Deep X-ray observation of NGC 3221: everything everywhere all at once

TL;DR

This study leverages a deep XMM-Newton/EPIC-pn observation of the nearby edge-on, star-forming galaxy NGC 3221 to disentangle the galaxy’s high-energy components. By combining spatial, temporal, and spectral analyses, the authors confirm a low-luminosity AGN, reveal diverse day-scale variability among six ULXs, and identify two new predominantly soft sources. Spectral modeling across bands demonstrates a multiphase ISM with two thermal components ($T\sim0.15$ keV and $T\sim0.55$ keV) and reveals a temperature inversion where the extraplanar gas is hotter than the disk gas; a significant fraction of the X-ray luminosity arises from unresolved diffuse emission rather than point sources. The results favor a Compton-thick AGN interpretation for the nucleus and suggest that the X-ray luminosity function in a given galaxy can be highly time-variable, emphasizing the need for broadband, multi-epoch observations to accurately characterize X-ray populations and feedback processes in star-forming galaxies. The study highlights the importance of future >10 keV follow-ups and high-resolution imaging to robustly separate AGN, XRBs, superbubbles, and hot gas in the disk–halo interface.

Abstract

We present a comprehensive analysis of 475 ks (438 ks unpublished & 37 ks archival) XMM-Newton/EPIC-pn observation of a nearby, highly inclined, star-forming, luminous infrared galaxy NGC 3221 through spatial, temporal, and spectral information. We confirm the presence of a low-luminosity (presumably Compton-thick) AGN. The 0.4$-$12 keV luminosity and the hardness ratio of the six ultra-luminous X-ray sources (ULX) previously identified in Chandra data exhibit diverse variability on day-scale. The collective emission from unresolved sources exhibits a different day-scale variability. We have also discovered two new predominantly soft ($<1$ keV) sources. One of these has an enigmatic spectral shape featuring a soft component, which we interpret as a superbubble in NGC 3221, and a variable hard component from a compact object, which is unresolved from the superbubble. We do not confidently detect any X-ray emission from SN 1961L. The hot gas in the ISM (out to $\pm$6 kpc from the disk plane) and the extraplanar region (6$-$12 kpc) both require two thermal phases at $\sim 0.15$ keV and $\sim 0.55$ keV. The $\sim 0.55$ keV component is fainter in the ISM than the $\sim 0.15$ keV component, but the emission from the latter falls off more steeply with disk height than the former. This makes the extraplanar region hotter and less dense than the ISM. The proximity of NGC 3221 and the occurrence of the underluminous AGN offer a unique observing opportunity to study the hot diffuse medium in conjunction with nuclear and disk-wide point sources.

Figures (6)

• Figure 1: SDSS $gri$ (blue-green-red) image of NGC 3221. Positions of point sources (previously known or detected in this paper) are labeled. See §\ref{['sec:analysis']} for more details.
• Figure 2: Top to bottom: supersoft (0.4--0.7 keV), soft (0.7--1.0 keV), medium (1-5 keV), hard (5-12 keV) and broadband (0.4-12 keV) contours from XMM-Newton/EPIC-pn data overplotted on GALEX-NUV images of NGC 3221. Contours are constructed after subtracting OOT and QPB and correcting for vignetting. Contours are drawn on a linear scale, the outermost contour is 3$\sigma$ above the background, and increases by 1$\sigma$ inside. Left to right: different obsIDs are shown chronologically, with the earliest to the left. Six point sources detected in the Chandra data and identified as ULX in Luangtip2015 are shown here with green circles. Blue '+' symbol denotes SN 1961L. Yellow 'x' marks the optical/IR center of the galaxy, i.e., the AGN. The red diamonds are drawn at the positions of two previously unpublished sources that are not detected in the Chandra data. The PSF of EPIC-pn is shown with a hatched circle at the bottom-left corner of each image. The scale of 10 kpc is shown at the top left corner.
• Figure 3: Top: Count rate and bottom left (ten panels) hardness ratios of the point sources from XMM-Newton data, and bottom right (six panels) unabsorbed 0.3--10 keV luminosity of the ULXs from XMM-Newton and Chandra data. The position of each object has been shown in Figure \ref{['fig:optical']} and \ref{['fig:contour']}. The count rates for supersoft (0.4--0.7 keV), soft (0.7--1.0 keV), medium (1--5 keV), hard (5--12 keV), and broad (0.4--12 keV) bands are displayed in red, orange, green, blue, and black, respectively. The hardness ratio for the SED that is assumed to estimate the luminosity is shown with gray horizontal dotted lines.
• Figure 4: Top left: the extraction region for the AGN ("on-source" region) is shown in a green circle, and the background region is shown in a black dashed circle with the regions of red circles masked - plotted over the 0.4--12 keV science image of 0922170301. 0.4-12 keV spectrum of the AGN is fitted with an absorbed powerlaw (and discarded; top right), and a combination of an absorbed disk blackbody and absorbed powerlaw (bottom left). A physically better (and statistically indistinguishable) alternative of an absorbed circumnuclear emission (apec) and the reflection of a Compton-thick AGN (pexrav) is shown in the bottom right. ObsIDs 0922170101, 0922170201, 0922170301, and 0922170401 are shown in red, green, black, and blue.
• Figure 5: Extraction region (top left) for the new source--1 is shown in green circle, and background region (bottom left) is shown in circle with the regions of red circles masked - plotted over 0.4--12 keV science image of 0922170301. 0.4--12 keV spectrum is fitted with an absorbed powerlaw (and discarded; middle), and with an additional absorbed apec at the redshift of NGC 3221 (right). ObsIDs 0922170101, 0922170201, 0922170301, and 0922170401 are shown in red, green, black, and blue.