Seismic imaging of mantle plumes

The mantle plume hypothesis was proposed thirty years ago by Jason Morgan to explain hotspot volcanoes such as Hawaii. A thermal diapir (or plume) rises from the thermal boundary layer at the base of the mantle and produces a chain of volcanoes as a plate moves on top of it. The idea is very attractive, but direct evidence for actual plumes is weak, and many questions remain unanswered. With the great improvement of seismic imagery in the past ten years, new prospects have arisen, Mantle plumes are expected to be rather narrow, and their detection by seismic techniques requires specific developments as well as dedicated field experiments. Regional travel-time tomography has provided good evidence for plumes in the upper mantle beneath a few hotspots (Yellowstone, Massif Central, Iceland). Beneath Hawaii and Iceland, the plume can be detected in the transition zone because it deflects the seismic discontinuities at 410 and 660 km depths. In the lower mantle, plumes are very difficult to detect, so specific methods have been worked out for this purpose. There are hints of a plume beneath the weak Bowie hotspot, as well as intriguing observations for Hawaii. Beneath Iceland, high-resolution tomography has just revealed a wide and meandering plume-like structure extending from the core-mantle boundary up to the surface. Among the many phenomena that seem to take place in the lowermost mantle (or D"), then are also signs there of the presence of plumes. In this article I review the main results obtained so far from these studies and discuss their implications for plume dynamics. Seismic imaging of mantle plumes is still in its infancy but should soon become a turbulent teenager.