Part 1. Two radiative transfer models with terrestial applications. Part 2. Testing the porcupine plate hypothesis

Citation

Gerstell, Marguerite F. (1995) Part 1. Two radiative transfer models with terrestial applications. Part 2. Testing the porcupine plate hypothesis. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/e7ex-9x13. https://resolver.caltech.edu/CaltechTHESIS:03302010-131151765

Abstract

Paper I: Goody's convolution theorem for obtaining the cumulative k-distribution of a gas mixture requires stronger assumptions than the multiplicative property of band transmission; thus new experimental investigations of its effectiveness were undertaken. The convolution was found to be a useful speed optimization of k-distribution calculations at high pressures. For low pressures a variety of mixing methods were compared, all taking advantage of the idea that stratospheric lines are too narrow to overlap. Appendix I discusses the context and application of k-distribution calculations. Paper II: We used a "quasi-random" radiative transfer model to estimate stratospheric radiative perturbations produced by SO_2 gas, silicate ash, and H_2SO_4 aerosols after the 1982 El Chichon eruptions. One week after the last eruption, net radiative heating perturbations exceeding 20 K/day were modeled at altitudes near 26 km. Silicate ash heating may have been balanced by global enhancement of stratospheric meridional circulation, with upward velocities of 1 cm/s near Chichon's latitude. Radiative forcing by silicate ash and SO_2 gas should be included in more comprehensive models of plume evolution. Particle size distributions inferred from ash fallout rates could be wrong if radiative heating is neglected. Paper III: Uncertainties in the solar spectrum can affect modeled net heating rates in the upper stratosphere by a factor of several. Variation among Antarctic surface albedo values in common use can affect modeled net heating rates in the lower stratosphere by tens of percent. Large uncertainties in polar cloud cover are less important to stratospheric heating models. I join Marcel Nicolet in urging support for a continuous solar observation program, and recommend that future intercomparisons of stratospheric radiation models prescribe a solar spectrum, to reveal other differences. Appendix 2 gives the details of some further validation and sensitivity tests for the quasi-random model. Paper IV: The Porcupine Plate was postulated in 1986 to explain difficulties in reconstructing anomalies 21 and 24 in the North Atlantic. Its main feature was thought to be a transpressive Eocene plate boundary along Charlie-Gibbs Fracture Zone. Eliminating data that could have been affected by subsequent movements of Greenland relative to North America leads to a picture that casts doubt on the Porcupine Plate hypothesis.

Thesis Files