Skip to content

The Geomagnetic Field and Its Variations

CSIR NET Earth Sciences Part C Geophysics: field elements, Gauss coefficients, dipole field, IGRF, secular variation, geodynamo, Sq, storms, Dst and pulsations.

  • 20 min read
  • Updated
On this page · 14 sections
  1. Prerequisites
  2. The Hook: Gauss Proves the Source Is Inside
  3. Elements and Sources
  4. Spherical Harmonic Description
  5. Variations of Internal Origin
  6. The Geodynamo
  7. Variations of External Origin
  8. Memory Tricks
  9. Don’t Confuse
  10. Common Mistakes
  11. Practice
  12. Quick Revision Table
  13. Key Takeaways
  14. What’s Next

Prerequisites

The Hook: Gauss Proves the Source Is Inside

Elements and Sources

The seven elements

ElementMeaningRelations
\( F \)total intensity\( F^2 = H^2 + Z^2 = X^2 + Y^2 + Z^2 \)
\( H \)horizontal intensity\( H = F\cos I \)
\( Z \)vertical intensity (positive down)\( Z = F\sin I \)
\( X \)north component\( X = H\cos D \)
\( Y \)east component\( Y = H\sin D \)
\( D \)declination (east positive)\( \tan D = Y/X \)
\( I \)inclination (dip, positive down)\( \tan I = Z/H \)

Field strength at the surface ranges from about 22,000 nT (South Atlantic) to about 67,000 nT (near the south magnetic dip pole). Observatories record \( H, D, Z \) (or \( X, Y, Z \)) continuously; plots against time are magnetograms.

Where the field comes from

SourceShare of surface fieldSpatial / temporal character
Core (main field)most of the field (dominant up to spherical-harmonic degree about 13)large scale; changes over years to millennia
Crust (lithospheric field)up to hundreds or thousands of nT locallyshort wavelength; static on human time scales; only rocks cooler than their Curie temperature contribute
External (ionosphere, magnetosphere)tens of nT quiet, hundreds to thousands in stormsseconds to the 11-year solar cycle
Induced (currents in Earth and oceans driven by external variations)fraction of the external partfollows external variations; basis of magnetotelluric and geomagnetic depth sounding

Because crustal magnetisation exists only above the Curie isotherm (about 580 °C for magnetite), the crustal field comes from the top few tens of kilometres at most.

Spherical Harmonic Description

In a source-free region (between the crust and the ionosphere, to a good approximation), \( \mathbf{B} = -\nabla V \) and \( \nabla^2 V = 0 \). The solution in colatitude \( \theta \), longitude \( \lambda \) and radius \( r \) is

$$V = a\sum_{n=1}^{N}\sum_{m=0}^{n}\left(\frac{a}{r}\right)^{n+1}\left[g_n^m\cos m\lambda + h_n^m\sin m\lambda\right]P_n^m(\cos\theta) + a\sum_{n,m}\left(\frac{r}{a}\right)^{n}\left[q_n^m\cos m\lambda + s_n^m\sin m\lambda\right]P_n^m(\cos\theta)$$
  • \( a = 6{,}371.2 \) km (the geomagnetic reference radius); \( P_n^m \) are Schmidt quasi-normalised associated Legendre functions.
  • \( g_n^m, h_n^m \) (Gauss coefficients, in nT) describe internal sources; \( q_n^m, s_n^m \) describe external sources. Gauss separated the two by using both horizontal and vertical components.
  • There is no \( n = 0 \) term: no magnetic monopole.
  • Internal terms fall off as \( r^{-(n+2)} \) in field strength. The higher the degree, the faster it decays upward, which is why the crustal (high-degree) field is weak at satellite altitude and the core field at the surface is dominated by low degrees.

The dipole terms

CoefficientMeaningApproximate value (IGRF, 2020, rounded)
\( g_1^0 \)axial dipole−29,405 nT
\( g_1^1 \)equatorial dipole along 0° longitude−1,451 nT
\( h_1^1 \)equatorial dipole along 90° E+4,653 nT

\( g_1^0 \) is negative because the dipole points towards geographic south: the field lines leave the southern hemisphere and enter the northern, so a compass’s north-seeking end points (roughly) north.

From the three dipole coefficients:

$$B_0 = \sqrt{(g_1^0)^2 + (g_1^1)^2 + (h_1^1)^2} \qquad \cos\theta_0 = \frac{-g_1^0}{B_0} \qquad M = \frac{4\pi a^3}{\mu_0}B_0$$

With the values above: \( B_0 \approx 29{,}806 \) nT, dipole tilt \( \theta_0 \approx 9.4° \), dipole moment \( M \approx 7.7\times 10^{22} \) A m\(^2\). The geomagnetic north pole (where the dipole axis cuts the northern hemisphere) is at about 80.6° N, 72.7° W.

Three kinds of “pole”

PoleDefined byAntipodal?
Geographicrotation axisyes
Geomagneticbest-fitting geocentric dipole axisyes, by construction
Dip (magnetic)where the observed field is vertical (\( I = \pm 90° \))no, because of the non-dipole field

The geomagnetic coordinate system (geomagnetic latitude and longitude) is built on the geomagnetic poles. The dip equator (\( I = 0 \)) crosses the southern part of peninsular India (Thumba, near Thiruvananthapuram, lay on it when the rocket range was sited there in the 1960s), which is why that region hosts studies of the equatorial electrojet. Its exact position drifts with secular variation.

The dipole field

At geomagnetic latitude \( \lambda \) on a sphere of radius \( a \):

$$Z = 2B_0\sin\lambda \qquad H = B_0\cos\lambda \qquad \boxed{\tan I = 2\tan\lambda} \qquad \boxed{F = B_0\sqrt{1 + 3\sin^2\lambda}}$$
  • Field at the poles is twice that at the equator.
  • A dipole field line satisfies \( r = L\,a\cos^2\lambda \), where \( L \) is the equatorial crossing distance in Earth radii (the L-shell). The line that reaches \( 4a \) at the equator meets the surface at \( \lambda = 60° \).

Non-dipole field, power spectrum and the IGRF

Subtracting the dipole leaves the non-dipole field, about a tenth of the surface field, with continental-scale features. The Lowes–Mauersberger spectrum, \( R_n = (n + 1)\sum_m[(g_n^m)^2 + (h_n^m)^2] \), falls steeply with degree up to about \( n = 13 \) (core field) and then flattens (crustal field). Continuing the core part down to the core–mantle boundary turns the steep fall into a nearly flat (“white”) spectrum, one way of locating the source depth.

The International Geomagnetic Reference Field (IGRF) is a set of Gauss coefficients agreed internationally under IAGA and revised every five years, with a predicted secular-variation model for the following five years. It is subtracted from survey data to obtain crustal magnetic anomalies.

Variations of Internal Origin

Secular variation

FeatureDescription
Westward driftnon-dipole features move west at roughly 0.2° of longitude per year on average (first suggested by Halley in the 1690s); not uniform, and not all features drift
Dipole decaythe dipole moment has been decreasing since systematic measurements began in the nineteenth century
Pole motionthe north dip pole has migrated across the Canadian Arctic towards Siberia, accelerating in recent decades
South Atlantic Anomalya growing region of weak field; trapped radiation comes closest to Earth there
Geomagnetic jerksabrupt changes (over about a year) in the trend of secular variation, seen in observatory records; a well-known one around 1969
Excursions and reversalslarge departures over thousands of years; treated with palaeomagnetism on the next page

The westward drift gives a rough speed for flow at the top of the core. At the core radius \( r_c = 3{,}480 \) km, 0.2° per year is

$$v = \frac{0.2\pi}{180}\times\frac{3.48\times 10^6\ \text{m}}{3.156\times 10^7\ \text{s}} \approx 3.8\times 10^{-4}\ \text{m s}^{-1}$$

about 0.4 mm s\(^{-1}\), or roughly 10 km per year: fast compared with plate motions or mantle convection by a factor of about \( 10^5 \).

The Geodynamo

Why not a permanent magnet or a fossil field?

  • Temperature: below a few tens of kilometres the Earth is hotter than the Curie temperature of any common magnetic mineral, so no permanent magnetisation can exist in the deep interior.
  • Ohmic decay: a field left in a conducting sphere of radius \( L \) and conductivity \( \sigma \) decays with time constant
$$\tau \approx \frac{\mu_0\sigma L^2}{\pi^2}$$

For the core (\( L = 3.48\times 10^6 \) m, \( \sigma \) of order \( 10^5 \)–\( 10^6 \) S m\(^{-1}\)) this is of order \( 10^4 \) years. Palaeomagnetism shows a field for at least 3 billion years, so it must be continuously regenerated.

The self-exciting dynamo

Moving conducting fluid in a magnetic field induces currents that sustain the field. The governing induction equation is

$$\frac{\partial\mathbf{B}}{\partial t} = \nabla\times(\mathbf{v}\times\mathbf{B}) + \eta\nabla^2\mathbf{B} \qquad \eta = \frac{1}{\mu_0\sigma}$$

The ratio of the advection term to the diffusion term is the magnetic Reynolds number

$$\boxed{R_m = \mu_0\sigma vL}$$

A dynamo needs \( R_m \) well above about 10–100. With \( v \approx 4\times 10^{-4} \) m s\(^{-1}\), \( L \approx 2.26\times 10^6 \) m (outer-core thickness) and \( \sigma = 5\times 10^5 \) S m\(^{-1}\), \( R_m \approx 570 \).

IngredientRole
Liquid iron outer coreconducting, convecting fluid
Thermal convectionheat flow across the core–mantle boundary
Compositional convectionlight elements released as the inner core freezes rise through the outer core
Coriolis forceorganises flow into columns parallel to the rotation axis, aligning the dipole roughly with the spin axis
\( \omega \)-effectdifferential rotation shears poloidal field into toroidal field
\( \alpha \)-effecthelical (twisting) flow converts toroidal field back to poloidal

Cowling’s theorem (1933): an axisymmetric magnetic field cannot be maintained by dynamo action. A dynamo-generated field must have non-axisymmetric parts, consistent with the observed tilt and non-dipole field. Numerical dynamo models since the mid-1990s (Glatzmaier and Roberts, 1995) have produced self-sustaining fields with spontaneous reversals.

Variations of External Origin

The setting: ionosphere and magnetosphere

  • The solar wind compresses the field on the dayside and stretches it into a long magnetotail on the night side. The boundary, the magnetopause, lies about 10 Earth radii upstream; a bow shock forms outside it.
  • The ionosphere (roughly 60 km to over 1,000 km) is conducting. In the E region (about 90–150 km) tidal winds drive charged particles across field lines: an atmospheric dynamo.
  • Trapped energetic particles form the Van Allen belts; westward-drifting ions and eastward-drifting electrons in the inner magnetosphere make the ring current.

The main external variations

VariationPeriod / durationTypical sizeSource
Sq (solar quiet daily)24 h and harmonics; daysidetens of nTE-region dynamo current vortices, one in each hemisphere on the sunlit side
L (lunar daily)lunar day, mainly semi-diurnala few nTlunar tidal winds in the ionosphere
Equatorial electrojet (EEJ)daytime, within a few degrees of the dip equatorenhances Sq in \( H \) several-foldeastward current concentrated where the field is horizontal (Cowling conductivity)
Magnetic storm1 to several daystens to hundreds of nT decrease in \( H \) at low latitudesring current intensification after a solar-wind disturbance
Substorm (bay)1–3 h, night side, high latitudehundreds to over 1,000 nTauroral electrojets fed by the magnetotail
Pulsations0.2 s to 10 minfractions of nT to tens of nTmagnetohydrodynamic waves in the magnetosphere
27-day recurrencesolar rotationmodulates activityrecurrent high-speed streams
Solar-cycle variationabout 11 yearsmodulates everythingsolar activity cycle

Anatomy of a magnetic storm

graph LR
    A["Solar wind shock or
coronal mass ejection
reaches Earth"] --> B["Sudden commencement
H jumps up by tens of nT
(magnetopause compressed)"] B --> C["Initial phase
H stays raised
for hours"] C --> D["Main phase
H falls sharply,
tens to hundreds of nT
(ring current grows)"] D --> E["Recovery phase
H returns over
one to several days
(ring current decays)"]

Not every storm starts with a sudden commencement; storms driven by recurrent high-speed streams often begin gradually. The intensity is measured by the minimum Dst.

Pulsation classes

Continuous (Pc)Period (s)Irregular (Pi)Period (s)
Pc10.2–5Pi11–40
Pc25–10Pi240–150
Pc310–45
Pc445–150
Pc5150–600

Activity indices

IndexMeasuresBuilt from
Kdisturbance at one observatoryquasi-logarithmic 0–9 scale of the range in each 3-hour interval, after removing the quiet-day curve
Kpplanetary disturbancestandardised K values from a network of mid-latitude observatories, 3-hourly, in thirds (0o, 0+, 1−, …, 9o)
ap / Apthe same on a linear scale3-hourly ap; daily average Ap
Dstring-current strengthhourly mean \( H \) disturbance from a few low-latitude observatories, corrected for latitude; negative during storms
AE (AU, AL)auroral electrojet strength\( H \) from a ring of auroral-zone observatories; substorms

Induction and depth of penetration

External variations induce currents in the conducting Earth. A variation of period \( T \) penetrates to about one skin depth

$$\delta = \sqrt{\frac{2\rho}{\mu_0\omega}} \approx 503\sqrt{\rho T}\ \text{m}$$

(\( \rho \) resistivity in \( \Omega \) m, \( T \) in s). In a 100 \( \Omega \) m half-space, a 1-hour variation reaches about 300 km and a daily variation about 1,480 km. Longer periods probe deeper: the principle of geomagnetic depth sounding and magnetotellurics.

Memory Tricks

Don’t Confuse

ThisvsThat
Geomagnetic pole (dipole axis, antipodal)Dip pole (where \( I = 90° \), not antipodal)
Geomagnetic latitude (from the dipole)Geographic latitude (from the spin axis)
Secular variation (internal, years to centuries)Sq and storms (external, hours to days)
Magnetic storm (global; ring current; Dst)Substorm (auroral zone; electrojets; AE)
Kp (3-hourly, quasi-logarithmic)Ap (daily, linear) and Dst (hourly, nT)
Sq (solar daily; tens of nT)L (lunar daily; a few nT)
Main field: core, low degreeCrustal field: lithosphere above Curie isotherm, high degree

Common Mistakes

Practice

All questions below are original practice questions written for this page. They are not from any CSIR NET paper. Take \( \mu_0 = 4\pi\times 10^{-7} \) H m\(^{-1}\) and \( a = 6{,}371.2 \) km where needed.

Level 1: Foundation

  1. Write the relation between inclination and geomagnetic latitude for a dipole.
  2. Which current system is measured by the Dst index?
  3. Give one reason why the main field cannot be due to permanently magnetised rock in the deep interior.
Answers
  1. \( \tan I = 2\tan\lambda \).
  2. The ring current in the inner magnetosphere.
  3. Temperatures below a few tens of kilometres exceed the Curie temperature of magnetic minerals (also: a fossil field would decay ohmically in about \( 10^4 \) years).

Level 2: Part B Style

Practice Q1. A field model has \( g_1^0 = -29{,}405 \) nT, \( g_1^1 = -1{,}451 \) nT and \( h_1^1 = 4{,}653 \) nT. The tilt of the dipole axis from the rotation axis and the dipole moment are closest to:

(a) 9.4°, \( 7.7\times 10^{22} \) A m\(^2\) (b) 2.8°, \( 7.5\times 10^{22} \) A m\(^2\) (c) 9.4°, \( 7.7\times 10^{15} \) A m\(^2\) (d) 15°, \( 3.0\times 10^{22} \) A m\(^2\)

Answer: (a) 9.4°, 7.7 x 10^22 A m^2

\( B_0 = \sqrt{29405^2 + 1451^2 + 4653^2} = 29{,}806 \) nT.

Tilt: \( \cos\theta_0 = 29405/29806 \Rightarrow \theta_0 = 9.4° \).

\( M = \frac{4\pi a^3}{\mu_0}B_0 = \frac{(6.3712\times 10^6)^3\times 2.9806\times 10^{-5}}{10^{-7}} = 7.71\times 10^{22} \) A m\(^2\).

(b) uses only \( g_1^1 \) for the tilt; (c) forgets \( 4\pi/\mu_0 = 10^7 \).

Practice Q2. For a geocentric dipole with equatorial field 30,000 nT, the total field and inclination at geomagnetic latitude 20° are:

(a) 31,900 nT, 20.0° (b) 34,870 nT, 36.1° (c) 34,870 nT, 20.0° (d) 60,000 nT, 36.1°

Answer: (b) 34,870 nT, 36.1°

\( F = 30000\sqrt{1 + 3\sin^2 20°} = 30000\sqrt{1 + 3(0.1170)} = 30000\times 1.1623 = 34{,}870 \) nT.

\( \tan I = 2\tan 20° = 0.7279 \Rightarrow I = 36.05° \).

Practice Q3. An observatory records \( H = 32{,}000 \) nT and \( Z = 24{,}000 \) nT. If the field were that of a geocentric dipole, the geomagnetic latitude would be:

(a) 10.3° (b) 20.6° (c) 36.9° (d) 53.1°

Answer: (b) 20.6°

\( \tan I = Z/H = 0.75 \Rightarrow I = 36.87° \); \( \tan\lambda = \tan I/2 = 0.375 \Rightarrow \lambda = 20.6° \). (Also \( F = 40{,}000 \) nT.) (c) is the inclination itself.

Practice Q4. Match each variation with its main source.

VariationSource
P. Sq1. Ring current
Q. Storm main phase2. Auroral electrojets
R. Substorm bay3. E-region dynamo
S. Secular variation4. Outer-core flow

(a) P-3, Q-1, R-2, S-4 (b) P-1, Q-3, R-2, S-4 (c) P-3, Q-2, R-1, S-4 (d) P-4, Q-1, R-2, S-3

Answer: (a) P-3, Q-1, R-2, S-4

Sq: ionospheric E-region dynamo. Storm main phase: ring current. Substorm: auroral electrojets. Secular variation: core flow (internal).

Level 3: Part C Style

Practice Q5. Non-dipole features drift westward at 0.2° per year. Take the core radius as 3,480 km, outer-core thickness 2,260 km and core conductivity \( 5\times 10^5 \) S m\(^{-1}\). Which statements are correct?

  1. The implied flow speed at the top of the core is about 0.4 mm s\(^{-1}\).
  2. The magnetic Reynolds number is of order \( 10^2 \)–\( 10^3 \), large enough for dynamo action.
  3. The free ohmic decay time of a field in the core, \( \mu_0\sigma r_c^2/\pi^2 \), is about \( 2\times 10^4 \) years.
  4. Because \( R_m \) is large, an axisymmetric field could be self-sustaining.

(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 true: \( v = (0.2\pi/180)\times 3.48\times 10^6/3.156\times 10^7 = 3.85\times 10^{-4} \) m s\(^{-1}\).
  • 2 true: \( R_m = 4\pi\times 10^{-7}\times 5\times 10^5\times 3.85\times 10^{-4}\times 2.26\times 10^6 \approx 550 \)–570.
  • 3 true: \( \tau = \frac{4\pi\times 10^{-7}\times 5\times 10^5\times(3.48\times 10^6)^2}{\pi^2} = 7.7\times 10^{11} \) s \( \approx 2.4\times 10^4 \) yr.
  • 4 false: Cowling’s theorem forbids an axisymmetric dynamo, whatever \( R_m \) is.

Practice Q6. Hourly Dst values (nT) during a disturbance:

Hour0146810142030
Dst+5+32−40−120−165−135−91−50−18

Assuming the recovery after hour 8 is exponential, which statement is correct?

(a) The event is a moderate storm with a recovery time constant of about 3 h. (b) The event is an intense storm; the jump at hour 1 reflects magnetopause compression; the recovery time constant is about 10 h. (c) The positive values at hour 1 show that the ring current was strongest then. (d) The recovery time constant is about 22 h, the time for Dst to return to −18 nT.

Answer: (b)
  • Minimum Dst −165 nT, below −100 nT: an intense storm on the usual classification.
  • The rise to +32 nT at hour 1 is the sudden commencement: solar-wind pressure compresses the magnetopause, increasing \( H \). It is not a ring-current effect, so (c) is wrong.
  • Recovery: \( \text{Dst}(t) = -165\,e^{-(t - 8)/\tau} \). At hour 20, \( -50 = -165\,e^{-12/\tau} \Rightarrow \tau = 12/\ln(3.3) = 10.0 \) h. Checking hour 14: \( -165e^{-0.6} = -91 \). Consistent.
  • (d) confuses the total duration with the e-folding time.

Practice Q7. A region has a uniform resistivity of 100 \( \Omega \) m. Which statements are correct?

  1. A variation with a 1-hour period penetrates to about 300 km.
  2. The Sq variation (24-hour period) penetrates to about 1,480 km.
  3. Pc3 pulsations (period about 30 s) sample deeper structure than Sq.
  4. If the resistivity were 10 \( \Omega \) m, the daily variation would penetrate only about 470 km.

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

Answer: (a) 1, 2 and 4 only

\( \delta \approx 503\sqrt{\rho T} \) m.

  • 1 true: \( 503\sqrt{100\times 3600} = 503\times 600 = 302 \) km.
  • 2 true: \( 503\sqrt{100\times 86400} = 503\times 2939 = 1{,}479 \) km.
  • 3 false: shorter periods penetrate less deeply (about 28 km for 30 s at 100 \( \Omega \) m).
  • 4 true: \( 503\sqrt{10\times 86400} = 468 \) km.

Quick Revision Table

IdeaFact or formula
Elements\( H = F\cos I \), \( Z = F\sin I \), \( X = H\cos D \), \( Y = H\sin D \)
Surface intensityabout 22,000–67,000 nT
Potential\( V = a\sum(a/r)^{n+1}[g\cos m\lambda + h\sin m\lambda]P_n^m \); \( \mathbf{B} = -\nabla V \)
Gauss (1838–39)internal and external parts separable; main field internal
Dipole (IGRF 2020, rounded)\( g_1^0 \approx -29{,}405 \) nT; tilt about 9.4°; \( M \approx 7.7\times 10^{22} \) A m\(^2\)
Moment\( M = 4\pi a^3B_0/\mu_0 \)
Dipole relations\( \tan I = 2\tan\lambda \); \( F = B_0\sqrt{1 + 3\sin^2\lambda} \); \( r = La\cos^2\lambda \)
Polesgeomagnetic poles antipodal; dip poles are not
Spectrumcore dominant to degree about 13; crust beyond
Westward driftabout 0.2° per year; core flow about 0.4 mm s\(^{-1}\)
Ohmic decay\( \tau \approx \mu_0\sigma L^2/\pi^2 \sim 10^4 \) yr
Dynamo\( R_m = \mu_0\sigma vL \) of order hundreds; Cowling: no axisymmetric dynamo
Sq / LE-region dynamo, tens of nT / lunar tides, a few nT
EEJalong dip equator; enhances \( H \) by day
StormSC (H up), main phase (H down, ring current), recovery
IndicesK, Kp (3-hourly), Ap (daily), Dst (hourly, ring current), AE (auroral)
PulsationsPc1–Pc5 (0.2–600 s); Pi1–Pi2 (1–150 s)
Skin depth\( \delta \approx 503\sqrt{\rho T} \) m

Key Takeaways

What’s Next

Advertisement
Advertisement