The outer shells of Earth - New Theory of the Earth - page 50

What is a Hotspot?
Morgan (1971) identified some 20 volcanic fea-
tures that he proposed were underlain by deep
mantle plumes equivalent in strength to the
hypothetical Hawaiian plume. Most were islands
or on land and most are now known to be on
major fracture zones. Modern seafloor maps
and global tectonic maps have many more
features, including fracture zones and fields of
seamounts, and there are many more active
or recently active volcanoes than were recog-
nized by Wilson and Morgan when they devel-
oped their ideas about hotspots and island
chains. Most proposed plumes/hotspots are in
tectonic locations that suggest lithospheric or
stress control, e.g., Yellowstone, Samoa, Afar,
Easter island, Louisville, Iceland, Azores and
Tristan da Cuhna. There is little evidence from
tomography, heat flow or magma temperatures
that the mantle under hotspots is particu-
larly hot [ mantleplumes, 'global hotspot
maps' ].
Wilson described his concept of hotspots as
follows; 'At a level of 400 to 700 km, the mantle
becomes opaque, so that heat slowly accumu-
lates until, due to local irregularities, cylindri-
cal plumes start to rise like diapirs in the upper
part of the mantle. These plumes reach the
surface, which they uplift, while their excess
heat gives rise to volcanism. The lavas at these
uplifts are partly generated from material ris-
ing from depths of several hundred kilome-
ters, which is thus chemically distinct from that
generated at shallow levels. The plumes are con-
sidered to remain steady in the mantle for
millions of years'. This hypothesis is very differ-
ent from Morgan's, and both are very different
from modern concepts of relatively weak, deep
mantle plumes, which are easily 'blown in the
mantle wind,' and that provide only a small frac-
tion of the Earth's heat flow and magma. The
common theme of hotspot hypotheses is that
they are caused by hot upwelling mantle from
a deep thermal boundary layer (TBL) well below
the upper mantle; the upwellings are active,
that is, they are driven by their own thermal
buoyancy rather than passively responding to
plate tectonics and subduction. A TBL instabil-
ity gives rise to narrow buoyant upwellings. The

concepts of 'plumes' and 'hotspots' have been
coupled since these early papers, even though vol-
canoes, volcanic chains, time-progressive volcan-
ism and swells can exist without plumes. Passive
upwellings such as those triggered by spread-
ing or by subduction are not considered to be
plumes, and neither are intrusions such as dikes
triggered by magma buoyancy, despite the fact
that all these phenomena are plumes in the fluid
dynamics sense.
The terms hotspot and plume refer to dif-
ferent concepts but they are generally used
interchangeably; they do not have well-defined
meanings that are agreed upon in the Earth sci-
ence community [see global hotspot maps ].
For some, hotspot is simply a region of magma-
tism that is unusual -- and sometimes not so
unusual -- in location, volume or chemistry, com-
pared with some segments of the ocean ridge
system. Geochemists use the term plume to refer
to any feature that has 'anomalous' geochem-
istry but anomalous is not used in any statisti-
cal or consistent sense. Geochemical plumes are
simply regions that provide basalts that differ
in some detail from what has been defined as
normal MORB; the phrase fertile blob could be
substituted for plume in isotope geochemistry
papers since there is no constraint on tempera-
ture, melting temperature or depth of origin. In
fluid dynamics a plume is a thermal upwelling or
downwelling. The word plume is now applied to
tomographic, geochemical and other 'anomalies'
that have little or no surface expression, thereby
decoupling the concepts of hotspots and melt-
ing anomalies from plumes. Some seismologists
use plume to refer to any region of the mantle
having lower than average seismic velocity. One
assumption behind this usage is that any low-
velocity structure is hot and buoyant. Early tomo-
graphic maps used red and other warm colors for
low seismic velocities and this was interpreted
by later workers as indicating that low-velocity
regions must be hot. A low-velocity region may,
however, be due to volatiles or a low melting
point or a different chemistry, and need not be
buoyant, or even hot compared with the sur-
rounding mantle. Dense eclogite sinkers from
delaminated continental crust, for example, can
show up as low shear wave velocity features.