Unsettled – Image for illustrative purposes only (Image credits: Unsplash)
Former Obama administration science official Steven Koonin has released an updated edition of his book Unsettled that examines how climate records are assembled and presented. The work focuses on patterns in temperature measurements and extreme weather reporting that he says can distort public understanding of long-term trends. Koonin draws on specific technical examples to illustrate where adjustments or choices in data handling appear to favor higher warming signals.
Early Temperature Records Show Sampling Effects
One central point involves the way historical temperature series are constructed. Koonin notes that record high temperatures appear more often in earlier portions of the data simply because fewer stations existed at the time. This sampling difference can make recent years look more extreme by comparison even when underlying conditions have not changed dramatically.
The Climate Science Special Report receives particular attention for how it handles these series. Koonin argues that the presentation overlooks the built-in bias created by expanding observation networks over decades. As a result, readers may receive an impression of accelerating records that stems partly from improved coverage rather than solely from physical changes.
Model Adjustments Influence Warming Projections
Koonin also reviews the process of tuning climate models. A detailed paper on the Max Planck Institute model describes how one subgrid parameter tied to atmospheric convection was adjusted after the initial value produced roughly twice the observed warming. Such calibration steps are common in complex simulations, yet Koonin questions whether they receive enough scrutiny when results are communicated to policymakers.
The example highlights a broader pattern in which model outputs are refined to align more closely with measured temperatures. While this improves short-term agreement, it can mask uncertainties in the underlying physics. Koonin suggests these choices deserve clearer documentation so that projections carry appropriate caveats about their sensitivity to parameter settings.
Weather Impact Data Receives Selective Treatment
Reporting on events such as tornadoes and heavy rainfall shows similar patterns according to the book. Koonin describes cases where data cuts or emphasis on certain periods amplify apparent increases in frequency or intensity. These choices can create stronger narratives of worsening extremes even when longer records or alternative metrics tell a more mixed story.
The number of record temperatures recorded in U.S. data further illustrates the issue. Koonin points out that the count depends heavily on the exact criteria used to define a record. Different thresholds or time windows produce noticeably different tallies, raising questions about how consistently such statistics are applied across studies.
Research Incentives Shape Presentation Choices
Throughout the updated edition, Koonin returns to the role of professional incentives. He observes that climate researchers often face pressure to highlight anthropogenic influences because funding, publications, and public attention tend to follow those findings. This dynamic does not imply deliberate misconduct in every case, yet it can encourage selective emphasis that downplays contradictory evidence.
The book stops short of claiming widespread fabrication. Instead it calls for greater transparency in data handling and model development so that debates rest on clearer technical foundations. Readers are left to weigh whether current practices sufficiently separate scientific assessment from institutional priorities.
