Heat Flow, the Gravity Field and the Crust

This chapter completes the heading “Introduction to Earth and planetary systems” with its last three items: “Heat flow within the Earth; Gravitational field of the Earth; Continental and oceanic crust – composition, structure and thickness”. It teaches Fourier’s law and the geothermal gradient, radiogenic heat and the steady conductive geotherm, the variation of g with latitude, height and depth, the geoid and the Bouguer slab, and the composition, layering and thickness of continental and oceanic crust. The numericals are temperature from heat flow, heat flow and temperature in a heat-producing layer, the Bouguer slab, the elevation correction, Moho depth from two-way time and the elevation implied by a crustal root.

1. Heat flow by conduction: Fourier’s law and the geothermal gradient

Heat leaves the Earth by conduction through the rigid lithosphere and by convection, the movement of the material itself, in the mantle and the outer core. Fourier’s law says that heat flow density is proportional to the temperature gradient. With depth z measured downward and temperature rising with depth, the upward heat flow is q = k dT/dz, where k is the thermal conductivity and dT/dz is the geothermal gradient. In SI units q is in W/m² (quoted as mW/m²), k in W m⁻¹ K⁻¹ and the gradient in K/km, the same as °C/km. The older heat-flow unit is 1 HFU = 10⁻⁶ cal cm⁻² s⁻¹ = 41.84 mW/m².

Crystalline crustal rocks have k of about 2 to 3 W m⁻¹ K⁻¹ and sediments less. A continental gradient of about 25 °C/km near the surface with k = 2.5 gives q ≈ 62 mW/m². Turned round, q = 60 mW/m² and k = 2.5 give a gradient of 0.060/2.5 = 0.024 K/m = 24 °C/km. Across layers of different k in steady state without sources, q is the same in all layers, so the low-conductivity layer has the steeper gradient. The mean heat flow is about 65 mW/m² on the continents, 101 mW/m² through the ocean floor and 87 mW/m² globally, a total loss near 44 to 47 TW.

Heat flow is measured on land from the temperature profile in a borehole together with the conductivity of the core, and at sea with a probe pushed into the sediment. The oceanic mean exceeds the continental mean because young, hot ocean lithosphere loses heat fast, while continents are insulated by thick crust and cratonic roots.

⚠️ The gradient is not constant with depth
Extending 25 °C/km to the core–mantle boundary at 2,890 km would give about 70,000 °C, far above the few thousand kelvin actually inferred. The gradient falls because the heat-producing elements sit in the upper crust and because below the lithosphere the mantle convects, so its temperature rises roughly along an adiabat of well under 1 °C/km. The base of the lithosphere is at about 1,300 °C.

2. Sources of heat, radiogenic heat and the conductive geotherm

The heat comes from three kinds of source. Radiogenic heat is released by the decay of ²³⁸U, ²³⁵U, ²³²Th and ⁴⁰K. Secular cooling is heat left from accretion and core formation and slowly given up. Core heat includes the latent heat and gravitational energy released as the solid inner core grows. The heat production per unit volume A is largest in felsic rocks, about 2 to 3 µW/m³ for granite, a few tenths of that for basalt and almost nothing for peridotite. The thin continental crust therefore holds a disproportionate share of the heat-producing elements, which is why continental heat flow depends so strongly on crustal composition.

In steady state a layer of uniform heat production A obeys k d²T/dz² = −A. Heat flow decreases with depth because the heat generated below must be added: q(z) = q0 − A z. Integrating once more, T(z) = T0 + (q0/k) z − A z²/(2k). The heat flow leaving the base of a layer of thickness h is q0 − A h, the contribution from deeper sources. Within a province the observed surface heat flow rises linearly with surface heat production, q0 = qr + D A0, where qr is the reduced (deep) heat flow and D, of order 10 km, is the depth scale over which A falls with depth.

ℹ️ Worked example
A 15 km granitic layer has A = 2.0 µW/m³, k = 2.5 W m⁻¹ K⁻¹, surface heat flow 70 mW/m² and surface temperature 10 °C. At its base q = 70 − (2.0×10⁻³ mW/m³ × 15,000 m = 30) = 40 mW/m². The temperature there is T = 10 + (0.070 × 15,000/2.5) − (2.0×10⁻⁶ × 15,000²)/(2 × 2.5) = 10 + 420 − 90 = 340 °C. The heat production lowers the temperature by 90 °C compared with a source-free layer.

Surface heat flow varies with tectonic setting. On the sea floor it is highest at ridges and falls as the lithosphere ages and cools, but seawater circulating through the young, fractured crust carries heat away, so conductive measurements on young floor fall below the conduction model. On the continents it is lowest in old stable cratons and higher in young orogens, rifts and regions of recent magmatism.

3. The gravitational field of the Earth and how g varies

For a sphere of mass M the attraction at radius R is g = GM/R². With G = 6.674×10⁻¹¹ N m² kg⁻², M = 5.97×10²⁴ kg and R = 6,371 km this gives about 9.8 m/s². The mean density of the Earth, about 5.5 g/cm³, is roughly twice that of surface rocks, which proves that the interior is dense. Gravity is measured in gal: 1 gal = 1 cm/s², so 1 mGal = 10⁻⁵ m/s² and the field is about 980,000 mGal. Anomalies of interest are a few to a few hundred mGal.

  • Latitude: g is about 9.78 m/s² at the equator and 9.83 m/s² at the poles, a difference near 0.5 per cent. Two effects add up: the equatorial radius is about 21 km larger, so the equator is farther from the centre, and the centrifugal acceleration ω²R cos²φ, which is zero at the poles, reduces g at the equator. With ω = 7.29×10⁻⁵ rad/s and R = 6,378 km, the equatorial value is ω²R = 0.034 m/s², about 1/289 of g.
  • Height: g = GM/(R + h)², so dg/dh = −2g/R, about −0.3086 mGal per metre (the free-air gradient). Gravity falls by about 0.31 mGal for every metre of elevation.
  • Depth: inside a uniform sphere g falls in proportion to r. The real Earth has a dense core, so g stays near 9.8 to 10 m/s² through most of the mantle, reaches a maximum of about 10.7 m/s² at the core–mantle boundary, and then falls almost linearly to zero at the centre.
  • Local density: excess or deficient mass beneath the surface changes g by small amounts, which is what gravity exploration maps.
⚠️ Gravity inside the Earth is not a straight line
The linear fall to the centre belongs to a uniform-density sphere. Because the Earth’s core is far denser than the mantle, most of the mass is enclosed at depth, and g first stays high or even rises slightly before it drops to zero.

4. Geoid, ellipsoid and the elevation corrections

The geoid is the equipotential surface of gravity plus rotation that coincides with mean sea level and is continued under the continents; it is everywhere perpendicular to the plumb line. It is not a surface of constant g, because g varies along it. The smooth reference ellipsoid, with a flattening of about 1/298, approximates its shape. The geoid departs from the ellipsoid by only about ±100 m, with a low of roughly −100 m south of India.

To compare stations at different heights, observed gravity is reduced to the geoid. The free-air correction adds 0.3086 mGal per metre of elevation to undo the loss with height. The Bouguer correction removes the attraction of the rock between station and datum, treated as an infinite horizontal slab of thickness h and density ρ: Δg = 2πGρh, which is about 0.0419 ρ h mGal with ρ in g/cm³ and h in metres, or 0.112 mGal/m for ρ = 2,670 kg/m³. For a slab 500 m thick this is 2π × 6.674×10⁻¹¹ × 2,670 × 500 = 5.6×10⁻⁴ m/s², that is 56 mGal. On a high plateau the two corrections act in opposite senses, so a rise of 1 m lowers g by about 0.3086 − 0.112 ≈ 0.197 mGal.

⚠️ Equipotential does not mean equal g
A common wrong statement is that the geoid is a surface of constant gravity. It is a surface of constant potential. Gravity differs along it by about 5,000 mGal between equator and pole, and the geoid is also not the ellipsoid: the ellipsoid is a mathematical reference and the geoid is a physical surface.

5. Continental crust: composition, structure and thickness

The continental crust averages about 35 to 40 km thick. It is 20 to 30 km in rifted and extended regions, around 35 to 45 km under old shields, and about 70 km beneath Tibet and the Himalaya where it has been doubled by collision. Its density is about 2.7 to 2.8 g/cm³, and it is old, because its buoyancy prevents it from being subducted: the oldest surviving rocks are about 4 Ga.

  • Upper crust (to about 10 to 15 km): granitic to granodioritic, rich in quartz and feldspar, felsic, with Vp about 6.0 to 6.4 km/s.
  • Lower crust: more mafic, granulite-facies rock rich in plagioclase and pyroxene, with Vp about 6.5 to 7.2 km/s. A weak Conrad discontinuity between the two is seen in some regions only.
  • Moho at the base, where Vp reaches 8.0 to 8.2 km/s in peridotite. The bulk crust is intermediate, about 60 wt% SiO₂, felsic above and mafic below.

6. Oceanic crust and how crustal thickness is found

The oceanic crust is thin, about 7 km on average (5 to 10 km), dense, about 2.9 to 3.0 g/cm³, and basaltic to gabbroic. Seismology divides it into Layer 1, sediment that is thin near ridges and thickens toward continents; Layer 2, about 1.5 to 2 km of pillow basalt over sheeted dykes; and Layer 3, about 4 to 5 km of gabbro with Vp about 6.5 to 7.0 km/s. Below the Moho lies peridotite with Vp about 8.1 km/s. On land, ophiolites such as the Oman ophiolite preserve the same sequence: pillow lava, sheeted dykes, gabbro, ultramafic rock. Oceanic crust is formed at ridges and consumed at trenches, so the floor of the main ocean basins is nowhere older than about 180 to 200 Ma.

Continental and oceanic crust compared
PropertyContinentalOceanic
ThicknessAbout 35–40 km (20–70 km)About 7 km (5–10 km)
Density (g/cm³)2.7–2.82.9–3.0
CompositionIntermediate bulk; felsic upper, mafic lowerBasalt, dykes and gabbro (mafic)
AgeUp to about 4 GaMain ocean basins younger than about 200 Ma
Fate at convergenceBuoyant, not subductedSubducted and recycled

Crustal thickness is found by several methods. Seismic refraction uses the Pn head wave and its crossover distance. Deep seismic reflection records the Moho reflection at normal incidence, so h = V T₂/2, where T₂ is the two-way time and V the average crustal speed: 11 s at 6.4 km/s gives 35.2 km. Receiver functions use P-to-S conversion at the Moho. Gravity with isostasy predicts thickness from elevation. In Airy balance a crust T thick under elevation h satisfies T = T0 + h + r = T0 + h (1 + ρc/(ρm − ρc)), so for ρc = 2,800, ρm = 3,300 and T0 = 35 km a 70 km crust implies h = 35/6.6 = 5.3 km, close to the 4.5 to 5 km elevation of the Tibetan plateau.

Key takeaways

  • Fourier’s law: q = k dT/dz, with q in mW/m², k about 2 to 3 W m⁻¹ K⁻¹ and a surface gradient near 25 °C/km. Mean heat flow is about 65 mW/m² on the continents, 101 on the ocean floor and 87 globally (44 to 47 TW). The gradient is not constant with depth: the mantle is convecting.
  • Heat sources are radiogenic decay of U, Th and ⁴⁰K, secular cooling and core heat. With uniform A in steady state q(z) = q0 − A z and T(z) = T0 + (q0/k) z − A z²/(2k); surface flow follows q0 = qr + D A0. Seawater circulation lowers measured heat flow on young ocean floor.
  • g = GM/R² gives about 9.8 m/s²; 1 mGal = 10⁻⁵ m/s². g rises about 0.5 per cent from equator to pole because of the bulge and the centrifugal term; it falls 0.3086 mGal per metre of height; inside the Earth it peaks near 10.7 m/s² at the core–mantle boundary.
  • The geoid is an equipotential surface at mean sea level, within about ±100 m of the ellipsoid, and not a surface of constant g. The Bouguer slab is 2πGρh (0.0419 ρ h mGal); free-air and Bouguer effects oppose, so the net elevation correction is about 0.197 mGal/m for ρ = 2,670.
  • Continental crust: 35 to 40 km, 2.7 to 2.8 g/cm³, felsic upper over mafic lower, intermediate bulk (about 60 wt% SiO₂), old (to 4 Ga), not subducted. Oceanic crust: about 7 km, 2.9 to 3.0 g/cm³, sediment, pillow basalt and dykes, gabbro, younger than about 200 Ma.
  • Thickness comes from refraction (crossover), reflection (h = V T₂/2), receiver functions and Airy isostasy (T = T0 + h (1 + ρc/(ρm − ρc))); a 70 km crust implies about 5.3 km elevation for 2,800 and 3,300 kg/m³.

Practice questions (17)

Attempt each one before opening the answer. Every explanation names the tempting wrong option as well as the right one, because that is where marks are lost.

  1. In SI units, thermal conductivity k and heat flow density q are expressed respectively in:

    1. W m⁻² for conductivity and W m⁻¹ K⁻¹ for heat flow density
    2. W m⁻¹ for conductivity and W m⁻² K⁻¹ for heat flow density
    3. J m⁻² K⁻¹ for conductivity and J s⁻¹ for heat flow density
    4. W m⁻¹ K⁻¹ for conductivity and W m⁻² for heat flow density
    Show answer

    Answer: D — W m⁻¹ K⁻¹ for conductivity and W m⁻² for heat flow density

    From q = k dT/dz, k = q/(dT/dz) has units (W m⁻²)/(K m⁻¹) = W m⁻¹ K⁻¹, and q is power per unit area, W m⁻². Swapping the two units is the usual confusion between a material property and a flux.

  2. Surface heat flow at a site is 54 mW/m² and the thermal conductivity of the rock is 3.0 W m⁻¹ K⁻¹. Assume a constant gradient with no heat sources and a surface temperature of 10 °C. The temperature at 8 km depth, in °C, is ____.

    Numerical answer — type the value.

    Show answer

    Answer: 154

    Gradient = q/k = 0.054/3.0 = 0.018 K/m = 18 °C/km. At 8 km the rise is 18 × 8 = 144 °C, so T = 10 + 144 = 154 °C. Forgetting to add the surface temperature gives 144.

  3. A 10 km thick crustal layer has uniform heat production 2.5 µW/m³ and conductivity 2.5 W m⁻¹ K⁻¹. Surface heat flow is 65 mW/m² and surface temperature is 10 °C. Assuming steady one-dimensional conduction, the temperature at the base of the layer, in °C, is ____.

    Numerical answer — type the value.

    Show answer

    Answer: 220

    T(z) = T0 + (q0/k) z − A z²/(2k). The first term: 0.065 × 10,000/2.5 = 260 °C. The heat-production term: 2.5×10⁻⁶ × (10⁴)²/(2 × 2.5) = 250/5 = 50 °C. So T = 10 + 260 − 50 = 220 °C. Ignoring the source term gives 270 °C, which overestimates, since part of the surface heat flow is made inside the layer.

  4. Which of the following statements about the Earth’s heat are correct?

    1. The main heat-producing isotopes in the Earth are ²³⁸U, ²³⁵U, ²³²Th and ⁴⁰K
    2. Heat is carried mainly by conduction in the lithosphere and mainly by convection in the mantle below it
    3. In a layer of uniform heat production in steady state, heat flow increases with depth
    4. Mean heat flow through the continents is greater than through the ocean floors
    Show answer

    Answer: A — The main heat-producing isotopes in the Earth are ²³⁸U, ²³⁵U, ²³²Th and ⁴⁰K; B — Heat is carried mainly by conduction in the lithosphere and mainly by convection in the mantle below it

    U, Th and K are the principal radiogenic sources, and conduction dominates the rigid lithosphere while the mantle convects. With sources in the layer q(z) = q0 − A z, so heat flow decreases with depth. The ocean-floor mean of about 101 mW/m² exceeds the continental mean of about 65.

  5. Measured conductive heat flow on very young ocean floor near a ridge crest is lower than conductive-cooling models predict. What is the best explanation?

    1. Circulating seawater in the young, fractured crust carries heat away, so only part is conducted through the sediment
    2. The lithosphere is thicker under the ridge than elsewhere, so less heat is able to reach the surface by conduction there
    3. Oceanic crust contains no radioactive elements at all, so the ridges can produce less heat than the continents do
    4. Mantle upwelling under the ridges stops once the crust has formed, which cuts off the supply of heat to the surface
    Show answer

    Answer: A — Circulating seawater in the young, fractured crust carries heat away, so only part is conducted through the sediment

    Hydrothermal circulation removes heat by advection through fractures, so the sediment probe sees only the conducted part. The lithosphere is in fact thinnest at the ridge, the total heat loss is highest there, and upwelling continues as plates spread.

  6. Why is the acceleration due to gravity at the poles about 0.5 per cent greater than at the equator?

    1. The polar crust is denser than the equatorial crust, so it attracts bodies at the surface more strongly
    2. The Coriolis force adds to gravity at the poles and subtracts from it at the equator, as the Earth rotates
    3. The poles are nearer the centre because of the bulge, and there the centrifugal acceleration is zero
    4. Most of the Earth’s mass is concentrated at the poles, where the core lies closer to the surface
    Show answer

    Answer: C — The poles are nearer the centre because of the bulge, and there the centrifugal acceleration is zero

    The equator is about 21 km farther from the centre than the poles and also loses about 0.034 m/s² to centrifugal acceleration, while the poles lose none. Both effects raise polar gravity. The Coriolis force acts only on moving bodies and has no role in the static value of g.

  7. Which description of how gravitational acceleration g varies from the surface to the centre of the real Earth is correct?

    1. It decreases in direct proportion to the distance from the centre, falling linearly from the surface to zero at the centre, as in a sphere of uniform density
    2. It stays near 9.8 to 10 m/s² through most of the mantle, peaks near 10.7 m/s² at the core–mantle boundary, then falls almost linearly to zero at the centre
    3. It increases steadily with depth and is largest at the centre of the Earth, because the enclosed mass is greatest there and the pull grows as one goes deeper
    4. It falls abruptly to a very small value at the core–mantle boundary, because the outer core is liquid and a liquid cannot attract bodies
    Show answer

    Answer: B — It stays near 9.8 to 10 m/s² through most of the mantle, peaks near 10.7 m/s² at the core–mantle boundary, then falls almost linearly to zero at the centre

    At radius r, g depends on the mass enclosed, M(r), divided by r². Because the dense core holds so much mass, M(r)/r² stays high through the mantle and peaks at the core boundary, then falls toward zero at the centre. The linear fall is the uniform-sphere result and does not fit a dense core. Enclosed mass shrinks toward the centre, so g cannot be largest there, and a liquid attracts just as a solid does.

  8. An infinite horizontal rock slab 500 m thick has density 2,670 kg/m³. Take G = 6.674×10⁻¹¹ N m² kg⁻² and 1 mGal = 10⁻⁵ m/s². The gravitational attraction of the slab (the Bouguer slab effect), in mGal, to the nearest integer, is ____.

    Numerical answer — type the value.

    Show answer

    Answer: 56

    Δg = 2πGρh = 2π × 6.674×10⁻¹¹ × 2,670 × 500 = 5.598×10⁻⁴ m/s² = 55.98 mGal, so 56 mGal. Using πGρh, or forgetting the 2π, gives 28 or 8.9 mGal.

  9. The geoid is best defined as:

    1. The surface on which the value of g is the same everywhere, because the geoid is defined by constant gravity
    2. The smooth mathematical ellipsoid that best fits the overall shape of the Earth, flattened at the poles
    3. The equipotential surface of gravity and rotation that coincides with mean sea level
    4. The topographic surface of the continents and of the sea floor, as measured by levelling and by satellites
    Show answer

    Answer: C — The equipotential surface of gravity and rotation that coincides with mean sea level

    The geoid is a surface of constant potential, always perpendicular to the plumb line. The reference ellipsoid is the mathematical approximation, and g varies along the geoid by thousands of mGal, so it is not a surface of constant g.

  10. A gravity station lies on a vast flat plateau 1,000 m above sea level, underlain throughout by rock of density 2,670 kg/m³. Compared with the value at sea level at the same latitude, and ignoring every other effect, by about how much is the observed g smaller? Use the free-air gradient 0.3086 mGal/m, G = 6.674×10⁻¹¹ N m² kg⁻², treat the plateau as an infinite slab and take 1 mGal = 10⁻⁵ m/s².

    1. About 112 mGal, which is the attraction of the rock slab alone, taken as the whole effect
    2. About 197 mGal, the free-air loss reduced by the slab attraction
    3. About 309 mGal, which is the free-air loss alone, with the rock beneath the station ignored
    4. About 421 mGal, the free-air loss plus the attraction of the slab, since the slab adds to the loss
    Show answer

    Answer: B — About 197 mGal, the free-air loss reduced by the slab attraction

    Rising 1,000 m lowers g by 0.3086 × 1,000 = 308.6 mGal. The rock between the station and sea level pulls downward by 2πGρh = 2π × 6.674×10⁻¹¹ × 2,670 × 1,000 = 112.0 mGal. The net decrease is 308.6 − 112.0 = 196.6, about 197 mGal. Quoting the free-air loss alone ignores the slab, and adding the slab to the loss gets its sign wrong, since the extra rock increases g.

  11. Which pair gives the typical average thickness of continental crust and of oceanic crust, respectively?

    1. About 7 km and 35 to 40 km
    2. About 100 km and 35 km
    3. About 15 km and 3 km
    4. About 35 to 40 km and 7 km
    Show answer

    Answer: D — About 35 to 40 km and 7 km

    Continental crust averages 35 to 40 km, and oceanic crust about 7 km. The pair that lists 7 km first reverses them, and 100 km is closer to the thickness of the lithosphere than of the crust.

  12. In a complete ophiolite sequence, which unit corresponds to oceanic seismic Layer 3, the thickest part of the oceanic crust?

    1. Pillow basalt lavas
    2. Sheeted dyke complex
    3. Gabbro
    4. Peridotite
    Show answer

    Answer: C — Gabbro

    Layer 3, about 4 to 5 km thick with Vp of 6.5 to 7.0 km/s, is gabbro, the slowly cooled magma chamber. Pillow lavas and sheeted dykes together form Layer 2, and peridotite lies below the Moho in the mantle.

  13. Which of the following statements comparing continental and oceanic crust are correct?

    1. Continental crust is on average thicker and less dense than oceanic crust
    2. The floors of the main ocean basins are younger than about 200 Ma, while continental rocks reach about 4 Ga
    3. Oceanic crust is dominated by granite and granodiorite, like continental crust
    4. Oceanic crust is about 35 km thick, more than twice as thick as continental crust
    Show answer

    Answer: A — Continental crust is on average thicker and less dense than oceanic crust; B — The floors of the main ocean basins are younger than about 200 Ma, while continental rocks reach about 4 Ga

    Continental crust is thicker (35 to 40 km against about 7 km) and lighter (2.7 to 2.8 against 2.9 to 3.0 g/cm³), and since oceanic crust is subducted it is young while buoyant continental crust survives to 4 Ga. Oceanic crust is basalt and gabbro, not granite, and it is the thinner of the two.

  14. A normal-incidence seismic reflection survey records the Moho reflection at a two-way travel time of 11.0 s. The average P-wave speed of the crust is 6.4 km/s. The depth to the Moho, in km, to the nearest integer, is ____.

    Numerical answer — type the value.

    Show answer

    Answer: 35

    The wave goes down and back, so h = V T₂/2 = 6.4 × 11.0/2 = 35.2 km, which is 35 km. Forgetting to halve the two-way time gives 70 km.

  15. In Airy isostatic balance, normal crust at sea level is 35 km thick, the crustal density is 2,800 kg/m³ and the mantle density is 3,300 kg/m³. A plateau is underlain by crust 70 km thick. The elevation of the plateau predicted by the model, in km, to one decimal place, is ____.

    Numerical answer — type the value.

    Show answer

    Answer: 5.3

    The crust is T = 35 + h + r with r = h × 2,800/500 = 5.6 h, so 70 = 35 + 6.6 h and h = 35/6.6 = 5.30 km. Treating all 35 km of extra thickness as root gives 35/5.6 = 6.25 km, which forgets that the elevation itself is part of the crust.

  16. Which statement best describes the bulk composition of the continental crust?

    1. Felsic and granitic throughout, because the upper crust, which is the part that is exposed, is granitic
    2. Intermediate, about 60 wt% SiO₂, with felsic upper crust over more mafic lower crust
    3. Mafic and basaltic, the same as oceanic crust, because both kinds of crust float on the denser mantle
    4. Ultramafic, like peridotite, because the continental crust is derived directly from the mantle by melting
    Show answer

    Answer: B — Intermediate, about 60 wt% SiO₂, with felsic upper crust over more mafic lower crust

    The upper crust is felsic but the lower crust is mafic granulite, so the average is intermediate near 60 wt% SiO₂. Judging from the exposed granitic rocks overstates silica. Oceanic crust is mafic and thin, and an ultramafic crust would show Vp near 8 km/s, which is the mantle.

  17. Which of the following statements about gravity and the shape of the Earth are correct?

    1. One mGal equals 10⁻⁵ m/s²
    2. Gravity is greater at the poles than at the equator, by about 0.5 per cent
    3. The geoid is a surface on which the value of g is the same everywhere
    4. The geoid coincides exactly with the reference ellipsoid at every point
    Show answer

    Answer: A — One mGal equals 10⁻⁵ m/s²; B — Gravity is greater at the poles than at the equator, by about 0.5 per cent

    1 mGal is 10⁻³ cm/s² = 10⁻⁵ m/s², and polar g (about 9.83) exceeds equatorial g (about 9.78) by about 0.5 per cent. The geoid is an equipotential, along which g varies, and it departs from the ellipsoid by up to about ±100 m.