| Summary: | This thesis consists of three studies which focus on characteristics of volcaniclastic particles to gain insight to eruptive and depositional processes involved in their formation. The combination of modal, grain-size, geochemical, and particle-shape analyses results in a multi-dimensional approach to studying volcaniclastic deposits. The first chapters deal with volcanic clasts deposited adjacent to hotspot ocean islands (Gran Canaria and Iceland). Clastic apron sediments can be especially useful, recording a complete geologic history including stages of volcanism absent from the subaerial record. Apron sediments from Gran Canaria, recovered at Site 953 (ODP Leg 157), were examined for insight to hotspot island evolution and clastic apron formation. Modal and morphological analyses of particles from Unit V (850 to 889 mbsf), representing the late phase of the shield-building process, characterize the transition from submarine volcanism to explosive shallow water and subaerial volcanism over time. The second chapter concerns volcaniclastic deposits from Iceland's southeast coast collected following the jökulhlaup (glacial flood) produced by the October 1996 Gjálp eruption beneath Vatnajökull glacier. Although jökulhlaups are potentially important sources of volcaniclastic sediment to the North Atlantic, this is the first study to examine their effects to the marine environment. Geochemical, grain-size, and modal analyses constrain the source of volcanic material transported by the jökulhlaup and the extent of deposition on the seafloor. The third study focuses on quantitative evaluation of the shapes of volcanic particles and its potential as a geologic tool. The morphologies of volcanic particles can yield information regarding eruptive parameters such as magma viscosity, volatile content, interaction with water, and modes of transport and deposition. This study has resulted in a new fractal method of shape analysis (based on dilation) which produces a full spectrum of fractal dimensions over a range of scales for each particle. Using samples produced by a variety of eruptive styles, the technique proves more effective than fractal methods yielding one or two dimensions per particle. Also, production of multiple fractal dimensions for each particle allows incorporation of multivariate statistical analysis, strengthening the differentiating power of the technique.
|
|---|