arXiv subjects
Publications and source records attributed to L. B. Lucy.
Exact analytic solutions are derived for radiation in time-dependent relativistic flows. The flows are spherically-symmetric homologous explosions or implosions of matter with a grey extinction coefficient. The solutions are suitable for testing numerical transfer codes, and this is illustrated for a fully relativistic Monte Carlo code.
The He I 5876A absorption line recently discovered in the spectrum of the companion to the millisecond pulsar PSR J1740-5340 is tentatively attributed to electron impact excitations due to the irradiation of its atmosphere by gamma-rays emitted by the pulsar's magnetosphere. Numerical calculations, similar to those carried out previously for Type Ib SNe, indicate that a pulsar beam with photon energies ~ 1 MeV gives rise to a 5876A line of the observed strength if the beam's spin-down conversion efficiency approaches 1%. However, a significant difficulty for the proposed mechanism is the strength of the singlet line at 6678A. Compared to the corresponding triplets, singlet lines are weak because of the loss of excitation when photons emitted in decays to the ground state ionize hydrogen atoms, an effect absent in the hydrogen-free atmospheres of Type Ib SNe.
The macroscopic quantizations of matter into macro-atoms and radiant and thermal energies into r- and k-energy packets initiated in Paper I is completed with the definition of transition probabilities governing energy flows to and from the thermal pool. The resulting Monte Carlo method is then applied to the problem of computing the hydrogen spectrum of a Type II supernova. This test problem is used to demonstrate the scheme's consistency as the number of energy packets N -> infinity, to investigate the accuracy of Monte Carlo estimators of radiative rates, and to illustrate the convergence characteristics of the geometry-independent, constrained Lambda-iteration method employed to obtain the NLTE stratifications of temperature and level populations. In addition, the method's potential, when combined with analytic ionization and excitation formulae, for obtaining useful approximate NLTE solutions is emphasized.
Two iterative techniques are described for decomposing a long-slit spectrum into the individual spectra of the point sources along the slit and the spectrum of the underlying background. One technique imposes the strong constraint that the spectrum of the background is spatially-invariant; the other relaxes this constraint. Both techniques are applicable even when there are numerous overlapping point sources superposed on a structurally-complex background. The techniques are tested on simulated as well as real long-slit data from the ground and from space.
Transition probabilities governing the interaction of energy packets and matter are derived that allow Monte Carlo NLTE transfer codes to be constructed without simplifying the treatment of line formation. These probabilities are such that the Monte Carlo calculation asymptotically recovers the local emissivity of a gas in statistical equilibrium. Numerical experiments with one-point statistical equilibrium problems for Fe II and Hydrogen confirm this asymptotic behaviour. In addition, the resulting Monte Carlo emissivities are shown to be far less sensitive to errors in the populations of the emitting levels than are the values obtained with the basic emissivity formula.
A strong emission line at 2.8935 mu discovered by Rubin et al. (2001) in an ISO SWS02 spectrum of the Orion Nebula is identified as the 4p - 4s multiplet of the O I triplets. Line formation is due to de-excitation cascades following UV-pumping of high triplet terms and occurs in the O I zone immediately behind the Hydrogen ionization front. This cascade mechanism also accounts for permitted O I triplet lines in the optical spectrum (Grandi 1975). An escape probability treatment of the O I cascades accounts for the strength of the 2.89 mu line and suggests interesting diagnostic possibilities for the optical lines.
Superlevel partitioning is combined with a simple relaxation procedure to construct an iterative technique for solving equations of statistical equilibrium. In treating an $N$-level model atom, the technique avoids the $N^{3}$ scaling in computer time for direct solutions with standard linear equation routines and also does not fail at large $N$ due to the accumulation of round-off errors. In consequence, the technique allows detailed model atoms with $N \ga 10^{3}$, such as those required for iron peak elements, to be incorporated into diagnostic codes for analysing astronomical spectra. Tests are reported for a 394-level Fe II ion and a 1266-level Ni I--IV atom.