Typescript. / Thesis (Ph. D.)--University of Hawaii, 1968. / Bibliography: leaves [119]-122. / ix, 122 l illus., tables

Studies of the evolution of hotspot swells, mantle plumes, and Hawaiian basalts are presented in three parts in this dissertation. In part 1, the evolution of mid-plate hotspot swells are simulated numerically as an oceanic plate rides over a hot, upwelling mantle plume. The transient heat transfer equations, with time- and space-dependent boundary conditions, are solved in cylindrical coordinates. Geophysical data are used to constrain the models. Formation of the Hawaiian swell requires a mechanism of convective thinning of the lithosphere. The models constrain the Hawaiian heat source to have a maximum anomalous temperature of 250-300°C, and a perturbing heat flux 5-6 times the background value. On the other hand, the Bermuda swell is likely produced by heat conduction due to weakness of the heat source. In part 2, an analytic model of axisymmetric mantle plumes is presented. Plume parameters beneath the lithosphere, which are constrained from the swell models, are used to infer the plume source regions. The Hawaiian plume likely originates near the core-mantle boundary, but other hotspots may have shallower sources. Chemical plumes are much narrower than thermal plumes because of low chemical diffusivity in the mantle. For mantle plumes driven by combined thermal-chemical diffusion, the chemical signature of the source regions may only be observed near plume centers. Finally, melt generation and extraction along the Hawaiian volcanic chain are discussed in part 3. As a part of the plate moves over the heat source, melting largely takes place in the region where the lithospheric material is engulfed and swept away by the flow of the heat source. At least three mantle components must be involved in the melt generation: the plume material, the asthenosphere, and the engulfed lithospheric material. Significant amount of melts may also come from direct melting of the upwelling plume at depths below the initial plate-plume boundary. Melt extracts continuously from an active partial melting zone of 10-20 km thick, which moves outward as heating and compaction proceed. The models explain quantitatively the general characteristics of Hawaiian volcanism as the result of plume-plate interaction.

Typescript. / Thesis (Ph. D.)--University of Hawaii, 1966. / Bibliography: leaves 156-168. / xiii, 168 l illus. (part mounted), tables

Using the controlled-source electromagnetic technique, resistivity soundings were obtained at 49 •locations around the summit caldera and upper rift zones of Kilauea volcano. Each sounding consisted of vector measurements of the magnetic field induced by a large-moment horizontal loop current source at discrete frequencies between 0.04 and 8 Hz. The source-to-sensor distances ranged from 2.5 to 13 km. The data have been computer-inverted to produce a best-fitting horizontally layered earth model. Although each sounding's interpretation is different in detail, the volcano’s structure appears simple and can be represented by four, subhorizontal layers. The surface layer is highly resistive and coincid.es with. the dry, basaltic overburden. At a depth of 500 to 1000 m, resistivities decrease abruptly to between 30 and 50 ohm-m, marking the top of the water-saturated zone. The third layer occurs between 2 and 3 km depth and has a resistivity of less than 10 ohm-m and a total conductance of about 200 mhos. This layer is underlain everywhere by highly resistive rock to a depth of at least 6 km, the estimated limit of penetration by this study. Pockets of low resistivity (less than 20 ohm--m) occur irregularly within the high-resistivity basement. Because of its widespread occurrence, 'the shallower conductive layer (layer 3) is probably water-saturated rock at high, temperature; however, the possibility of thin, intruded sills of magma contributing to the low resistivities cannot be refuted, The pockets of low resistivity within layer 4 occur at a depth of 5 km and are believed to be magma chamber 2 to 3 km deeper than models derived from earthquake hypocenter location and surface deformation studies. / ill