Roadmap | The review process | Home
   Transition zone

Variations in Global Transition Zone Thickness

G. R. Foulger1 & B. R. Julian2

1Dept. Geological Sciences, Univ. Durham, U.K.,

2U.S. Geological Survey, Menlo Park, CA 94025, U.S.A.,


The mantle has complex structure. The mantle transition zone (TZ) is the layer between two discontinuities in seismic wave-speed that lie at depths of approximately 410 km and 650 km [Anderson, 1989]. These discontinuities are polymorphic phase changes, caused by pressure-induced changes of crystal structure in certain minerals [Anderson, 1967]. In thermodynamic equilibrium, their depths depend on temperature and composition, and they may also be significantly affected by departures from equilibrium in subducted lithospheric slabs. Consequently, the detailed structure of the transition region contains the key to a number of geophysical problems [Birch, 1952] and is a subject of intense investigation.

Figure 16 from Presnall (1995)


The TZ was defined by Bullen as a diffuse region of high seismic wave-speed gradient extending from 400 to 1,000 km. In Bullen's nomenclature this was Region C and the lower mantle (Region D) started at 1,000 km. Birch suggested that the Repetti discontinuity near 1,000 km marked the top of the lower mantle and that high seismic wave-speed gradients are caused by polymorphic phase changes. The early models of Jeffreys and Gutenberg were smooth and had high wave-speed gradients without abrupt discontinuities, but in the 1960s it was discovered that there are abrupt jumps in seismic velocity at depths of approximately 400 and 650 km. During that decade, various investigations, including detailed studies of the travel times of both first- and later-arriving body waves, seismic array measurements of apparent velocities, observations of reflected waves such as precursors to the core phase P'P', and analysis of the dispersion of fundamental and higher-mode surface waves, all confirmed the existence of the discontinuities.

Thermodynamic considerations were used to argue that the discontinuities are abrupt phase changes of olivine to the spinel crystal structure, and then to a “post-spinel” phase, not chemical changes, and that the deeper one has a negative Clapeyron slope [Anderson, 1967]. This means that cold subducting material of the same composition as the surrounding mantle would depress the 650-km discontinuity, inhibiting vertical motion, and would change to the denser phase only after warming up or being forced to greater depth.
There is now confusion in the geochemical literature about whether 650 or 1,000 km depth is the boundary between the upper and lower mantles and whether there are chemical changes deeper than 1,000 km depth. Geochemical studies, and many convection simulations, assume that the 650-km phase change separates the “depleted convecting upper mantle” from the “primordial undegassed lower mantle”. Geodynamic modelers assume that if the 650-km discontinuity is not a chemical change then there can be no deeper chemical change and the mantle is chemically homogeneous. The TZ thus holds the key to whether there is whole-mantle or layered-mantle convection.
Unusually low temperatures, e.g., as expected in a downgoing slab, will warp the 410-km discontinuity up by about 8 km per 100 K, and the 650-km discontinuity down by about 5 km per 100 K. Conversely, unusually high temperatures will depress the 410-km discontinuity and elevate the 650-km discontinuity. This means that the TZ is predicted to thicken by about 13 km per 100 K where temperatures are low and to thin by a similar amount where temperatures are high [Bina & Helffrich, 1994]. The Clapeyron slopes are uncertain but most estimates for the total thickening/thinning of the TZ lie in the range 1 km per 6-8 K, or 12-17 km per 100 K.

Lateral temperature variations in normal mantle, resulting from plate tectonics and mantle convection, are predicted to be at least ~ 200 K, and thus thinning of the TZ of the order of 20-35 km as a result of deflection of both discontinuities is expected without plume heating. Plumes in the lower mantle need to be much hotter than this to become buoyant, rise through the high viscosity lower mantle, penetrate the TZ and erupt in mid-plate environments [Cordery et al., 1997].

Chemistry also affects the depths and sharpness of the discontinuities. Increasing FeO content decreases the depths of the discontinuities and generally increases their thicknesses. Water content is predicted to increase the abruptness of the 410 km discontinuity. The pyroxene-“garnet” transitions also contribute to velocity gradients in the TZ [Anderson, 1989].


Figure 5.10 from Davies (1999)

Several seismological techniques exist for determining the structure of the TZ, and its regional variation:

  1. Seismic rays from shallow earthquakes observed at distances between about 1,500 and 3,000 km turn in the transition zone, and are particularly sensitive to its structure. Studying the travel times (especially of later arrivals) and waveforms of these waves can reveal fine structural detail, but requires unusually dense instrumental coverage.
  2. Upward-traveling seismic waves are partially reflected downward by TZ discontinuities, and these echoes may appear on otherwise quiet portions of seismograms at distant stations. These waves take the form of “precursors” to surface-reflected seismic phases such as SS, multiple ScS, and P’P’.
  3. Similarly, upward-traveling waves generate converted phases that may be recognizable on seismograms recorded above the conversion point. Usually though, these waves reverberate to produce complicated seismograms, and a special “receiver-function” technique is used to interpret them.

Receiver functions

A widely used tool for studying the TZ is the somewhat inscrutably named “receiver function”, which is derived from seismograms of teleseismic earthquakes recorded at a three-component seismometer in an area of interest and describes the reverberation of P- and S-waves between boundaries beneath the seismometer [Ammon et al., 1990]. For nearly vertically incident P waves, the vertical-component seismogram approximates the incident waveform, whereas the horizontal-component seismograms are dominated by converted shear waves generated at interfaces beneath the station [Langston, 1979]. The receiver function is the mathematical operator that describes this process, and it is obtained by deconvolving the vertical-component seismogram from the two horizontal-component seismograms. Determining the structure beneath a station from a receiver function is a nonlinear inverse problem, which may be solved by computing the theoretical receiver function for an initial estimate of the structure and then iteratively perturbing the structure to improve the agreement between theory and observation.

The receiver-function method suffers from three major problems:

  1. It is sensitive to velocity discontinuities but insensitive to absolute wave speeds; “trading off” wave speeds and interface depths in a way that preserves the arrival times of reverberating waves has little effect on receiver functions. Independent constraints on absolute wave speeds from other sources are thus needed to provide starting models that are close to the truth. Surface-wave dispersion measurements are a convenient source of such information [e.g., Du & Foulger, 1999].

Vertical cross-section of receiver functions along a NW-SE profile across the eastern Snake River Plain, near the Yellowstone hotspot. Colors: zones of rapid increase (red) and decrease (blue) of seismic wave speed with depth. Black horizontal lines: nominal depths of 410- and 650-km discontinuities. From the website of Ken Dueker, Univ. Wyoming.

  1. Horizontal-component seismograms, which contain the converted waves, are frequently noisy. To improve signal/noise ratios, it is often necessary to “stack” (add up) seismograms of several earthquakes with similar locations. Another approach is to determine structures for closely-spaced arrays of stations where structural similarity is expected and can be used to guide data processing and interpretation. An example of the latter approach is shown in the figure at left.



  1. Current receiver-function inversion methods assume the structure beneath a station is laterally homogeneous. The observations usually show that it actually is not, because the receiver functions derived using earthquakes in different directions differ substantially. Comparing results from different earthquake source regions, and from nearby seismometers, can help in identifying cases of especially strong lateral variation.
Notwithstanding these problems, several hotspot locations have been studied in detail using receiver functions. If the TZ beneath hotspots is hot, as the plume hypothesis predicts, the TZ should thin as a result of downwarping of the 410-km discontinuity and upwarping of the 650-km discontinuity.


The following is an update on global TZ thickness and interpretations. The average thickness of the TZ is ~ 242 km, and uncertainties are typically ±7 km for good data and ±15 km for noisy data. Typical thicknesses beneath high-heat-flow areas are 220-230 km. The topology of the relevant phase diagrams predicts antisymmetry in the directions of deflection of the discontinuities for the cases of both cold downwellings and hot upwellings [Anderson, 1989]. However, the depths of the 410- and 650-km discontinuities are uncorrelated on a global scale, even when three-dimensional variations in wave speed are allowed for. This decorrelation is a robust global feature [Gu et al., 2003]. They also sometimes show steps, which makes interpretation in terms of temperature not straightforward. Some places show rapid lateral changes in TZ thickness that may indicate non-thermal effects. It must therefore be borne in mind that temperature may not be the only control. The expected effects of temperature on the depths of the discontinuities are also based on uncertain laboratory calibrations, may be in error by a factor of two, and chemical effects may be stronger than generally supposed.

Continents are typically underlain by a thick TZ and oceans by a thin TZ, in good agreement with the relative temperatures broadly expected there. There are also clear correlations in the case of subduction zones, where thickening of the TZ is generally observed. There is abundant tomographic evidence for the accumulation of slabs at the 650-km discontinuity [Fukao et al., 2001], which is expected to cool the TZ and the overlying and underlying mantle.

Thickness maps of the TZ. Residuals are interpolated using a spherical harmonic expansion up to degree 12. The long-wavelength features are fairly consistent among these maps, which are dominated by low-degree harmonics (from Gu & Dziewonski, 2001). See also Gu et al., 2001).


Click to enlarge

There is, however, no global correlation between TZ thickness and the locations of surface hotspots and lower-mantle “superplumes”. For example, TZ thickness is normal beneath southern Africa (245 km) and the East African Rift and Afar (244 ± 19 km), which are underlain by the South Atlantic superplume and the postulated Afar plume. TZ thickness beneath hotspots is generally within the range for normal oceans and often close to the global average. Some examples are Pitcairn [226 km; Niu et al., 2000], Eifel (227 km, flat 650-km discontinuity), Hawaii (228.8 ±16.8 km) and Easter Island (236.8 ±12.2 km). Neither hotspots at the surface, nor lower mantle superplumes, thus appear to correlated with anomalously high temperature in the TZ. The TZ thus appears to buffer the upper and lower mantles from each other [Anderson, 2001].

The observed topography on the discontinuities does not seem to be explicable by thermal effects to the extent expected. The observations are consistent with the decorrelation of seismic anomalies between the upper and lower mantles observed both in tomographic images and revealed by matched filtering using plate/slab reconstructions. A few anomalies do extend from the surface through the TZ and into the lower mantle, and it would be interesting to calculate if they are more numerous than would be expected by chance.

Detailed study of some specific regions have yielded surprises. The thinnest TZ region, 181 km, is found in Sumatra, where a thick accumulation of cold slabs is thought to exist in the TZ. Within a 1500 x 800 km region of the western USA, the thickness of the TZ varies from 220 to 270 km, with an average thickness of 246 ± 9 km. There is 20 to 30 km relief on each discontinuity, and no correlation is found with surface geology, topography or between the discontinuities. According to experimental studies, the 410-km discontinuity is expected to have 150% of the relief of the 650-km discontinuity, if thermal effects dominate, but the opposite is observed in the western USA.

In spite of these complications, TZ thickness may still prove to be a useful thermometer and an important part of any plan to map lateral variations of temperature in the mantle.


Bullen’s Region C may be terminated by a variable-depth discontinuity ranging from 800 to 1,200 km, which may be the Repetti discontinuity [Birch, 1952]. The 650-km discontinuity is undoubtedly mainly a result of phase changes in mantle silicates, and probably deepens as the temperature decreases. It appears to be a barrier to mantle convection [Hamilton, 2002] although slabs may penetrate it locally. The top of the lower mantle (Bullen’s Region D) is at about 1,000 km depth and there is evidence from the scattering of seismic waves that there may be a chemical boundary near that depth [Anderson, 2002; Niu & Kawakatsu, 1998]. Much of modern mantle geochemistry is based on the conjecture that the 650-km phase change is also a major chemical change, and that this is the boundary between the upper and lower mantles. Mantle geodynamics is also based on the assumption that if slabs can penetrate the phase change they will sink to the core-mantle boundary. The TZ therefore continues to be a critical region for investigation.

References cited

Recent Transition Zone Studies

  • Flanagan, M.P. and P.M. Shearer, Global mapping of topography on transition zone velocity discontinuities by stacking SS precursors, J. Geophys. Res., 103, 2673-2692, 1998.
  • Flanagan, M.P. and P.M. Shearer, Topography on the 410-km seismic velocity discontinuity near subduction zones from stacking of sS, sP, and pP precursors, J. Geophys. Res., 103, 21,165-21,182, 1998.
  • Flanagan, M.P. and P.M. Shearer, A map of topography on the 410-km discontinuity from PP precursors, Geophys. Res. Lett., 26, 549-552, 1999.
  • Shearer, P.M., M.P. Flanagan and M.A.H. Hedlin, Experiments in migration processing of SS precursor data to image upper mantle discontinuity structure, J. Geophys. Res., 104, 7229-7242, 1999.
  • Shearer, P.M. and M.P. Flanagan, Seismic velocity and density jumps across the 410- and 660-kilometer discontinuities, Science, 285, 1545-1548, 1999.
  • Shearer, P.M., Upper mantle seismic discontinuities, in Earth's Deep Interior: Mineral Physics and Tomography from the Atomic to the Global Scale, AGU Geophysical Monograph 117, 115-131, 2000.
  • Gilbert, Hersh J.; Sheehan, Anne F.; Dueker, Kenneth G.; Molnar, Peter, Receiver functions in the western United States, with implications for upper mantle structure and dynamics, J. Geophys. Res., 108, B5, 10.1029/2001JB001194, 2003.
  • Gilbert, Hersh., A. F. Sheehan, D. A. Wiens, K. G. Dueker, L. M. Dorman, J. Hildebrand, and S. Webb, Upper mantle discontinuity structure in the region of the Tonga Subduction Zone, Geophys. Res. Lett., 28, 1855-1858, 2001.
  • Gu, Y. J., Upper mantle transition zone: structure and topography of discontinuities, Ph.D. thesis, 227 pp, Harvard University, 2001.
  • Gu, Y. J., A.. M. Dziewonski, and G. Ekström, Simultaneous inversion for mantle shear velocity and topography of transition zone discontinuities, Geophys. J. Int., in press, 2003.
  • Antolik, M., Y. J. Gu, Adam M. Dziewonski, and G. Ekström, a new joint model of compressional and shear velocity in the mantle, Geophys. J. Int., in press, 2003.
  • Li, X., R. Kind, X. Yuan, S. V. Sobolev, W. Hanka, D.S. Ramesh, Y. J. Gu, and A. M. Dziewonski, Seismic detection of narrow strong oceanic plumes and relation to mantle transition zone temperature, Geophys. Res. Lett., 2002.
  • Gu, Y. J., and A. M. Dziewonski, Global variability of transition zone thickness, J. Geophys. Res., 107, B7, 2135, doi:10.1029/2001JB000489, 2002.
  • Gu, Y. J., A.. M. Dziewonski, and G. Ekström, Preferential detection of the Lehmann discontinuity beneath continents, Geophys. Res. Lett., 28, 4655-4658, 2001.
  • Gu, Y. J., and A.. M. Dziewonski, Variables in thickness of the upper mantle transition zone, 2nd OHP/ION Joint Sym. Contribution, edited by B. Romanowicz, K. Suyehiro, and H. Kawakatsu, pp175-180, 2001.
  • Gu, Y. J., A.. M. Dziewonski, W.-J. Su, and G. Ekström, Models of the mantle shear velocity and discontinuities in the pattern of lateral heterogeneities, J. Geophys. Res., 106, 11,169-11,199, 2001.
  • Antolik, M., G. Ekström , A. M. Dziewonski, Y. J. Gu, J. Pan, and L. Boschi, A new joint P and S velocity model of the mantle parameterized in cubic B-splines, 22nd Ann. DoD/DoE Seism. Res. Sym. Proc., 2, 15-23, 2000.
  • Gu, Y. J., A. M. Dziewonski, and C. B. Agee, Global de-correlation of the topography of transition zone discontinuities, Earth Planet. Sci. Lett., 157, 57-67, 1998.
  • Li, X., Kind, R., Yuan, X., Sobolev, S.V., Hanka, W., Ramesh, D.S., Gu, Y. & Dziewonski, A.M., Seismic observation of narrow plumes in the oceanic upper mantle, Geophys. Res. Lett., 30, 10.1029/2002GL015411, 2003.

Discovery papers: The history of transition zone studies

  • Niazi, M., and Anderson, D. L., Upper mantle structure of western North America from apparent velocities of P waves, J. Geophys. Res., 70, 4633-4640, 1965.
  • Anderson, D. L., Recent evidence concerning the structure and composition of the Earth’s mantle, Physics and Chemistry of the Earth, 6, 1-131, Pergamon Press, Oxford, 1966.
  • Toksoz, M., and Anderson, D. L., Phase velocities of long-period surface waves and structure of the upper mantle, 1. Great circle Love and Rayleigh wave data, J. Geophys. Res., 71, 1649-1658, 1966.
  • Anderson, D. L., Latest information from seismic observations, Ch. 12, in The Earth’s Mantle, Academic Press Inc., London, p. 355-420, 1967.
  • Anderson, D. L., Phase changes in the upper mantle, Science, 157, 1165-1173, 1967.
  • Johnson, L., Array measurements of P velocities in the upper mantle, J. Geophys. Res., 72, 6309-6325, 1967.
  • Tokoz, M. N., Chinnery, M. A., and Anderson, D. L., Inhomogeneities in the earth’s mantle, Geophys. J. Royal astron. Soc., 13, 31-59, 1967.
  • Julian, B. R., and Anderson, D. L., Travel times, apparent velocities and amplitudes of body waves, Bull. Seismol. Soc. Am., 58, 339-366, 1968.
  • Archambeau, C.B. et al., Fine structure of the upper mantle, J. Geophys. Res., 74, 5825, 1969.
  • Whitcomb, J. H., and Anderson, D. L., Reflection of P’P’ seismic waves from discontinuities in the mantle, J. Geophys. Res., 75, 5713-5728, 1970.
  • Anderson, D.L., and J.D. Bass, Transition region of the Earth's upper mantle, Nature, 320, 321-328, 1986.
last updated March 2nd, 2004