Massive NLTE models for X-ray novae with PHOENIX

TL;DR

This paper develops advanced NLTE expanding atmosphere models for X-ray novae in the super-soft source phase, improving spectral fits and physical insights compared to previous hydrostatic models.

Contribution

The authors introduce new methods and improvements to the PHOENIX code, enabling more accurate NLTE modeling of expanding nova atmospheres during the SSS phase.

Findings

Models show significant effects of expansion on spectra.

Spectra are sensitive to atmospheric structure details.

Expanding models match observed spectra better than hydrostatic models.

Abstract

X-ray grating spectra provide the confirmation of continued mass loss from novae in the super-soft source (SSS) phase of the outburst. In this work expanding nova atmosphere models are developed and used to study the effect of mass loss on the SSS spectra. The very high temperatures combined with high expansion velocities and large radial extension make nova in the SSS phase very interesting but also difficult objects to model. The radiation transport code PHOENIX was applied to SSS novae before, but careful analysis of the old results has revealed a number of problems which lead to new methods and improvements to the code: 1) an improved NLTE module (a new opacity formalism, rate matrix solver, global iteration scheme, and temperature correction method); 2) a new hybrid hydrostatic-dynamic nova atmosphere setup; 3) the models are treated in pure NLTE (no LTE approximation for any opacity). With the new framework a modest amount of models (limited by computation time) are calculated. These show: 1) systematic behaviour for various atmospheric conditions, 2) the effect of expansion on the model spectrum is significant, and 3) the spectra are sensitive to the details of the atmospheric structure. The models are compared to the ten well-exposed grating spectra presently available: 5x V4743 Sgr, 3x RS Oph, and 2x V2491 Cyg. Although the models are on a coarse grid they do match the observations surprisingly well. Also, hydrostatic models are computed. The reproduction of the data is clearly inferior to the expanding models and, more importantly, their interpretation with hydrostatic models leads to conclusions opposite to those from expanding models. The models enable the derivation of accurate constraints on the physical conditions deep in the nova atmosphere that are revealed only in the SSS phase.