cost of renewables makes them an increasingly large driver of power sector decarbonization. 14/14
