T32C-01 INVITED

Testing exhumation models – the roles of field data and of dating structures

* Wheeler, J johnwh@liv.ac.uk, University of Liverpool, Dept. Earth and Ocean Sciences, Liverpool, L69 3GP, United Kingdom

What distinguishes different models for HP/UHP exhumation, and how may they be tested? All models for exhumation involve the relative movements of bodies of rock. The dynamics – the driving force – for exhumation has often been given prominence in the description of exhumation models over the past 20 years. However the driving force – or the past distribution of stress during exhumation – is almost impossible to constrain by observation. There is instead much more scope for constraining the kinematics. Structures and kinematic indicators can give relative movements. Such structures can be dated directly, or dated indirectly by comparing PTt paths around them. It is instructive, then, to compare different exhumation models with reference to their kinematic implications. Many interpretations now involve thrusting below an HP/UHP unit (or "sliver"), with "normal sense" faulting or ductile shear above it but with the opposite movement sense. Consequently the HP/UHP unit moves up relative to rocks both above and below it. This has been proven for example in the Alps. It is important to note that proof of such a mechanism must involve dating both the normal sense movement and the thrusting – it is not sufficient to just deploy movement sense, because the ages might be different. The kinematics of this model are the same as for "channel flow" as invoked in the Himalayas. Apart from the fact the exhumed core region is not HP in the Himalayas, are there criteria that, objectively, distinguish the two models? I have asserted that the Alpine exhumation was driven by buoyancy of the exhumed material, plausible if it was buried below a dense mantle wedge. Channel flow is driven by a push from behind and squeezing. I will discuss whether the overall size and internal strain of bodies of exhumed material could be used to distinguish different driving forces.

T32C-02 INVITED

Exhumation of the Ultrahigh-Pressure Western Gneiss Region: Structural Geology, Petrology and Geochronology

* Hacker, B hacker@geol.ucsb.edu, Earth Science, University of California, Santa Barbara, CA 93106,
Andersen, T B t.b.andersen@geologi.uio.no, Geosciences, Universitetet i Oslo, Oslo, 0316, Norway
Barth, N ncbarth@umail.ucsb.edu, Earth Science, University of California, Santa Barbara, CA 93106,
Johnston, S scojoh@email.arizona.edu, Physics, California Polytechnic State University, San Luis Obispo, CA 93407,
Kylander-Clark, A R akylander@umail.ucsb.edu, Earth Science, University of California, Santa Barbara, CA 93106,
Peterman, E epeterman@umail.ucsb.edu, Earth Science, University of California, Santa Barbara, CA 93106,
Walsh, E ewalsh@cornellcollege.edu, Geology, Cornell College, Mount Vernon, IA 52314,
Young, D David.Young@utsa.edu, Geology, University of Texas, San Antonio, TX 78249,

The Western Gneiss Region of Norway includes ~40,000 km² of high-pressure rocks overlying ~10,000 km² of ultrahigh-pressure rocks. The Caledonian HP-UHP metamorphism spanned 420- 400 Ma and reached 800° C at ~125 km depth. Exhumation was near-isothermal up to depths as shallow as 15–20 km. HP deformation diminishes toward the foreland: chaotic folds in the core of the orogen give way within ~20 km distance to moderately strong, but well-organized Caledonian fabrics, and within 100 km to only weakly overprinted Precambrian fabrics complete with Precambrian sphene and garnet; the stretching direction rotated from an early SE trend (e.g., the Stadlandet area) to a late NE trend (e.g., the Møre-Trøndelag Fault Zone and Nordfjord Mylonitic Shear Zone). At ~50 km depth, high-pressure minerals in the bulk gneiss reverted to a symplectite formed of low-pressure phases. Subsequent deformation was concentrated into specific zones: top-E motion along the eastern edge of the WGR and top- W extension along the western edge of the WGR (including the Nordfjord-Sogn Detachment Zone).

http://www.geol.ucsb.edu/faculty/hacker/#Norway

T32C-03

Foundering and Exhumation of UHP Terranes: Race Car or School Bus?

* Kylander-Clark, A R kylander@geol.ucsb.edu, UC Santa Barbara, Department of Earth Science, Santa Barbara, CA 93106-9630, United States
Hacker, B R hacker@geol.ucsb.edu, UC Santa Barbara, Department of Earth Science, Santa Barbara, CA 93106-9630, United States

Recent geochronologic data from the giant ultrahigh-pressure (UHP) terrane, in the Western Gneiss Region of Norway, indicate that subduction and exhumation were relatively slow (a few mm/yr), and that the terrane was exhumed to the surface as a relatively thick, coherent body. These conclusions are in stark contrast to those reached in previous studies of some of the best-studied, smaller UHP terranes and suggest that the processes that form and/or exhume small UHP terranes are fundamentally different from the processes that affect large UHP terranes. These differences may be the result of variations in the buoyancy forces of different proportions of subducted felsic crust, mafic crust, and mantle lithosphere. Initial collision occurs via the subduction of smaller portions of continental material, such as microcontinents or ribbon continents. Because the proportion of continental crust is small, the processes involved in early UHP terrane formation are dominated by the oceanic slab; subduction rates are fast because average plate densities are high, and, as a result, subduction angles are steep. Because these smaller, thinner portions of crust are weak, they deform easily and mix readily with the mantle. As the collision matures, thicker and larger portions of continental material—such as a continental margin—are subducted, and the subduction regime changes from one that was ocean dominated to one that is continent dominated. The increased buoyancy of the larger volume of continental crust resists the pull of the leading oceanic lithosphere; subduction shallows and plate rates slow. Because the downgoing continent is thick, it is strong, remains cohesive and has limited interaction with the mantle. Although the subduction regime during early orogenesis is distinct from that during late orogenesis, the degree of mountain building and crustal thickening may be similar in both stages as small volumes and fast flow rates of buoyant material give way to large volumes and slow flow rates.

T32C-04

Extension and Doming in (Ultra-)High-Pressure Terranes: Relationship to Syn-Convergent Exhumation of (U)HP Nappes

Jamieson, R A beckyj@dal.ca, Department of Earth Sciences, Dalhousie University, Halifax, NS B3H 4J1, Canada
* Warren, C J c.warren@open.ac.uk, Department of Earth and Environmental Sciences, The Open University, Milton Keynes, MK7 6AA, United Kingdom
Beaumont, C chris.beaumont@dal.ca, Department of Oceanography, Dalhousie University, Halifax, NS B3H 4J1, Canada
Butler, J P butlerjp@dal.ca, Department of Earth Sciences, Dalhousie University, Halifax, NS B3H 4J1, Canada

Many ultra-high-pressure (UHP) terranes are characterised by structural domes cored by nappe stacks comprising UHP eclogites in tectonic contact with overlying high-pressure (HP) and lower-grade rocks. Contacts between UHP and HP nappes are generally interpreted as thrusts, whereas contacts with overlying lower-grade rocks are typically normal-sense shear zones. Medium- and high-resolution (nested) numerical models for subduction and exhumation of continental crust in a subduction channel during the onset of collision provide an internally consistent explanation for these observations, including syn-convergent extension. The models involve subduction of a weak continental margin and stronger continental interior, and include strain weakening and reversible density changes accompanying metamorphic phase transformations. Results are interpreted in terms of the exhumation number, E, which expresses the relative contributions of Couette (down-channel) and Poiseuille (up-channel) flow components. During subduction, margin material progressively weakens, leading to diachronous detachment and exhumation of UHP rocks. Decoupling initiates at depth along a shear zone that propagates upwards. The buoyant UHP plume tunnels up the shear zone until it encounters HP material in the upper part of the subduction channel; the HP nappe is then folded over and exhumed by the rising UHP plume. The resulting structural dome is cored by the (U)HP nappe stack, with attenuated remanants of the suture zone overlying the UHP plume and its HP carapace. During the later stages of exhumation, a prominent low-angle extensional shear zone forms above the rising nappe stack, creating foreland-directed normal faults that separate the (U)HP nappes from lower- grade cover. Other nappe geometries can be produced by models significantly stronger or weaker than the reference model, but all include doming and accompanying extension that are driven by the buoyant rise of a substantial volume of UHP material. Model geometries and PTt paths are consistent with observations from Tso Morari and other natural examples.

T32C-05

Were the world's youngest eclogites (NW D'Entrecasteaux Islands, Papua New Guinea) exhumed in rising gneiss domes or by shear on a deep-seated fault?

* Little, T A tim.little@vuw.ac.nz, School of Geography, Environment & Earth Sciences, Victoria University, Wellington, 6140, New Zealand
Hacker, B hacker@geol.ucsb.edu, Institute of Crustal Studies, University of California, Santa Barbara, CA 93106-9630, United States
Seward, G seward@geol.ucsb.edu, Institute of Crustal Studies, University of California, Santa Barbara, CA 93106-9630, United States

The up to ~2.5 km-high gneiss domes of the NW D'Entrecasteaux Islands of Papua New Guinea host the world's youngest terrane of HP (eclogite-facies, ~2-4 Ma) to UHP (coesite-bearing) gneissic rocks (~8 Ma). Previous models for their exhumation at >2 cm/yr have called upon: 1) buoyant rise of crustal diapers, or 2) normal-slip on deeply penetrating faults. A recent variant of the latter suggests that a paleo- subduction zone near the southern edge of the Solomon Sea has been inverted as a result of microplate tectonics. We present structural, microstructural, and electron back-scatter diffraction data of lattice preferred orientations (LPO's) from gneisses of Goodenough and Fergusson Islands to further explore mechanisms of exhumation. Relict eclogite-facies assemblages occur in mafic dikes and boudins, but most HP deformational fabrics are overprinted. The enclosing felsic gneisses are pervaded by amphibolite-facies ductile fabrics formed during their exhumation from the lower crust. These migmatitic rocks (metatexites) were partially molten during their deformation at temperatures of 570-730°C and pressures of 7-11 kb, but today are dominated by solid-state fabrics. The gneisses are capped by remnants of an ultramafic sheet that did not experience HP metamorphism. Below the ultramafics is a ~1 km-thick carapace zone. These high-strain gneisses generally have domal fabrics parallel to, and gradational to, those in the underlying core zone, which they locally rework. Active NE-dipping normal faults on the NE flank of the domes cut across the ultramafic contact and are underlain by a m-thick zone of pseudotachylite-bearing S/C fabrics. A sweeping pattern of stretching lineations reveals a 3-D pattern of ductile flow. In both the carapace and upper core zone, lineations are mostly EW: subparallel to the long dimension of the domes and perpendicular to plate motion in the Woodlark Rift. At greater structural depth, within the core zone, they deflect to become more nearly plate-motion parallel. Shear indicators diverge across the dome crests, suggesting of an inward flow of deeper rocks into the dome; or are locally variable, consistent with bulk irrotational deformation. In the gneisses (both core and carapace), conjugate shear-band microstructures and near-orthorhombic quartz LPOs, and back-rotation of mantled porphyroclasts indicate that ductile strain in domes was near plane, but that it was not simple shear (and included significant vertical shortening). The LPO's of the deepest rocks record activity of the high-T prism-[c] and prism- slip systems, whereas the outermost carapace rocks record basal- and rhomb- slip. The data reveal that deformational temperatures increased toward the dome centers, rather than outwardly into the carapace. Quartz LPO's in both dome and carapace are of uniformly modest intensity (~2-3 times random). Feldspar LPO's suggest slip on the (010)[001] or (010)[100] systems, and in some cases a shear sense opposite to quartz. While we cannot resolve how the eclogitic rocks ascended isothermally from the mantle into the lower crust, the simplest model invokes diapiric ascent (with decompression melting), ponding and lateral spreading along the Moho during early Woodlark Basin rifting. Subsequent exhumation of these rocks from the lower crust involved continued upward movement and vertical shortening of the gneisses combined with subhorizontal rift-parallel flow. Finally, normal faulting and minor erosion exhumed these rocks through the ultramafic cover to their present levels.