HOW SCIENCE WORKS · EARTH SCIENCE · SUBJECT LIBRARY · BATCH 12
Geophysics uses physical measurements at Earth’s surface, in boreholes, from aircraft and from satellites to infer structures and processes we cannot observe directly. Seismic waves reveal elastic structure. Gravity responds to mass distribution. Magnetic fields record electrical currents and magnetic minerals. Geodesy measures deformation. Heat flow constrains the planet’s thermal state.
Wait, what? Scientists know Earth’s outer core is liquid without drilling anywhere near it. They can measure plates moving only millimetres to centimetres per year. A tiny gravity anomaly can reveal a buried density contrast. Geophysics works by turning fields and waves into inverse problems: observe a response, propose a physical model, and ask which hidden structure could have produced it.
This article owns the physical-inference layer between Earth Science, Geology and Physics. It does not replace hazard warnings, engineering-site investigations or operational earthquake procedures.
Reading route: Read seismic waves → Infer Earth structure → Use gravity and magnetism → Measure deformation → Solve inverse problems → Learn and test understanding.
1. The scientific job is to infer hidden structure from physical response
Geophysics begins with a forward question: if Earth had a specified density, elasticity, temperature or conductivity structure, what signal should an instrument record? It then turns the problem around: given the recorded signal, which structures remain plausible?
This reverse step is the inverse problem. It is powerful because the planet is mostly inaccessible, but it is also dangerous because several hidden structures can sometimes produce similar measurements.
2. A CivDJ lens: field, disturbance, receiver and reconstruction
A geophysical explanation can be organised around a physical field or medium, a disturbance, a receiver and a reconstructed hidden state. An earthquake launches waves; Earth modifies their speeds and paths; seismometers record ground motion; inversion reconstructs possible internal structure.
The same pattern appears in gravity, magnetics and geodesy. The framework helps keep the measured signal separate from the inferred Earth model.
3. P waves and S waves respond differently to material properties
Compressional P waves alternately compress and expand material along the propagation direction. Shear S waves deform material perpendicular to propagation. Their speeds depend on density and elastic moduli.
S waves require shear rigidity and therefore do not propagate through an ordinary liquid in the same way they propagate through solids. That difference is central to evidence about Earth’s deep interior.
4. Travel time converts wave arrival into structure
A seismic station records when different phases arrive after an event. Travel time depends on path length and wave speed along the path.
If Earth were uniform, travel-time curves would be simple. Their observed curvature and phase changes reveal that wave speed varies strongly with depth.
5. Worked example: separate distance from speed
Original teaching example. Suppose a simplified wave travels 300 km through a uniform model region in 50 s. Its average path speed is 6.0 km/s.
If another path of the same geometric length takes 60 s, the average speed is 5.0 km/s. The delay could reflect a slower medium, but before making that claim the path geometry and phase identity must also be checked.
6. Refraction bends seismic rays through velocity gradients
When wave speed changes with depth, seismic rays refract. Smooth gradients curve paths; sharp boundaries can reflect and refract distinct phases.
This is why deeper paths can sometimes arrive sooner than a naive straight-line model predicts. They may spend part of their journey in faster material.
7. Reflection reveals boundaries
Changes in seismic impedance can reflect part of an incoming wave. Reflection methods therefore map boundaries such as sediment layers, crustal interfaces or other contrasts.
A reflection image is not a photograph of rock. It is a reconstruction from recorded waveforms, source timing, assumed velocities and processing choices.
8. Earthquake location is a timing inversion
An earthquake origin has an unknown location and start time. Stations at different positions record P and S arrivals. The source parameters are adjusted until predicted and observed arrival times fit sufficiently well.
More stations with better geometry reduce ambiguity. Stations all on one side can locate an event less reliably than a network surrounding it.
9. The Moho is inferred from a seismic speed contrast
The Mohorovičić discontinuity marks a strong increase in seismic velocity between crust and mantle. It was discovered because distant seismic arrivals could not be explained by one uniform crustal speed.
The boundary is therefore an inference from consistent travel-time behaviour across many earthquakes and stations.
10. S-wave absence reveals the liquid outer core
Because ordinary liquids cannot sustain shear stress statically, direct S waves do not pass through Earth’s liquid outer core. Their shadow zones are a major line of evidence for the core’s state.
P waves do cross the core but change speed and direction strongly, providing additional constraints on core boundaries.
11. Worked example: why one missing phase matters
Original conceptual example. Suppose a model predicts an S-wave phase through a deep region, but no station beyond the corresponding angular range records it while P-wave phases continue to arrive.
The pattern is stronger than one missing trace. A systematic absence across many paths supports a material-state explanation rather than a single instrument failure.
12. Seismic tomography turns many paths into a three-dimensional model
Tomography compares observed travel times with predictions from a reference Earth. Paths that arrive early can indicate faster-than-reference regions; late arrivals can indicate slower regions.
Thousands or millions of path constraints are combined to estimate a 3D velocity field. Resolution varies strongly with ray coverage, so every tomography image should be read with its resolution map.
13. Seismic velocity is not temperature alone
Temperature affects elastic properties, but composition, pressure, partial melt, water and crystal orientation can also change seismic speed.
A slow seismic anomaly therefore should not automatically be coloured “hot”. Translating velocity into temperature requires mineral-physics assumptions and independent constraints.
14. Earthquake magnitude and intensity answer different questions
Magnitude estimates the size of the source from seismic observations. Intensity describes the effects of shaking at a particular place.
One earthquake has one assigned magnitude within a chosen scale and method, but many intensities across a region because distance, local geology and building response vary.
15. Gravity maps mass distribution
Gravity measured at Earth’s surface or from satellites depends on the distribution of mass beneath and around the observation point.
After correcting for elevation and known geometry, residual gravity anomalies can indicate density contrasts such as sedimentary basins, mountains with deep roots or dense igneous bodies.
16. A gravity anomaly is not a unique underground picture
A shallow small dense body and a deeper larger dense body can sometimes produce similar surface gravity signals.
This non-uniqueness is fundamental. Geological mapping, seismic data or drilling may be needed to distinguish the alternatives.
17. Magnetics reveal currents and magnetic minerals
Earth’s main magnetic field is generated by electrical currents in the liquid outer core. Crustal rocks can also carry induced or remanent magnetisation.
Magnetic surveys therefore contain signals from multiple sources at different depths and scales. Separating them is part of the interpretation.
18. Paleomagnetism records past field direction in rocks
As some rocks cool or sediments settle, magnetic minerals can acquire remanent magnetisation related to the ambient field.
Patterns of magnetic polarity on the seafloor became major evidence for seafloor spreading when they matched the known history of geomagnetic reversals on both sides of mid-ocean ridges.
19. Heat flow constrains Earth’s thermal engine
Geothermal heat reaches the surface through conduction, advection and volcanic or hydrothermal processes. Surface heat-flow measurements combine temperature gradients with thermal conductivity.
The measured heat reflects both primordial heat and ongoing radioactive decay, while mantle convection transports energy on large scales.
20. Geodesy measures Earth’s changing shape
GNSS stations, satellite radar, levelling and other geodetic methods measure positions and changes in position.
Millimetre-scale annual motions become visible because modern reference frames and repeated observations are extraordinarily precise.
21. Worked example: plate motion accumulates slowly but measurably
Original scaling example. A station moving 30 mm per year relative to a stable reference accumulates 0.30 m of displacement in ten years if the rate stays constant.
A real velocity estimate should include uncertainty, reference-frame definition and possible transient motion from earthquakes, groundwater change or volcanic processes.
22. InSAR maps deformation from radar phase differences
Interferometric synthetic aperture radar compares the phase of radar observations from repeated satellite passes. Changes in path length can reveal ground motion along the radar line of sight.
Atmospheric water vapour, orbital errors and surface changes can contaminate the phase. Beautiful colour fringes become evidence only after those effects are controlled.
23. Elastic strain stores energy before some earthquakes
Tectonic loading can slowly deform rocks around faults. When frictional resistance is overcome, part of the stored elastic energy is released as seismic waves, fracture and permanent deformation.
This does not permit deterministic short-term prediction of ordinary earthquakes from strain alone. Fault systems contain heterogeneous stress, friction and geometry.
24. Inverse problems require regularisation
Because geophysical data are finite and noisy, the mathematically best-fitting model can become unrealistically complicated.
Regularisation adds constraints such as smoothness, simplicity or prior geological information. The chosen regularisation influences the image, so it must be reported rather than hidden.
25. Resolution tells us what the data can actually distinguish
A model grid can contain tiny cells even when the data cannot resolve features that small.
Resolution tests, synthetic recovery experiments and uncertainty estimates reveal which structures are supported and which merely reflect the parametrisation.
26. Common geophysics failure modes
- Image equals photograph: forgetting inversion and processing.
- One anomaly equals one cause: ignoring non-uniqueness.
- Slow seismic velocity equals hot: ignoring composition and melt.
- Magnitude equals local damage: confusing source size with site intensity.
- Dense grid equals high resolution: confusing model pixels with data information.
- One instrument proves mechanism: ignoring the value of independent fields and measurements.
27. How to think like a geophysicist
Define the physical property to be inferred. Write the forward model. Identify the measurement geometry. Estimate uncertainty and resolution. Test alternative structures. Use geology, geochemistry or geodesy as independent constraints where possible.
Most importantly, distinguish the recorded signal from the hidden-Earth model reconstructed from it.
28. A staged learning route
First encounter: distinguish wave, field and deformation observations. Use simple travel-time and gravity analogies.
Secondary-to-JC bridge: add P/S waves, refraction, earthquake location, gravity anomalies, plate motion and magnetic stripes.
Higher resolution: add elasticity, tomography, potential theory, geodesy, inverse methods, regularisation and resolution analysis. This is a learning route, not a syllabus claim.
29. Checkpoints with answers
Why do S-wave shadow zones support a liquid outer core? Because ordinary liquids cannot support the shear deformation required for direct S-wave propagation.
Does a gravity anomaly uniquely determine underground structure? No. Different density geometries can produce similar surface fields.
Can a 1 km model grid prove 1 km resolution? No. Resolution is controlled by data coverage, noise and model sensitivity.
Why combine seismic and geodetic evidence? They respond to different physical aspects of Earth and reduce ambiguity when their inferences agree.
30. The final skill is making the inaccessible Earth testable
A complete geophysical explanation names the property being inferred, the physical law connecting it to the signal, the instrument response, the inversion assumptions, the resolution and the independent evidence capable of falsifying the reconstruction.
Sources and connected subjects
Useful public foundations include the USGS Earthquake Hazards Program, EarthScope, and NASA Earth-observation resources on gravity and geodesy. Numerical examples above are original teaching constructions.
Continue to Geochemistry, Geology, Earth Science and Physics.
Return to How Science Works or the How X Works Hub.