Modern climate models are used to support the claim that recent warming is unprecedented and largely man-made. This claim is based on the assumption that the models can accurately reproduce Earth’s past climate behavior. But they don´t.

A closer look at Nicola Scafetta’s new study shows that climate models repeatedly fail to reproduce known warm periods of the past, including the Medieval Warm Period (MWP) between about 900 and 1300 AD.

Scafetta highlights the Medieval Warm Period (MWP, ~900–1300 CE) as a prime example of model shortcomings.

Proxy-based reconstructions (e.g., from earlier studies like Lamb 1965, and more recent ones such as Moberg et al. 2005, Ljungqvist 2010, Christiansen and Ljungqvist 2012) show a pronounced warm phase in the Northern Hemisphere, often estimated as warmer than the first half of the 20th century in some records, followed by the Little Ice Age (LIA, ~1300–1850 CE). These form part of quasi-millennial oscillations (e.g., the Eddy cycle).

However, CMIP GCMs (especially those in CMIP6 used for IPCC AR6) produce relatively flat or minimally varying temperatures during this medieval interval. They do not capture the significant warming.

Nicola Scafetta is an Italian physicist and researcher (affiliated with institutions like the University of Naples and previously with Duke University/NASA-related labs) known for his work on climate variability, solar influences, and critiques of mainstream climate models.

His research frequently discusses the Medieval Warm Period (MWP, roughly 800–1350 CE), a period of relative warmth documented in various paleoclimate records, particularly in the Northern Hemisphere (e.g., Europe, North Atlantic regions). Scafetta argues that the MWP (along with earlier periods like the Roman Warm Period) reflects significant natural climate variability, often linked to a quasi-millennial oscillation in climate records.

Key points from his publications:

• He highlights that current global climate models (e.g., CMIP6 models used in IPCC reports) fail to reproduce the MWP. For instance, in his 2021 and 2024 papers, he points out that these models show little to no warming during the medieval era because they assume low secular variability in total solar irradiance (TSI) and lack mechanisms to capture natural millennial-scale cycles. He references IPCC AR6 figures acknowledging larger model-reconstruction disagreements before ~1300 CE.
• Scafetta attributes the MWP (and contrasting cold periods like the Little Ice Age) primarily to solar activity variations, including higher solar output during the medieval era, possibly amplified by non-radiative mechanisms (e.g., solar magnetic modulation of cosmic rays or other particle fluxes). He contrasts this with models that minimize solar forcing, leading them to attribute nearly all modern warming to anthropogenic factors.
• In reconstructions (e.g., his 2021 millennial global surface temperature work), he incorporates natural oscillations (including ~60-year, ~1000-year cycles) and suggests these explain a substantial portion of historical and recent temperature changes, implying lower equilibrium climate sensitivity (ECS around 1–2 °C) than many GCMs assume.
• Broader context in his work: He uses multi-proxy solar records and empirical analyses to argue that natural variability (solar/astronomical) has been underestimated, affecting interpretations of both the MWP and modern warming.

“Detection, attribution, and modeling of climate change: Key open issues” is a 2025 (published 2026 in print) open-access review paper by Nicola Scafetta, published in Gondwana Research (Volume 152, April 2026, Pages 92-128; DOI: 10.1016/j.gr.2025.05.001). An earlier preprint appeared on arXiv in May 2025 (arXiv:2506.13994).

This work builds on Scafetta’s longstanding research emphasizing natural climate variability (especially solar and astronomical influences), critiques of mainstream Coupled Model Intercomparison Project (CMIP) global climate models (GCMs), and implications for lower climate sensitivity and moderated future warming projections.

Main Highlights and Arguments

• Core Thesis: Climate science faces unresolved “key open issues” in detection (accurately identifying changes), attribution (assigning causes), and modeling (simulating past and future climates). Scafetta argues that CMIP GCMs (used heavily in IPCC assessments) struggle significantly with these, particularly in reproducing natural multi-timescale variability observed throughout the Holocene, including periods like the Medieval Warm Period and Little Ice Age.
• Primary Scientific Challenges Identified:
  1. Uncertainties in observational data: Global surface temperature records may be affected by non-climatic biases, such as urban heat island (UHI) effects, station moves/relocations, instrumentation changes, and land-use alterations. Scafetta suggests these could lead to overestimation of recent warming trends, citing evidence like rural-only station subsets showing less warming and the post-1980 “divergence problem” between instrumental records and tree-ring proxies.
  2. Underestimation of natural variability: GCMs assume low secular solar irradiance variability and lack adequate mechanisms for millennial-scale cycles (e.g., Eddy ~1000-year cycle). This causes poor hindcasting of pre-industrial warm/cold periods and over-attribution of recent warming to anthropogenic factors.
  3. Model-observation discrepancies: Examples include:
     • Lack of observed tropical upper-tropospheric “hot spot” predicted by models.
     • Satellite tropospheric records showing less warming than surface records (especially in the Northern Hemisphere), suggesting surface data inflation (~21% excess globally since ~1980–2020).
     • Divergences in proxy vs. instrumental trends post-1980.
• Climate Sensitivity and Projections:
  • Equilibrium climate sensitivity (ECS) to CO₂ doubling may be lower than IPCC ranges (often >3 °C), with empirical estimates suggesting ~1.1 ± 0.4 °C or below 3 °C.
  • Incorporating natural oscillations (e.g., 60-year, millennial cycles) in harmonic/empirical models explains much of 20th–21st century warming, projecting only moderate additional warming (1 °C or less by 2100 under mid-range SSP scenarios).
  • This challenges the urgency of aggressive “Net-Zero” mitigation policies, favoring adaptation and refined modeling over drastic emissions cuts.
• Policy and Broader Implications:
  • Scafetta calls for more refined approaches that better integrate natural drivers (solar, astronomical) and address data uncertainties.
  • He argues GCM limitations mean projections may overestimate risks, urging open discussion of alternatives for informed, sustainable strategies.