Last modified December, 2013;
TC1: Lithosphere age and temperature
The site is being
Hosting by Yahoo Web Hosting
Number of page visits
The Lithosphere book by Artemieva. Cambridge
Artemieva I.M., 2011.
An interdisciplinary

Cambridge University Press,
794 pp., ISBN 9780521843966.

Download Contents and Preface
Download CV
Download papers
TC1: Global 1°x1° model for the continental lithosphere:
age, temperatures, and
implications for lithosphere secular evolution
Artemieva I.M.
416, 245-277, 2006
Data files can be downloaded
or requested from the author.
Please send me email :
indicating which data you are
interested in
and for which region.
Double-click on images to enlarge them
Surface heat flow measurements allow us to constrain the thermal state of the
upper mantle only for about 40% of the continents (
Figure 1), which is sufficient
to perform a statistically significant analysis of lithospheric geotherms for
continental terranes with different tectonic settings and different geological
ages. New compilation of crustal ages (
Figure 2) together with previously
reported geotherms for stable continental regions (Artemieva and Mooney,
2001) form the basis for this analysis. These data were supplemented by
xenolith P-T arrays and electrical conductivity profiles for cratonic regions. A
new global thermal model TC1 for the continental lithosphere is constrained on
a 1ox1o grid.
Map of heat flow measurements
Figure 1. Global heat flow data
coverage (updated after Pollack et al.,
heat flow measurements; diamonds –
locations of mantle xenoliths discussed
in the text.
Global 1 deg x 1 deg compilation of tectono-thermal ages of the continents
Age of the crust and lithosphere
Figure 2.  Geological ages of the continents on a
1ox1o grid (based on Goodwin, 1996; Fitzgerald,
2002; Condie, 2005, and numerous regional
crust-forming events (see Table 2), rather than ages
of the juvenile crust, and forms the basis for the global
thermal model for the continental lithosphere TC1.

Note: this map can be slightly different from the
published one due to on-going updates of the

A major assumption for the analysis is that lithospheric mantle has the same age as the overlying crust.
Download databases
Lithospheric thermal thickness linearly decreases with time from Mesoarchean to present (Fig. 3a). However, this correlation can be biased by the uneven
distribution of heat flow measurements as indicated by the lower value of area-weighted correlation (r=0.68) calculated on a 1ox1o grid
(Fig. 3b).
Contrary to the global trend,
lithosphere is thicker in young Archean than in old Archean (>3.0 Ga) cratons (Figure 3c). Thick lithospheric roots (250-350
km) are found solely for 3.0-2.5 Ga cratons.
  • In several Archean cratons regions with low heat flow, thick crust and thick lithosphere spatially coincide with paleoterrane boundary(ies). I propose
    that in such regions an exceptionally thick lithosphere has a relatively small lateral extent and was formed during Archean-early Proterozoic thrust-
    thickening collisions of pre-existing continental nuclei.
  • Heat from the mantle during episodes of high tectonic activity could be effectively diverted by such “lithospheric teeth”, forming belts of anorogenic
    magmatism at their edges.
Lithosphere thickness versus age
Global statistics: variations of lithosphere thickness with age:
Figure 3. Correlation between lithospheric thermal
thickness and geological age of terranes. The bars
(a) Global correlations for continents based on
individual heat flow measurements; gray area shows
the general trend of lithospheric thickness variations
with age (Artemieva and Mooney, 2001);
(b) Global correlations for continents, area-weighted
on a 1ox1o grid for terranes of different ages;
(c) correlations for Archean terranes
continental geotherms from heat flow and xenolith data
Global 1 deg x 1 deg thermal model for the continental lithosphere
Global statistics: continental geotherms from heat flow and xenolith data
Figure 4. Typical continental geotherms constrained by
heat flow data for stable regions Five groups of typical
geotherms include (from the coldest to the warmest):
  • (a) Archean terranes younger than 3.0 Ga;
  • (b) older Archean cratons (which includes mostly
    cratons of Gondwana-continents)and early
    Proterozoic terranes;
  • (c) reworked Archean cratons that have mantle
    temperatures similar to middle Proterozoic
    terranes (shown by line with bars);
  • (d) Paleozoic and late Proterozoic regions;
  • (e) Meso-Cenozoic regions.
Thin lines – conductive geotherms of Pollack and
Chapman (1977), values are surface heat flow in
who visits my web-site

view my analytics
continental geotherms from xenolith data
Figure 5. Xenolith P-T arrays confirm the results of the thermal model and suggest that there are two
groups of Archean cratons with significantly different thermal regimes.
  • Cratons with lower mantle temperatures follow a ca. 35-38 mW/m2 conductive geotherm and include
    Archean terranes of the northern hemisphere (Slave craton, Fennoscandia (central Finland and
    Arkhangelsk region), and Siberia).
  • Xenolith P-T arrays for Kaapvaal, Namibia, Superior and Wyoming cratons, and Somerset Island
    follow a 40-45 mW/m2 conductive geotherm.
  • The difference between the 37 and 42 mW/m2 geotherms implies ca. 80-100 km difference in the
    lithospheric thickness between the cratons in the two groups.
Global correlations between thermal regime, xenolith geotherms, and electrical structure of the cratonic upper mantle
lithospheric thickness from heat flow and xenolith geotherms
thickness of electric lithosphere
thickness of electric lithosphere vs heat flow
Figure 6. Correlation between lithospheric thickness constrained by heat flow and xenolith geotherms for stable continental terranes. Note that the
bars show the range of thickness estimates for different cratonic blocks, but not an uncertainty.  Because the deepest known kimberlite magmas originated at ca. 250 km depth,
xenolith and thermal constrains on lithospheric thickness diverge at greater depth. However, xenoliths from Finland derived from 240 km depth do not show shearing and
suggest lithospheric thickness greater than 240 km (shown by an arrow).

Figure 7. Electrical conductivity profiles for different cratonic regions of the world: Baltic shield, East European craton, Ukrainian Shield, Siberia,
India, Slave craton, and Superior Province. Also shown are synthetic conductivity curves calculated for conductive continental geo¬therms of 37 and 50 mW/m2 for the standard
olivine model SO2 (Constable et al., 1992) with 90% of forsterite (thick solid lines) and conductivity along a-axis of anisotropic wet olivine (1000 ppm H/Si) for mantle adiabats of
1300 to 1400 oC (Karato, 1990) (thick dashed lines). All cratonic conductivity curves fall between theoretical conductivity estimates for 37 and 50 mW/m2 conductive geotherms.

Figure 8. Correlation between surface heat flow and depth to the high-conductive layer in the mantle (electric asthenosphere).
The curve can be used as a proxy to the depth of the HCL for regions with a known heat flow.

The strong correlation between the thermal state and the age of the lithosphere allows us to build a new global thermal model TC1 of the continental
lithosphere on a 1ox1o grid, which shows strong thermal heterogeneity of the continental mantle. Maximum temperature anomalies (up to 800 oC) in
shallow mantle can produce seismic velocity anomalies of up to 3%. A map of the depth to the 550 oC isotherm provides a proxy for the thickness of the
magnetic crust and for the thickness of a mechanically strong layer in >200 Ma lithosphere with olivine rheology of the mantle.
Temperature at 50 km depth
Temperature at 100 km depth
Temperature at 150 km depth
Lithosphere thickness
Temperature at 50 km depth
Temperature at 100 km depth
Temperature at 150 km depth
Figure 9. Lithosphere thickness
Growth and preservation rate of the continental lithosphere; correlation with global tectonic events
Figure 10. Growth, preservation, and recycling of the continental lithosphere.

(after Condie, 2004, 2005), and Archean and Proterozoic anorthosites (data from Ashwal, 1993).

(b) Growth/preservation rate of the present-day continental lithosphere (crust + lithospheric
mantle) calculated on a 1ox1o grid from the global model of lithospheric thickness TC1 (Figure 9) with
age steps of 0.2 Ga (black line) and 0.3 Ga (gray line). The peak in lithospheric growth at ca. 1.7-2.0
Ga is a robust feature of the model. Note that this peak is followed by massive extraction of Proterozoic
anorthosites suggesting that they were produced by edge-driven convection and not by mantle plumes.

(c) Growth rate of the continental crust; dashed line – based on isotope ages of the juvenile crust
(data from Condie, 1998); solid line – based on integrated interpretation of seismic data on the crustal
structure and models of mantle melting (based on the average of the two models proposed by Abbott et
al., 2000).  Although there is a general agreement between Figures (b) and (c), the peak in the growth
rate of juvenile crust at 2.7-2.5 Ga, associated with superplume events, is not evident in the curves of
lithosphere growth/preservation.

(d) Volume of recycled lithosphere calculated as the difference between the volume of juvenile
lithosphere and the present-day volume of preserved continental lithosphere. The latter is based on the
TC1 model. The former is derived  from the volume of juvenile crust under an assumption that the ratio
of juvenile crust to juvenile lithospheric mantle produced by mantle differentiation is constant through the
calculated for the upper envelope of typical curves for the volume of juvenile crust (approximately the
upper of the curves of McLennan and Taylor (1982) and Abbott et al. (2002)); “Min” calculated for the
lower envelope (the lower of the curves of Abbott et al. (2002) and Veizer and Jansen (1985));  bold line
is based on the average of the two models proposed by Abbott et al., 2000).

1) A global map of lithospheric thickness (Figure 12), constrained by the TC1 model, was used to estimate the volume of the preserved continental lithosphere,
which is ca. 27.8(± 7.0)x10^9 km3 (excluding submerged terranes with continental crust such as oceanic plateaus and shelves). The average growth rate of the
continental lithosphere was ca. 5-8 km3/year in the Archean and twice higher at 2.1-1.7 Ga. About 50% of the present continental lithosphere was formed by 1.8 Ga.

2) The growth rate of the lithosphere since the Archean (as manifested by its present-day lithospheric volume for terranes of different ages) (
Figure 10b) does not
reveal a peak in lithospheric volume at 2.6-2.7 Ga as expected from growth curves for juvenile crust (
Figure 10c).
A nearly zero rate of lithosphere recycling (
Figure 10d) was calculated for the Archean-early Proterozoic lithosphere, reflecting stabilization of cratonic lithosphere
by the late Archean and its enhanced survivability.

3) The major peak in the rate of lithospheric growth correlates with one of the major world-wide recognized crust-forming episodes at 2.0-1.7 Ga and is followed by a
global extraction of Proterozoic massif-type anorthosites. I propose that large-scale variations in lithospheric thickness at cratonic margins and at paleoterrane
boundaries controlled anorogenic magmatism and, in particular, the extraction of Precambrian anorthosites which were produced by vigorous small-scale
convection at the margins of continental lithospheric keels formed at 2.0-1.7 Ga.
  • The hypothesis is based on the observed correlations
  • (a)  between the spatial distribution of Proterozoic anorthosites and lithospheric structure and
  • (b) between the time of their extraction and lithosphere growth rate.
More common and more voluminous occurrence of middle Proterozoic massive anorthosites in the northern hemisphere (where xenolith and thermal data suggest
an exceptionally thick lithosphere) supports the hypothesis that their emplacement can be associated with a presence of >250 km lithospheric roots.
Figure 8.
free visitor counters
International Lithosphere
Programme ILP
Irina M. Artemieva
Geology Section, IGN
University of Copenhagen
Øster Voldgade 10
Copenhagen DK-1350