Abstract It is increasingly clear that present-day dynamic topography on Earth, which is generated and maintained by mantle convective processes, varies on timescales and length scales on the order of 1–10 m.y. and 103 km, respectively. A significant implication of this behavior is that Phanerozoic stratigraphic records should contain indirect evidence of these processes. Here, we describe and analyze a well-exposed example of an ancient landscape from the Grand Canyon region of western North America that appears to preserve a transient response to mantle processes. The Surprise Canyon Formation lies close to the Mississippian-Pennsylvanian boundary and crops out as a series of discontinuous lenses and patches that are interpreted as remnants of a westward-draining network of paleovalleys and paleochannels within a coastal embayment. This drainage network is incised into the marine Redwall Limestone whose irregular and karstified upper surface contains many caves and collapse structures. The Surprise Canyon Formation itself consists of coarse imbricated conglomerates, terrestrial plant impressions including Lepidodendron, and marine invertebrate fossils. It is overlain by marine, fluvial, and aeolian deposits of the Supai Group. These stratal relationships are indicative of a transient base-level fall whose amplitude and regional extent are recognized as being inconsistent with glacio-eustatic sea-level variation. We propose that this transient event is caused by emplacement and decay of a temperature anomaly within an asthenospheric channel located beneath the lithospheric plate. An analytical model is developed that accounts for the average regional uplift associated with landscape development and its rapid tectonic subsidence. This model suggests that emplacement and decay of a ~50 °C temperature anomaly within a channel that is 150 ± 50 km thick can account for the observed vertical displacements. Our results are corroborated by detrital zircon studies that support wholesale drainage reorganization at this time and by stratigraphic evidence for spatially variable regional epeirogeny. They are also consistent with an emerging understanding of the temporal and spatial evolution of the lithosphere-asthenosphere boundary.