In the above figure arrows refer to NUVEL-1A plate motions with respect to a stable Eurasia (green=Indo-Australian, pink=Pacific and Caroline, yellow=Philippine Sea Plate), dots to earthquakes (yellow=shallow, blue=intermediate, magenta=deep) and contours to the age of oceamic lithosphere.

**Regional tectonics**

From a tectonic point of view Southeast
Asia (SEA) is a very complex area. It accommodates the convergence of three
major plates (the Eurasian (EAP), Indo-Australian (IAP) and Pacific (PaP)
plates, of which the latter drives the Caroline (CP) and Philippine Sea
(PP) plates) and reveals a variety of tectonic processes.

Along the western and central boundary
between the IAP and the EAP oceanic material subducts under Sundaland (the
stable interior of SEA). At Sumatra (WSS and ESS) the oblique component
of convergence is partly taken up by a large strike-slip fault that cuts
Sumatra parallel to the trench. Java (JS) exhibits subduction in its most
classic form.

Continental material of Australia has
reached the trench at the eastern side of this boundary (TBS) resulting
in almost completed accretion of the IAP and Timor. Here, convergence has
been transferred from the fore-arc trench to back-arc thrusts.

At Irian Jaya the largly oblique component
of convergence between the IAP and the CP is taken up by crustal shortening
and elevation and by the left-lateral Sorong-Yapen-Ransiki fault. This
fault runs from eastern New Guinea through the Bird's Head at northwestern
Irian Jaya towards Halmahera (HS), where it possibly connects to the Philippine
trench.

Left-lateral translation along the Matano-Palu-Koro
fault system and subduction of the Celebes Sea (CS) at the North Sulawesi
trench accommodate the clockwise rotation of Sulawesi.

The left-lateral Philippine Fault (PF)
system that runs from the southern to the northern Philippines is a result
of the oblique component of convergence between the PP and Sundaland. The
accommodation of the normal component is divided between two trench systems,
located at the western and eastern side of and roughly parallel to the
Philippines.

**Data**

The figure above shows the different GPS velocity data that we use in our model, shown within their original reference frame.

The model parameterization consists of
great circle segmentation of the major faults and Delaunay triangulation
of the areas bounded by faults or model boundary segments.

Our inversion model solves (in a least
squares approximation) for the velocity gradient tensor (VGT) at the nodes
of the triangles and for the fault slip vectors at the fault segments simultaneously.
The inclusion of several data sets that are defined with respect to different
reference frames requires an extra model parameter, i.e. a uniform rotation
vector per extra data set. This vector projects all data of the same set
onto the reference frame of the GEODYSSEA data.

**Results**

The qualitative aspects of the results match the observations and derivations of others quite well:

**Fault slip solution:**

The figure above shows the results for
fault slip. Red, blue, yellow and green arrows denote thrust, normal, left-
and right-lateral fault slip, respectively. The general features are thrusting
along the Sunda Arc, right-lateral slip at Sumatra, backarc thrusting instead
of fore-arc subduction along the Banda Arc, strong left-lateral slip at
Irian Jaya, Sulawesi and the Philippines and moderate thrusting magnitudes
along both sides of the boundary between Sundaland and the PP.

**Continuous deformation:**

The figure on your left-hand side below
shows the results for the strain rate field (or the symmetric part of the
VGT). Contours denote the effective strain rate (second invariant of the
strain rate tensor) and arrows are the principal axes of the strain rate
tensor. We find insignificant strain rates at the plate interiors, arc-parallel
dilatation at the Sumatran fore-arc, strong strike-slips at the Timor fore-arc,
contraction at the Banda Sea, arc-parallel dilatation at Irian Jaya and
strong strike-slip regimes at Sulawesi, the Philippines and Taiwan.

The figure on your right-hand side below
depicts the resulting rotation rate field (anti-symmetric part of the VGT).
To enable comparison with recent paleomagnetic data we have translated
the rotation rates into degrees/Ma. The results that can be verified with
existing (Quaternary) paleomagnetic data are the neighbouring regimes of
counterclockwise rotation of Irian Jaya and clockwise rotation of Sulawesi.

This stress map of Greece is provided for
by the World Stress Map
(WSM) project. We have applied our method to the Aegean area by inverting
a preliminary GPS data set of the SING
project. This integrated set consists of different GPS data sets that
were obtained during the last 12 years and 100 year old triangulation data.
We inferred the rotation rate field to compare the present crustal deformation
field with recent paleomagnetic data (about 1 Ma), sampled and interpreted
by Charon
Duermeijer.

The figure below shows the distribution of the
ChRM (characteristic remnant magnetization) and AMS (anisotropy of the
magnetic susceptibility) data of Plio-Pleistocene age on the Aegean outer-arc
[Duermeijer et al., *EPSL*, **176**, 509-525]. The
solid lines in the ChRM plots (half circles) represent the mean declination.
The solid lines in the AMS data (full circles) indicate the mean lineation
direction per area. Further examination of these results indicates that
a clockwise rotation phase took place between ~0.8 Ma and Recent on Zakynthos
and in the central and Eastern Peloponessos. The anticlockwise rotation
phase in the southeastern arc may be equally young (< 1.8 Ma), although
dating is insufficiently accurate.

The figure above shows the velocity vectors used in the inversion with their 3-sigma error ellipses from a combination of several GPS campaigns. These GPS derived velocities are from a preliminary combination of the SING data (SING-project) defined with respect to a reference frame fixed to Eurasia. The figure below shows the resulting contoured geodetic rotation rates scaled to degrees per Ma. The numbers refer to local (50 - 100 km scale) averages of rotation rates with their formal errors. In general, the geodetically derived rotations agree with the young paleomagnetic data: we find (strong) clockwise rotations along the western arc and (weaker) anti-clockwise rotations along the central and eastern arc.