Exploring the conditions conducive to convection within the Greenland Ice Sheet

Numerical modeling suggests that convection drives large englacial plumes in north Greenland by exploiting softer, more stable ice, a finding that implies significantly lower ice viscosity and reduced basal sliding than previously assumed, thereby offering a pathway to improve future ice-sheet mass balance projections.

The Gist

Imagine the Greenland Ice Sheet not just as a giant, solid block of frozen water, but as a slow-moving, thick river of ice. For decades, scientists have been trying to map the history of this river by looking at its layers, much like reading the rings of a tree. But in the northern part of Greenland, they've found something strange: massive, chaotic "plumes" or bubbles of ice near the bottom that look like they've been stirred up, scrambling the neat layers of history.

This paper is like a detective story trying to figure out how these plumes got there.

The Mystery: The Stirred-Up Ice

Think of the ice sheet as a giant pot of soup. Usually, the soup sits still, with layers forming neatly over time. But in the north, the bottom of the pot seems to be bubbling up, creating giant swirls that stretch for miles. Scientists have seen these "plumes" on radar images, but they didn't know what caused them. Was it the bedrock underneath? Was it water melting and refreezing?

The authors of this paper propose a new theory: Thermal Convection.

The Analogy: The Lava Lamp

To understand convection, imagine a lava lamp.

• The wax at the bottom gets heated by the light bulb.
• As it gets hot, it becomes lighter (less dense) and floats upward.
• When it reaches the top, it cools down, gets heavy again, and sinks back down.
• This creates a continuous, rolling loop of movement.

The authors suggest that the bottom of the Greenland Ice Sheet is doing something similar. The Earth's core is warm, heating the ice from below. If the ice is soft enough and thick enough, that warm bottom layer tries to float up, creating these giant, rising plumes of ice that disrupt the layers above.

The Recipe for a "Lava Lamp" Ice Sheet

The researchers ran computer simulations to see what conditions are needed to make this "ice lava lamp" work. They found that you need a very specific recipe, and most of Greenland doesn't have it. Only the North seems to fit the bill:

1. The Ice Must Be Thick: You need a deep pot (over 2,000 meters deep) for the bubbles to grow large enough to see.
2. The Ice Must Be "Soft": Think of the ice like honey. If the honey is cold and hard (like rock), it won't flow. But if it's warm and soft, it can move. The paper suggests the ice in North Greenland is 9 to 15 times softer than scientists previously thought. It's like the difference between trying to stir cold molasses versus warm honey.
3. The Ice Must Be Slow: If the ice sheet is moving too fast (like a rushing river), the "bubbles" get stretched out and torn apart before they can form. North Greenland is very slow-moving, giving the bubbles time to grow.
4. No Heavy Rain (Snow): If it snows too much, the weight of the new snow pushes the ice down, squashing the rising bubbles. North Greenland is dry; South Greenland is wet. This explains why we see these plumes in the North but not the South.

Why Does This Matter?

You might ask, "So what? It's just some weird ice bubbles."

Here is the big picture:

• Predicting the Future: Scientists use computer models to predict how much ice will melt and how much sea levels will rise. These models rely on knowing how "stiff" or "soft" the ice is.
• The Correction: If the ice in North Greenland is actually 15 times softer than we thought, our current models are wrong. They are likely overestimating how much the ice is sliding on the ground and underestimating how much it is deforming internally.
• The Fix: By updating the models to include this "soft ice" and the "convection" process, we can get much more accurate predictions about future sea-level rise.

The Bottom Line

This paper suggests that the Greenland Ice Sheet isn't just a static block of ice. In the quiet, cold, and thick north, it's actually churning like a slow-motion lava lamp. This churning is caused by the ice being much softer than we realized. Recognizing this hidden movement helps us understand the ice sheet's true behavior and gives us a better crystal ball for predicting how our coastlines might change in the future.

Technical Summary

Here is a detailed technical summary of the paper "Exploring the conditions conducive to convection within the Greenland Ice Sheet" by Law et al.

1. Problem Statement

The Greenland Ice Sheet (GrIS) contains large, plume-like features (>1/3 of local ice thickness) that disrupt radiostratigraphy (isochrones), complicating the reconstruction of past ice dynamics and paleoclimate.

• Observation: These "basal plumes" are predominantly found in northern Greenland but are scarce in the south.
• Current Hypotheses: Previous explanations included basal freeze-on or traveling basal slippery spots, both requiring thawed bed conditions. However, these mechanisms struggle to explain the specific geometry, ubiquity in the north, and absence in the south.
• Rheological Uncertainty: Numerical models predicting future mass loss rely on accurate ice rheology (flow laws). A critical parameter is the enhancement factor ($E$), which accounts for grain size, impurities, and fabric. $E$ is poorly constrained in situ, and errors in $E$ lead to compensatory errors in basal traction inversions, reducing model accuracy.
• Research Question: Can thermal convection explain the formation of these large englacial plumes, and what constraints does this place on the rheology of the GrIS?

2. Methodology

The authors employed numerical modeling using the geodynamics software ASPECT 2.5.0, chosen for its ability to handle buoyancy forces and convection benchmarks, which are often lacking in standard ice-sheet models.

• Domain Setup:
  • 2D along-flow slices (25 km length) and 3D cuboids (22 km $\times$ 18 km).
  • Uniform ice thickness (default 2.5 km).
  • Surface mass balance set to zero (except in specific sensitivity runs).
  • Basal boundary condition: Rigid ($v_x = 0$) to simulate a frozen bed, isolating internal deformation.
• Physics & Governing Equations:
  • Solved conservation of mass, momentum, and energy under the Boussinesq approximation (density variations only affect buoyancy).
  • Rheology: Simplified to Newtonian flow ($n=3$) where viscosity is controlled by the enhancement factor ($E$) and temperature. Effective stress ($\tau_e$) was prescribed based on ISSM simulations to approximate realistic stress states.
  • Thermal Structure: Two baseline temperature profiles were used:
    • NEEM (North): Colder, representing the northern interior.
    • DYE-3 (South): Warmer, representing the southern interior.
  • Initial Conditions: A temperature perturbation (simulating a fold) was introduced to trigger potential convection.
• Parameter Sweep: The study systematically varied four key parameters to determine their effect on the maximum upward vertical velocity ($max(v_z)$):
  1. Enhancement factor ($E$).
  2. Surface shear velocity ($v_{x,s}$).
  3. Ice thickness ($H$).
  4. Snow accumulation rate ($v_{z,s}$).
• Classification: Convection behavior was categorized into three zones based on $max(v_z)$ over 20 kyr:
  • Suppressed: $max(v_z) < 0.01$ m/yr or decaying.
  • Sustained: Intermediate growth.
  • Amplifying: $max(v_z) > 0.4$ m/yr or significant growth.

3. Key Contributions

• Mechanism Identification: Proposes local thermal convection as the primary mechanism for forming large (>1/3 thickness) englacial plumes, distinct from full-thickness convection or basal sliding mechanisms.
• Rheological Constraint: Provides a novel method to constrain the enhancement factor ($E$) by matching modeled convection viability with observed plume distributions.
• Spatial Explanation: Explains the north-south dichotomy of plume distribution in Greenland based on the interplay of ice thickness, flow speed, and accumulation rates.
• Modeling Framework: Demonstrates the utility of geodynamics codes (ASPECT) for investigating ice-sheet internal processes that standard models cannot resolve.

4. Results

The modeling identified four critical thresholds required for convection to occur and sustain plume-like structures:

1. Ice Thickness: Must be $> \sim 2,200$ m. Thinner ice prevents the development of sufficient buoyancy forces.
2. Horizontal Shear: Surface velocity must be low ($< \sim 1$ m/yr). High shear disrupts the vertical flow required for convection.
3. Snow Accumulation: Must be low ($< \sim 0.15$ m/yr). High accumulation (downward advection) suppresses upward convective motion.
4. Enhancement Factor ($E$): Must be significantly higher than standard assumptions.
   • For the NEEM (North) profile, convection becomes viable when $45 \le E \le 75$.
   • Standard models typically assume $E \approx 4-6$.
   • This implies the ice in northern Greenland is 9–15 times softer (lower viscosity) than commonly assumed.

Spatial Correlation:

• Northern Greenland: Satisfies all conditions (thick ice, slow flow, low accumulation, old/soft ice). This aligns with the observed concentration of plumes.
• Southern Greenland: Fails conditions due to faster flow, higher accumulation, and younger ice. This explains the dearth of plumes in the south.

Plume Geometry:

• 3D simulations produced symmetric overturning and roll-over features that closely match the geometry of observed radar plumes (Figs. 2 and 4).
• The heat flux required to sustain these temperatures ($\sim 40-70$ mW m$^{-2}$) is consistent with geothermal heat flux estimates for the GrIS.

5. Significance and Implications

• Revised Ice Rheology: The finding that northern GrIS ice may be 9–15 times softer than standard models suggests that internal deformation (rather than basal sliding) may be the dominant mode of ice transport in these regions.
• Impact on Future Projections:
  • Current models assuming higher viscosity likely overestimate basal traction (friction) to compensate, leading to biased projections of mass loss.
  • Implementing this "softer" rheology could reduce compensatory errors and improve the accuracy of future sea-level rise projections.
• Paleoclimate Reconstruction: Recognizing convection as a driver of stratigraphic disruption allows for better interpretation of radiostratigraphy, potentially improving the reconstruction of past ice dynamics and climate history.
• Antarctic Context: The relative lack of such plumes in Antarctica (outside specific mountainous regions) may be due to colder temperatures (higher viscosity) or sampling bias, rather than a fundamental difference in ice physics.

In conclusion, the study posits that thermal convection is a viable and likely dominant mechanism for the formation of large englacial plumes in northern Greenland, necessitating a significant revision of assumed ice rheology parameters in numerical models.