Title: On convective memory
Plate tectonics is the surface expression of mantle convection, but many aspects of the present-day tectonic setting depend on the history of how this system has evolved over time. I review recent work across spatio-temporal scales addressing how such convective memory can be used to validate tectonic scenarios and to better understand and predict how plate boundaries evolve consequently. Seismic anisotropy in the upper mantle is one recorder of convective deformation and the duration over which textures are reworked controls the lifespan of memory. This means that the asthenospheric depths show us present-day to recent convection, while shallower regions may allow distinguishing between different plate tectonic scenarios over the last ~50 Ma. Uncertainties about those scenarios and slab rheology mean that our understanding of the important issue of the path that slabs take into the deep mantle is still incomplete, leading to a range of estimates of mantle memory of past subduction. One particular question is how slabs are deformed upon bending in the trench, and I show results from convection models with rheological memory that can reproduce large offset faulting at the trench. Those models also show emergent slab segmentation, which may be relevant for the interpretation of deep mantle anomalies such as under Japan. The same sort of lithospheric rheologies with damage produce interesting effects in global mantle convection models which generate plate-like convection, and I close with an exploration of the typical timescales of fluctuations in convective heat transport as a function of plate boundary dynamics.