Upper-mantle velocities below the Scandinavian Mountains from P- and S- wave traveltime tomography

The relative traveltime residuals of more than 20 000 arrival times of teleseismic P and S waves measured over a period of more than 10 yr in five separate temporary and two permanent seismic networks covering the Scandinavian (Scandes) Mountains and adjacent areas of the Baltic Shield are inverted to 3-D tomograms of P and S velocities and the VP/VS ratio. Resolution analysis documents that good 3-D resolution is available under the dense network south of 64° latitude (Southern Scandes Mountains), and patchier, but highly useful resolution is available further north, where station coverage is more uneven. A pronounced upper-mantle velocity boundary (UMVB) that transects the study region is defined. It runs from SE Norway (east of the Oslo Graben) across the mountains to the Norwegian coast near Trondheim (around the Møre−Trøndelag Fault Complex), after which it follows closely along the coast further north. Seismic velocities in the depth interval 100−300 km change significantly across the UMVB from low relative VP and even lower relative VS on the western side, to high relative VP and even higher relative VS to the east. This main velocity boundary therefore also separates relatively high VP/VS ratio to the west and relatively low VP/VS to the east. Under the Southern Scandes Mountains (most of southern Norway), we find low relative VP, even lower relative VS and hence high VP/VS ratios. These velocities are indicative of thinner lithosphere, higher temperature and less depletion and/or fluid content in a relatively shallow asthenosphere. At first sight, this might support the idea of a mantle buoyancy source for the high topography. Under the Northern Scandes Mountains, we find the opposite situation: high relative VP, even higher relative VS and hence low VP/VS ratios, consistent with thick, dry, depleted lithosphere, similar to that in most of the Baltic Shield area. This demonstrates significant differences in upper-mantle conditions between the Southern and Northern Scandes Mountains, and it shows that upper-mantle velocity anomalies are very poor predictors of topography in this region. An important deviation from this principal pattern is found near the topographic saddle between the Southern and Northern Scandes Mountains. Centred around 64°N, 14°E, a zone of lower S velocity and hence higher VP/VS ratio is detected in the depth interval between 100 and 300 km. This ‘Trøndelag−Jämtland mantle anomaly’ (TJMA) is still interpreted as part of relatively undisturbed lithosphere of shield affinity because of high relative P velocity, but the relatively low VP/VS ratios indicate lower depletion, possibly higher fluid content, and most likely lower viscosity relative to the adjacent shield units. We suggest that this mantle anomaly may have influenced the collapse of the Caledonian Mountains, and in particular guided the location and development of the Møre−Trøndelag Fault Complex. The TJMA is therefore likely to have played an important role in the development of the ‘two-dome architecture’ of the Scandes Mountains.