So EPA Administrator Scott Pruitt is gone, the environmentalists’ noirest bete noire since James Watt ran the Interior Department early in the Reagan Administration. He will be replaced (at least temporarily) by Deputy Administrator Andrew Wheeler, a longtime Washington insider with a keen knowledge of the Agency. He served as chief of staff for the Senate Environment and Public Works Committee, which oversees all EPA activities, and he also has extensive and diverse experience as a lobbyist.

He’s likely to be lower key than Pruitt, but also very effective, given his thick rolodex. He’s certain to continue EPA’s efforts to separate its science advisors from its research largess. He’ll likely continue to prioritize and streamline remediation of the 1000+ “superfund” sites nationwide, a key Pruitt program. And he most assuredly will not bring back anything like the Obama Administration’s clean power plan with its “renewable” solar and wind power mandates. Instead, the switch from coal to more efficient gas‐​fired power plants is likely to continue.

The linchpin of EPA’s regulatory apparatus for global warming is the 2009 “Endangerment Finding” from emissions of carbon dioxide. According to the 2007 Supreme Court decision in Massachusetts v. EPA, if the Agency finds that resultant global warming endangers health and welfare, it can then regulate those emissions under the Clean Air Act Amendments of 1990.

The Finding is in turn based upon prospective climate change generated by large computer models both for climate and an estimate of its “social cost”. In 2016, Science magazine carried a news story by Paul Voosen noting that all of the climate models used by the United Nations Intergovernmental Panel on Climate Change have been “tuned” to simulate 20^th century climate changes. This includes a substantial warming from 1910 to 1945, despite the fact that there was very little additional carbon dioxide in the atmosphere when it began.

In 2017, France’s Frederic Hourdin and 14 coauthors published a landmark paper called “The Art and Science of Model Tuning” in the Bulletin of the American Meteorological Society. Among other things, they revealed that there is an “anticipated acceptable range” of output. Who decides that?

It turns out that the forecasts of 21^st century climate change are subjectively bounded. Voosen’s article describes the problems encountered at the Max‐​Planck Institute when their model produced seven degrees (C) of warming for doubled carbon dioxide. “They had to get that number down”, Voosen wrote.

Then there’s the whole problem of the “social cost of carbon” calculated by the Obama Administration to justify its sweeping “Clean Power Plan”. As shown by Kevin Dayaratna and colleagues at the Heritage Foundation, that cost is extremely sensitive to changes in prospective warming, and conceivably can be negative (i.e. a benefit) for modest warming, because carbon dioxide itself is a boon to plant growth, which includes crops.

Just last month, EPA proposed a revised method to calculate the cost, which cut the previous estimate by around 85%.

The model tuning, and the subjective estimates of both the amount and costs of future warming, indeed may endanger the Endangerment Finding itself. The new Acting Administrator should revisit that critical 2009 document.