Jump to Search Jump to Navigation Jump to Content

CO2 production during subduction zone initiation and evolution, and intracratonic rift setting



cd20ma_6 Model shows the effects of increasing velocity on subduction dynamics of a young (20 Ma) slab. The model is characterized by “double subduction”, or subduction of the lithospheric mantle beneath the continent adjacent to the subducting oceanic slab. A mélange composed of sediments, basaltic, and lower crustal components is subducted to approximately 60 km. At about 19.44 Ma, fast dehydration and melt weakening of the overriding continental crust results in partial extension and asthenospheric inflow. Consequently, asthenospheric inflow into lower portions of the mantle wedge results in high temperature gradient across the slab at shallow depths driving partial melting of sediment and altered basalt rocks.

The mélange originally extended to 60 km depth, however it is destroyed leaving behind a small, thickened crustal wedge extending to ~40 km depth. The restite of the partially molten mélange mixture is relaminated onto the continental lower crust and underlain by hydrated mantle lithosphere. The hydrated mantle lithosphere releases H2O fluid as it is relaminated under the lower continental crust. At approximately 48.09 Ma extension ceases due to the strong dry olivine rheology of the asthenosphere that flowed in and subduction returns to a compressional regime. Molten sediments, basalts, and parts of the lower crust sourced from the initial mélange are peeled away by partial melting and relaminated back onto the lower continental crust.

Translithospheric diapirism (cd_60ma3 and cd_40ma4)


Translithospheric diapirism is observed in a number of subduction models spanning various plate ages and velocities. Common between these models is that diapirism is often quick (shortest cumulative time is cd_40ma4 spanning 14 Ma). Each run is characterized by an initially steep slab angle with basal erosion of the accretionary wedge and a significant amount of sediments entering the trench as seen in model cd_60ma3. The sediments partially melt and the column of rocks above the diapir weakens according to the low λmelt(.001). In cd_60ma3, continental lithospheric necking begins by 10.62 Ma and the sedimentary melt begins to intrude and emplace into the continental lithosphere. At 14.28 Ma, asthenosphere flows into the weakened wedge melting mélange material. This results in a larger thermal gradient at shallow depths. Initial subduction of the sediment load and formation of the translithospheric diapir causes extension in the continental crust. The subsequent extension results in nappe stacking in the accretionary wedge as the trench retreats.

Consequently, a plug forms in the wedge promoted by underscraped basalt, which controls the sediment supply into the subduction zone via basal erosion processes. Sustained translithospheric diapirism ceases due to a reduction in the amount of sediments entering the trench. Extensive serpentinization of the forearc at 26.08 Ma occurs due to H2O devolatilization of sediments and altered basalts. The subduction angle relaxes from steep to shallow as a consequence of serpentinization, which reduces the rheological strength of the mantle wedge. Stable subduction sets in without any noticeable arc compression or extension. CO2 mass transfer in this system is predominantly through the advection of sediments and sediment melt which intrudes into the crust.

The second instance involves decarbonation of the slab in the forearc mantle. CO2 fluids derive from sediments and altered basaltic crust and are absorbed by the peridotitic mantle when the fluids percolate through. In the cold serpentinized mantle, CO2 can remain stable within the peridotite without further decarbonation unless later perturbed by a thermal anomaly. An example of this occurs in cd_60ma_3 after a translithospheric diapir advects from the subducting slab and impinges on the continental lithosphere. The first 30 Ma is dominated by the effects of the diapir and incipient influx of hot asthenosphere into the colder part of the mantle wedge. The first episode of lithospheric carbonation begins at 41.51 Ma. Small decarbonation episodes occur, which result in the carbonation of forearc mantle peridotite between 47 and 62.23 Ma. In cd_60ma3, the forearc mantle is carbonated at or near its limit of ~1.5 CO2 wt.%



Initially, this study investigated subduction age and velocity with a fixed value for λfluid to simulate diapirism as effectively as possible. However, this parameterization does not allow for extension within the active margin, a common feature in subduction zones. Therefore, we ran an additional model to explore parameter space in such a case where extensional regimes are present for a fixed velocity of 4 cm year-1. We vary plate age in an attempt to examine decarbonation response within a much more dynamic system. This is achieved by decreasing λfluid to .001, which decreases the overriding plate strength, and allows extension to occur within the arc area. The initial stages of subduction in model irep80 are similar to the reference model cd_60ma4. However, at 7.33 Ma notable extension in the arc occurs and slab rollback begins. Consequently, asthenosphere rushes in to replace the retreating slab, where decompression melting occurs in upwelling mantle at ~2675 km. As slab rollback intensifies at 14.41 Ma, hot asthenosphere exploits the weak, hydrated and serpentinized mantle resulting in steepening of the subduction angle. As the subduction angle steepens during the rollback phase, a large mélange forms in the forearc mantle region. A large flux of high-temperature asthenosphere intrudes at 14.41 Ma. This results in the final stage of extension at ca. 18.21 Ma. A portion of the mélange detaches and begins sustained decarbonation for ~8 Ma due to buoyancy differences between the underlying depleted asthenosphere and the mélange mixture. After extension subsides, high temperatures in the forearc mantle persist for ~8 Ma and wane with a return to a stable subduction regime. Heat loss due to conduction quickly returns the thermal structure back to the cold, stable regime by ~30 Ma. Model irep80 exhibits the importance of incorporating CO2 fluids into dynamic models despite our use of such simplistic methods, in comparison to those applied in elsewhere (e.g., phase fractionation and open vs. closed system devolatilization). Future opportunities are open for implementation of more sophisticated and fully coupled fluid interactions into this code.

Archon – Tecton extension models

(Failed rifting events:
Ma_Archon (Tecton) 7Ma_Archon.mp4
Ma_Archon (Tecton) 7Ma_Tecton.mp4
Ma_Archon (Tecton) 8Ma_Archon.mp4
Ma_Archon (Tecton) 8Ma_Tecton.mp4
Ma_Archon (Tecton) 9Ma_Archon.mp4
​Ma_Archon (Tecton) 9Ma_Tecton.mp4
Ma_Archon (Tecton) 10Ma_Archon.mp4
Ma_Archon (Tecton) 10Ma_Tecton.mp4
Ma_Archon (Tecton) 11Ma_Archon.mp4
Ma_Archon (Tecton) 11Ma_Tecton.mp4
Ma_Archon (Tecton) 12Ma_Archon.mp4
Ma_Archon (Tecton) 12Ma_Tecton.mp4

Reference models:

Higher Moho Temperatures & felsic lower crust:

Higher Moho Temperatures & Mafic Granulite lower crust:

Cratons form the stable core roots of the continental crust. Despite their long term stability, they have failed spectacularly in the past. Cratonic destruction has been attributed to rejuvenation at the base of the lithospheric mantle in some cases, but still remains a poorly constrained process. These models use a 2D petrological--thermomechanical modeling code to elucidate the effects a carbonate metasomatized lithospheric mantle has on full cratonic breakup and failed rifting events. New thermodynamic look--up tables are used for Archon (chemically depleted) and Tecton (young and fertilized) tectothermal ages. Each lithospheric look--up table is also supplemented by an addition of 2 CO2wt.% to the bulk composition, which represents a carbonate metasomatized lens. Thermodynamic modeling shows density differences (>-60 kg m-3) between the two tectothermal ages and also between the dry and metasomatized portions (-10 to -20 kg m-3). The following parameters were tested in the simulations: (1) -- rifting duration, (2) -- lithospheric age, (3) -- Moho temperature, and (4) -- crustal composition. Models extended until cratonic breakup were similar, exhibiting many of the same rifting features and decarbonation trends. We show that convective removal of the metasomatized layer in the Tecton composition whereas the Archon lithosphere remains stable. Three regimes of failed rifting was also observed: (1) -- failed rifting without decarbonation, (2) -- failed rifting with decarbonation, and (3) -- semi{failed rifting with mantle melting. Decarbonation trends were greatest in the failed rifts. Convective removal of the metasomatized lithospheric mantle was only observed in the semi--failed case. Increased Moho temperatures did not show any overall changes in rifting or decarbonation rates. However, the higher temperatures did affect the stratigraphic sequences at the rift flanks. Additionally, higher temperatures also resulted in an increase in time required to experience dry mantle melting. Lastly, crustal composition greatly affects the rifting style. We show increased asymmetry with a felsic lower crustal composition, which also led to an 8 Ma increase in the amount of time it took for cratonic breakup.