Firstly, can you outline the context from which your research has emerged? The mantle represents about 84 per cent of the Earth’s volume, with the core representing around 15 per cent and the crust just 1 per cent. The mantle only rarely reaches the surface and as a result, it is mostly un-sampled, except for basalts or xenoliths (solid pieces of the mantle). In fact, there have been numerous attempts to directly sample and analyse the mantle, but they have been mostly unsuccessful. This leaves room for exciting work to be done, uncovering the composition of the largest part of the Earth. Your latest investigation focuses on the intraplate volcanism of the Hawaiian Islands. Can you explain why you study mantle plumes? Mantle plumes are hotter than the surrounding material and rise from deep in the mantle, all the way from the core-mantle boundary in some cases, a movement of around 2,890 km. When the plume head begins to melt near the surface, at a depth of about 100 km, it forms a large igneous province or oceanic plateau comprising huge volumes of magma. Oceanic islands are another manifestation of mantle plumes; the study of their composition tells us much about the Earth’s mantle. Mantle plumes are probes of the deep mantle, and this is why I study them. You have observed significant geochemical differences between volcanoes of Kea and Loa trends. Can you explain what is meant by this? The first geographically-derived evidence for the existence of two chains of volcanoes dates back to Dana in 1849, then Jackson in 1972. It was later shown that these two chains also show distinct lead isotopic differences, but without an explanation for the source of these differences. The disparities were puzzling, especially considering that the two chains of volcanoes are separated by only about 50 km, while the sampling zone of the plume is roughly 100 km. Temperature, upwelling rate and more general physical parameters of the plume are concentrically zoned, so a lateral compositional and isotopic difference between the two sides did not make sense. With a significant database of high-precision analyses of around 600 samples of shield lavas, we have been able to show that there are also differences in other isotopic systems, including strontium, neodymium and hafnium. We have also made a connection with the presence of an anomalous zone (low seismic velocities) at the core-mantle boundary to account for the different signature of Mauna Loa, the largest volcano on Earth. Why do late-stage volcanoes allow for a different understanding of the isotopic systematics of the Earth’s mantle? On Hawaii, the evolution of a volcano presents a typical cycle, usually over 1-1.5 million years, with: a pre-shield stage with alkalic compositions (Loihi seamount); a shield stage with tholeiitic compositions that represent over 95 per cent of the volume of the volcano (Mauna Loa and Kilauea are in this stage now) when the volcano grows above sea-level, and; a post-shield stage with a return to alkalic compositions. Sometimes, there is a resurgence of volcanism long after the volcano has moved away from the centre of the hotspot, and this stage is also alkalic. Post-shield and rejuvenated volcanism corresponds to significantly lower degrees of partial melting in the mantle than shield volcanism and because of that, they are more Dr Dominique Weis explains the foundations of her research into the geochemistry of the Earth’s mantle, via mantle plumes, which probe the deep, hidden interior of the planet Taking up the mantle 116 INTERNATIONAL INNOVATION DR DOMINIQUE WEIS KILAUEA CALDERA ON MAY 10 2009: HALEMA’UMA’U CRATER IS APPROXIMATELY 100 M DEEP