Tectonics studies the movement and deformation of Earth’s lithosphere. It connects the arrangement of plates, faults, folds, mountains and basins with the forces and material behaviour that produced them. The gold standard is an explanation that can travel from a map to a measurement, from a measurement to a mechanism, and from that mechanism to an honest account of uncertainty.
Start with an apparently modest observation: two survey stations slowly change their separation. A few millimetres seem small over a year. Sustained over geological time, related motions can rearrange oceans and continents. The scientific challenge is to connect those timescales without assuming that the present rate has remained unchanged forever.
This subject root brings the deformation system together. Geology provides the wider account of Earth materials and history. Seismology develops earthquake signals and fault processes; volcanology explains volcanic systems; geomorphology follows landforms and surface processes. Tectonics supplies the framework linking plate motion, deformation and their geological record.
The numerical cases below are invented teaching models, not measurements of a named plate or forecasts of an earthquake. Their educational purpose follows The Gold Standard of Education: understand why a method works, then recognise the assumptions that need changing in a new setting.
Find your next question
Begin with the route that fits your purpose. Read sequentially for the full argument, or use the chapter index to return to a worked example, an evidence check or a connected guide.
Full chapter index · Tectonics: mechanism library · How X Works
Chapter index
Read movement and deformation · Chapters 1–4
Follow plate interactions · Chapters 5–8
Test signals and histories · Chapters 9–12
Connect landscapes and learning · Chapters 13–14
Check and continue · Chapters 15–18
CHAPTER 1 OF 18 · Read movement and deformation
1. Read movement as a relationship between places
Motion requires a reference. A plate can move relative to another plate, a chosen reference frame or a point represented in a model. Saying that a plate moves “east at a particular speed” is incomplete unless the frame and relevant location are understood.
Imagine two stations moving east at 20 and 35 millimetres per year in the same reference frame. Their relative eastward velocity is 15 millimetres per year. If a third station moves northward instead, relative motion becomes a vector problem, not a subtraction of speed magnitudes alone.
On a spherical Earth, rigid-plate motion is commonly represented as rotation about an axis. The local velocity depends on position relative to that axis. A single arrow drawn on a continent is a useful simplification, but it does not mean that every point on the plate has an identical direction and speed in a flat-map description.
Before interpreting a map, read the scale, frame, dates and arrow definition. An arrow might describe total displacement over a period or velocity per year. It might include or exclude a regional background motion. These distinctions affect the conclusion even when the plotted symbols look similar.
The first learning checkpoint is therefore relational. Ask which two objects or frames are being compared, what quantity is changing and over what interval. This habit is familiar from transport problems in mathematics, yet it becomes especially important when the objects are continents and the map projection makes the surface seem simpler than it is.
CHAPTER 2 OF 18 · Read movement and deformation
2. Lithosphere and asthenosphere describe mechanical behaviour
The lithosphere includes the crust and the relatively strong uppermost mantle. It is not simply another word for crust. The asthenosphere is a weaker part of the mantle that can deform over geological timescales. Earth’s plates do not float on a global ocean of molten magma.
The US Geological Survey explicitly addresses this common misconception: much of the mantle is solid, while its long-term behaviour permits slow flow. Material can be solid and still deform over a sufficiently long interval under suitable temperature and stress. Everyday categories such as solid and liquid are not enough to describe every relevant timescale.
A classroom analogy might use a material that responds differently to a quick push and a sustained load. The analogy reveals rate dependence, but it does not establish the mantle’s actual composition or mechanism. Keep its explanatory job narrow and return to geological and geophysical evidence.
Plate thickness and strength vary. Oceanic lithosphere generally cools and thickens as it moves away from a spreading centre, while continental lithosphere has a more varied history. A simple uniform slab is a starting representation, not a complete description of every region.
For a changed-context question, ask why a cold rock specimen on a desk does not reproduce mantle flow in a short classroom demonstration. Temperature, pressure, time and scale differ. The correct response is to identify those controlling conditions, rather than conclude that solid mantle flow is impossible because a hand specimen does not visibly bend during a lesson.
CHAPTER 3 OF 18 · Read movement and deformation
3. Stress and strain separate causes from deformation
Stress describes force distributed over area. Strain describes deformation relative to the original geometry. Compression, extension and shear refer to different aspects of deformation or loading, and the relationship between stress and strain depends on material properties and conditions.
For a simple invented bar, length increases from 100 to 101 metres. The extension is one metre, while the longitudinal strain is 1 ÷ 100 = 0.01, or 1%. Strain is dimensionless. If the same one-metre extension occurs across an original kilometre, strain is 0.1%. The absolute movement is the same; its relative deformation differs.
Elastic deformation can be recovered when a load is removed within the relevant range. Brittle failure involves fracturing or sliding on discontinuities. Ductile deformation can accumulate without the same discrete fracture behaviour. Depth, temperature, fluids, strain rate and rock type influence which response is important.
A map of surface strain does not by itself identify a unique fault geometry at depth. Several subsurface arrangements may be compatible with similar surface observations. Additional evidence from geology, seismicity and other methods can narrow the possibilities.
The practical learning task is to label each statement correctly. “The ground shortened by two centimetres” describes a displacement. “The shortening was distributed across ten kilometres” adds the geometry needed for a strain estimate. “A force acted on the region” is still insufficient to calculate stress without area. Tectonic reasoning becomes clearer when these quantities are connected without being treated as interchangeable.
CHAPTER 4 OF 18 · Read movement and deformation
4. Plate boundaries are systems with different relative motions
Divergent boundaries involve relative separation, convergent boundaries involve approach and transform boundaries involve substantial lateral relative motion. Real boundary zones can combine components and distribute deformation over a broad region. The categories organise the first explanation; the actual geometry determines the details.
At a spreading centre, new oceanic crust forms as mantle material rises and melts under appropriate conditions. At a subduction zone, one plate descends beneath another. Continental collision can thicken and deform crust. Transform systems accommodate lateral movement, although local bends and steps can introduce compression or extension.
Do not assign one landform to every boundary of a given class. A continent–continent collision differs from an oceanic subduction setting. A continental rift differs from an established ocean ridge. Shared relative motion does not remove differences in material, geometry and history.
The USGS plate-motion guide supplies a useful institutional overview of these boundary relationships. For a particular region, the next step is evidence: plate velocities, geological mapping, seismic patterns and structures. A coloured boundary line on a global map is not a measurement of every local process along it.
As an independent task, give learners three relative-velocity arrows across an imaginary boundary. Ask them to separate normal and sideways components before naming the setting. Then rotate the boundary while preserving the velocity vector. The components change even though the original arrow does not. This makes oblique convergence intelligible and exposes the limitation of classifying a boundary from speed alone.
CHAPTER 5 OF 18 · Follow plate interactions
5. Divergence connects mantle processes with the age of ocean floor
Ocean ridges are places where plates separate and new oceanic crust is generated. Mantle rising beneath a ridge can partially melt because pressure decreases during ascent. The mechanism is decompression melting, rather than a claim that the entire mantle becomes a freely flowing liquid beneath the ocean.
As newly formed oceanic material moves away and cools, its properties evolve. Seafloor ages, heat-flow patterns and magnetic records can provide related evidence for spreading. Each observation tests a different part of the account, making the argument stronger than a visual comparison of coastline shapes alone.
Imagine a fictional ridge with rocks dated at one million years old 25 kilometres from its axis on each side. If spreading was steady, perpendicular and symmetric, each side’s average half-spreading rate is 25 millimetres per year. The full separation rate between the two sides is 50 millimetres per year. Confusing those definitions doubles or halves the inferred motion.
The rate follows from distance divided by time. Twenty-five kilometres is 25 million millimetres; dividing by one million years gives 25 millimetres per year. State the conversion instead of hiding it inside a calculator result.
Now change the assumptions: one side spreads faster, or the ridge position shifts. The age–distance pattern becomes asymmetric. That observation need not disprove seafloor spreading; it challenges the simplified symmetric model. A good learner can revise the model while preserving the established mechanism supported by the wider evidence.
CHAPTER 6 OF 18 · Follow plate interactions
6. Subduction combines sinking, fluids and deformation
At a subduction zone, a plate descends into the mantle. The descending slab, the overriding plate and the intervening region form a coupled system involving deformation, thermal change, fluids and sometimes magma generation. Treating it as one slab simply melting into a sea of magma loses the important processes.
Water and other volatiles released from the descending material can promote partial melting in the mantle above the slab. The locations of earthquakes, volcanic activity and surface deformation therefore need interpretation within the geometry and thermal structure, rather than being assumed to lie on a single vertical line.
The sinking of cold, relatively dense lithosphere can exert a substantial force on the plate. Other forces and resistances also matter. A cartoon of convection cells dragging passive plates may introduce the connection with mantle circulation, but it is not a complete force balance for every plate.
For a fictional cross-section, draw a descending slab at an angle and mark surface observations at stated distances. Ask what depth the simple geometry predicts at a chosen horizontal position. Then ask which real conditions could make that straight-line estimate unsuitable: curvature, changing dip and uncertainty in the mapped boundary are examples.
The point of the exercise is to translate a representation into a testable consequence, not to design a real hazard assessment. Seismology and volcanology supply the specialist routes for the signals and processes. Tectonics keeps their relationships visible while recognising that the same broad subduction category includes settings with different behaviour.
CHAPTER 7 OF 18 · Follow plate interactions
7. Collision, folding and mountain building require a material budget
Continental collision can shorten and thicken crust, with deformation distributed through faults, folds and deeper flow. Mountain height then reflects interactions among tectonic thickening, buoyancy, erosion and other processes. High topography is not a direct ruler measuring one tectonic variable.
Use a deliberately simple area-conservation exercise. A rectangular cross-section of crust is initially 100 kilometres wide and 30 kilometres thick, giving 3,000 square kilometres of cross-sectional area. If its width shortens to 75 kilometres with no addition or removal of material in the model, average thickness becomes 3,000 ÷ 75 = 40 kilometres.
The ten-kilometre thickening does not mean the surface rises by ten kilometres. The deeper geometry and density relationships affect buoyancy, and erosion can remove material. The model conserves cross-sectional area; it does not directly predict topographic height. That distinction is the central lesson.
Folds reveal deformation of layered material, but their interpretation depends on orientation, continuity and later changes. A single photograph of a fold may not establish the complete regional shortening direction. Structures need mapping in three dimensions and placement within a sequence of events.
Ask students to change the model by removing 300 square kilometres of cross-sectional material through a stated process. The final average thickness becomes 2,700 ÷ 75 = 36 kilometres under the same simplified geometry. The changed result makes erosion and material transfer part of the explanation, instead of decorative additions to a story in which collision alone determines every mountain.
CHAPTER 8 OF 18 · Follow plate interactions
8. Transform motion can create local compression and extension
Transform boundaries accommodate lateral relative motion, but real faults are not perfectly straight lines in a uniform material. Bends, steps and interacting segments can change local deformation. Some geometries favour compression; others favour extension. The regional movement and local response must be distinguished.
A useful paper model represents two blocks sliding past each other along a straight boundary. It demonstrates lateral relative motion. Add a bend and the fit between the blocks changes. The model can suggest why local geometry matters, but its rigid edges and friction do not reproduce every feature of a natural fault zone.
Measure offset features carefully. A displaced stream or rock unit can provide a record of cumulative motion if the feature’s history and correlation are understood. Its present geometry may also reflect erosion, channel migration or several deformation events. Matching two shapes by eye is a hypothesis, not automatically a displacement measurement.
For an invented case, a mapped feature has 300 metres of offset and a constrained age of 10,000 years. The average rate is 0.03 metres per year, or 30 millimetres per year, if the offset accumulated during that full interval. It is an average over the geological marker’s history, not a claim that the fault slipped smoothly every year.
Changed-context practice can introduce age uncertainty or several possible marker correlations. The answer should become a range or a conditional estimate. Precision in the arithmetic cannot remove ambiguity in the evidence used to choose the inputs.
CHAPTER 9 OF 18 · Test signals and histories
9. Earthquakes connect accumulated deformation with sudden motion
Some faults accumulate elastic deformation while portions remain locked, then release part of that stored deformation through rapid slip. Other portions may creep or behave differently. Plate movement, fault slip and earthquake occurrence are related but are not identical quantities.
In a toy locked-fault model, relative motion is 20 millimetres per year and the assumed locked interval is 100 years. Multiplication gives two metres of accumulated relative displacement. Calling that a slip deficit requires the model’s assumptions about locking and accommodation elsewhere. It does not establish that the next event will slip by exactly two metres.
Nor does dividing an assumed event slip by a long-term rate predict a calendar date. Real fault systems have variable histories, interacting segments and incomplete records. A simple recurrence calculation can clarify a scale while remaining unsuitable as a deterministic earthquake forecast.
The USGS distinguishes earthquake prediction from probabilistic hazard assessment. The scientific value of plate tectonics does not depend on promising an exact time, location and magnitude for an impending earthquake. A theory can explain mechanisms and support useful hazard analysis while leaving that specific prediction unavailable.
For readers, the evidence check is to ask what a statement actually reports: a measured velocity, a historical event, a modelled strain accumulation or a probability over a defined interval. These statements should not be substituted for one another. Practical preparedness should follow relevant local authorities and engineered standards, rather than a classroom multiplication presented as a countdown.
CHAPTER 10 OF 18 · Test signals and histories
10. Several independent observations support plate motion
Plate tectonics became convincing through converging evidence, not one attractive diagram. Geological relationships, fossils and past climates, magnetic records, seafloor ages, earthquake distributions and modern geodetic measurements contribute different constraints. Their strength lies partly in testing the account through different methods and timescales.
For example, an age pattern across a ridge asks where and when crust formed. Magnetic patterns ask how cooling rocks recorded past field orientations and were transported. Earthquake locations and depths help reveal active deformation and descending slabs. Geodesy measures present movement. Each has its own uncertainties and interpretation steps.
Agreement does not require every observation to be perfect. Instead, ask whether the proposed mechanism explains several patterns together better than alternatives. If one observation disagrees, investigate calibration, representation and the model’s scope before either ignoring it or abandoning the entire framework.
A good classroom investigation can compare supplied datasets rather than stage a vote about whether the continents look as if they fit. Give students a simplified age map, velocity arrows and a cross-section of earthquake depths. Ask which inferences have support from more than one source and where the evidence remains incomplete.
Keep the distinction between independent evidence and repeated presentation. Three websites copying the same map are not three independent measurements. The same discipline serves research across subjects: trace the claim back to the observation and method. Tectonics teaches how a large explanatory system can be robust while particular local reconstructions remain open to revision.
CHAPTER 11 OF 18 · Test signals and histories
11. Geodetic measurements need reference frames and a time history
Satellite navigation measurements and radar interferometry can reveal small changes in Earth’s surface. Turning those signals into tectonic movement requires appropriate processing and an account of other influences. A surface displacement is an observation to interpret, not a label that automatically identifies its cause.
Groundwater changes, landslides, volcanic activity, surface loading and instrument conditions can affect measurements. Radar interferometry generally measures a component along the instrument’s line of sight. It does not directly provide a full three-dimensional displacement vector from one viewing geometry.
Suppose a fictional radar measurement detects 10 millimetres of line-of-sight movement. A student cannot call it 10 millimetres of vertical uplift without additional geometric assumptions. The physical surface could have moved in a direction with a different vertical and horizontal combination. More viewing directions or other observations can help.
For a station time series, distinguish a gradual trend from seasonal variation and sudden offsets. An earthquake may produce an abrupt displacement; later deformation can follow a different rate from the preceding trend. Fitting one line across all those behaviours may produce an average with little physical meaning.
The practical method is to read metadata before interpreting the curve. Identify coordinates, reference frame, component, sampling interval and processing notes. Then ask which model explains the observed shape and which independent evidence supports that model. This is a professional habit that can be introduced through simple graphs long before a student studies advanced geodesy.
CHAPTER 12 OF 18 · Test signals and histories
12. Rocks preserve a sequence that must be reconstructed
A landscape can contain several generations of deformation. A fault may cut folded layers; a younger deposit may cover an older structure; erosion may remove part of the record. Tectonic history is therefore an inference about event order, rather than a description of the final arrangement alone.
Cross-cutting relationships can constrain relative timing: a feature cutting another is younger than the material or structure it cuts, subject to correct identification. Stratigraphy supplies methods for organising layered records and recognising gaps. Geochronology adds age constraints whose meaning depends on the event actually dated.
Consider a fictional sequence: sedimentary layers form, the layers are folded, a fault cuts the folds and an undeformed deposit covers the faulted surface. The observations support that order if the mapped relationships are reliable. They do not yet provide exact durations between events or establish that every nearby valley shares the same history.
An age measured in a mineral might date crystallisation, cooling or another process, depending on the method and context. It cannot automatically be assigned as the age of the nearest fault. The geological relationship between sample and deformation is part of the dating argument.
Ask learners to separate observations from inferred events in two columns. “The fault ends below this deposit” is an observation. “Faulting predates deposition” is an interpretation with stated assumptions. This small discipline produces stronger histories and prevents a fluent narrative from hiding a missing piece of evidence. The Stratigraphy and Geochronology guides develop these evidence questions in greater depth.
CHAPTER 13 OF 18 · Connect landscapes and learning
13. Tectonics interacts with erosion, basins and resources
Uplift and subsidence change the space available for erosion and deposition. Erosion redistributes mass and can influence the evolving landscape. Sedimentary basins preserve parts of this history, while their geometry and fill reflect more than one process. Tectonics and surface systems therefore need to be studied together.
A basin can deepen as the crust extends or bends under a load. Sediment can partly fill that space. A thick sedimentary sequence does not by itself measure an identical amount of tectonic subsidence, because compaction, water depth and loading also enter the reconstruction.
For an invented budget, suppose a basin floor subsides by 100 metres while 70 metres of uncompacted sediment accumulates and the reference water level remains fixed. The simple accommodation not filled by that sediment is 30 metres. A real reconstruction needs additional terms; the arithmetic merely makes the accounting visible.
Tectonic history can influence mineral systems, geothermal settings and hydrocarbon basins, but it does not guarantee a resource at every favourable-looking structure. Detailed geological, geochemical, economic and environmental assessment is necessary. A classroom model explains why investigators ask a question, not why a deposit must exist.
The useful everyday transfer is to resist one-cause stories. A mountain, valley or basin has a history of interacting processes. Start with a proposed tectonic mechanism, then check which observations would distinguish it from erosion, deposition or another cause. This makes a field description more informative than attaching a dramatic geological word to a shape.
CHAPTER 14 OF 18 · Connect landscapes and learning
14. Learn through maps, models and changed assumptions
Young learners can start with materials that bend, break and move, while recognising that a small model is an analogy. The first goal is to describe relative motion and keep a record of what changed. A model earns its place by making one relationship visible.
At secondary level, connect vectors, rates, graphs, density, forces and geological sequences. Teach a half-spreading calculation beside a full-spreading calculation, then ask why their numbers differ. Compare a strain estimate with a displacement estimate. Change the reference interval and require the units to change correctly.
Advanced study introduces mechanics, rheology, geophysics, numerical methods and field interpretation. Professional practice also requires careful mapping, instrument processing, uncertainty analysis and communication with other disciplines. Recognising a boundary type is a beginning, not a qualification to make engineering or emergency decisions.
A strong teaching sequence moves from a worked model to a partially scaffolded case and then to an unfamiliar map. Ask the learner what stayed the same: conservation of material, relative motion or the geometry of a component. Then ask what changed and which inference no longer follows.
Progress appears when students request a reference frame, distinguish a rate from a displacement and refuse to turn an average into an event prediction. An error log can record whether a mistake concerned units, geometry, time, material behaviour or evidence. The next lesson should repair that specific connection. Repeating a colourful plate-boundary poster will not repair a hidden confusion between centimetres per year and kilometres of accumulated motion.
CHAPTER 15 OF 18 · Check and continue
15. Independent checks with explained answers
| Quantity | Calculation or observation | Interpretation limit |
|---|---|---|
| Relative velocity | Subtract compatible vector components. | Speeds alone omit direction. |
| Half-spreading rate | Distance from ridge divided by age. | Full separation requires both sides. |
| Strain | Deformation divided by original dimension. | Displacement alone lacks the reference geometry. |
| Slip deficit | A specified locking and motion model. | It is not an earthquake countdown. |
Check one: two stations move north at 12 and 27 millimetres per year in the same frame. Their relative northward rate is 15 millimetres per year. If one moves east instead, subtract vector components rather than speed magnitudes. The direction is part of the information.
Check two: rocks one million years old lie 18 kilometres from a fictional ridge on each side. Under steady, symmetric spreading, half-spreading rate is 18 millimetres per year and full separation is 36. If only one flank is observed, symmetry remains an assumption requiring evidence.
Check three: a 50-kilometre-wide section, initially 20 kilometres thick, shortens to 40 kilometres with cross-sectional area conserved. Initial area is 1,000 square kilometres, so final average thickness is 25 kilometres. This does not imply five kilometres of surface uplift.
Check four: a fault marker has 600 metres of offset over a constrained 20,000-year interval. The average is 30 millimetres per year. The calculation neither establishes smooth annual slip nor dates the next earthquake. It also depends on the marker correlation and age interpretation.
Check five: a radar product shows six millimetres of movement along its line of sight. Without additional geometry, it is not a six-millimetre vertical movement measurement. State the missing information and propose an independent observation. The gold standard combines correct arithmetic with a clear account of what the evidence does and does not determine.
Is tectonics the same as geology?
Geology is the wider study of Earth materials, processes and history. Tectonics focuses on movement and deformation of the lithosphere and their larger organisation. The subjects connect closely, while their root guides serve different questions.
Do plates float on liquid magma?
No global magma ocean underlies the plates. Much of the mantle is solid and can deform over geological time. Local partial melting occurs in particular conditions. The distinction between material state and long-term mechanical behaviour is essential.
Can plate movement tell us when the next earthquake will occur?
A measured movement rate contributes to understanding deformation and hazard. It does not provide an exact event date by simple division. Fault behaviour, incomplete histories and model uncertainty prevent that shortcut from becoming a reliable prediction.
Why do geological and present-day rates sometimes differ?
They average different intervals and may concern different geometries or processes. Motion can vary, markers can be uncertain and a short record can include transient effects. Investigate those possibilities before treating disagreement as an arithmetic problem.
What is the most useful transferable lesson?
Keep the reference frame, interval and model assumptions beside the number. Then connect several independent observations before building a large story. That discipline helps with maps, environmental claims, engineering explanations and many comparisons beyond geology.
Choose a route by the question you want to answer. These connected guides develop the specialist mechanisms and adjacent subjects; they are companions to this explanation rather than substitutes for checking a particular real-world case.
Follow movement and deformation
How Science Works | Tectonics — Plate Motion, Faults, Folds, Mountain Building and the Deformation of Earth’s Lithosphere — Develop the complete existing mechanism library for plates, faults, folds and deformation.
How Science Works | Seismology — Earthquakes, Seismic Waves, Fault Rupture, Ground Motion and the Evidence Inside Earth — Examine earthquake signals and the fault processes they constrain.
How Science Works | Geophysics — Seismic Waves, Gravity, Magnetism, Deformation and the Hidden Earth — Connect physical measurements with Earth’s interior and structure.
How Science Works | Geology — Rocks, Tectonics, Deep Time, Earthquakes and the Dynamic Crust — Place deformation within Earth materials and geological history.
Read the changing landscape
How Science Works | Geomorphology — Rivers, Slopes, Erosion, Sediment and the Changing Shape of Landscapes — Trace interactions between uplift, erosion and landforms.
How Science Works | Volcanology — Magma Ascent, Eruptions, Lava, Ash, Hazards and the Evidence Inside Volcanoes — Investigate magma, volcanic systems and their tectonic settings.
How Science Works | Sedimentology — Weathering, Transport, Deposition, Sedimentary Structures and the Evidence of Ancient Environments — Connect sediment transport and deposition with basins and changing accommodation.
Continue across Gold Standard subjects
The Gold Standard of Geology provides the connected geology perspective.
The Gold Standard of Geophysics provides the connected geophysics perspective.
The Gold Standard of Seismology provides the connected seismology perspective.
The Gold Standard of Volcanology provides the connected volcanology perspective.
The Gold Standard of Geomorphology provides the connected geomorphology perspective.
The Gold Standard of Sedimentology provides the connected sedimentology perspective.
The Gold Standard of Geochronology provides the connected geochronology perspective.
The Gold Standard of Stratigraphy provides the connected stratigraphy perspective.
The Gold Standard of Geodesy provides the connected geodesy perspective.
For the wider collection, return to How X Works and How Science Works.
These primary institutional and research sources were checked on 21 October 2026. They support the scientific framing and specialist references; the numerical cases, teaching sequence and interpretation checklists are original educational examples.
USGS: what is a tectonic plate? — The lithospheric plate and its material setting.
USGS: are plates floating on magma? — Distinguish solid mantle deformation from a global magma-ocean misconception.
USGS: understanding plate motions — Boundary motion, geological markers and the evidence for spreading and convergence.
USGS: unanswered questions — Forces including slab pull, and the limits of simple driving-force cartoons.
USGS: seafloor-spreading and subduction model — An explicit teaching model for magnetic records, spreading and subduction.
USGS: can you predict earthquakes? — The distinction between exact prediction and probabilistic hazard assessment.
Return to the source appropriate to the specific measurement, method and context before using a classroom model to make a professional decision. Where the article states a conditional example, its conclusion depends on the assumptions explicitly given with it.
Contents · Previous chapter · Continue into the subject mechanisms