Interesting study that shits on the climate change narratives pretty heavily: SCC-Grok-3-Review-V6.pdf (785.8 KB)
Lay summary:
This paper says the mainstream “CO₂ drives the climate” story is built on adjusted datasets and models that do not track reality, and that when you look at raw measurements and alternative solar records, most recent warming can be explained by natural processes rather than human emissions. The authors argue human CO₂ is a small, fast‑cycled slice of a massive carbon system, temperature changes tend to come first with CO₂ following, climate models miss real‑world wiggles and overheat the future, and “homogenization” steps inflate warming in the historical record. They also claim the Intergovernmental Panel on Climate Change (IPCC) picked a subdued solar record from several competing options; choosing higher‑variability solar reconstructions, plus known feedbacks, matches observed trends without needing large CO₂ effects. Net: they call for a reset of climate attribution that centers natural variability, solar forcing, and unadjusted observations over CO₂‑centric modeling.Salient points:
Data choice matters: They lean on UAH (University of Alabama in Huntsville) satellite temperatures, the pristine USCRN (U.S. Climate Reference Network) surface network, raw USHCN (U.S. Historical Climatology Network)/GHCN (Global Historical Climatology Network) rural stations, and NSIDC (National Snow & Ice Data Center) sea‑ice extent, emphasizing “unadjusted” records over homogenized global products.
CO₂ share and turnover: Human emissions are framed as ~10 GtC/year (~4% of a ~230 GtC annual carbon flux), with an effective atmospheric residence time of ~3–5 years, not 100+ years; see the residence‑time comparison table on PDF page 6 (journal p.18) summarizing bomb‑pulse ¹⁴C, Harde’s modeling, and the IPCC Bern model.
Temperature leads CO₂: Using stochastic causality, they report temperature changes precede CO₂ by ~6–12 months in modern data, and by centuries in ice‑core records, arguing CO₂ largely responds to temperature via oceanic outgassing and biospheric fluxes.
COVID‑19 “natural experiment”: A ~7% drop in 2019–2020 human emissions allegedly left no discernible dent in the Mauna Loa CO₂ curve, interpreted as sinks rapidly rebalancing the system.
Model‑data mismatch: CMIP5/6 runs are said to over‑warm and poorly track actual month‑to‑month variability (low R² versus UAH), miss pauses and spikes, and overstate Arctic sea‑ice loss relative to a post‑2007 plateau.
Sensitivity and feedbacks: The paper contends models inflate water‑vapor/cloud feedbacks and thus climate sensitivity, while observations favor lower or offsetting feedbacks.
Solar forcing choice is pivotal: They catalog 27 TSI reconstructions and argue the IPCC’s favored low‑variability PMOD (Physikalisch-Meteorologisches Observatorium Davos/World Radiation Center**)** understates solar influence; higher‑variability composites (e.g., ACRIM, or Active Cavity Radiometer Irradiance Monitor**)** plus cloud‑albedo feedbacks purportedly explain 50–100% of observed warming.
Adjustments and “hockey sticks”: Homogenization steps (e.g., cooling the 1930s, warming recent decades) are presented as manufacturing steeper trends that fit models; they also critique proxy reconstructions that flatten the Medieval period.
Scenario realism: Among IPCC SSPs (Shared Socioeconomic Pathways), they call SSP2 the most plausible baseline and label SSP5/RCP8.5 increasingly implausible, yet overused for impacts framing.
“Hens and eggs” framing: Causality is cast as bidirectional and stochastic, not a one‑way CO₂→temperature chain, which they say deterministic GCMs (General Circulation Models or Global Climate Models) cannot capture.
Methodological critique: They flag shared model code bases, tuning to adjusted datasets, and inadequate representation of ocean–atmosphere oscillations (ENSO (El Niño-Southern Oscillation), AMO (Atlantic Multidecadal Oscillation), PDO (Pacific Decadal Oscillation)) as structural weaknesses.
Meta‑note: Authored with Grok 3 (AI) as lead and named coauthors (including Soon and Legates), accepted in 2025 by Science of Climate Change; the authors themselves highlight heavy editorial correction of references and attributions.