Numerical modelling of the craton destruction process, with a type-example from the North China Craton

The phenomenon of cratonic destruction has attracted attentions among numerical modelers. However, existing modelling work appears to lack enough constraints from geological and geophysical observations. To better combine the observations with the modelling approach, and to further understand the destruction of the North China Craton (NCC), I firstly reviewed current knowledge about the evolution of the North China Craton, and then outlined the strength and weakness associated with classic delamination and convective erosion models. By integrating new observational data and 2-D numerical modeling results, I put forward a two-stage model to account for the destruction of the NCC, that is, cratonic keel delamination followed by localized convective erosion and lithospheric extension.As revealed in the evolution of the NCC, its destruction appears to be accompanied by the inversion of a pre-existing North China Cratonic Basin (NCCB), which is partly preserved in southeastern Ordos. This >300 Myr subsidence of the NCCB can most probably be explained by assuming a pre-existing dense keel for the eastern NCC (ENCC), consistent with the petrological and geochemical observations from mantle xenoliths/xenocrysts hosted in kimberlite/basalt in the ENCC.As the bottom was dense, cratonic keel stability would have only been sustained by the buoyancy of the overlying lithospheric portion. Hereafter, if some weak intra-cratonic regions had existed above the dense bottom and were reactivated, intra-cratonic decoupling may finally occur, leading to keel delamination. Coincidentally, some seismically visible Mid-Lithospheric Discontinuity Layers (MLDLs) exist at depths of ~80-100 km within modern cratons, including the remaining portions of the NCC and the Wyoming Craton (WC),. Interestingly, these depths are also where the relic lithospheric bottom appears to remain beneath the keelless sub-regions of the eastern NCC and the WC. Because the MLDLs are suggested to be regions of preferential accumulation of metasomatic minerals and/or anomalously wet (>1000 ppm) peridotites (both of which would lead to a relatively weak rheology), we accordingly propose that the weak MLDLs may have been the intra-cratonic weak zone, which allowed delamination of the dense cratonic keels of the NCC and the WC. This delamination model is studied in detail using 2-D analytical and numerical modelling, and the modelling results may help constrain the MLDL rheology needed for explaining the keel delamination rate estimated on the NCC and the WC. One can speculate that the delaminated cratonic keels may still exist or stagnate within the upper mantle. This suggestion may explain some, if not all, of the high velocity bodies beneath North China and Iowa, USA (which was the location caped by the Wyoming Craton before the Laramide Orogeny).After keel delamination along the MLDLs, a protracted (ca. 50-100 Myr) period of small-scale convective erosion and/or lithospheric extension thinned the remaining ~90 km thick lithosphere and continuously reworked the former cratonic lithospheric mantle beneath the Moho. As indicated in the modeling results, this second stage evolution may explain the episodic magmatism and crustal deformation associated with the destruction of the NCC.