Skip to content

Plate Boundaries, Kinematics and Seismicity

CSIR NET Geophysics Part C: Euler vectors, plate circuits, triple-junction stability, earthquakes at ridges, transforms, subduction zones, collisions.

  • 20 min read
  • Updated
On this page · 13 sections
  1. Prerequisites
  2. The Hook: A Thousand Kilometres of Fault in Ten Minutes
  3. Plate Kinematics on a Sphere
  4. Triple Junctions
  5. Seismicity at Plate Boundaries
  6. India’s Plate Boundaries and Their Seismicity
  7. Memory Tricks
  8. Don’t Confuse
  9. Common Mistakes
  10. Practice
  11. Quick Revision Table
  12. Key Takeaways
  13. What’s Next

Prerequisites

The Hook: A Thousand Kilometres of Fault in Ten Minutes

Plate Kinematics on a Sphere

Euler vectors

Euler’s fixed-point theorem says any displacement of a rigid cap on a sphere is a rotation about an axis through the centre. For present-day (instantaneous) motion of plate B relative to plate A we use an angular velocity vector \( {}_A\boldsymbol{\omega}_B \), the Euler vector: its direction gives the Euler pole, its length the rotation rate.

The velocity of a point at position \( \mathbf{r} \) on plate B, relative to plate A, is

$$\boxed{\mathbf{v} = \boldsymbol{\omega}\times\mathbf{r}} \qquad \boxed{|\mathbf{v}| = \omega R\sin\Delta}$$

where \( \Delta \) is the angular distance from the Euler pole and \( R = 6{,}371 \) km. Remember \( \omega \) must be in radians per unit time.

Angular distance between a pole \( (\lambda_p, \phi_p) \) and a point \( (\lambda, \phi) \) (latitude, longitude):

$$\boxed{\cos\Delta = \sin\lambda_p\sin\lambda + \cos\lambda_p\cos\lambda\cos(\phi - \phi_p)}$$

Direction. The velocity is tangent to the small circle about the Euler pole through the point, perpendicular to the great circle joining the point to the pole. So:

  • transform faults lie along small circles about the pole;
  • spreading direction is along small circles, and the ridge axis, where it is orthogonal, lies along great circles through the pole;
  • earthquake slip vectors at a boundary point along the local small circle.

Finding an Euler pole from data

Three kinds of observation are inverted together (global models such as NUVEL-1 and MORVEL were built this way):

DataConstrainsGeometry used
Transform-fault azimuthspole positiongreat circles drawn perpendicular to transforms meet at the pole
Spreading rates from magnetic anomaliespole position and raterate varies as \( \sin\Delta \) along the ridge: smallest near the pole
Earthquake slip vectorsdirectionslip vectors lie along small circles

Space geodesy (GPS/GNSS, VLBI, satellite laser ranging) gives present-day motions over years to decades; the anomaly-based rates average about the last few million years. Their close agreement shows that plate motion is steady on geological time scales.

Plate circuits

Euler vectors add like vectors:

$$\boxed{{}_A\boldsymbol{\omega}_C = {}_A\boldsymbol{\omega}_B + {}_B\boldsymbol{\omega}_C} \qquad {}_A\boldsymbol{\omega}_B = -\,{}_B\boldsymbol{\omega}_A$$

Going round a closed circuit of plates the sum is zero. This is how motion across a boundary that has no spreading record (for example a subduction zone) is found: from the motions across ridges elsewhere in the circuit.

One boundary, several types

Because velocity is fixed by the Euler vector and the boundary can wander in direction, a single boundary can change character along its length: divergent where it is perpendicular to the relative motion, transform where it is parallel to the motion, convergent where the motion has a component towards it. The rate also changes as \( \sin\Delta \): a ridge near its Euler pole spreads slowly; the same ridge 90° away spreads at its maximum.

Triple Junctions

A triple junction is where three plates meet. Boundaries are coded R (ridge), T (trench; the overriding plate is shown) and F (transform fault). McKenzie and Morgan (1969) asked whether a junction keeps its geometry as the plates move.

The velocity-triangle test

  1. Draw the velocity triangle: the three relative velocities \( {}_A\mathbf{v}_B \), \( {}_B\mathbf{v}_C \), \( {}_C\mathbf{v}_A \) form a closed triangle in velocity space because they sum to zero.
  2. For each boundary, draw the locus of velocities with which a point can move while staying on that boundary:
    • ridge (symmetric spreading): the perpendicular bisector of the relative-velocity side;
    • trench: a line through the overriding plate’s velocity point, parallel to the trench;
    • transform: the line through the side itself (parallel to the fault).
  3. The junction is stable if the three loci meet at a single point. That point is the junction’s own velocity.
JunctionStability
RRRalways stable (the perpendicular bisectors of a triangle always meet, at the circumcentre)
FFFnever stable
TTT, RTF, FFT, RTT, FFR and othersstable only for particular orientations or velocities

Unstable junctions do not last: they change into a stable configuration or migrate, changing the boundaries around them. When the Pacific–Farallon ridge met the North American trench, the junctions that formed migrated along the margin, and the transform between them grew into the San Andreas system.

Seismicity at Plate Boundaries

Global seismicity outlines the plates. Most earthquakes, and most seismic moment, are released at plate boundaries; the circum-Pacific belt dominates the total, followed by the Alpine–Himalayan belt, with ridges and transforms contributing many smaller events.

Boundary by boundary

SettingDepthMechanismStress axesSize
Mid-ocean ridgeshallow (a few km to about 10 km below the sea floor)normal faulting on ridge-parallel planesT axis horizontal, perpendicular to the ridgemoderate; larger at slow ridges (thicker brittle layer); swarms with dyking
Oceanic transformshallow, deepening with the age (temperature) of the lithospherestrike-slip; one nodal plane parallel to the transformP and T horizontal, at about 45° to the faultmoderate to large; much slip is aseismic
Outer rise (seaward of trench)upper part of the bending platenormal (bending), sometimes thrust deeper in the platehorizontal extension at the topoccasionally large
Subduction interface (megathrust)from near the trench to about 40–50 kmlow-angle thrust, plane dipping landwardP axis roughly along convergencelargest earthquakes on Earth (Chile 1960, Mw about 9.5; Sumatra–Andaman 2004; Tohoku 2011)
Intraslab, intermediate depthabout 70–300 kmvariable; down-dip extension or compressionaxes along the slab dipup to about Mw 8
Intraslab, deep300 to about 700 kmmostly down-dip compressionP along the slab dipup to about Mw 8
Continental collisionshallow, spread over a wide zonethrusting at the front, normal faulting on the high plateau, strike-slip faultsvaries across the beltgreat thrust earthquakes at the front

Subduction zones in detail

Wadati–Benioff zone. Hypocentres define a dipping zone, the seismically active cold core of the subducting slab, reaching about 700 km depth at most. Its dip \( \delta \) from a cross-section:

$$\boxed{\tan\delta = \frac{\Delta(\text{depth})}{\Delta(\text{horizontal distance})}}$$

Arc volcanoes typically stand where the top of the slab is about 100 km deep (with a range from slab to slab), where fluids released from the slab lower the melting point of the mantle wedge. For a planar slab of dip \( \delta \), that places the arc at a horizontal distance of about \( 100/\tan\delta \) km from the point where the slab top would reach the surface, so steeper slabs give narrower arc–trench gaps.

Stress in the slab (Isacks and Molnar). Where a slab is short and hangs freely, its intermediate-depth earthquakes show down-dip extension: the slab is being pulled by its own weight. Where a slab reaches the resistant transition zone or lower mantle, the deeper part shows down-dip compression, and the stress can switch with depth. Many slabs show a gap in seismicity at intermediate depths and resume at greater depth.

Double seismic zones. In some slabs (northeast Japan is the classic case) intermediate-depth earthquakes form two parallel planes some tens of kilometres apart, with different mechanisms in the upper and lower planes, consistent with unbending of the slab.

Why earthquakes at such depth? At high pressure ordinary frictional failure should be suppressed. Proposed mechanisms:

Depth rangeLeading hypotheses
Intermediate (about 70–300 km)dehydration embrittlement: water released from hydrous minerals raises pore pressure and allows faulting
Deep (about 300–700 km)transformational faulting in metastable olivine persisting in the cold core of the slab; also thermal runaway shear instabilities

Thermal parameter. How deep a slab stays cold enough to break depends on how fast cold material is fed downward:

$$\boxed{\Phi = t\,v\sin\delta}$$

with \( t \) the age of the subducting plate (Myr), \( v \) the convergence rate (km Myr\(^{-1}\)) and \( \delta \) the dip; \( \Phi \) comes out in km. Old, fast, steep slabs have large \( \Phi \): they are cold to great depth and tend to have the deepest earthquakes (Tonga is the standard example). Young, slow slabs (for example beneath parts of the Cascades) have small \( \Phi \) and shallow maximum seismic depths.

Oblique subduction and slip partitioning. When convergence is oblique, the motion is often split: near-trench-normal thrusting on the megathrust and margin-parallel strike-slip on a fault in the overriding plate. Sumatra, with the Great Sumatran Fault behind the trench, is the textbook case.

Seismic coupling and the moment budget

A locked fault segment of length \( L \) and down-dip width \( W \) accumulates a slip deficit at the plate rate \( v \). The seismic moment that could be released after time \( t \) is

$$\boxed{M_0 = \mu L W (c\,v\,t)} \qquad \boxed{M_w = \tfrac{2}{3}\left(\log_{10}M_0 - 9.1\right)}$$

with \( \mu \) the rigidity (about \( 3\times 10^{10} \) Pa), \( M_0 \) in N m, and \( c \) the seismic coupling coefficient: the fraction of plate motion released in earthquakes (0 for fully creeping, 1 for fully locked and released seismically). Oceanic transforms and some subduction zones have low \( c \); others are strongly coupled.

Ridge and transform earthquakes and temperature

Earthquakes in oceanic lithosphere stop below a limiting isotherm, so the maximum focal depth deepens with lithospheric age. At ridges and young transforms only the uppermost few kilometres are brittle; old oceanic lithosphere can have earthquakes many tens of kilometres deep. Slow ridges have colder, thicker brittle layers than fast ridges, and so larger normal-faulting earthquakes.

India’s Plate Boundaries and Their Seismicity

BoundaryTypeSeismicity
Himalayan front (Main Himalayan Thrust and its surface splays)continent–continent convergencegreat earthquakes of the Himalayan belt: 1905 Kangra, 1934 Bihar–Nepal and 2015 Gorkha (Nepal) on the Himalayan thrust system; 1897 Shillong (Shillong Plateau) and 1950 Assam in the eastern Himalaya and Assam region
Tibet and the Himalayan interiorcollision, plateaunormal faulting on north–south grabens (east–west extension), strike-slip faults
Andaman–Sumatra arcoceanic subduction, obliquemegathrust events (2004), back-arc spreading in the Andaman Sea, strike-slip faults
Western margin (Chaman fault system, Owen Fracture Zone)transformstrike-slip; Owen Fracture Zone separates the Indian and Arabian plates
Carlsberg and Central Indian ridgesdivergentshallow normal-faulting events
Central Indian Oceandiffuse boundary between the Indian and Australian platesbroad zone of earthquakes and folded, faulted oceanic lithosphere: plates are not perfectly rigid
Peninsular Indiaintraplatemoderate events on old weak zones and reactivated rifts (Koyna 1967, reservoir-associated; Latur 1993; Jabalpur 1997; Bhuj 2001)

The 1950 Assam earthquake (Mw about 8.6) was among the largest instrumentally recorded continental earthquakes. The Himalayan arc is also known for “seismic gaps”: segments with no great earthquake in the historical record, where strain is expected to have accumulated.

graph LR
    A[Euler vector and plate circuit] --> B[Relative velocity at each boundary]
    B --> C{Boundary orientation}
    C -->|perpendicular, apart| D[Ridge: shallow normal faults]
    C -->|parallel| E[Transform: strike-slip]
    C -->|towards, oceanic| F[Subduction: outer rise, megathrust, slab]
    C -->|towards, continental| G[Collision: wide thrust belt]

Memory Tricks

Don’t Confuse

ThisvsThat
Euler vector (instantaneous angular velocity; adds as a vector)Finite rotation (reconstruction; order matters)
Small circle about the Euler pole (transforms, flow lines)Great circle through the pole (orthogonal ridge segments)
Megathrust earthquake (interplate, low-angle thrust)Intraslab earthquake (within the slab, deeper, varied mechanisms)
Outer-rise normal faulting (bending, seaward of the trench)Ridge normal faulting (plate separation)
Down-dip extension (freely hanging slab)Down-dip compression (slab meeting resistance)
Seismic gap (segment that has not ruptured recently)Shadow zone (seismic-wave term, nothing to do with hazard)
Diffuse plate boundary (broad deforming zone)Intraplate seismicity (isolated events within a plate interior)

Common Mistakes

Practice

All questions below are original practice questions written for this page. They are not from any CSIR NET paper.

Level 1: Foundation

  1. Along which kind of curve about the Euler pole do transform faults lie?
  2. What focal mechanism is typical of earthquakes on a mid-ocean ridge axis?
  3. Which triple junction is stable for all orientations of its boundaries?
Answers
  1. Small circles about the Euler pole.
  2. Normal faulting on ridge-parallel planes, with the T axis horizontal and perpendicular to the ridge.
  3. RRR (ridge–ridge–ridge).

Level 2: Part B Style

Practice Q1. Two plates rotate relative to each other at 0.5° Myr\(^{-1}\). The relative speed at a point 60° from the Euler pole is about:

(a) 2.8 cm yr\(^{-1}\) (b) 4.8 cm yr\(^{-1}\) (c) 5.6 cm yr\(^{-1}\) (d) 9.6 cm yr\(^{-1}\)

Answer: (b) 4.8 cm/yr

\( \omega = 0.5\times\pi/180 = 8.727\times 10^{-3} \) rad Myr\(^{-1}\); \( \omega R = 55.6 \) km Myr\(^{-1}\).

\( v = 55.6\times\sin 60° = 55.6\times 0.866 = 48.1 \) km Myr\(^{-1}\) \( \approx 4.8 \) cm yr\(^{-1}\).

(c) is the speed 90° from the pole; (a) uses \( \cos 60° \).

Practice Q2. Earthquakes at a slow-spreading mid-ocean ridge axis are best described as:

(a) deep-focus thrusts with P axes along the spreading direction (b) shallow normal-faulting events with T axes perpendicular to the ridge (c) shallow strike-slip events with nodal planes parallel to the ridge (d) intermediate-depth events with down-dip compression

Answer: (b)

Plates separate across the axis, so the least compressive stress is horizontal and perpendicular to the ridge. (c) describes transforms, and their nodal plane is parallel to the transform, not to the ridge.

Practice Q3. In a cross-section through a subduction zone, earthquakes at 150 km depth lie 260 km horizontally from the trench, and the zone is planar from the trench. The dip of the Wadati–Benioff zone is about:

(a) 22.5° (b) 30° (c) 35° (d) 60°

Answer: (b) 30°

\( \tan\delta = 150/260 = 0.577 \Rightarrow \delta \approx 30° \).

(d) is \( \tan^{-1}(260/150) \), the angle from the vertical.

Practice Q4. Which statement about oceanic transform faults is correct?

(a) Their earthquakes are deep-focus because the faults are long. (b) One nodal plane of their focal mechanisms is parallel to the transform, and slip vectors give the direction of relative plate motion. (c) They lie along great circles passing through the Euler pole. (d) They are seismically fully coupled, releasing all plate motion in earthquakes.

Answer: (b)

(a): transform earthquakes are shallow, limited by temperature. (c): transforms follow small circles about the pole. (d): a large share of transform slip is aseismic.

Level 3: Part C Style

Practice Q5. A megathrust segment 500 km long with a locked down-dip width of 100 km converges at 4 cm yr\(^{-1}\). Take \( \mu = 3\times 10^{10} \) Pa and assume the segment has been fully locked for 200 years. Which statements are correct?

  1. The accumulated slip deficit is 8 m.
  2. The seismic moment available is about \( 1.2\times 10^{22} \) N m.
  3. Released in one event, it corresponds to about Mw 8.7.
  4. If only half the deficit is released seismically, the magnitude falls by about 0.5.

(a) 1 and 2 only (b) 1, 2 and 3 only (c) 2, 3 and 4 only (d) 1, 2, 3 and 4

Answer: (b) 1, 2 and 3 only
  • 1: \( 0.04\times 200 = 8 \) m. True.
  • 2: \( 3\times 10^{10}\times 5\times 10^{5}\times 10^{5}\times 8 = 1.2\times 10^{22} \) N m. True.
  • 3: \( \log_{10}(1.2\times 10^{22}) = 22.08 \); \( M_w = \frac{2}{3}(22.08 - 9.1) = 8.65 \). True.
  • 4: False. Halving \( M_0 \) lowers \( M_w \) by \( \frac{2}{3}\log_{10}2 = 0.20 \), to about 8.45.

Practice Q6. Two subduction zones:

SlabAge of subducting plate (Myr)Convergence rate (cm yr\(^{-1}\))Dip
X130950°
Y20430°

Which statements are correct?

  1. The thermal parameter of slab X is about 9,000 km.
  2. The thermal parameter of slab Y is 400 km.
  3. Slab Y is the more likely of the two to have earthquakes deeper than 300 km.
  4. Slab X’s thermal parameter is more than 20 times that of slab Y.

(a) 1, 2 and 4 only (b) 1 and 3 only (c) 2, 3 and 4 only (d) 1, 2, 3 and 4

Answer: (a) 1, 2 and 4 only

Convert rates: 9 cm yr\(^{-1}\) = 90 km Myr\(^{-1}\); 4 cm yr\(^{-1}\) = 40 km Myr\(^{-1}\).

  • X: \( \Phi = 130\times 90\times\sin 50° = 11{,}700\times 0.766 \approx 8{,}960 \) km. Statement 1 true.
  • Y: \( \Phi = 20\times 40\times\sin 30° = 400 \) km. Statement 2 true.
  • 3: False. The old, fast, steep slab X stays cold to greater depth; X is the candidate for deep-focus earthquakes.
  • 4: \( 8{,}960/400 \approx 22 \). True.

Practice Q7. Along a ridge, the full spreading rate is 3.0 cm yr\(^{-1}\) at a point 30° from the Euler pole of the two plates. Which statements are correct?

  1. The rotation rate is about 0.54° Myr\(^{-1}\).
  2. At a point 60° from the pole the full rate is about 5.2 cm yr\(^{-1}\).
  3. The transform faults offsetting this ridge lie along great circles through the Euler pole.
  4. The greatest possible rate on this boundary is about 6 cm yr\(^{-1}\), 90° from the pole.

(a) 1 and 3 only (b) 1, 2 and 4 only (c) 2, 3 and 4 only (d) 1, 2, 3 and 4

Answer: (b) 1, 2 and 4 only
  • \( \omega R\sin 30° = 30 \) km Myr\(^{-1}\), so \( \omega R = 60 \) km Myr\(^{-1}\) and \( \omega = 60/6371 = 9.42\times 10^{-3} \) rad Myr\(^{-1}\) \( = 0.54° \) Myr\(^{-1}\). Statement 1 true.
  • At 60°: \( 60\times 0.866 = 52 \) km Myr\(^{-1}\) = 5.2 cm yr\(^{-1}\). Statement 2 true.
  • 3: False. Transforms are parallel to the relative motion, which follows small circles about the pole.
  • 4: \( \omega R = 60 \) km Myr\(^{-1}\) = 6 cm yr\(^{-1}\) at \( \Delta = 90° \). True.

Quick Revision Table

IdeaFact or formula
Plate velocity\( \mathbf{v} = \boldsymbol{\omega}\times\mathbf{r} \); \( v = \omega R\sin\Delta \) (\( \omega \) in rad)
Angular distance\( \cos\Delta = \sin\lambda_p\sin\lambda + \cos\lambda_p\cos\lambda\cos(\phi - \phi_p) \)
Directionstransforms and slip vectors on small circles; perpendiculars to transforms meet at the pole
Plate circuit\( {}_A\boldsymbol{\omega}_C = {}_A\boldsymbol{\omega}_B + {}_B\boldsymbol{\omega}_C \); finite rotations do not commute
Triple junctionsvelocity triangle; loci meet at one point → stable; RRR always, FFF never
Ridge quakesshallow, normal, T axis ⊥ ridge
Transform quakesshallow, strike-slip, nodal plane ∥ transform, much aseismic slip
Subductionouter rise (normal), megathrust (thrust, largest events), intraslab (to about 700 km)
Slab stresshanging slab: down-dip extension; resisted slab: down-dip compression
Deep quake mechanismsdehydration embrittlement (intermediate); transformational faulting (deep)
Thermal parameter\( \Phi = t\,v\sin\delta \) (km); large \( \Phi \) → deep seismicity
Benioff dip\( \tan\delta = \Delta z/\Delta x \); arc above slab at about 100 km
Moment budget\( M_0 = \mu L W c v t \); \( M_w = \frac{2}{3}(\log_{10}M_0 - 9.1) \)
IndiaHimalaya (1897, 1905, 1934, 1950, 2015), Andaman–Sumatra (2004), Owen FZ, Carlsberg Ridge, diffuse Central Indian Ocean

Key Takeaways

What’s Next