Skip to content

Gravimeters, Magnetometers and Data Acquisition

Absolute and relative gravimeters, zero-length springs, fluxgate, proton, Overhauser and optically pumped magnetometers, and survey design for CSIR NET.

  • 22 min read
  • Updated
On this page · 13 sections
  1. Prerequisites
  2. The Hook: Weighing a Buried Ore Body with a Spring
  3. What Is Being Measured
  4. Gravimeters
  5. Magnetometers
  6. Data Acquisition
  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: Weighing a Buried Ore Body with a Spring


What Is Being Measured

MethodQuantity measuredUnitsTypical anomaly
Gravityvertical component of the acceleration due to gravity, \( g_z \)mGal (\( 10^{-5} \) m s\(^{-2}\)); g.u. (\( 10^{-6} \) m s\(^{-2}\) = 0.1 mGal)0.01 to tens of mGal
Gravity gradiometrycomponents of the gradient tensor \( \partial g_i/\partial x_j \)Eötvös, E (\( 10^{-9} \) s\(^{-2}\) = 0.1 mGal/km)a few to hundreds of E
Magnetics (total field)magnitude \( \lvert \mathbf{F} \rvert \) of the field vectornT (gamma; \( 10^{-9} \) T)0.1 nT to thousands of nT
Magnetics (vector)components \( X, Y, Z \) or \( H, Z \)nTsame

Two physical properties drive these methods: density \( \rho \) for gravity and magnetisation \( \mathbf{M} \) (induced, \( \chi \mathbf{H} \), plus remanent) for magnetics. Gravity anomalies come from density contrasts \( \Delta\rho \); magnetic anomalies from magnetisation contrasts.


Gravimeters

graph TD
    A[Gravity measurement] --> B[Absolute: g itself]
    A --> C[Relative: difference in g between stations]
    B --> B1[Free-fall: dropped or thrown corner cube, laser interferometer]
    B --> B2[Pendulum: historical]
    C --> C1[Stable spring balance]
    C --> C2[Unstable / astatic: zero-length spring]
    C --> C3[Quartz spring with electronic feedback]
    C --> C4[Superconducting: levitated sphere, fixed station]
    C --> C5[Marine and airborne: stabilised platform]

Absolute measurement

Free-fall instruments. A corner-cube reflector is dropped in a vacuum chamber and its position is tracked by a laser interferometer against an atomic clock. Fitting

$$z(t) = z_0 + v_0 t + \tfrac{1}{2} g t^2$$

to many position–time pairs gives \( g \). In a rise-and-fall variant the mass is thrown up and timed passing two levels on the way up and down; for levels separated by \( h \) with time intervals \( T_1 \) (lower level) and \( T_2 \) (upper level) between the up and down passages,

$$g = \frac{8h}{T_1^2 - T_2^2}$$

Absolute instruments are heavy and slow, so they are used to set base values for networks, not to survey.

Pendulums. For a simple pendulum \( T = 2\pi\sqrt{L/g} \). Used relatively (same pendulum at two sites):

$$\frac{\Delta g}{g} = -\frac{2\,\Delta T}{T}$$

A precision of 0.01 mGal needs the period to about 5 parts in \( 10^9 \). Pendulums were the standard before spring gravimeters and are now of historical interest.

Relative gravimeters: the stable spring

A mass \( m \) on a spring of stiffness \( k \) stretches by \( mg/k \). A change \( \Delta g \) changes the length by

$$\boxed{\Delta L = \frac{m\,\Delta g}{k} = \frac{\Delta g}{\omega_0^2} = \frac{T_0^2}{4\pi^2}\,\Delta g}$$

where \( T_0 = 2\pi\sqrt{m/k} \) is the natural period of the system. Sensitivity grows as the square of the period: a long-period (soft) system is a sensitive one.

Unstable (astatic) gravimeters and the zero-length spring

An astatic (unstable, labilised) gravimeter adds a force that grows with displacement and opposes the restoring force of the spring. The net restoring torque is small, the effective period is long, and sensitivity is high. The system is operated close to, but not at, the point of instability.

The classic design is the LaCoste–Romberg gravimeter. A hinged beam carries the mass and is held up by a zero-length spring:

  • A zero-length spring is pre-tensioned during winding so that its tension is proportional to its total length, not to its extension. It would shrink to zero length if nothing held its coils apart.
  • With this spring attached between the beam and a point above the hinge, the torque of the spring balances the torque of gravity on the beam at every angle for one value of \( g \). For gravity close to that value the beam has a very long period, so a tiny change in \( g \) gives a large rotation.
  • The reading is made by a null method: a micrometer screw raises the top of the spring until the beam returns to a fixed reference position. The dial reading is converted to mGal with the instrument’s calibration table.

Other designs: the Worden gravimeter uses a fused-quartz spring system in a vacuum flask with temperature compensation; modern quartz-spring instruments with capacitive feedback read the force needed to hold the mass at null electronically and record digitally, often with automatic tilt and tide corrections.

PropertyStable (spring balance)Astatic (zero-length spring)
PrincipleHooke’s law, extension \( \propto \Delta g \)near-unstable equilibrium, very long effective period
Sensitivitylowhigh, field precision of order 0.01 mGal
Readingdeflectionnull (return beam to reference)
Rangenarrowwide, with resetting screw
Weaknesstoo insensitive for explorationdrift, tares, needs careful levelling and transport

Drift, tares and calibration

All spring gravimeters drift: the spring creeps and the reading changes slowly even at a fixed site. Drift is partly linear with time and partly due to temperature and handling. A sudden offset after a jolt is a tare. Both are handled by reoccupying a base station (see Data Acquisition below). Relative gravimeters also need calibration: the dial-to-mGal factor is checked by reading at two stations of known gravity difference (a calibration line), or by tilting methods in the laboratory.

Superconducting, gradiometer, marine and airborne gravity

  • Superconducting gravimeter: a niobium sphere levitated in a magnetic field from superconducting coils; very low drift and very high precision; used at fixed observatories for tides, free oscillations of the Earth and long-term changes, not for field surveys.
  • Gravity gradiometer: measures differences in gravity over a short baseline, so common accelerations of a moving platform cancel. Gradients are less affected by aircraft motion than gravity itself, which is why gradiometry is flown.
  • Marine and airborne gravimeters sit on gyro-stabilised platforms and are heavily damped; the platform’s vertical accelerations are removed by long time-averaging (filtering), so short wavelengths are lost.

For any moving platform the Eötvös effect matters. Moving east adds to the Earth’s rotation and increases the centrifugal acceleration, so measured gravity is lower; moving west, higher. The correction to be added is

$$\boxed{E = 7.503\,V\cos\phi\,\sin\alpha + 0.004154\,V^2\ \text{mGal}}$$

with \( V \) in knots, \( \phi \) the latitude and \( \alpha \) the heading measured clockwise from north. The first term is \( 2\omega V\cos\phi\sin\alpha \) with \( \omega = 7.292\times 10^{-5} \) rad s\(^{-1}\); the second is \( V^2/R \).


Magnetometers

InstrumentMeasuresPhysical principleNotes
Fluxgatea component (vector)two high-permeability cores driven into saturation by an AC field; an external field makes the saturation asymmetric and produces a signal at twice the drive frequency, proportional to the field along the core axisneeds orientation; used in observatories, borehole and satellite work, gradiometers; drift and temperature effects
Proton precessiontotal field (scalar)protons in a hydrocarbon or water sample are polarised by a strong DC field; when it is switched off they precess about the Earth’s field at the Larmor frequency, which is countedabsolute (depends only on a physical constant); no orientation needed; reads in cycles, not continuously; fails in steep gradients
Overhausertotal fieldproton precession enhanced by RF excitation of free electrons in the fluid (dynamic nuclear polarisation)continuous polarisation, faster sampling, lower power than a plain proton instrument
Optically pumped (caesium, potassium, helium)total fieldatoms are pumped by circularly polarised light into one sublevel; an RF field at the Larmor frequency of the electrons depopulates it and changes the light transmittedvery sensitive and fast; standard in airborne surveys; has heading error and dead zones in some sensor orientations
SQUIDcomponent or gradientquantised flux in a superconducting loop with Josephson junctionsextremely sensitive; needs cryogenic cooling; laboratory rock magnetism, some field use

The proton precession magnetometer in numbers

The precession frequency is proportional to the field:

$$\boxed{f = \frac{\gamma_p}{2\pi}\,B} \qquad \frac{\gamma_p}{2\pi} \approx 0.042577\ \text{Hz nT}^{-1} \quad\Longleftrightarrow\quad B \approx 23.487\,f\ \text{nT}$$

where \( \gamma_p \approx 2.675\times 10^{8} \) rad s\(^{-1}\) T\(^{-1}\) is the gyromagnetic ratio of the proton.

Because the frequency depends only on \( \gamma_p \), a proton magnetometer is absolute and does not drift. Its weaknesses: the signal decays during the count, so strong field gradients (near steel, near very magnetic rocks) dephase the protons and kill the signal; and readings come in cycles (polarise, then count). Overhauser and optically pumped instruments remove much of this limitation.

Fluxgate in one picture

Two identical cores are wound in opposition and driven by an AC current strong enough to saturate them each half-cycle. With no external field the two cores’ outputs cancel. A field component along the axis biases one core towards saturation earlier than the other, and the secondary coil picks up a voltage at twice the driving frequency whose amplitude is proportional to that component. The fluxgate is therefore a vector instrument; three orthogonal fluxgates give the full field.

Units and the field being measured

  • \( 1\ \text{nT} = 10^{-9}\ \text{T} = 1\ \gamma \). The main field ranges from about 25,000 nT near the magnetic equator to about 65,000 nT near the poles.
  • Strictly, a magnetometer measures \( \mathbf{B} \) in tesla; \( \mathbf{H} = \mathbf{B}/\mu_0 \) in A m\(^{-1}\) outside magnetic materials.
  • Induced magnetisation is \( \mathbf{M}_i = \chi\mathbf{H} = \chi\mathbf{F}/\mu_0 \), with \( \chi \) the dimensionless SI volume susceptibility.

Data Acquisition

Gravity surveys on land

1. Base network and ties. Every survey is tied to a base station whose absolute gravity is known, usually transferred from a national or international network of absolute and relative stations. Within the survey, local bases are set up and tied to it.

2. Looping. Readings are taken in loops that start and end at a base (A, 1, 2, 3, A, 4, 5, A, …). The base is reoccupied every one to two hours. The difference between the opening and closing base readings, divided by the elapsed time, gives the drift rate, which is assumed linear and removed from the readings in between (see Corrections and Regional–Residual Separation).

3. Position and height. The corrections need each station’s latitude and elevation, and gravity is very sensitive to elevation:

QuantitySensitivityAccuracy needed for 0.01 mGalfor 0.1 mGal
Elevation (free-air + Bouguer, 2.67 g cm\(^{-3}\))about 0.197 mGal per mabout 5 cmabout 0.5 m
North–south positionabout \( 0.81\sin 2\phi \) mGal per km; largest (0.81 mGal/km) at 45°about 12 m at 45°about 120 m at 45°

So elevation is the controlling survey requirement. Differential GNSS or levelling is used; a barometric altimeter is adequate only for regional work.

4. Station spacing. A body at depth \( z \) produces an anomaly whose width is of the order of \( z \) (the half-width of a sphere’s anomaly is \( 0.77z \)). Stations must sample the shortest wavelength of interest at least twice per wavelength (Nyquist), and in practice several times across an anomaly. Regional crustal surveys use spacings of kilometres; mineral and engineering surveys (microgravity for cavities) use metres.

5. Terrain information. A digital elevation model or field sketches out to some distance, for the terrain correction.

Magnetic surveys on the ground

  • Diurnal variation. The external field changes by tens of nT through a quiet day and by hundreds or more during a magnetic storm. A base-station magnetometer records continuously during the survey; its variation is subtracted from the field readings at the same times. On a ground survey with one instrument, the base is reoccupied, as for gravity. Surveys are suspended during magnetic storms.
  • Cultural noise. Readings are kept away from fences, pipes, vehicles, power lines; the operator carries no steel.
  • Sensor height. A pole lifts the sensor above near-surface soil and debris; two sensors on one pole form a vertical gradiometer, which needs no diurnal correction (diurnal changes are the same at both sensors) and sharpens shallow anomalies.

Aeromagnetic and airborne surveys

ParameterWhat it controls
Flight-line spacingthe shortest wavelength resolved across lines; lines run perpendicular to the geological strike
Tie linesflown perpendicular to the flight lines; differences at crossings are used for levelling (removing residual diurnal and heading errors)
Terrain clearancethe height above ground; lower clearance gives higher resolution of shallow sources. A draped survey follows topography; a barometric survey flies at constant altitude
Sampling interval along lineset by aircraft speed and sampling rate; usually much finer than line spacing
Compensationthe aircraft’s own permanent and induced fields and eddy currents are removed using a calibration flight with manoeuvres
Lag correctionshifts readings to the true position of the sensor, which trails the navigation antenna
Heading correctionremoves the reading difference between opposite flight directions

Airborne gravity uses the same layout but with long filters for aircraft accelerations, which limits resolution to wavelengths of kilometres; gradiometry recovers shorter wavelengths.

Satellite data

Satellite gravity and magnetic missions map the longest wavelengths: the geoid and crustal magnetic field at wavelengths of hundreds of kilometres. They complement, not replace, ground and airborne surveys.


Memory Tricks

Don’t Confuse

ThisvsThat
Absolute gravimeter: measures \( g \) (free fall)relative gravimeter: measures differences, needs a base tie and drift control
Stable gravimeter: Hooke’s law deflectionastatic gravimeter: near-unstable equilibrium, null reading
Drift: slow, continuous, removed by loop closuretare: sudden offset from a jolt
Fluxgate: vector, relative, can driftproton precession: scalar total field, absolute
Proton precession: polarise then count, cyclicOverhauser and caesium: continuous, faster, more sensitive
Diurnal correction (magnetics): time-dependent external fielddrift correction (gravity): instrument behaviour
Total-field anomaly \( \Delta T \): projection on \( \hat{\mathbf{F}} \)vertical-field anomaly \( \Delta Z \)
mGal: \( 10^{-5} \) m s\(^{-2}\)gravity unit g.u.: \( 10^{-6} \) m s\(^{-2}\); Eötvös: \( 10^{-9} \) s\(^{-2}\) (a gradient)

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. Which magnetometer measures a single component of the field using saturation of high-permeability cores?
  2. Express 1 mGal in m s\(^{-2}\) and in gravity units.
  3. Why is a base station reoccupied during a gravity survey?
Answers
  1. Fluxgate magnetometer. The second-harmonic output is proportional to the field component along the core axis.
  2. \( 10^{-5} \) m s\(^{-2}\) = 10 g.u. (1 g.u. = \( 10^{-6} \) m s\(^{-2}\)).
  3. To measure and remove instrument drift (and detect tares) by comparing repeated base readings, and to tie all stations to a common datum.

Level 2: Part B Style

Practice Q1. A proton precession magnetometer records a precession frequency of 2,300 Hz. Taking \( \gamma_p/2\pi = 0.042577 \) Hz nT\(^{-1}\), the total field is closest to:

(a) 48,000 nT (b) 54,000 nT (c) 58,000 nT (d) 98,000 nT

Answer: (b) 54,000 nT

\( B = f/(0.042577) = 2300 \times 23.487 \approx 54{,}020 \) nT.

Option (d) multiplies the frequency by 42.577 (the constant in Hz per microtesla) instead of dividing: \( 2300\times 42.577 \approx 97{,}900 \). That is also unphysical, since the main field at the surface does not exceed about 65,000 nT.

Practice Q2. A stable spring gravimeter is redesigned so that its natural period rises from 2 s to 10 s. Its mechanical sensitivity (displacement per unit change in \( g \)) changes by a factor of:

(a) 5 (b) 10 (c) 25 (d) 100

Answer: (c) 25

\( \Delta L = T_0^2\Delta g/4\pi^2 \), so sensitivity \( \propto T_0^2 \): \( (10/2)^2 = 25 \).

Option (a) assumes sensitivity proportional to the period.

Practice Q3. For a microgravity survey that must detect anomalies of 0.02 mGal, the elevation of each station must be known to about (use 0.197 mGal m\(^{-1}\) for the combined elevation effect):

(a) 1 cm (b) 10 cm (c) 1 m (d) 10 m

Answer: (b) 10 cm

\( 0.02/0.197 \approx 0.10 \) m. To keep elevation error below the anomaly, heights must be good to about 10 cm (in practice better).

Practice Q4. Which statement about an optically pumped caesium magnetometer is correct?

(a) It measures the vertical component of the field only (b) It measures the total field and can sample rapidly, making it suitable for airborne surveys (c) It needs liquid-helium cooling (d) It works by counting proton precession after a polarising field is removed

Answer: (b)

Optically pumped magnetometers measure the total field via the Larmor frequency of atomic electrons, continuously and quickly. (a) describes no standard scalar instrument; (c) describes a SQUID; (d) describes a proton precession magnetometer.

Level 3: Part C Style

Practice Q5. A gravity loop gives these readings (instrument units already converted to mGal):

StationTimeReading (mGal)
Base B08:001250.40
P109:001253.10
P210:301248.95
Base B12:001250.80

Assuming linear drift, the drift-corrected gravity differences P1 − B and P2 − B are:

(a) 2.70 and −1.45 mGal (b) 2.60 and −1.70 mGal (c) 2.80 and −1.20 mGal (d) 2.30 and −1.85 mGal

Answer: (b) 2.60 and −1.70 mGal

Drift = \( (1250.80 - 1250.40)/4\ \text{h} = 0.10 \) mGal h\(^{-1}\) (readings rising).

Base value expected at 09:00: \( 1250.40 + 0.10 = 1250.50 \). P1 − B = \( 1253.10 - 1250.50 = 2.60 \) mGal.

Base value expected at 10:30: \( 1250.40 + 0.25 = 1250.65 \). P2 − B = \( 1248.95 - 1250.65 = -1.70 \) mGal.

The distractors are the usual slips:

  • (a) ignores drift: \( 1253.10 - 1250.40 = 2.70 \), \( 1248.95 - 1250.40 = -1.45 \).
  • (c) applies the drift with the wrong sign: \( 1253.10 - 1250.30 = 2.80 \), \( 1248.95 - 1250.15 = -1.20 \).
  • (d) uses the closing base reading for every station: \( 1253.10 - 1250.80 = 2.30 \), \( 1248.95 - 1250.80 = -1.85 \).

Practice Q6. An airborne magnetic survey is planned over a granite–greenstone terrain with steeply dipping units striking north–south. Consider:

  1. Flight lines should run east–west.
  2. Lowering the terrain clearance improves the resolution of near-surface sources.
  3. Tie-line crossings are used to level the data.
  4. A single total-field sensor needs no diurnal correction because the aircraft moves quickly.

Which are correct?

(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. Lines run across strike to sample the shortest wavelengths. Correct.
  2. Anomaly wavelength scales with source-to-sensor distance; flying lower sharpens shallow anomalies. Correct.
  3. Differences at crossings are minimised to remove residual diurnal and heading errors. Correct.
  4. Wrong. The aircraft samples one line over many minutes and the whole survey over days; diurnal changes of tens of nT occur over those times. Base-station records are still needed. Only a gradiometer is largely immune.

Practice Q7. A ship records gravity while steaming at 8 knots on a heading of 270° (due west) at latitude 60° N. The Eötvös correction is approximately:

(a) +30.3 mGal (b) −29.8 mGal (c) −30.3 mGal (d) +29.8 mGal

Answer: (b) −29.8 mGal

\( E = 7.503\times 8\times\cos 60°\times\sin 270° + 0.004154\times 64 \) \( = 7.503\times 8\times 0.5\times(-1) + 0.266 = -30.01 + 0.27 = -29.75 \) mGal.

Westward motion reduces the centrifugal acceleration, so the ship reads too much gravity and the correction is negative. Option (c) subtracts the \( V^2 \) term instead of adding it (\( -30.01 - 0.27 = -30.28 \)); ignoring that term altogether would give \( -30.0 \), still not (b). Option (a) has the sign of the main term wrong (\( +30.01 + 0.27 \)), and (d) reverses both signs.


Quick Revision Table

IdeaFormula or fact
Units1 mGal = \( 10^{-5} \) m s\(^{-2}\) = 10 g.u.; 1 E = \( 10^{-9} \) s\(^{-2}\); 1 nT = 1 γ
Free fall\( z = z_0 + v_0t + \tfrac12 gt^2 \); rise-and-fall \( g = 8h/(T_1^2 - T_2^2) \)
Pendulum (relative)\( \Delta g/g = -2\Delta T/T \)
Stable spring\( \Delta L = T_0^2\Delta g/4\pi^2 \)
Astatic gravimeternear-unstable, zero-length spring, null reading, drift and tares
Proton precession\( f = (\gamma_p/2\pi)B \); 0.042577 Hz/nT; \( B = 23.487f \) nT
Fluxgatecomponent, second harmonic
Optically pumpedtotal field, fast, airborne standard, heading error
Total-field anomaly\( \Delta T \approx \Delta\mathbf{B}\cdot\hat{\mathbf{F}} \)
Eötvös\( 7.503V\cos\phi\sin\alpha + 0.004154V^2 \) mGal, V in knots
Height sensitivityabout 0.197 mGal/m (2.67 g cm\(^{-3}\))
N–S position sensitivityabout \( 0.81\sin 2\phi \) mGal/km
Survey designbase ties, loops, across-strike lines, tie lines, diurnal base

Key Takeaways

What’s Next