The Frontier of Climate Science: Solar Variability, Natural Cycles and Model Uncertainty by Nicola Scafetta

Clintel Foundation
Date: 12 July 2026

The central theme of the presentation is that current global climate models (GCMs), do not adequately represent natural climate variability. Rather than arguing that human activities play no role in climate change, Scafetta maintains that the contribution of natural processes has been substantially underestimated, leading to an overestimation of climate sensitivity to greenhouse gases.

As he summarizes: “The IPCC assessments… have serious limitations because these models do not really capture natural variability across multiple time scales.”

Scafetta begins by placing modern climate change within Earth’s long geological history. Climate, he argues, has never been static. Over hundreds of millions of years, Earth has experienced alternating warm and cold periods that greatly exceeded the magnitude of modern warming. More importantly, climate records reveal oscillations occurring over a wide range of time scales—from annual and multidecadal cycles to millennial and even multimillion-year variations. Many of these cycles correspond to astronomical phenomena, including orbital variations, solar activity and oceanic oscillations.

Historical perspective

This historical perspective forms the foundation of his criticism of current climate attribution studies. Before scientists can determine how much of recent warming is caused by human activities, they must first understand the full extent of natural climate variability. While acknowledging that rising greenhouse gas concentrations are expected to influence climate, Scafetta believes that the natural background signal remains incompletely understood.

He then turns to the attribution framework used by the IPCC. He explains that the latest IPCC assessment concludes that virtually all observed warming since the late nineteenth century is attributable to human influence, while natural factors such as solar variability, volcanic activity and internal climate variability contribute little or nothing to the long-term warming trend.

This conclusion depends almost entirely on simulations produced by global climate models. These models are run under two different conditions: one including both anthropogenic and natural forcing, and another including only natural forcing. In the latter case, the models produce almost no warming since 1850, whereas warming appears only when anthropogenic greenhouse gas emissions are included. The IPCC therefore concludes that recent warming is almost entirely human-induced.

Scafetta argues that this reasoning suffers from a fundamental methodological weakness. The ‘natural forcing only’ simulations cannot be independently validated because obviously there is no observational record of an Earth without human influence against which the simulations can be tested. Consequently, the attribution is ultimately based on model assumptions rather than direct empirical verification.

This criticism becomes one of the central arguments of the lecture. Scafetta repeatedly emphasizes that model outputs should not be confused with experimentally verified scientific facts. Instead, he characterizes them as hypotheses whose credibility depends on how well they reproduce observed climate behaviour.

Climate sensitivity

One major source of uncertainty concerns equilibrium climate sensitivity—the amount of long-term warming expected after atmospheric carbon dioxide concentrations double. While the greenhouse effect of carbon dioxide is well established, large uncertainties remain in estimating climate feedbacks, particularly those involving clouds.

Cloud processes, he explains, strongly influence how much additional warming occurs following an initial radiative perturbation. Yet cloud dynamics remain difficult to model accurately, and after several decades of climate model development the estimated range of climate sensitivity has not narrowed substantially.

This persistent uncertainty raises questions about the confidence that can be placed in high-sensitivity model projections. He notes that different climate models produce equilibrium climate sensitivities ranging from approximately 1.8°C to more than 5.5°C for a doubling of atmospheric CO2, a remarkably wide spread for models intended to represent the same physical climate system.

He also cites studies suggesting that several of the newest generation of global climate models produce warming that is systematically higher than observed. Models with lower climate sensitivity generally agree better with the available observational datasets, whereas the higher-sensitivity models tend to exaggerate warming trends.

Urban heat island

Scafetta continues by discussing observational evidence that exposes additional weaknesses in current climate models. One example is the discrepancy between satellite temperature measurements and surface temperature records. Satellite observations of the lower troposphere generally indicate less warming than surface-based datasets, particularly over land. Since satellite and sea-surface observations agree much more closely over the oceans, part of the stronger warming over land probably results from non-climatic influences such as the urban heat island effect.

Scafetta’s broader criticism concerns the ability of global climate models to reproduce well-known climate variations from the past. He argues that if the models cannot accurately simulate major historical warm periods that occurred before industrialization, then they are likely missing important physical mechanisms that also influence the modern climate.

He focuses first on the Medieval Warm Period. Temperature reconstructions indicate that significant regional warmth occurred roughly between AD 900 and 1300. Yet, he argues, current climate models fail to reproduce this event convincingly. If such naturally occurring warm periods cannot be generated by the models, he reasons, then the models must be omitting important sources of natural variability: “The models are completely unable to reproduce natural climate variability, in particular the warm periods of the past.”

He extends the same argument to the Holocene Thermal Maximum approximately 6,000–8,000 years ago, when many paleoclimate reconstructions suggest temperatures in several regions exceeded those of the twentieth century. Because atmospheric CO2 concentrations during that period were lower than today, these warm conditions cannot easily be explained by greenhouse forcing alone. He concludes that additional natural mechanisms must have contributed substantially to long-term climate evolution.

Solar variability

A recurring theme throughout the presentation is the role of solar variability. According to Scafetta, the IPCC assigns only a very small contribution to changes in solar activity since the nineteenth century. But numerous paleoclimate records display strong correlations between reconstructed solar activity and long-term climate changes over centuries and millennia.

One reason for this disagreement, is that climate models rely on solar irradiance reconstructions that show relatively little long-term variability. Other published reconstructions suggest considerably larger changes in solar output. Because the choice of solar dataset directly affects model results, Scafetta contends that the contribution of the Sun may have been underestimated from the outset.

He also argues that solar influences may operate through mechanisms beyond changes in total solar irradiance alone. In particular, he discusses the hypothesis that variations in the Sun’s magnetic activity modulate the flux of galactic cosmic rays reaching Earth’s atmosphere. These changes could, in turn, influence cloud formation and thereby alter Earth’s energy balance. While acknowledging that this mechanism remains the subject of ongoing scientific debate, Scafetta argues that it deserves far more attention than it currently receives in mainstream climate modelling.

Cycles

Building on these ideas, he presents a series of empirical climate models that incorporate natural oscillations, including multidecadal and millennial solar-related cycles. According to Scafetta, these empirical models reproduce observed temperature fluctuations more accurately than conventional global climate models, especially features such as the warming during the early twentieth century, the mid-century cooling and the approximately 60-year oscillation visible in several observational datasets.

He argues that incorporating stronger natural variability reduces the proportion of recent warming attributed to anthropogenic greenhouse gases while still acknowledging a human contribution. In his interpretation, natural and anthropogenic influences are both important, whereas current IPCC models assign too little weight to natural drivers.

Summarizing his position, Scafetta states: “It’s possible to model climate change by assuming the natural cycles… the anthropogenic component is much lower than what the IPCC assumes.”

These scientific arguments ultimately lead to policy conclusions. Because Scafetta believes current climate models systematically overestimate future warming, he argues that achieving the objectives of the Paris Agreement may not require immediate global net-zero emissions. Using what he considers more realistic emission pathways together with climate models that incorporate stronger natural variability, he concludes that future warming could remain below 2°C without the most stringent mitigation scenarios, Consequently, he advocates a greater emphasis on adaptation rather than rapid decarbonization.

Scafetta’s critique is not primarily directed at the existence of anthropogenic warming, but at the magnitude assigned to it by existing models. He argues that unresolved uncertainties surrounding cloud feedbacks, climate sensitivity, historical temperature reconstructions, solar forcing and multidecadal climate oscillations indicate that today’s models remain incomplete representations of the climate system.

The article you have just read was made possible by our donors.

Every day, Clintel publishes articles on climate, energy and science. We also translate and share international analyses in multiple languages, produce videos, publish reports and organize conferences and events around the world.

We receive no government funding and rely entirely on the support of our donors. Your support helps us promote independent research and contribute to a more open and balanced debate on climate and energy issues.

Would you like to support our work? Choose the option that suits you best:

• Become a Friend of Clintel – support us with an annual contribution
• Make a recurring donation – provide ongoing support and help us plan ahead
• Make a one-time donation – every contribution helps

Thank you for your support.