Cosmic & Galactic Backgrounds

The cosmic background, even if not the object of the observation, must be modeled explicitly as it is a significant background in all directions and at all energies, and it varies significantly both in intensity and spectral shape over the sky. The cosmic background can usually be modeled by four or fewer components.

First, a cool ($E\sim0.1$ keV), unabsorbed thermal component representing emission from the Local Hot Bubble and heliosphere. It may be useful to look at the work of Snowden et al (1998) or Liu et al. (2017) to get an understanding of the strength of the LHB in the direction of your field. The exact temperature of this component will depend on the model (Raymond & Smith, APEC, MeKaL, etc.) and abundances (Anders & Grevesse, Wilms, etc.) used. (Note that the temperature quoted by Liu et al. is not appropriate for the current version of APEC!)

Second, a cool ($E\sim0.1$ keV), absorbed thermal component representing emission from the cooler halo. This component may have an intrinsically zero intensity or may be effectively absorbed if the Galactic column is greater than a few times $10^{20}$ H I cm$^{-2}$. In either case the component can be deleted from the fitted model.

Third, a higher temperature ( $E\sim0.25-0.7$ keV), absorbed thermal component representing emission from the hotter halo and/or intergalactic medium. At high Galactic latitudes, the 0.25 keV component tends to dominate. In the plane of the Galaxy, the 0.7 keV component can be quite strong. For an observation near the Galactic plane, this component may also end up representing distributed thermal emission through the Galactic disk.

Fourth, an absorbed power law with $\alpha\sim1.45$ representing the unresolved background of cosmological sources.

Since getting some of these normalizations correct is important, consider a fit of the model


to the RASS data for $b>85\arcdeg$ and $45\arcdeg<\ell<270\arcdeg$. We found

$N_{LHB}=8.63\times10^{-7}$ for $kT=0.0901$ keV
$N(H)=0.0136\times10^{22}$ cm$^{-2}$
$N_{soft~halo}=2.57\times10^{-6}$ for $kT=0.1074$ keV
$N_{hard~halo}=1.06\times10^{-6}$ for $kT=0.2373$ keV
$N_{\Gamma}=1.11\times10^{-6}$ for $\Gamma=1.45$

Note that since this is RASS data, the normalizations will be per square arcminute. In this fit the absorbing column was held fixed since the absorption is due mainly to atomic gas which is well traced by HI 21 cm emission. At high Galactic latitude there is no perceptible 0.75 keV component. These values are representative of those found at high Galactic latitudes. We can now consider what values might be expected for other observations.

The normalization for the Local Hot Bubble emission (which also contains some of the heliospheric solar wind charge exchange emission) is somewhat higher than usual. Reference to Snowden et al (1998) shows why; the path length through the Local Hot Bubble is, in general, greater towards the poles than it is in the Galactic Plane.

The normalization of the Galactic halo components is typical for high Galactic latitudes; at lower Galactic latitudes the emission may be so strongly absorbed that the normalization is very uncertain. Kaaret et al (2020) have shown that the $3\over4$ keV Galactic halo emission is patchy, so the normalization of the harder component can be quite variable.

The normalization of the extragalactic power law has been studied by a number of authors and is $\sim10.5$ keV cm$^{-2}$ s$^{-1}$ sr$^{-1}$ keV$^{-1}$. This normalization translates to $8.88\times10^{-7}$ photon cm$^{-2}$ s$^{-1}$ keV$^{-1}$ arcmin$^{-2}$ at 1 keV, the standard xspec units for a single square arcminute. Note that the normalization for your particular observation is likely to be lower as you have removed point sources. In principle, given the cosmic normalization, the point source detection limit, and the luminosity function for extragalactic sources (suitably corrected to the absorbing column over the field of view), one can calculate the expected normalization. The XMM-ESAS package provides the task point-source which will calculate the normalization using several different models and user-selected parameters. In practice one finds significant variation. However, it is important to get this normalization correct because errors in this normalization are (usually) highly correlated with errors in the normalization of the SPF component; fixing the value of $N_{\Gamma}$ to its expected value may help to better determine the normalization of the soft proton component.

The absorbing column density can be found using the HEASARC NH Tool:,
the HEASARC X-ray Background Tool:,
or the Planck maps. What one desires is the total column density of hydrogen, in all of its forms: H I, H II, and H$_2$. One assumes that the ratio between the hydrogen column density and the column density of the actual absorbing atoms is reasonably set by the chosen abundances. There is a certain amount of complexity here, as the H I 21 cm measurements do not capture the molecular or ionized components of the absorbing ISM, and the abundances may vary significantly from one line of sight to another. Conversely, the translation from Planck dust emission to the actual column density of absorbing atoms is not as straight-forward as it might seem.