A new study published in the Proceedings of the National Academy of Sciences (PNAS) in 2026 proposes that the famous “Snowball Earth” events—particularly the long Sturtian glaciation (~717–658 million years ago)—may have involved a far stranger, oscillating climate cycle than the traditional “one long freeze followed by rapid thaw” model.

Traditional Snowball Earth View

During the Cryogenian Period (part of the Neoproterozoic), Earth experienced extreme glaciations where ice sheets reached the tropics, and the planet was largely (or entirely) encased in ice for millions of years.

The classic mechanism:

Runaway cooling: Increased ice cover raised Earth’s albedo (reflectivity), reflecting more sunlight and causing further cooling in a positive feedback loop.

Escape from snowball: Volcanic CO₂ built up over millions of years because silicate weathering (a major CO₂ sink) largely stopped under ice. Eventually, the greenhouse effect overwhelmed the ice, leading to rapid deglaciation and extremely hot “hothouse” conditions afterward.

This explained glacial deposits at low latitudes and cap carbonates (unusual rock layers formed during rapid warming). However, it struggled with some observations, such as the very long duration of the Sturtian (~56–60 million years) and evidence that life (including oxygen-dependent processes) persisted.

The New “Limit Cycle” Model: A Stranger, Repeating Climate Oscillation

The recent modeling study suggests the Sturtian wasn’t one monolithic deep freeze. Instead, it featured repeated “limit cycles”—oscillations between glacial (cold, icy) and interglacial/hothouse (warmer) states driven by interactions between volcanism, the Franklin Large Igneous Province (LIP), and silicate weathering.

Key dynamics:

• Weathering of fresh basaltic rocks from the Franklin LIP (a massive volcanic event) acted as a powerful CO₂ sink, drawing down atmospheric carbon and triggering or prolonging glaciation.
• During full glaciation, weathering slowed or halted → CO₂ from ongoing volcanism built up → warming eventually occurred.
• Upon partial deglaciation, remaining unweathered basalt from the LIP became available again for rapid weathering → CO₂ drawdown resumed → cooling restarted the cycle.

This back-and-forth continued until the “weathering power” of the LIP was largely exhausted. It better explains the prolonged duration of the Sturtian and how biogeochemical cycles (including oxygen levels supporting life) could continue operating, rather than everything shutting down for tens of millions of years.

_____________________________________________________________________________________

Repeated snowball–hothouse cycles within the Neoproterozoic Sturtian glaciation

“Repeated snowball–hothouse cycles within the Neoproterozoic Sturtian glaciation” is the exact title of a new paper published in the Proceedings of the National Academy of Sciences (PNAS) in late April 2026 by Charlotte Minsky and colleagues.

Published: Proceedings of the National Academy of Sciences (PNAS)

Authors: Charlotte Minsky, Robin Wordsworth, David T. Johnston and Andrew H. Knoll

DOI: 10.1073/pnas.2525919123

Abstract

The Neoproterozoic Era was a pivotal era of Earth’s history in which multiple severe glaciations profoundly influenced the biosphere, but explaining the duration and nature of these events remains a major challenge. Notably, geochronology indicates that the Sturtian glaciation lasted for ∼56 Myr—far longer than can be accommodated by canonical “Snowball” or “Slushball” models. Here we use a coupled box model of the Neoproterozoic climate and carbon cycle to develop a hypothesis for Sturtian climate evolution. We show that weathering of the Franklin igneous province would have caused Earth to enter a limit cycle regime, alternating between Snowball and hothouse states for the duration of the Sturtian. This scenario resolves the duration problem, is allowable given currently observed patterns of sedimentation, and predicts syn-Sturtian oxygen stability.

Key Finding

Instead of viewing the Sturtian glaciation (~717–661 Ma, lasting roughly 56 million years) as one continuous, unbroken “hard snowball” state, the authors propose it involved repeated limit cycles — oscillations between extreme glacial (snowball) conditions and warmer hothouse intervals.

This dynamic cycling better explains several longstanding puzzles:

The unusually long duration of the Sturtian (far longer than the later Marinoan glaciation).

How biogeochemical cycles (especially oxygen) and life could persist over tens of millions of years.

Evidence that the climate system was not statically frozen for the entire interval.

The Role of the Franklin Large Igneous Province (LIP)

The cycles were driven by the massive Franklin LIP, a huge volcanic province emplaced around 719–716 Ma in what is now northern Canada and Greenland, just before or overlapping the onset of the Sturtian.

Here’s how the limit cycle worked in the model:

Initial cooling and snowball onset: Weathering of the fresh, calcium- and magnesium-rich basalts from the Franklin LIP rapidly drew down atmospheric CO₂ (silicate weathering is a major CO₂ sink). Combined with other factors (low mid-ocean ridge outgassing, continental configuration), this pushed Earth past a tipping point into runaway ice-albedo feedback, triggering global glaciation.

During full snowball: Ice cover greatly reduced or halted subaerial silicate weathering → volcanic CO₂ (from ongoing volcanism) began to accumulate in the atmosphere.

Deglaciation and hothouse phase: Rising CO₂ eventually overwhelmed the high albedo, causing rapid warming and partial or full deglaciation. During this warmer “hothouse” interval, the remaining unweathered portions of the Franklin basalts were exposed again to intense tropical weathering.

Repeat: Renewed rapid CO₂ drawdown from the fresh basalt triggered cooling and another return to snowball conditions.

This back-and-forth continued until most of the weatherable basalt volume of the LIP was exhausted, eventually allowing a more stable exit from the glacial regime. The model shows that if only a portion of the LIP was weathered in the first snowball, enough remained to fuel subsequent cycles.

Why This Is a “Stranger” Climate Cycle

Traditional “canonical” snowball models assumed a single long freeze followed by one dramatic escape driven by volcanic CO₂ buildup.

The new limit-cycle hypothesis introduces internal oscillations within the overall glacial epoch, driven by the interaction of LIP emplacement, weathering feedback, and the carbon cycle.

It reconciles the extreme length of the Sturtian without requiring implausibly slow CO₂ buildup or constant subglacial weathering (though other 2026 papers explore continued subglacial weathering as an additional factor that could prolong glaciation).

This framework also helps explain how oxygen-dependent life and other processes could continue operating, as full shutdown for 56 million years would have been biologically challenging.

The study highlights how large igneous provinces can drive not just one-off climate shifts but prolonged, self-sustaining oscillatory behavior through carbon cycle feedbacks. It refines our understanding of Earth’s climate sensitivity in extreme states and provides a testable hypothesis for future geological and geochemical studies of Cryogenian rocks.