Dripping or delamination? A range of mechanisms for removing the lower crust or lithosphere. Issue 2 (10th May 2017)

• Record Type: Journal Article
• Authors: Beall, Adam P.; Moresi, Louis; Stern, Tim
• Is Part Of: Geophysical journal international. Volume 210:Issue 2(2017:Aug.)
• Page Start: 671
• Page End: 692
• Publication Date: 2017-05-10
• Subjects: Instability analysis -- Numerical modelling -- Dynamics: gravity and tectonics; Dynamics of lithosphere and mantle; Intra-plate processes; Geophysics -- Periodicals; 550
• DOI: 10.1093/gji/ggx202
• Languages: English
• ISSNs: 0956-540X

Abstract

Abstract: Under some conditions, dense parts of the lower crust or mantle lithosphere can become unstable, deform internally and sink into the less dense, underlying asthenosphere. Two end-member mechanisms for this process are delamination and dripping. Numerical calculations are used to compare the time taken for each instability to grow from initiation to the point of rapid descent through the asthenosphere. This growth period is an order of magnitude shorter for delamination than dripping. For delamination, the growth rate varies proportionally to the buoyancy and viscosity of the sinking material, as with dripping. It also depends on the relative thickness ($L^{\prime }_c$ ) and viscosity ($\eta ^{\prime }_c$ ) of the weak layer which decouples the sinking material from the upper crust, varying proportionally to $L_c^{\prime 2}/\eta ^{\prime \frac{2}{3}}_c$ . As instabilities commonly resemble a mix of dripping and delamination, the analysis of initial instability growth includes a range of mechanisms in-between. Dripping which begins with a large perturbation and low $\eta ^{\prime }_c$ reproduces many of the characteristic features of delamination, yet its growth timescale is still an order of magnitude slower. Previous diagnostic features of delamination may therefore be ambiguous and if rheology is to be inferred from observed timescales, it is important that delamination and this 'triggered dripping' are distinguished. Transitions from one mechanism or morphology to another, during the initial growth stage, are also examined. 3-D models demonstrate that when $\eta ^{\prime }_c$ is small, a dripping, planar sheet will only transition into 3-D drips if the initial triggering perturbation is less than a third of the dense material's thickness. This transition occurs more easily at large $\eta ^{\prime }_c$, so rheological heterogeneity may be responsible for morphological transitions through time. We also calculate the rates at which delamination grows too slowly to outpace cooling of the upwelling asthenosphere, resulting in stalling and switching to dripping. Common lithospheric viscosities and observed timescales indicate that both instability transitions are feasible. Overall, the timescale and persistence of delamination depends on three parameter groups, which characterize the properties of the anomalously dense material, weak lower crustal layer and relative rate of thermal diffusion. These scalings appear to unify the varying results of previous delamination studies.