Resmini, R.G., 1994, Dynamics of magma within the crust: A study using crystal size distributions: Baltimore, Maryland, The Johns Hopkins University, Ph.D. dissertation, 342 p.
Crystal size distributions (CSDs) have been measured for plagioclase and olivine in comagmatic suites of basaltic and andesitic lavas from Dome Mountain and Crater Flat, Nevada, and from Atka Volcano, Alaska, with the goal of understanding near-surface magmatic residence time and overall magmatic system behavior. Residence times calculated from the CSDs using a growth rate of 10-10cm/sec range from 1.5 to 10.8 years for Dome Mountain; 2.8 to 11.0 years for Crater Flat; and 6.0 to 13.0 years for Atka. Residence times increase with decreasing age of the lavas at Crater Flat, possibly indicating waning magmatism. Residence times vs. flow stratigraphy for Dome Mountain and Atka indicate that each system may have been derived from: (1) an open system that may have operated at a steady-state of magma chamber replenishment and withdrawal; or (2) a batch of magma that was sampled by several temporally closely spaced eruptions between which little cooling occurred. A linear relationship between CSD slope and intercept favors the latter interpretation. The well-known increase in solids with decreasing temperature is used as a closure condition with a nucleation model to predict crystal growth rates and CSDs in igneous bodies. With nucleation rate as a linear function of cooling rate, a conductive cooling model allows an estimate of nucleation rate and density within the solidification fronts of a 40-m-thick sill and the production of synthetic CSDs. The synthetic CSDs are remarkably linear; they closely resemble those from actual rocks and show no dependence on position in the sill. Although maximum crystal size changes considerably, the slope of the CSDs is uniform. Through-sill relations of log (growth rate) to log (distance from contact) and log (nucleation rate) to log (cooling rate) mimic the same relations described by Cashman (1993; Contrib. Mineral. Petrol., v. 113, p. 126-142) for natural systems. The results indicate that crystallization is more accurately modeled with cooling rate than with undercooling. Analytical and numerical solutions to the two-dimensional Poisson equation are presented to better understand the linear relationship between surface heat flow and the radiogenic heat production of rocks. Though agreement between analytical and numerical solutions is excellent, model heat fluxes and radioelement enrichment depths are over- and underestimated, respectively. The discrepancies are due to horizontal heat conduction. A method of removing the effects of horizontal heat conduction and thus obtaining the true basal heat flux and radioelement enrichment depth is presented, as are preliminary results of analytical solutions to the three-dimensional Poisson equation.
Theses and Dissertations
