Geology Reference
In-Depth Information
accepted kinematic models for the modern
plates, which represents a refinement of a
model published in 1990, known as NUVEL-
1 (Northwestern University VELocity model ver.
1, DeMets et al.
1990
). In this model, the average
spreading rates used in the least squares fitting
procedure are determined through the analysis
of marine magnetic anomalies spanning the last
3.2 Myrs. The model includes 12 large plates,
and the components of the Euler vectors are
expressed in a reference frame fixed to the Pacific
plate. The difference between the two versions is
in the geomagnetic polarity time scales used to
analyse the marine magnetic anomalies during
the determination of the spreading rates, so
that the angular velocities of NUVEL-1A are
95.62 % of the corresponding velocities listed in
NUVEL-1.
One of the main problems of the classic mod-
els is represented by the very different time inter-
vals associated with the input data. The spreading
rates along the world's mid-ocean ridges, which
are estimated through the analysis of marine mag-
netic anomalies, represent averages over the last
3.2 Myrs. These averages strongly depend from
the choice of a geomagnetic polarity time scale.
Conversely, earthquake slip vectors average di-
rections of relative motions over much shorter
time intervals (decades to centuries). Another
problem is represented by the relatively small
number of plates that are considered in these
models, which limits their capability to repre-
sent the internal deformation of some continents.
Therefore, there is not much surprise in seeing
that inconsistencies often emerge when the linear
velocities predicted on the basis of the Euler
vectors are compared to velocities estimated from
Global Positioning System (GPS) techniques and
other geodetic methods. In fact, the latter data
are consistent averages performed over a few
decades, which are not necessarily representative
of the long-term geological processes. Finally,
the most serious issue of NUVEL-1A and its
predecessors is probably the failure to satisfy
the closure rule (Eq.
2.35
) along some three-
plate circuits. In particular, NUVEL-1A does not
satisfy Eq.
2.36
at the Galapagos triple junction
(Pacific-Cocos-Nazca circuit) and at the Bouvet
triple junction (Africa-South America-Antarctic
circuit) at the desired level of confidence.
A major improvement to NUVEL-1A, which
tries to overcome the difficulties mentioned
above, has been proposed in recent times by
DeMets et al. (
2010
). The new model, which
has been called MORVEL (Mid-Ocean Ridge
VELocity), extends the data set to the 25 plates
shown in Fig.
2.38
. With respect to the system
of 23 plates shown in Fig.
2.16
, this model
decomposes the eastern part of Africa in two
sub-plates (Lwandle and Somalia), separates
two sub-plates (Capricorn, and Macquarie)
from Australia, introduces the Yang-Tze plate
in eastern Asia and Sur in the South Atlantic,
but incorporates the Anatolian block in Eurasia,
Easter in Nazca, and the Okhotsk plate in N.
America. Using Eq.
2.34
, we see that this
model includes 46 triple junctions and 69 plate
boundaries, 23 of which must be free boundaries.
Differently from its predecessors, MORVEL is
based on few earthquake slip directions. In this
model, about 75 % of the input data are sea floor
spreading rates and strikes of transform faults.
The very limited usage of earthquake slip vectors
(
2 % of the total data set) has minimized the
possibility of biased estimates of relative velocity
directions along the world's subduction zones,
which are usually caused by forearc deformation
(e.g., Jarrard
1986
; McCaffrey
1992
). Finally,
it has been avoided a mix between long-term
geological data and geodetic velocities in the
estimation of Euler vectors, the usage of GPS
data having been limited to the determination
of the motion of six small plates, for which no
other data were available. The 24 Euler vectors
of MORVEL, relative to the Pacific plate, are
listed in Table
2.3
, while the resulting linear
velocity fields between adjacent plates are shown
in Fig.
2.38
.
The kinematic models described so far furnish
the Euler vectors of the major modern tectonic
plates relative to the Pacific. The components
of these vectors are expressed in the geographic
reference frame (where London, Eurasia, has a
fixed longitude). By vector summation, we can
calculate the Euler vector of relative motion be-
tween any pair of plates, assess closure conditions
