Evidence from Spectra of Bright Fireballs

Abstract

Spectral data with dispersions from 11 to 94 Å/mm on 4 fireballs of actual brightness of —4 to —12 magnitude and with velocities of about 30 km/s at 70 to 80 km heights are used for studies of meteor radiation problems. Previously published analyses need revision for two main reasons: (a) the absolute values of oscillator strengths of Fe I lines from laboratory data were recently recognized to be 1 order of magnitude lower, (b) the luminous efficiency factor τ of Fe I is now much better known from several different experiments. The radiation of fireballs is found to be strongly affected by self-absorption. But if the emission curve of growth is used for correction oj the self-absorption of Fe I lines, a great discrepancy between spectral data and efficiency data for total Fe I light is found. If one assumes that the self-absorption is superposed on another effect, a decrease of the dimensions of the radiating volume with increasing lower potential E1, the spectral data on Fe I lines will be in agreement with the luminous efficiency of total Fe I meteor radiation. Formulas for emission curve of growth and Boltzmann distribution including this effect are derived. This effect is important for fireballs brighter than about —1 or — 2 magnitude, while self-absorption seems to be important even for fainter meteors. The optically thin radiation of all Fe I lines might be expected for meteors fainter than +5 magnitude. Excitation temperature of 5500° K and relaxation time of 0.02 s were found as typical values for the Fe I radiation of fireballs studied. The light of fireballs is emitted during a relatively long relaxation time, which is many orders of magnitude longer than the time necessary for spontaneous radiation of excited Fe I atoms. The dimensions of the radiating volume of Fe I gas for lines with E1 = 0 were found to be 0.3X9 m at 0 absolute magnitude and 2×60 m at —10 absolute magnitude. It ivas not possible to determine any realistic abundances of other elements due to small numbers of lines for an analysis independent of Fe I, while the Fe I curve of growth cannot be used for other elements, because the radiation originates mainly from the effective surface of the radiating volume. A general formula for meteor radiation is also derived and compared with the conventional luminosity equation.