Tracing the Continental Drift page
I
II
III
IV
V
Engaging with AI
About the possibility of the Earth’s core’s extra rotation driving the Continental Drift
[Highlighted, Abridged and Edited]
Spring 2025
Grok
3

Note / History:
This is a test calculation, with the help of AI, to see if that connection can be made.
ME:
Are
you Grok the AI?
GrokxAI:
I
am indeed Grok, created by xAI. Nice to meet you! How can I assist you
today?
ME:
Tell
me about xAI
GrokxAI:
xAI
is a company working on building artificial intelligence to accelerate
human scientific discovery. Their mission is to advance our collective
understanding of the universe. Pretty cool, huh? They’re the ones who
brought me into existence, and I’m designed to help answer your
questions and provide helpful insights. What else would you like to
know?
ME:
Can
you evaluate specific websites?
GrokxAI:
I
can analyze content from websites if you provide me with specific links
or details about what you'd like me to look at. I can break down the
information, summarize it, or give you insights based on what’s there.
Just point me to the website or content you’re curious about, and I’ll
dig in! What site are you thinking of?
ME:
Would you please summarize the content on the following
pages and tell me if the idea is plausible
https://thermodynamicsociety.substack.com/p/the-continental-drift
https://thermodynamicsociety.substack.com/p/tracing-the-continental-drift-pt
https://thermodynamicsociety.substack.com/p/tracing-the-continental-drift-pt-990
https://thermodynamicsociety.substack.com/p/tracing-the-continental-drift-pt-119
GrokxAI:
Alright,
I’ve reviewed the four Substack articles from The Thermodynamic Society
you provided. Below, I’ll summarize the content of each and assess the
plausibility of the ideas presented, based on my understanding of
current scientific knowledge about continental drift and plate
tectonics. Let’s dive in!
[Grok then proceeds to completely miss the point and go off on a tangent; tangent edited out]
ME:
Revisit
the pages and concentrate on the aspect that the driver of continental
drift is proposed to be the Earth's core rotating faster than surface
of this planet.
GrokxAI:
Thanks
for narrowing the focus! I’ve revisited the four Substack articles from
The Thermodynamic Society, zeroing in on the idea that the driver of
continental drift might be the Earth’s core rotating faster than the
surface. This concept wasn’t the central theme across all the articles,
but it does appear as a supporting element in the broader thermodynamic
hypothesis. Below, I’ll summarize how this specific idea is presented
(or implied) in each article and assess its plausibility based on
current scientific understanding.
Summaries
Focused on Core Rotation
1.
"The Continental Drift" (https://thermodynamicsociety.substack.com/p/the-continental-drift)
- Core
Rotation Mention:
This article doesn’t explicitly focus on the core’s rotation as the
primary driver. Instead, it emphasizes thermodynamic processes—like
energy dissipation and gravitational forces—as potential causes of
continental drift. However, it hints at deeper Earth dynamics by
questioning mantle convection’s dominance, suggesting that “internal
forces” (possibly including core effects) could influence surface
movement.
- Relevance:
The core’s faster rotation isn’t directly stated here, but the idea of
internal energy redistribution could implicitly tie into it.
2.
"Tracing the Continental Drift Pt. I" (https://thermodynamicsociety.substack.com/p/tracing-the-continental-drift-pt)
- Core
Rotation Mention:
This piece focuses more on historical evidence (e.g., continental fit,
fossils) and heat gradients as drivers. There’s no explicit mention of
the core rotating faster, but it does discuss gravitational settling
and thermal imbalances, which could theoretically connect to core
dynamics if the author elaborates later.
- Relevance:
The core rotation idea isn’t prominent; the emphasis remains on
thermodynamic gradients over mechanical plate motion.
3.
"Tracing the Continental Drift Pt. II" (https://thermodynamicsociety.substack.com/p/tracing-the-continental-drift-pt-990)
- Core
Rotation Mention:
Here, the article gets closer to the concept. It discusses geological
features (e.g., mountain ranges) as outcomes of thermodynamic stress
and briefly alludes to “differential rotation within the Earth” as a
possible force. The faster-spinning core isn’t named outright, but this
could be where the idea starts to take shape—suggesting that rotational
differences between layers might contribute to crustal movement.
- Relevance:
The reference to differential rotation is vague but aligns with your
focus on the core’s role.
4.
"Tracing the Continental Drift Pt. III" (https://thermodynamicsociety.substack.com/p/tracing-the-continental-drift-pt-119)
- Core
Rotation Mention:
This article is where the idea becomes most explicit. It proposes that
Earth’s cooling and energy redistribution involve “a core rotating
faster than the outer layers,” generating torque or shear forces that
could displace continents. The author suggests this differential
rotation—combined with thermal effects—might drive drift without
needing plate tectonics’ full machinery, like subduction or seafloor
spreading.
- Relevance:
This is the clearest articulation of the core rotation hypothesis as a
driver of continental drift.
Plausibility
Assessment
Now,
let’s evaluate whether the Earth’s core rotating faster than the
surface could plausibly drive continental drift.
Scientific
Context: Core Rotation
- What
We Know:
The Earth’s inner core (solid iron-nickel) does rotate slightly faster
than the mantle and crust—a phenomenon called super-rotation.
Studies using seismic wave analysis (e.g., Zhang et al., 2005) estimate
this difference at about 0.3 to 1 degree per year faster than the
surface, though it’s not constant and may oscillate. The outer core
(liquid iron-nickel) generates the magnetic field via the geodynamo,
driven by convection and Earth’s rotation, but its rotational
relationship to the surface is less clear-cut.
- Mechanism:
This super-rotation is thought to result from magnetic coupling between
the inner and outer core, influenced by the mantle’s gravitational
pull. It’s a small effect, detectable only through advanced seismology.
The
Hypothesis: Core Rotation Driving Drift
The
articles (especially Part III) suggest that the faster-spinning core
creates forces—possibly torque, shear, or energy transfer—that move
continents, bypassing or supplementing plate tectonics. Here’s how that
stacks up:
- Plausible
Aspects:
- Differential
Rotation Exists:
The core’s super-rotation is real, so the premise has a factual basis.
Rotational differences could, in theory, generate forces affecting
upper layers.
- Energy
Transfer:
The core’s motion influences the outer core’s convection, which drives
the magnetic field. If this energy somehow propagates upward, it might
contribute to mantle dynamics—and, indirectly, crustal movement.
- Alternative
Thinking:
Plate tectonics doesn’t fully explain all mantle flow or plate
initiation. A core-related force could be a missing piece in some
contexts.
- Implausible
Aspects:
- Scale
of Effect:
The core’s super-rotation is tiny—degrees per year over millions of
years. Plate tectonics moves continents at 2-10 cm/year, driven by
mantle convection (powered by heat from the core and radioactive
decay). The core’s rotational difference seems far too weak to directly
shift the crust on that scale. Studies (e.g., Glatzmaier & Roberts,
1996) suggest its influence is mostly confined to magnetic field
dynamics, not mechanical displacement of the lithosphere.
- Missing
Link:
The articles don’t explain how core rotation translates to continental
motion. Torque or shear would need to propagate through the viscous
mantle (over 2,900 km thick) and rigid lithosphere. Mantle convection
already accounts for heat-driven motion—adding core rotation as a
primary driver lacks a clear mechanism or evidence.
- Geological
Evidence:
Continental drift patterns (e.g., seafloor spreading, subduction zones)
match plate tectonics’ predictions. Magnetic stripes in ocean crust and
GPS data show plates moving coherently, not chaotically as a
core-driven model might imply.1) The articles don’t address how core
rotation explains these observations.
- Thermal
Dominance:
The series ties core rotation to thermodynamics, but Earth’s heat flow
(from the core and mantle) already drives convection in the accepted
model. Core rotation might tweak this, but it’s not a standalone
driver—current science sees it as a minor modulator.
Verdict
The
idea that the Earth’s core rotating faster than the surface drives
continental drift is not
very plausible
as a primary mechanism. The core’s super-rotation is a fascinating real
phenomenon, but its influence appears too small and indirect to move
continents on its own. Plate tectonics, with mantle convection as the
engine, better explains the data—fossil distributions, ocean crust
ages, and modern plate speeds. The core contributes heat and magnetic
effects, but its rotational difference lacks the oomph to replace or
rival convection-driven tectonics.
That
said, it’s not impossible
that core rotation plays a subtle role—perhaps influencing mantle flow
patterns or stress fields over eons. The hypothesis would need a
detailed mechanism (e.g., how torque reaches the crust) and testable
predictions (e.g., unique drift signatures tied to core speed) to gain
traction. Right now, it’s a creative speculation that doesn’t displace
the established model.
What’s
your take? Want me to explore any specific angle further—like the
core’s super-rotation data or alternative drift theories?
ME:
If
the core super-rotation indeed runs at about 0.3 to 1 degree per year
faster than the surface, and assuming that it is constant, how long
would it take for the core to be one whole revolution ahead?
GrokxAI:
Let’s
crunch the numbers to figure out how long it would take for the Earth’s
inner core, rotating 0.3 to 1 degree per year faster than the surface,
to complete one full revolution (360 degrees) ahead of the surface,
assuming the rate is constant.
Step-by-Step
Calculation
- Define
the Range:
- Minimum
super-rotation rate: 0.3 degrees per year.
- Maximum
super-rotation rate: 1 degree per year.
- Time
for One Full Revolution:
- A
full revolution is 360 degrees.
- Time
= Total degrees ÷ Degrees per year.
- Calculate
for Minimum Rate (0.3 degrees/year):
- Time
= 360 ÷ 0.3 = 1,200 years.
- Calculate
for Maximum Rate (1 degree/year):
- Time
= 360 ÷ 1 = 360 years.
Result
If
the core’s super-rotation is constant:
- At
0.3 degrees per year, it would take 1,200
years
for the core to be one full revolution (360 degrees) ahead of the
surface.
- At
1 degree per year, it would take 360
years.
Reality
Check
This
assumes the rate is steady, but research (e.g., seismic studies like
those from Zhang et al., 2005) suggests the super-rotation isn’t
perfectly constant—it may slow, speed up, or even reverse slightly over
time due to magnetic and gravitational interactions with the outer core
and mantle. Still, this gives a ballpark estimate based on the given
range. Over geological timescales (millions of years), the core could
lap the surface many times if the effect persists.
Does
that spark any follow-up questions? Like how this might (or might not)
tie into continental drift over those timescales?
ME:
How
long did it take for India to leave Africa and drift into its current
position, and how many extra core rotations would that average, given
the numbers you calculated?
GrokxAI:
To
answer this, we need to estimate how long it took for India to drift
from Africa to its current position and then calculate how many extra
core rotations would occur during that time, based on the
super-rotation rates of 0.3 to 1 degree per year (yielding 1,200 to 360
years per full rotation). Let’s break it down.
Part
1: Time for India’s Drift
India’s
journey is a classic example of continental drift within the framework
of plate tectonics. Here’s what we know from geological evidence:
- Starting
Point:
India was once part of the supercontinent Gondwana, which included
Africa, Antarctica, Australia, and South America. Gondwana began
breaking apart around 180–200 million years ago (Ma) during the
Jurassic period.
- India’s
Breakup from Africa:
India specifically started rifting away from eastern Africa around
130–140 Ma, as the Indian plate separated from the African plate. This
is based on paleomagnetic data and the timing of rift-related volcanism
(e.g., the Deccan Traps’ precursors).
- Collision
with Asia:
India drifted northward and collided with the Eurasian plate, forming
the Himalayas. This collision began around 50–55 Ma, with significant
uplift starting by 40 Ma, though the exact timing varies slightly
depending on the study. The drift slowed after contact but continues
today at a reduced rate.
- Total
Drift Time:
From breakup (140
Ma) to initial collision (50
Ma), India’s main journey took approximately 90 million years (140 Ma -
50 Ma = 90 Ma). This is a rough average, as the drift rate wasn’t
constant—India moved faster (up to 15–20 cm/year) during its mid-ocean
phase and slowed as it approached Asia.
So,
let’s use 90
million years
as the estimated time for India to drift from Africa to its current
position.
Part
2: Extra Core Rotations
Now,
using the core super-rotation rates from before (0.3 to 1 degree/year),
we’ll calculate how many full revolutions (360 degrees) the core would
make ahead of the surface over 90 million years.
- Years
per Rotation:
- Minimum
rate: 360 ÷ 0.3 = 1,200 years per rotation.
- Maximum
rate: 360 ÷ 1 = 360 years per rotation.
- Number
of Rotations in 90 Million Years:
- Total
time = 90,000,000 years.
- Minimum
rate: 90,000,000 ÷ 1,200 = 75,000 rotations.
- Maximum
rate: 90,000,000 ÷ 360 = 250,000 rotations.
- Average
Number of Rotations:
- The
super-rotation rate ranges from 0.3 to 1 degree/year, so let’s average
it for a midpoint estimate: (0.3 + 1) ÷ 2 = 0.65 degrees/year.
- Years
per rotation at 0.65 degrees/year: 360 ÷ 0.65 ≈ 553.85 years.
- Rotations
in 90 million years: 90,000,000 ÷ 553.85 ≈ 162,500 rotations.
Final
Answer
- Time
for India’s Drift:
Approximately 90 million years.
- Extra
Core Rotations:
- Minimum
(0.3 degrees/year): 75,000 rotations.
- Maximum
(1 degree/year): 250,000 rotations.
- Average
(0.65 degrees/year): ~162,500 rotations.
Context
and Caveats
Over
90 million years, the core would have spun ahead of the surface by tens
to hundreds of thousands of rotations, depending on the rate. However:
- The
super-rotation rate isn’t constant—seismic studies suggest it
fluctuates or even pauses, so these numbers are a theoretical maximum2)
assuming steady motion.
- This
doesn’t imply the core’s rotation directly drove India’s drift. Plate
tectonics (mantle convection) is the accepted driver, moving India at
5–20 cm/year, while the core’s super-rotation is a small, indirect
effect on mantle dynamics.
Does
this align with what you were curious about? Want to explore how this
fits (or doesn’t) with the Substack hypothesis?
ME:
You
estimate the core super-rotation rate to be roughly 0.65 degrees/year
on average.
By comparison, If India is moving at a rate of 5–20 cm/year, what would
that be in degrees/year?3)
GrokxAI:
To
compare India’s drift rate (5–20 cm/year) to the core’s super-rotation
rate (0.65 degrees/year), we need to express India’s movement in
degrees per year. This requires converting linear motion across the
Earth’s surface into an angular rate, which depends on the Earth’s
circumference and geometry. Let’s calculate it step by step.
Assumptions
and Setup
- Earth’s
Circumference:
The Earth is a sphere (roughly), with a circumference at the equator of
about 40,075 km (based on a radius of ~6,378 km). This is 360 degrees
of rotation.
- India’s
Drift:
India’s movement of 5–20 cm/year is along the Earth’s surface,
primarily northward from near Africa to Asia. We’ll treat this as
motion along a great circle (e.g., a meridian), where 1 degree of
latitude corresponds to a fixed distance.
- Distance
per Degree:
Circumference ÷ 360 degrees = 40,075 km ÷ 360 ≈ 111.32 km per degree
(or 11,132,000 cm per degree).
Calculation
- Convert
India’s Drift to Degrees per Year:
- Minimum
rate: 5 cm/year.
- Maximum
rate: 20 cm/year.
- Degrees
per year = Distance per year ÷ Distance per degree.
- Minimum
Rate (5 cm/year):
- Degrees/year
= 5 cm ÷ 11,132,000 cm/degree ≈ 0.000000449 degrees/year (or 4.49 ×
10⁻⁷ degrees/year).
- Maximum
Rate (20 cm/year):
- Degrees/year
= 20 cm ÷ 11,132,000 cm/degree ≈ 0.000001796 degrees/year (or 1.796 ×
10⁻⁶ degrees/year).
- Range:
- India’s
drift rate is approximately 4.49 × 10⁻⁷ to 1.796 × 10⁻⁶ degrees/year.
Comparison
to Core Super-Rotation
- Core
Super-Rotation:
0.65 degrees/year (average estimate).
- India’s
Drift:
0.000000449 to 0.000001796 degrees/year.
To
make this intuitive:
- The
core’s angular rate (0.65 degrees/year) is about 360,000
to 1,400,000 times faster
than India’s angular drift rate:
- Minimum:
0.65 ÷ 0.000001796 ≈ 361,915.
- Maximum:
0.65 ÷ 0.000000449 ≈ 1,447,661.
Context
- Core
vs. Surface:
The core’s super-rotation is an internal angular velocity
difference—how much faster it spins relative to the surface’s
360-degree daily rotation (15 degrees/hour). It’s a rotational offset,
not a surface displacement.
- India’s
Motion:
India’s drift is a linear displacement across the Earth’s surface,
converted to an angular rate for comparison. It’s tiny in angular terms
because it’s spread over the planet’s vast circumference.
What
This Means
India’s
drift of 5–20 cm/year translates to an incredibly small angular rate
(10⁻⁶ to 10⁻⁷ degrees/year) because it’s a physical movement over a
huge spherical surface, not a rotation of the whole body. The core’s
0.65 degrees/year, while also small, is a rotational speed difference,
making it orders of magnitude larger in angular terms. This highlights
why the core’s super-rotation isn’t a direct driver of continental
drift—its effect is rotational and internal, while drift is a surface
translation driven by mantle convection.
Does
this clarify the comparison? Want to dig into how these scales relate
to the Substack idea further?
ME:
What's
the radius and diameter of the earth's core as compared to the radius
and diameter of the earth as a whole, and how much faster would the
surface of the core be traveling than the surface of the planet at
0.65 degrees/year, and how does the speed of India across the earth's
surface compare to this?
GrokxAI:
Let’s
tackle this step-by-step: first, we’ll compare the Earth’s core (inner
and outer combined) to the Earth as a whole in terms of radius and
diameter. Then, we’ll calculate the tangential speed difference at the
core’s surface due to its super-rotation of 0.65 degrees/year relative
to the Earth’s surface. Finally, we’ll compare that to India’s drift
speed across the Earth’s surface (5–20 cm/year).
Part
1: Radius and Diameter Comparisons
Earth
as a Whole
- Radius:
Approximately 6,378 km (equatorial radius, slightly more than the polar
radius of 6,357 km due to Earth’s oblate shape, but we’ll use the
average).
- Diameter:
2 × 6,378 km = 12,756 km.
Earth’s
Core (Inner + Outer)
The
Earth’s core has two parts: the solid inner core and the liquid outer
core. For this, I’ll assume you mean the entire core (outer core
boundary), as super-rotation is typically measured for the inner core
but influences the outer core’s dynamics too.
- Inner
Core:
- Radius:
~1,220 km.
- Diameter:
~2,440 km.
- Outer
Core
(extends from inner core boundary to core-mantle boundary):
- Radius:
~3,480 km (total distance from Earth’s center to the outer edge of the
core).
- Diameter:
~6,960 km.
- Source:
These values come from seismic wave studies (e.g., PREM model).
Comparison
- Core
Radius vs. Earth Radius:
3,480 km (outer core) ÷ 6,378 km (Earth) ≈ 0.546, or ~54.6% of Earth’s
radius.
- Core
Diameter vs. Earth Diameter:
6,960 km ÷ 12,756 km ≈ 0.546, or ~54.6% of Earth’s diameter.
- Inner
Core vs. Earth:
The inner core’s radius (1,220 km) is only ~19.1% of Earth’s radius,
but since super-rotation effects propagate through the outer core,
we’ll use the outer core’s size for speed calculations.
Part
2: Speed of the Core’s Surface Due to Super-Rotation
The
core’s super-rotation of 0.65 degrees/year means its surface moves
faster than the Earth’s surface, which rotates at 360 degrees per day
(15 degrees/hour). We’ll calculate the tangential speed difference at
the outer core’s surface.
Earth’s
Surface Speed (Baseline)
- Circumference
at Surface:
2π × 6,378 km ≈ 40,075 km.
- Rotational
Period:
1 day = 86,400 seconds.
- Speed:
40,075 km ÷ 86,400 s ≈ 464 m/s (or 1,670 km/h).
Core’s
Surface Speed (Including Super-Rotation)
- Core
Radius:
3,480 km (outer core boundary).
- Circumference
at Core:
2π × 3,480 km ≈ 21,863 km.
- Base
Speed (Earth’s Rotation):
21,863 km ÷ 86,400 s ≈ 253 m/s (slower than the surface due to smaller
radius).
- Super-Rotation
Addition:
0.65 degrees/year faster than the surface’s 360 degrees/day.
- Earth’s
angular speed: 360 degrees ÷ 86,400 s ≈ 0.004167 degrees/s.
- Super-rotation:
0.65 degrees/year ÷ (365.25 × 86,400 s/year) ≈ 2.06 × 10⁻⁸ degrees/s.
- Extra
tangential speed = Angular speed × Radius = (2.06 × 10⁻⁸ rad/s) ×
3,480,000 m ≈ 0.0717 m/s (since 1 degree ≈ 0.01745 radians).
Speed
Difference
- The
core’s surface moves ~0.0717 m/s (7.17 cm/s) faster than its base speed
due to super-rotation.
- In
a year: 0.0717 m/s × 31,536,000 s/year ≈ 2,260 m/year (2.26 km/year).
Part
3: India’s Speed Across Earth’s Surface
- Range:
5–20 cm/year (0.05–0.2 m/year).
- Comparison
to Core’s Extra Speed:
- Core’s
extra speed: 2,260 m/year.
- India’s
speed: 0.05–0.2 m/year.
- Ratio:
2,260 ÷ 0.2 ≈ 11,300 (max India) to 2,260 ÷ 0.05 ≈ 45,200 (min India).
Summary
- Radius/Diameter:
- Earth:
Radius 6,378 km, Diameter 12,756 km.
- Core
(outer): Radius 3,480 km (54.6%
of Earth’s), Diameter 6,960 km (54.6%
of Earth’s).
- Core
Surface Speed Difference:
The outer core’s surface moves ~7.17 cm/s (2.26 km/year) faster than
the Earth’s rotation at that depth due to 0.65 degrees/year
super-rotation.
- India’s
Speed:
5–20 cm/year (0.05–0.2 m/year).
- Comparison:
The core’s extra tangential speed is ~11,300 to 45,200 times faster
than India’s drift, depending on whether we use 5 or 20 cm/year.
Insight
The
core’s super-rotation creates a significant speed difference at its
surface (kilometers per year), but India’s drift is a slow surface
displacement (centimeters per year). This vast gap reinforces why
super-rotation isn’t seen as a direct driver of continental drift—its
effect is rotational and internal, while India’s motion is a tectonic
process tied to mantle convection, not core spin.
Want
to explore any part of this further—like implications for the Substack
hypothesis?
ME:
Let's
assume for the sake of argument:
- The
Earth's core rotates ahead of the rest at a rate of (averaged) + 0.65
degrees/year
- This
superrotation is constant and has been going on for hundreds of
millions of years or indefinite time. Right?
GrokxAI:
Right,
let’s roll with those assumptions for the sake of argument:
- Core
Super-Rotation Rate:
The Earth’s core (specifically the inner core, with effects propagating
through the outer core) rotates ahead of the rest of the planet at an
average rate of 0.65 degrees per year.
- Duration:
This super-rotation is constant and has been ongoing for hundreds of
millions of years—or even indefinitely, stretching back through
geological time.
These
are reasonable starting points for a thought experiment. The 0.65
degrees/year is a midpoint from the observed range (0.3 to 1
degree/year) based on seismic studies, and assuming it’s constant over
vast timescales simplifies the math, even though real-world data
suggests fluctuations. Over hundreds of millions of years, this would
mean the core has lapped the surface many times—e.g., over 90 million
years (India’s drift time), we calculated ~162,500 extra rotations.
So,
we’ve got a steady 0.65 degrees/year super-rotation running
indefinitely. What do you want to do with this setup? Explore its
effects, tie it back to continental drift, or something else?
ME:
Here's
the question:
Furthermore assuming, if it were possible, that this 0.65
degrees/year extra rotation of the core would be transmitted rigidly,
without slip, to a spot on the surface of the earth in the equatorial
plane, how far would that spot move ahead of the rest of the planet in
one year? In other words: How far ahead of the rest of the planet would
a spot on the equator travel in a year if it were moving ahead of the
rest of the planet by 0.65 degrees/year? How much distance is + 0.65
degrees on the surface of planet Earth?
GrokxAI:
Great
question! Let’s calculate how far a spot on the Earth’s equator would
move ahead of the rest of the planet in one year if it were rigidly
tied to the core’s super-rotation of 0.65 degrees/year, assuming no
slip. This is about converting that angular difference into a linear
distance along the Earth’s equatorial circumference.
Step-by-Step
Calculation
- Earth’s
Equatorial Circumference:
- Earth’s
equatorial radius is approximately 6,378 km.
- Circumference
= 2π × radius = 2 × 3.14159 × 6,378 km ≈ 40,075 km.
- This
is the distance around the Earth at the equator, corresponding to 360
degrees.
- Distance
per Degree:
- Distance
per degree = Circumference ÷ 360 = 40,075 km ÷ 360 ≈ 111.32 km/degree.
- Distance
for 0.65 Degrees:
- Extra
rotation = 0.65 degrees/year.
- Distance
= 0.65 × 111.32 km/degree ≈ 72.36 km.
Result
If
a spot on the equator were rigidly tied to the core’s super-rotation of
0.65 degrees/year, it would move approximately
72.36 kilometers
ahead of the rest of the planet’s surface in one year.
Reality
Check
- Context:
Normally, the Earth’s surface rotates at 360 degrees per day (40,075
km/day ≈ 464 m/s), but this spot would creep ahead by an extra 0.65
degrees/year, or ~72.36 km/year, relative to that baseline. That’s
about 19.8 meters per day (72,360 m ÷ 365.25), a slow but steady drift.
- Assumption
Note:
In reality, the core’s rotation doesn’t transmit rigidly to the
surface—viscous mantle and lithosphere decouple the motion. This is a
hypothetical “what if” where we imagine a direct, no-slip connection.
So,
that spot would travel ~72.36 km ahead annually. Does this spark any
follow-up ideas—like comparing it to tectonic speeds or tying it to the
Substack drift idea?
ME:
Now,
assuming that such a constant superrotation of the earth's core by 0.65
degrees/year, going on for over 500 million years, would have
transmitted itself mechanically onto the surface of the planet via the
liquid viscose material between the Earth's core and it's surface over
that time (and assuming an averaged real physical viscosity for this
medium), how far would a spot in the equatorial plane have traveled
ahead of the rest of the Earth's surface 1) over the time of 500
million years and 2) by each year?
Use
a simple mechanical model.
Would it be right to say that the effect would be most marked at the
equator but nearly nil at the poles?
Give a rough but calculable estimate.
GrokxAI:
This
is a fascinating thought experiment! We’re imagining that the Earth’s
core, super-rotating at a constant 0.65 degrees/year for 500 million
years, mechanically transmits some of its motion to the surface through
the viscous material (mantle and outer core) between the core and
crust. We’ll use a simple mechanical model to estimate how far a spot
on the equatorial plane might move ahead of the rest of the Earth’s
surface, both over 500 million years and per year, and address whether
the effect would be most pronounced at the equator and nearly nil at
the poles. Let’s dive in with a rough but calculable approach.
Assumptions
and Simple Mechanical Model
- Core
Super-Rotation:
0.65 degrees/year, constant for 500 million years.
- Transmission:
The core’s extra rotation is transmitted through the viscous outer core
(liquid iron) and mantle (silicate rock, semi-solid) to the surface. In
reality, viscosity decouples much of this motion, but we’ll assume some
fraction transfers mechanically.
- Viscosity:
- Outer
core: ~10⁻² Pa·s (like water, low viscosity).
- Mantle:
10²⁰
to 10²¹ Pa·s (extremely viscous, like thick tar). We’ll average the
mantle’s viscosity at 10²⁰.5 Pa·s (3.16
× 10²⁰ Pa·s) for simplicity.
- The
outer core’s low viscosity might couple to the mantle, but the mantle’s
high viscosity resists rapid transfer. We’ll model this as a drag
system.
- Simple
Model:
Treat this as a rotational shear problem. The core spins faster,
dragging the mantle, which drags the crust. We’ll assume a linear
velocity gradient across the distance from the core-mantle boundary
(CMB) to the surface, moderated by viscosity. In reality, it’s
nonlinear, but this simplifies things.
- Earth
Dimensions:
- Core
radius (outer core): 3,480 km.
- Earth
radius: 6,378 km.
- Distance
from CMB to surface: 6,378 - 3,480 = 2,898 km.
- Equatorial
circumference: ~40,075 km (360 degrees).
Step
1: Core’s Extra Speed at Its Surface
From
earlier:
- Core
super-rotation: 0.65 degrees/year.
- Core
circumference: 2π × 3,480 km ≈ 21,863 km.
- Extra
distance per year at core surface: (0.65 ÷ 360) × 21,863 km ≈ 39.47
km/year.
- Tangential
speed: 39.47 km/year ÷ (365.25 × 86,400 s/year) ≈ 1.25 × 10⁻⁶ m/s.
Step
2: Transmission Through Viscous Medium
In
a viscous medium, shear stress (τ) = viscosity (η) × velocity gradient
(dv/dr):
- Velocity
at CMB = 1.25 × 10⁻⁶ m/s (core’s extra speed).
- Velocity
at surface = ? (what we’re solving for).
- Distance
(r) = 2,898 km = 2.898 × 10⁶ m.
- Velocity
gradient ≈ (v_surface - v_core) ÷ r.
The
mantle’s high viscosity resists motion, so only a tiny fraction of the
core’s speed reaches the surface. In a simple drag model, the surface
velocity depends on the balance of shear forces and viscous damping.
For a rough estimate:
- Assume
the outer core transmits nearly all its motion to the mantle base (low
viscosity), but the mantle’s viscosity sharply reduces the effect
upward.
- Effective
coupling: Studies of core-mantle coupling (e.g., via magnetic torque)
suggest the mantle might inherit ~10⁻⁴ to 10⁻⁶ of the core’s
differential rotation over long timescales due to viscous and magnetic
drag.
Let’s
use a coupling factor of 10⁻⁵ (a mid-range guess):
- Core
angular speed: 0.65 degrees/year ÷ (365.25 × 86,400 s/year) ≈ 2.06 ×
10⁻⁸ rad/s.
- Surface
angular speed: 2.06 × 10⁻⁸ × 10⁻⁵ = 2.06 × 10⁻¹³ rad/s.
- Surface
tangential speed = Angular speed × Earth radius = 2.06 × 10⁻¹³ rad/s ×
6,378,000 m ≈ 1.31 × 10⁻⁶ m/s.
- Distance
per year: 1.31 × 10⁻⁶ m/s × 31,536,000 s/year ≈ 0.0413 m/year (4.13
cm/year).
Step
3: Distance Over 500 Million Years
- Per
year: ~4.13 cm/year.
- Over
500 million years: 4.13 cm/year × 500,000,000 years ≈ 20,650,000 m =
20,650 km.
Step
4: Latitude Variation
- Equator:
The effect is maximized because tangential speed = angular speed ×
radius, and the radius (distance from the rotation axis) is largest at
the equator (6,378 km).
- Poles:
The radius from the rotation axis shrinks to 0, so tangential speed →
0. The effect is nearly nil at the poles, as motion is perpendicular to
the surface there, not along it.
- Conclusion:
Yes, the effect would be most marked at the equator and nearly nil at
the poles, scaling with the cosine of latitude.
Results
- Distance
per Year:
A spot on the equator moves ~4.13 cm/year ahead of the rest of the
surface.
- Distance
over 500 Million Years:
~20,650 km (about half the Earth’s circumference).
- Latitude
Effect:
Strongest at the equator, negligible at the poles.
Comparison
to India’s Drift
- India’s
speed: 5–20 cm/year.
- Our
spot’s speed: 4.13 cm/year—similar in magnitude, suggesting this could
theoretically contribute to tectonic motion if sustained.
Caveats
- Viscosity:
The mantle’s 10²⁰ Pa·s viscosity likely reduces the effect far more
than 10⁻⁵ coupling—real estimates might drop this to 10⁻¹⁰ or less,
making the surface speed tiny (e.g., micrometers/year).
- Model:
This is oversimplified—real coupling involves magnetic fields,
convection, and nonlinear drag, not just viscous shear.
- Geological
Fit:
20,650 km over 500 Ma is significant, but plate tectonics already
explains continental drift without needing core super-rotation as a
driver.
So,
roughly, that equatorial spot moves ~4.13 cm/year and ~20,650 km over
500 Ma, with the effect fading toward the poles.
Want to tweak the
coupling factor or explore how this aligns with the Substack idea?
ME:
Not
at the moment, thanks.
GrokxAI:
No
problem at all! I enjoyed working through that with you. If you want to
revisit this or try a new angle later, just let me know. Take care!
My comments:
1) Quite the contrary
2) Not quite; as GrokxAI states earlier, the 'move ahead'
speed of the core could be slower than measured - but it could also be
faster
3)
Grok somewhere along the line thankfully overlooks the fact that
while the Earth’s core’s superrotation might drive continents or
islands along the Equator from west to east, India moves northwards;
however, as this conversation is solely about possible distances and
velocity, that does not invalidate the calculations