Cratons, the Archean (≥ 2.5 Ga) cores of the continents are the longest-lived features of Earth’s surface. They are underlain by cold, iron-depleted, thick (>250 km), seismically fast lithospheric mantle
roots or keels that have survived thermal and chemical erosion over multiple Wilson cycles. Recently however, numerous studies have shown that cratonic mantle lithosphere can undergo modification andmay be removed entirely. I use passive-source seismology to investigate this further for the Superior Craton of North America and its abutting Proterozoic platform. Initially, I use P- and S-wave relative arrival-time tomography to investigate the regional seismic structure of southeast Canada. Here, three broad zones of decreasing seismic wavespeed span the Archean Superior, Proterozoic Grenville and Phanerozoic Appalachian provinces, respectively. A vertical boundary in wavespeed beneath the Grenville Front is interpreted as evidence for subduction-driven metasomatic enrichment of the Laurentian margin. Due to the loss of the background mean velocity structure, relative arrival-time datasets are not easily combined and are also unsuitable for estimation of the physical properties of the mantle. To address this, I develop the Absolute Arrival-time Recovery Method (AARM) to retrieve absolute arrival-times from temporary seismograph networks, whose data are often noisy. Tests indicate that AARM picks are accurate to ≤0.25 s, akin to uncertainties in ISC bulletins. As a result, small aperture, temporary deployments now represent an exploitable resource with which to fill gaps in global seismic tomographic studies. Finally, I incorporate multiple regional networks from eastern Canada into a continental P-wave absolute arrival-time tomographic inversion to produce the most up-to-date compressional wavespeed image of the North American lithosphere. These results show along-strike variability in upper lithospheric velocity structure within the Proterozoic Grenville Orogen (1.3 − 1.0 Ga). When reconciled with the geological record, these results may provide the first seismological evidence for delamination of subcontinental lithospheric mantle in a Proterozoic large, hot, orogen.