X-ray spectroscopy method of white dwarf mass determination in intermediate polars. External systematic uncertainties

TL;DR

This study evaluates systematic uncertainties in using X-ray spectra to determine white dwarf masses in intermediate polars, introduces a new spectral model grid, and applies it to observational data, confirming the method's validity for high-luminosity systems.

Contribution

The paper develops a new X-ray spectral model grid incorporating updated mass-radius relations and systematically assesses uncertainties affecting WD mass measurements in IPs.

Findings

Average WD mass in sample is 0.82 M_sun.

Universal spectral grid is valid for high-luminosity IPs.

Low-luminosity IPs require individual modeling.

Abstract

The masses of white dwarfs (WDs) in intermediate polars (IPs) can be determined from the shape of their hard X-ray spectra. Here we study the importance of all possible systematic uncertainties in this X-ray spectroscopy method, including finite radii and rotation of magnetospheres, finite accretion column height and accretion-flow inclination relative to the WD surface. We also investigate the importance of accretion-heated envelopes on WD surfaces in IPs which are increasing WD radii. Their presence changes the commonly used mass-radius relation for cold white dwarfs. As a first approximation we use thick ($10^{-4}M_\odot$) hydrogen envelope models with a surface temperature of 30 kK. We present a new model grid of hard X-ray spectra of high-luminous IPs computed among other things with using a new mass-radius relation. This grid is used for fitting Swift/BAT spectra of 47 IPs. The average WD mass in this sample is 0.82 $M_\odot$ and coincides with the average WD mass in cataclysmic variables obtained by optical methods. This means that the calculated hard X-ray spectra and the assumptions made that the magnetospheric radii in IPs are close to the corotation radii, and the relative heights of the accretion columns are small are basically correct, because most IPs have high luminosities. But this universal grid (as well as previous universal grids) cannot give correct results for the low-luminous IPs with probably relatively tall accretion columns on the WD surfaces. Such IPs have to be investigated with individual accretion column models.