The proposed study will address two problems in solid Earth dynamics: (I) How to reconcile the geochemical distinctions between mid-ocean ridge basalts (MORBs) and hotspot ocean island basalts (OIBs) which are widely interpreted in terms of a chemically layered mantle, with geophysical evidence for whole mantle (i.e., single layer) convection. (II) Determining the origin of the geographic variations in lava geochemistry observed at hotspots. The hypothesis is that variations in melting and ductile flow of a compositionally heterogeneous mantle cause many of the key geochemical characteristics of OIBs and MORBs. This hypothesis will be tested using numerical models that simulate the three-dimensional viscous flow, heat transport, and melting in the upper mantle. Models will be constructed based on the assumption that hotspots are formed by buoyant upwellings, or mantle plumes, interacting with the base of the lithosphere. The melting calculations will incorporate laboratory experimental data on the pressure-temperature conditions for melting of materials with different compositions. Computational results will quantify the relationships between specific physical conditions and predicted geographic variability of trace-element and Sr, Nd, Pb, and He isotope geochemistry in erupted magmas at three tectonic scenarios where adequate data sets exist for comparisons: an intraplate hotspot (Hawaii), a hotspot centered on a mid-ocean ridge (Iceland), and a hotspot off of, but near a mid-ocean ridge (Galapagos). Problem (I) will be addressed by determining whether models do or do not require mantle plumes--which could come from the lower mantle--to be compositionally distinct from the upper mantle. Problem (II) will be addressed by determining the physical conditions needed to explain the nature of spatial geochemical variability observed at the above types of hotspots. The origin of hotspot islands and the question of whole- versus two-layer mantle convection are of broad scientific interest, as evidenced by their treatment in high school and college textbooks. The investigators have given, and will continue giving, lectures and extracurricular seminars on these topics to local high school and undergraduate students. Most of the requested budget will support a Ph. D. student, providing funds needed for a comprehensive, cross-disciplinary dissertation.
