Steven Koonin’s Unsettled: Much to Like and to Dislike

Steven Koonin’s book Unsettled is a thoughtful and important synthesis of climate science but there are a few things about it that are just wrong.

Koonin eloquently explains how some researchers may have misrepresented or distorted the interpretation of climate science. He skillfully discusses the limitations of climate data and climate modeling. He describes how the policy response by government is commonly illogical, counter-productive and even absurd.

He never questions that climate change is real but many of his views are informed by overly generalized or incorrectly interpreted data.  This leads him to dismiss much the urgency of climate change and to attribute it to natural causes rather than to human activity. He dismisses CO2 as a meaningful factor for climate change and finds little evidence that polar ice loss is reason for alarm.

These three statements less than 250 words into his introduction encapsulate both the strengths and shortcomings of his work:

Humans have had no detectable impact on hurricanes over the past century.

Greenland’s ice sheet isn’t shrinking any more rapidly today than it was eighty years ago.

The net economic impact of human-induced climate change will be minimal through at least the end of this century.

He is correct that there is little evidence to link hurricanes and other extreme events to climate change. Not yet anyway.

His comment on Greenland’s ice loss is is patently false.

His assessment that climate change will have little effect on economic growth is only reasonable if ice loss is no cause for concern and if CO2 is not a major cause of warming.

Let’s start with Greenland Ice loss. Koonin just published an op-ed last week in the Wall Street Journal that provides more detail that he presents in Unsettled. In “Greenland’s Melting Ice is No Cause for Climate-Change Panic,” he presents an argument that “the annual loss has been decreasing in the past decade even as the globe continues to warm.” He does not dispute that a lot of ice has been lost but argues that this is normal and uses the chart below to make his case.

He points our that ice loss is not continuous and that it is no greater today than it was in the 1930s. That is just wrong.

“Since human warming influences on the climate have grown steadily—they are now 10 times what they were in 1900— you might expect Greenland to lose more ice each year. Instead there are large swings in the annual ice loss and it is no larger today than it was in the 1930s, when human influences were much smaller. Moreover, the annual loss of ice has been decreasing in the past decade even as the globe continues to warm.”

Steven Koonin

His ice loss curve is a 10-year average although that is not shown on the chart. A ten-year average of annual data is statistically problematic despite the fact that decadal time-averaging might be appropriate for addressing longer-term climate change.

I reproduced Koonin’s ice loss curve in Figure 2 below. I also included the annual data (gray curve) and the cumulative ice loss (gold curve).

Koonin’s 10-year average ice-loss curve leaves out much of the story. Summing the annual data above and below Koonin’s (blue) curve shows that ice loss far exceeded the 10-year average for the entire period of the chart. The volume of ice loss above the 10-year average (13 trillion metric tons) is 8 times more than loss below the 10-year average (1.5 trillion metric tons). That’s a trend that cannot be seen from Koonin’s single curve in Figure 1.

More importantly, the cumulative gold curve shows that ice loss has almost continuously increased and now accounts for almost 15 trillion metric tons of ice since 1900. The variation in the data is almost all loss—in other words, it varies between greater and lesser loss instead of between loss and gain. That’s another trend that cannot be seen from Koonin’s single curve in Figure 1.

Figure 2. Steven Koonin’s Greenland ice data leaves out much of the story. Greenland ice loss volume above the 10-year average (13 trillion metric tons) is 8 times greater than ice loss below the 10-year average (1.5 trillion metric tons). Source: Program for Monitoring of the Greenland Ice Sheet (PROMISE) & Labyrinth Consulting Services, Inc.

The data table in Figure 2 shows that Koonin’s bullet point above is false (“Greenland’s ice sheet isn’t shrinking any more rapidly today than it was eighty years ago”). 2020 ice loss was 399% greater than ice loss in 1940. Ice loss for the ten years ending in 2020 was 31% greater than the ten years ending in 1940. Ice loss for the three years ending in 2020 was 55% greater than for the three years ending in 1940. Any way you work the data, Koonin has Greenland ice loss wrong.

His statement that ice loss data should reflect the constant warming of the global temperature curve violates one of the key principles he discusses in Unsettled; namely, that global average temperature curves are imprecise and do not honor local variations.

Figure 3 shows local temperature variations in Greenland from 1998 to 2010. It indicates that the ice loss variations in figures 1 and 2 correlate well with changes in ice volume. In other words, what he calls “inconsistent” is normal and fully explained by seasonal temperature variations in Greenland.

Koonin states that “While a warming globe might eventually be the dominant cause of Greenland’s shrinking ice, natural cycles in temperatures and currents in the North Atlantic that extend for decades have been a much more important influence since 1900.”

Climate is complex and there is no single cause for Greenland ice loss. At the same time, the increase in world fossil fuel consumption provides a reasonable proxy for Greenland ice loss although it has the same potential global vs local disconnect that I mentioned for temperature. Nonetheless, the correlation is at least interesting (Figure 4). Ninety-five percent of ice loss has occurred since 1900 and 96% of total fossil fuel consumption has also been since 1900. Twenty-seven percent of ice loss has taken place since 2000 and 31% of fossil fuel has been used since 2000.

Koonin states in Unsettled that he doesn’t know of any expert who disputes the rise in CO2 concentration over the past 150 years is due to human activities (p. 65). That said, the way that he minimizes the affect of CO2 on warming is inconsistent with the data that I will show below.

Koonin writes that ” human influences on the climate were negligible prior to 1900…Human influences remained quite small as late as 1950…Variations in the climate before 1950, then, show that other phenomena must have been at play (p.36).”

Figure 5 below is from Unsettled (p. 67) and reveals why Koonin dismisses CO2 as a key factor in climate change.  He states, “By geological standards, today’s Earth is starved for atmospheric CO2.” His graph shows CO2 concentrations compared to 1950 levels. It indicates that current CO2 concentrations are at a 600 million year minimum only equalled during the Permian Period 300 million years ago.

Each division in Figure 4 represents 40 million years. Homo sapiens did not appear until after the last data marker on the right-hand side of the graph just before the number zero. Koonin talks a lot about the importance of choosing the correct scaling in science. Figure 5 is a textbook example of choosing the wrong scaling.

Figure 6 shows CO2 concentrations over the last 800,000 years using the same technique as in Koonin’s graph.  It shows that indeed CO2 levels fluctuated before the advent of Homo sapiens about 300,000 years before present. It also shows that those concentrations exceeded the 800,000 year maximum by around the year 1800. That seems to contradict Koonin’s assertion that human activity had negligible effect on climate before 1950. It seems like more than a coincidence that fossil energy consumption began in earnest in the 1700s.

Zooming in a bit more to the last 2,000 years, we see that his interpretation falls apart completely. Figure 7 shows that CO2 levels were quite flat through most of the last two millennia. Concentrations compared to 1950 increased above the 1750-year average around 1800. Contrary to Koonin’s assertion that human influence was negligible before 1950, it appears that human activity had considerable affect by at least 1850.

Steven Koonin is an accomplished scientist with a distinguished career mostly as an academic teaching theoretical physics. He spent five years at BP working on renewable energy strategies and then served as under-secretary of energy in the Obama administration before returning to academia. Unsettled is informative, engaging and provocative. He is right to question the science and motivation behind the current mainstream of climate science and climate change.

At the same time, Koonin finds himself on the wrong side of paradigm change. He seems unaware of Thomas Kuhn’s 1962 classic The Structure of Scientific Revolutions. Much of what Koonin complains about is simply part of how paradigms rise and fall.

A paradigm is a ruling theory. It is not speculative but is a model or framework that best links and explains existing data. A paradigm is not static. It changes as new information becomes available. It is a testable hypothesis. At some point it can no longer accommodate new observations and a new paradigm begins to compete and ultimately replaces it.

The previous paradigm was that climate change is dominated by natural forces and processes, and that human activities have negligible effect on climate. Koonin is that paradigm’s defender of the faith. Cracks began to appear in that paradigm in the 1950s. New data began to show that emissions from burning fossil fuels were modifying the seasonal exchange of CO2 between the atmosphere, biosphere and ocean. As more data was collected, a new paradigm emerged suggesting that human activities had become a key factor in climate change. That has become the new paradigm.

Kuhn was clear that a paradigm need not—and in fact never does—explain all the facts with which it is confronted. Paradigms gain their status because they are more successful than their competitors in solving problems that practitioners recognize as important. Once a paradigm is established, theoretical alternatives are strongly resisted. That’s a big part of Koonin’s objection but that’s also the way paradigms evolve.

Plate tectonics was a hotly debated subject in earth science when I was in undergraduate school in the late 1960s. The ruling paradigm explained the dynamic earth with a model of giant folds called geosynclines that collapsed as the earth cooled and contracted. The new plate tectonic model suggested that crustal plates were put in motion by oceanic spreading centers and that their collisions and separations better explained earth’s structural history. Today, no one talks much about geosynclines.

When I was in graduate school in the 1970s there were two competing paradigms for how oil was formed. One camp thought that oil was generated from organic carbon that came mostly from fossil algae. Another believed that it came from inorganic sources in the earth’s mantle. Today no one talks much about the inorganic origin of oil.

Geosynclines and the inorganic origin of oil were paradigms that didn’t survive the new data that became available as scientific research progressed.

Koonin has placed himself in the unfortunate position of defending a failed paradigm. Just because tens of thousands of scientists support the current climate change paradigm doesn’t mean that it’s right. At the same time, just because Steven Koonin and a much smaller number of scientists disagree doesn’t mean it’s wrong either.

Koonin distinguishes between The Science (upper-case) vs the science (lower-case). The former is how science is represented by politicians, the press and activists; the latter is the actual scientific research and publication work done by scientists. He writes that scientific “institutions frequently seem more concerned with making the science fit a narrative than with ensuring the narrative fits the science” (p.189). But, like it or not, that’s how science works!

“A striking feature of doing research is that the aim is to discover what is known in advance. When the outcome of a research project does not fall into this anticipated result range, it is generally considered a failure. Normal science does not aim at novelties of fact or theory and, when successful, finds none. Fundamental novelties of fact and theory bring about paradigm change.”

Thomas Kuhn

Most paradigm changes in science take place out of the public view because they concern things that do not directly affect most people’s lives. Climate change is different. People with no real understanding of science or climate are now involved in the discussion. Few of us want to hear that our behavior may be destroying the ecosystem or that our children and grandchildren may inherit a damaged planet.

I often hear non-scientists remark that climate is always changing and that natural processes have caused climate from the beginning. They describe periods in history that were hotter or cooler and use these as evidence that nothing new is really happening today. Do these people honestly think that professional scientists hadn’t thought of those things before they did?

Similarly, people often refer to something written twenty or thirty years ago that didn’t turn out as stated, and use that to discredit climate science. Perhaps they don’t understand that new data changes how scientists interpret things.

This brings me to the title of Dr. Koonin’s book—Unsettled. It’s a good title because we have all heard someone say that “The Science is settled.” Scientists know better. Science is never settled. It’s not about settling anything. As Koonin’s former Cal Tech colleague recently wrote:

“Science doesn’t prove anything.”

Sean Carroll

Science is mainly about observing and describing. Explanation comes later. Explanation doesn’t mean determining cause. It means connecting what is known into patterns or trends that can be tested for validity. I know that this conflicts with public understanding of science but that’s because the public has a distorted idea of what science really is and what scientists really do. Perhaps critics of climate science should take a moment to ask questions so they might learn what science is.

I completely agree with Koonin’s observation that policies designed to slow or reverse climate change will probably fail. The costs of these government action plans may not justify the climate benefits in the long run. I also agree that complex energy systems and human behavior patterns are unlikely to change as quickly as needed even if money were not an issue. He is also correct that many climate emergency movements are at least as unrealistic as most government policies.

There’s a lot to like in Unsettled but there’s also plenty to dislike. Koonin sees no reason for panic over climate change and believes that humans will adapt.  I suspect that adaptation will be traumatic and will probably involve widespread death and civil disorder.

Whether or not we can do much about climate change, Unsettled fails to provide people realistic expectations about potential future outcomes. Worse, his incorrect interpretations of polar melting and the role of CO2 in warming give skeptics justification to dismiss climate change altogether. For its many merits, people deserve better information and guidance than Unsettled provides.