According to the Intergovernmental Panel on Climate Change (IPCC), “It has been clear for decades that the Earth’s climate is changing, and the role of human influence on the climate system is undisputed.” The IPCC’s most recent report, published in August 2021, warns that increasing greenhouse gases, due to the combustion of fossil fuels, have not only caused global warming, but also have changed the Earth’s ice cover, precipitation, sea level, and ocean acidity, and have intensified extreme climate events such as cyclones.

In light of the intense academic and popular media attention on climate change, there is a serious need for a scientifically informed book that addresses the purported crisis, and Steven E. Koonin has produced just such a book. In Unsettled: What Climate Science Tells Us, What It Doesn’t, and Why It Matters, Koonin shows that there is substantial uncertainty concerning the extent to which global temperatures will rise as a result of burning coal, oil, and natural gas. He demonstrates that there has been considerable exaggeration to support climate change projections. He also highlights the social and political challenges inherent in reducing harmful emissions on a global scale.

Koonin brings impeccable credentials to the climate science debate. He worked as a professor of theoretical physics at Caltech for nearly thirty years, served as Caltech’s vice president and provost, and worked as part of JASON, an elite group of scientists who advise the U.S. government on scientific matters related to national security. He also served as Undersecretary for Science in the Department of Energy during the Obama administration. With such deep experience, including extensive research using computers to model complex problems, he is highly qualified to comment on the validity of global climate models. (Full disclosure: I provided some scientific data to Professor Koonin as he was writing this book, but I had no input into the shape of his arguments, nor any chance to see the completed book before agreeing to write this review.)

The Limits of Climate Models and Statistics

Start your day with Public Discourse

Sign up and get our daily essays sent straight to your inbox.

The IPCC’s climate change claims rely on elaborate computer programs called global climate models (GCMs). Koonin describes how these models are formulated and the numerous assumptions they require. In particular, he discusses the tuning of models by varying a large number of parameters to avoid a physically implausible future climate. The different GCMs come up with a wide range of future warmings. The Coupled Model Intercomparison Project (CMIP) analyzes the results of dozens of GCMs in an attempt to make a more accurate prediction.

Koonin notes two problems with this approach. First, the Earth’s average surface temperature varies among the models by about 3oC (5.7oF), three times more than the observed warming during the twentieth century. The models disagree less about the warming, defined as the temperature change relative to the average temperature. However, Koonin writes “one stunning problem is that in the years after 1960, the spread of GCM results considered by CMIP5 is larger than that of the models in CMIP3—in other words, the later generation of models is actually more uncertain than the earlier one.”

Indeed, the spread of GCM results considered by CMIP6, used by the most recent 2021 IPCC report, has also not decreased. In science, one normally expects results to become more accurate as our understanding improves and faster computers become available. The lack of such progress in climate modeling indicates the science is not well understood.

There is substantial uncertainty concerning the extent to which global temperatures will rise as a result of burning coal, oil, and natural gas.


Second, Koonin illustrates how statistics have been manipulated to misrepresent global warming effects.  For example, although the observed average global temperature has risen by about 1oC (1.8oF) from 1895 to 2018, the number of high temperature extremes has not increased in recent decades. One study examined the daily high and low temperatures observed during that time interval at 725 stations in the continental U.S. The year in which the highest and lowest temperatures occurred was found for each calendar day. The most record daily high temperatures occurred in the 1930s, while the number of record daily low temperatures decreased slightly in recent decades. As such, the data do not support claims of more frequent heat waves in recent decades.

Similarly, Koonin casts doubt on claims of impending huge sea level increases, catastrophic droughts, and more extreme events such as hurricanes and tornadoes. Though persuasive, his presentation is at times somewhat wordy, given that it is possible to graphically and succinctly make the point. For example, Dr. Ryan N. Maue has plotted the frequency of observed global hurricanes from 1980 to 2022, such that a reader immediately sees no increase in either the total number of hurricanes or the number of major hurricanes defined as having wind speed exceeding 96 knots, or 110 miles per hour.

Elusive Climate Solutions

Koonin punctures the balloon claiming that a transition to a green economy would be easy. He points out that about 80 percent of the world’s energy comes from fossil fuels. This simply cannot be replaced within a few decades, especially if one is also against nuclear power. Consider Germany, where coal and nuclear plants have been closed. Wind farms and solar cells are unable to supply the country’s electrical needs. Only natural gas from Putin’s Russia is available. Is that wise?

Koonin also points to the nations of the developing world. They see the Western standard of living, and want it. For this reason, greenhouse gas emissions will increase in the coming decades. He notes that U.S. emissions have fallen on average 1 percent a year since 2005 primarily because natural gas has replaced coal for electricity generation while global emissions have increased about one-third. In 2020, the U.S. only accounted for 13 percent of global greenhouse gas emissions.

However, the question that Koonin does not clearly answer is what should be done. He notes additional research is unlikely to provide a technological fix that reduces greenhouse gas emissions in the near term. For example, battery research has been ongoing for more than a century but we still don’t have a car battery that can be recharged in minutes. Conservation can curb emissions, but higher taxes to limit energy use can be politically challenging. Any transformation to energy systems will be an enormous undertaking that will require huge investments and take decades to accomplish.

Scientific integrity and independence are essential for continued technological progress that benefits society.


Koonin also expresses some skepticism about various geoengineering proposals such as increasing the Earth’s reflectivity by injecting aerosol particles into the upper atmosphere or schemes to sequester carbon dioxide underground. In short, there is no easy answer. The book concludes by noting that mankind has adapted to changes in the climate for millennia.

In fact, on the point of adaptability, Koonin soft-pedals the benefits of greater agricultural productivity resulting from more carbon dioxide in the atmosphere. Satellite measurements by NASA show that during the period of 1982 to 2001 more than 30 percent of the Earth’s land area greened for a 14 percent overall increase in total gross productivity. This is very beneficial to farmers who provide food for the world’s growing population. Plants extract carbon dioxide from the atmosphere through small holes in their leaves called stomata. At higher carbon dioxide levels, the stomata become smaller, reducing water loss from the plant. This is believed to be part of the reason for the notable greening of the Earth’s arid regions such as the area bordering the Sahara Desert.

The Importance of Scientific Independence 

Overall, Koonin’s most serious and significant criticism focuses on attempts to limit scientific inquiry. He cites legislation introduced in March 2019 by Senator Schumer and twenty-five other senators “to prohibit the use of funds to Federal agencies to establish a panel, task force, advisory committee, or other effort to challenge the scientific consensus on climate change.” Fortunately, the bill did not become law.

Science students are taught that their work should be independent of outside pressure. Koonin cites Galileo’s courage standing up to the Catholic Church as an example to follow. Disturbingly, he describes instances of colleagues resorting to name-calling and suggests that he probably would have been fired from either his private sector job or from government service if he publicly stated his conclusion that stabilizing human influences on climate would be “essentially impossible.”

Scientific integrity and independence are essential for continued technological progress that benefits society. Persecution of scientific thought is wrong. But even so, scientists—especially those advising governments—need to speak up.