The Formation of the Molucca Sea
The Molucca Sea currently the only present day example of ocean basin closure as a result subduction of oceanic lithosphere in two opposite directions. The Molucca sea is situated at the convergence of the Eurasian, Philippine Sea and Australian Plate in Indonesia. The geometry of the plates is highly complex, which reflects the formation of micro-continents and broken fragments from the collision of the 3 major plates (Widiwijayanti et al, 2004). The Molucca Sea is currently an isolated basin that is well advanced in double subduction. The Molucca Sea Plate is subduction in two directions; in the east at the Halmahera subduction zone and to the west in the Sangihe subduction zone. The oceanic lithosphere has virtually disappeared and the forearc crust is being dragged down towards the Eurasion plate. The subduction of the oceanic lithosphere has forced the Sangihe and Halmahera arcs to collide, forming a collision complex. Figure 4 illustrates that this more advance in the northern region of the Molucca Sea (cross-section A). The collision complex consists of deformed sediments and ophiolites that are currently being thrusted over both volcanic arcs.
The Molucca Sea Plate was originally part of the Philippine plate when the Australian Plate was much further south during the mid Miocene, 13 million years ago. The Sangihe trench was active at this time, subducting oceanic lithosphere of the Molucca Sea to the west (Hinschberger et al, 2005). The Sangihe subduction is believed to date back to the Oligocene and is much older than the Halmahera in the west. The subduction zones are dated by the depth of the downgoing slabs with is much less at the Halmahera subduction zone. There are conflicting ideas on the date that double subduction began and the Molucca Sea became isolated from the Philippine Sea Plate. In a plate kinematic model by Hinschberger et al, 2005, subduction at the Halmahera trench begin only 3.5 Ma, whereas Baker and Malaihollo (1996) dated volcanic rock from the Halmahera Ridge as old as 11 Ma. Hinschberger et al (2005) propose that the Halmahera trench did not exist until much later and followed a massive rearrangement of the tectonic plates in the middle of the Pliocene. Subduction at the Halmahera trench formed 3.5 Ma to accommodate part of the convergence between that Philippine Sea Plate and the East Asian Plate. The Molucca Basin was isolated and double subduction began and the Sangihe subduction remained dominant. Some discrepancy may be caused by errors in the radiometric dating of the rocks and the kinematic model may not be a full representation of the past, which is very difficult to calculate. The ages are significantly different and it is uncertain which is more correct. However, Hinschberger et al (2005) is the most recent of the two and should be the most up to date.
The cross-section in figure 4 illustrates the deformation caused by the collision of the arcs. The backarc of the Halmahera arc has been thrusted up over it and the Sangihe forarc is currently overthrusting the Halmahera arc. Volcanic activity along the Halmahera arc has ceased north of Morotai Island due to this overriding and volcanism continues further north along the opposite Sangihe arc. The thickness of the collision complex has been significantly increased by the overthrusting of one arc on top of the other. The collision of the Sangihe and Halmahera has produced a 'melange wedge' of deformed sediment over the two trenches (Widiwijayanti et al, 2004). Seismic reflection profiles confirm that there is 15 km of low-velocity material beneath the Molucca Sea associated with the varying sediments in the wedge. Central Ridge runs north to south between the two arcs. The islands that make up the ridge are blocks that have 'popped up' from the collision complex and been exposed to the surface. They consist of periodite, gabbros, metamorphic rocks and sediment which is typical of the collision zone. The positive density anomaly around the islands is strong evidence that these blocks were uplifted (Widiwijayanti et al, 2004). The arc-arc collision is at different stages along the Molucca Sea. Cross-section (A) through Talaud Island shows that the volcanic arc and forearc has almost been entirely overridden by the Sangihe forearc along this latitude. Snellius ridge is regarded as the backarc region of the Halmahera ridge and consists of a layer of carbonates on top of ophiolites and arc sediments (Hall, 2000).
The Sangihe arc has almost overridden the entire Halmahera arc north of Morotai island and is being thrusted up on to the flanks of the active Halmahera forearc in the south of the Molucca Sea. The active volcanic Halmahera arc was the zone of weakness during the collision indicative of a high quartz content at depth. Quartz has a low activation energy to overcome and achieve plastic deformation (flow). The overthrusting has greatly thickened the accretionary sediment and the convergence has thrusted up blocks from the collision complex. The islands of Central Ridge were formed by this uplifting of periodite, gabbros, metamorphic rocks and sediment from the collision complex or 'melange wedge.' The Sangihe Arc will continue to override the Halmahera arc and in 5 Ma and eventually the Molucca Sea will also disappear leaving little evidence that it was the unique site of double subduction.