Page 46 - Fundamentals of Physical Volcanology

9780632054435_4_002.qxd 12/10/2007 12:18PM Page 23

MAGMA GENERATION AND SEGREGATION 23

Cape Verde African Plate

Fernando
South American
Plate

Etendeka
St Helena
Martin Vas
Parana

Fig. 2.6 Hot spots near the mid-
Atlantic ridge that helped to initiate
spreading there. The Parana basalts
in South America and the Etendeka Tristan da Cunha
basalts in southwest Africa were both
erupted from the same hot spot, now
marked by Tristan da Cunha, before
Mid-Atlantic Ridge
the Atlantic ocean opened. (After fig. 3
Shona
in Duncan, R.A. and Richards, M.A.
(1991) Hotspots, mantle plumes,
flood basalts, and true polar wander.
Bouvet
Reviews of Geophysics, 29, 31–50.)
mantle material rises it experiences lower confining follows a path like that between points A and B in
pressures and so will expand slightly in volume as Fig. 2.1. Consider what happens to mantle material
it ascends. This expansion is adiabatic (i.e., occurs rising from point A to point B. The rise speed is

without the addition of any external heat) and so sufficiently great that the temperature declines only
causes a decrease in temperature of the rising slightly between points A and B, but during this
−1
mantle of ∼0.5–1.0°C km . In addition, the mater- ascent the reduction in confining pressure experi-
ial is moving into a zone of lower temperature enced by the rising mantle material means that its
and so there is the opportunity for heat loss to the melting temperature is considerably lower at depth
surrounding mantle. If the rise of the mantle mater- B than it was at depth A. At depth B the temperature
ial is slow then heat loss to the surroundings will of the rising mantle material matches its solidus
dominate and the rising mantle material will cool temperature and melting commences. Further asc-
sufficiently to ensure that, even at the progressively ent and reduction in confining pressure causes
lower pressures it experiences, no melting will further melting.
occur. However, if ascent is sufficiently rapid, con- Not all the material in the upwelling mantle
duction of heat to the surrounding mantle will be melts. Experimental studies in which mantle mate-
minimal and the cooling which occurs is limited to rial is melted at high pressures suggest that 20–
that caused by the adiabatic expansion of the 25% melting of “typical” mantle material produces
plume material. In practice it seems that most com- tholeiitic basalts like those produced at MORs and
monly the rise rate of mantle material, while not leaves behind a depleted mantle residuum from
strictly adiabatic, is sufficiently rapid to minimize which it would then be hard to produce further
loss by conduction and so the rising mantle material melts. Such studies thus support the idea that the