Slab strength and trench length influence lowermost mantle flow directions

The geometry of slabs within the upper mantle has been relatively well-imaged by regional tomography and seismic studies; however, slab deformation in the lower mantle remains poorly understood. Although tomography models reveal that the lower mantle beneath paleo-subduction regions are faster-than-average, the resolution is not high enough to image how slabs are actually deforming there. Geodynamical studies of lower mantle slabs have shown a wide range of possible deformational behaviors, ranging from stiff, buckling slabs to more-ductile accumulations of slab material undergoing pure shear (e.g., Loubet et al., 2009). Observations of seismic anisotropy due to crystal preferred orientation (CPO) may ultimately provide valuable insight into the manner of slab deformation in the lower mantle; however, the parameters and processes that cause CPO are not well-constrained. The development of CPO depends upon the intrinsic anisotropy and slip systems of lower mantle minerals, under the influence of the mantle flow field (i.e., strain). As a first step toward better understanding how CPO is generated in lowermost mantle slab regions, we employ 3D spherical geodynamical calculations to investigate how slab strength and trench length influence lowermost mantle lateral flow patterns. We find that strong slabs dominantly lead to stiff, buckling slabs in the deep mantle, resulting in mostly unidirectional lateral flow patterns in the lowermost mantle. In contrast, weaker slabs are dominantly smooshed under pure shear, resulting in more-azimuthally divergent lowermost mantle lateral flow patterns. Furthermore, we find that decreasing trench length is effectively similar to decreasing slab strength in terms of resultant deformational behavior and lowermost mantle lateral flow patterns. In summary, strong and/or wide slabs are expected to have different lowermost mantle flow patterns (more-unidirectional flow) than weak and/or narrow slabs (more azimuthally-divergent flow). This difference will likely result in a characteristically-different directionality to the CPO that would develop, possibly allowing observations of seismic anisotropy to provide constraints on slab strength.