The miniJPAS Survey: Detection of the double-core Ly$\alpha$ morphology for two high-redshift QSOs

The Javalambre-Physics of the Accelerated Universe Astrophysical Survey (J-PAS, Benitez et al. 2014) is a wide-area photometric survey soon to be conducted at the Javalambre Astrophysical Observatory (OAJ), located at the Sierra de Javalambre (Teruel, Spain). It is poised to obtain a deep, sub-arcsec spectrophotometric map of the northern sky across 8000 deg². The survey will use the 2.5 m JST/T250 telescope and obtain multi-band imaging in optical bands with an effective field of view of 4.2 deg² and a plate scale of 0.225 arcsec pixel^-1. The J-PAS filters provide low-resolution spectroscopy (resolving power, R $\sim$ 60; also referred as J-spectra in this paper) of a large sample of astronomical objects with 54 NB filters (3780 to 9100 Å; $\Delta$$\lambda$= 145 Å), one blue medium-band (MB) filter (3497 Å; $\Delta$$\lambda$= 509Å), one red MB filter (9316 Å; $\Delta$$\lambda$= 635Å), and three Sloan Digital Sky Survey (SDSS) Broad-Band (BB) filters. J-PAS nominal depth at signal-to-noise (S/N)$\sim$ 5 is between 22 to 23.5 AB mag for the NB filters and 24 for BB filters Bonoli et al. (2021). With a large set of NB filters, J-PAS makes it suitable for an extensive search for extended Ly$\alpha$ emission around QSOs from redshifts $z=2.11$ to 6.66. Meanwhile, the wide spectral coverage of the J-spectra can cover Ly$\alpha$ , SIV+OIV, CIV, HeII, CIII], CII, and Mg II at $z>2$, which allows us to study the properties of high redshift QSOs.

The miniJPAS survey (Bonoli et al. 2021) observed four All-wavelength Extended Groth Strip International Survey (AEGIS; Davis et al. 2007) fields in 2018-2019 with the pathfinder camera and detected more than 60,000 objects in a total area of 1 deg². The miniJPAS survey demonstrates the capabilities and unique potential of the upcoming J-PAS survey (Hernán-Caballero et al. 2021; González Delgado et al. 2021; Martínez-Solaeche et al. 2021; Queiroz et al. 2022; Martínez-Solaeche et al. 2022; Chaves-Montero et al. 2022). More details about the observation and data reduction of miniJPAS are given in Bonoli et al. (2021). We obtained the miniJ-PAS images, flux catalogs, filter curves, and all other instrument information of these QSOs from the J-PAS database (CEFCA archive)¹¹1http://archive.cefca.es/catalogues/minijpas-pdr201912.

With miniJPAS, we report two interesting QSOs, SDSS J141935.58+525710.7 (RA: 214.8983, Dec: 52.953, hereafter QSO1) at a redshift of $z_{\mathrm{spec}}=3.218$ and SDSS J141813.40+525240.0 (RA: 214.5559, Dec: 52.8778, hereafter QSO2) at redshift $z_{\mathrm{spec}}=3.287$, with the double-core Ly$\alpha$ morphology (Fig. 1). We selected these two QSOs in a systematic search on the Ly$\alpha$ nebulae around 59 (SDSS and miniJPAS cross-matched targets) spectroscopically confirmed high redshift QSOs ($z>2$) using miniJPAS (Rahna et al., in prep). In the first step, we selected spectroscopically confirmed QSOs at z$>$2 from SDSS survey DR16 and cross-matched them with miniJPAS 4 AEGIS fields, yielding 61 QSOs and after discarding the QSOs with bad warning flag of spectra, the final sample consists of 46 QSOs at 2$<$z$<$4.29. We utilized the contiguous NB imaging of miniJPAS for our selection criteria of Ly$\alpha$ nebulae. Our target selection of Ly$\alpha$ nebulae is based on the excess of Ly$\alpha$ flux and morphology in the filter covering the Ly$\alpha$ line and continuum. A QSO with a size of Ly$\alpha$ morphology greater than its continuum morphology at a 2 sigma level of Background is considered a Ly$\alpha$ nebula, after accounting for its point spread function (PSF). PSFs used in the analysis are downloaded from the JPAS database, where PSFs are created from PSFex (Moffat profiles) using nearby stars. Only these two targets from the parent sample show double-core morphology in their Ly$\alpha$  NBs. There are four other QSOs in the same redshift range of these two targets and their Ly$\alpha$ morphology doesn’t show double-core morphology. More details about the parent sample are explained in Rahna et al., in prep. Spectroscopic redshifts were available for these two QSOs thanks to SDSS/BOSS (Dawson et al. 2013; Pâris et al. 2017).

QSO1 has broadband observations with the Hubble Space Telescope (HST) Advanced Camera for Surveys (ACS/WFC; AEGIS Survey (Davis et al. 2007)). The HST/ACS F606W and F814W wide band images of QSO1 show a single point-like morphology (Fig. 2). The X-ray image of QSO1 from stacked Chandra observation of 800 ks (AEGIS-X survey (Laird et al. 2009; Nandra et al. 2015) and the XDEEP2 survey (Goulding et al. 2012)) shows QSO1 as a point-like source ²²2QSO1 was located at $\approx$7 arcmin from the aimpoint of ACIS-I: at that distance, the ACIS-I PSF is elongated and distorted, with a 50% radius of about 2 arcsec (= 50% of the counts fall within 2 arcsec). Combining this large PSF with the moderately small number of counts (about 500 counts in the stacked image), means that some faint, extended X-ray structures on scales $<$2 arcsec would be easily missed, especially with automated classifications of sources in large surveys.. The radio observation of QSO1 with VLA (Very Large Array) in 6cm band (Ivison et al. 2007) shows weak but single-source detection. QSO2 has not been observed by HST, Chandra, XMM-Newton (X-ray Multi-mirror Mission), nor VLA.

The NB images covering the Ly$\alpha$ line for the QSOs analyzed in this paper show a serendipitous double-cored morphology of Ly$\alpha$ (Fig. 1). The redshifted Ly$\alpha$ emission lines of the two QSOs fall into two NBs of J-PAS (J0510 and J0520 for QSO1 and J0520 and J0530 for QSO2). QSO1 shows features of double core morphology in both Ly$\alpha$ images, albeit more significant in J0520. QSO2 shows double-core morphology only in J0520. This may be due to the fact that J0530 covers mostly the NV line of QSO2 and the NV contamination may cause different morphology (Fig. 3). Furthermore, J0530 also shows an elongated core at the center with extended morphology compared to its compact and circular PSF. On the other hand, the continuum image only shows a compact single-core morphology. We find that the double-core morphology is not caused by the PSF although it is elongated in the J0520 filter. This also confirms with the morphology of nearby stars, which only exhibit single-core morphology with elongated PSFs. Based on the morphology in the continuum bands, especially the single point structures revealed in HST F606W and F814W images (Fig. 2), we infer that the double-core structures in the J0520 image of the two QSOs are caused by the diffuse and resonant-scattered Ly$\alpha$ radiation (see Section 4).

To quantitatively analyze the double-core morphology of the Ly$\alpha$ image of the two QSOs, we modeled the double-core with two PSF profiles through GALFIT (Peng et al. 2002). First, we subtracted the continuum flux from the NB image by using a PSF (of the NB image) scaled to the continuum flux to create the Ly$\alpha$ only image. The continuum value is taken from the nearby filter with the best SNR, which is consistent with the convolution of the SDSS spectra and the JPAS filters. Then, we convolved the two PSFs with the Ly$\alpha$ flux and subtract it from the Ly$\alpha$-only image to get the residual (Fig. 1). The contour levels of the model created with two PSFs clearly show the double-core morphology compared to the actual elongated single-core PSF. This demonstrates that the double-core is not due to the PSF effects. Input and output parameters of GALFIT fitting are given in Table 1. We note that the center of the two GALFIT input PSFs (blue crosses) in J0520 doesn’t coincide with the peak of the Ly$\alpha$ emission in the NB images because of the distorted structure of the PSF. Two cores are separated by a distance of 11.07$\pm$2.26 kpc in QSO1 and 9.73$\pm$1.55 kpc by QSO2 and SB peak at 18 and 9 sigma level of background standard deviation (STD).

Besides the double-core morphology in the Ly$\alpha$  images, both QSOs also show diffuse and spatially extended Ly$\alpha$ radiation. The spatial extent was measured at $2\sigma$ isophote above the background STD (10^-16 – 10^-17 erg s^-1 cm^-2 arcsec²) from each image (Fig. 4). Compared to the size of the QSOs in the J0520 filter, QSO1 is more extended in J0510 ($\sim$ 45.21 kpcs), with a size difference of $\sim$ 15.07 kpcs, and QSO2 is more extended in J0530 ($\sim$ 44.89 kpcs), with a size difference of $\sim$ 22.44 kpcs (see APER size column of Table 2). This may be due to the difference in the depth of the two filters (in Table 2). However, the current miniJPAS has limited ability to estimate the actual size of Ly$\alpha$ nebulae due to shallow observation depth (480 secs) and PSF effects.

We present the J-spectra (pseudo spectra) and the SDSS spectra of our two QSOs in the upper panels of Fig. 3. The SDSS spectra are obtained from the final data release of Baryon Oscillations Spectroscopic Survey (BOSS; Dawson et al. 2013) in SDSS DR12 (Pâris et al. 2017). QSO1 had been observed many times as part of the SDSS reverberation mapping project (Hemler et al. 2019) (59 epochs from 2014 to 2020) and the spectrum in Fig. 3 was taken on March 19, 2019, while the QSO2 spectrum was observed on June 6, 2013. In BOSS, the spectrum was observed through 2 arcseconds (diameter) fibers. The J-spectra in Fig. 3 is composed of the atmosphere-extinction corrected photometry (APER2-2^′′ diameter) of all J-PAS NB filters obtained from the CEFCA archive (López-Sanjuan et al. 2019). The prominent emission lines such as Ly$\alpha$, SIV+OIV, CIV, HeII, and CIII] detected in SDSS are visible in J-spectra. The properties of these filters are presented in Table 2. These two QSOs are type 1 QSOs characterized by their broad emission lines. The spectroscopic Ly$\alpha$ profile of each QSO shows an asymmetric double-peaked profile and is dominated by the red peak, which is consistent with the general picture of IGM and CGM absorption Gurung-López et al. (2020). The broadened Ly$\alpha$ profile indicates radiative transfer effects. The double peak in the CIV line of QSO1 may be due to the presence of CIV broad absorption lines (BALs), with high outflow velocity (vmax = 3243 km/s: Hemler et al. 2019).

By utilizing miniJPAS contiguous NB imaging, we can study the emission line morphology in various wavelengths (as shown in Fig. 4). We note that other emission lines observed through the NB and BB images do not exhibit such prominent double-core structures as in the Ly$\alpha$ image.

In Table 2, we list the emission line filter properties, APER size used for estimating luminosity in Col. 11 and we estimate the emission-line luminosities from miniJPAS images and SDSS spectra. The APER size is the spatial extent measured at $2\sigma$ isophote above the background STD from each image. In addition to Ly$\alpha$, QSO1 also exhibits spatially extended emission in HeII (J0690), CIV (J0650), and CIII (J0690), while QSO2 shows extended morphology in CIV (J0660) (Fig. 4). The UV-continuum emission line (e.g., J0500, J0740) and BBs (gSDSS, rSDSS, iSDSS) are $\sim$ 15kpc in size in both QSO1 and QSO2 and do not show spatially extended emission, as in Ly$\alpha$ and CIV. The extended PSF in some emission line filters also contributes to the extended morphology. The comparison of 2$\sigma$ contours on NB image (green) with PSF (white) in Fig. 4 shows how much the nebula is extended in different emission lines.

To derive the emission line luminosities of the extended nebulae from miniJPAS images, we adopted a method that exploits contiguous narrow band filters of J-PAS. In this method, we use two adjacent NB filters covering the emission line and the adjacent continuum (by assuming that the source continuum is approximately constant in the nearest NB filter covering the continuum) to calculate continuum-subtracted emission line fluxes. For Ly$\alpha$ , we used two filters (one on each side of the Ly$\alpha$ line: J0490, J0560 for QSO1 and J0460, J0570 for QSO2; Fig. 3) to estimate the continuum because, at z$\sim$3, the blue part of the spectrum may be attenuated by the IGM (Inoue et al. 2014) or any intervening NHI gas. The emission line luminosity estimated from such flux is tabulated in Table 2. The spectral line widths of QSO1 and QSO2 are significantly wider than the filter width of J-PAS/miniJPAS NBs, which will result in a loss of flux in the luminosity calculated from miniJPAS images in a single filter comparison to SDSS spectra. The line luminosity calculated from SDSS spectra (by convolving J-PAS filter curves with SDSS spectra) and from miniJPAS NB images in the same aperture size of 2 arcsecs are within 1$\sigma$ error limit (of L calculated from miniJPAS), except L${}_{Ly\alpha\leavevmode\nobreak\ }$and L${}_{Ly\alpha\leavevmode\nobreak\ +NV}$ in QSO1 and L_CIV in QSO2. This variation might be due to the PSF effects. The luminosity estimated within the 2$\sigma$ isophote size (col. 10) shows a significant increase in Ly$\alpha$ , SIV+OIV, CIV, and HeII (weak) in QSO1 and Ly$\alpha$ in QSO2 as aperture size increases. Besides any possible intrinsic variability, the different line flux measurements may also contribute to the flux difference.

The relative strength of the three important brightest UV emission lines such as Ly$\alpha$, CIV, and HeII provide a powerful probe of the thermodynamic properties of the gas, its ionization state, gas density, and metallicity. These emission lines can also disentangle the powering mechanism for the Ly$\alpha$ emission. For example, line ratio diagnostic diagrams can be used to distinguish the photo-ionization scenario and shock scenario (e.g., Allen et al. 1998; Arrigoni Battaia et al. 2015). In the cooling zone behind high-velocity shocks, UV collisionally excited lines are strongly excited at high temperatures (2 $\times 10^{4}-10^{5}$ K), compared with photoionized plasmas, where these species are excited at $10^{4}$ K (Allen et al. 1998). Therefore, UV line intensities are predicted to be much stronger in shock scenarios than in simple photo-ionization scenarios. Arrigoni Battaia et al. (2015) demonstrated the CIV/Ly$\alpha$ versus HeII/Ly$\alpha$ diagram as a method to discriminate between shock and photoionization models with strong upper limits calculated above 10^-18 erg s^-1 cm^-2 arcsec² (at 2$\sigma$).

The line ratios of CIV/Ly$\alpha$ and HeII/Ly$\alpha$ are 0.262 $\pm$ 0.008, 0.028 $\pm$ 0.005 for QSO1 and 0.48 $\pm$ 0.024, 0.048 $\pm$ 0.022 for QSO2, which is estimated by the ratio of fluxes calculated from SDSS spectra. These values are consistent with the AGN photoionization (optically thick) and shock model for both QSOs explained by Arrigoni Battaia et al. (2015). But we would need to reach SB level of 10^-18 to 10^-19 erg s^-1 cm^-2 arcsec² at 2$\sigma$ to trace the full extent of Ly$\alpha$ nebulae to discriminate these models. However, due to the limited depth (10^-16 erg s^-1 cm^-2 arcsec² at 2$\sigma$) and spatial-spectral power of miniJPAS, we cannot constrain further details about different models. A deep J-PAS like NB surveys or IFU observation with an 8-10m telescope would help to give a more definitive conclusion on this aspect.

There is no clear definition for different types of Ly$\alpha$ nebulae. It can be roughly classified into Ly$\alpha$ halos (LAH), Ly$\alpha$ blobs (LABs), and enormous Ly$\alpha$ nebulae (ELAN) based on their luminosity and size in the Ly$\alpha$ wavelength (Ouchi et al. 2020). The LAHs are spatially extended Ly$\alpha$ nebulae with a physical scale of 1 to 10 kpc and a Ly$\alpha$ luminosity of $\sim 10^{42}-10^{43}$ erg s ^-1 (Hayashino et al. 2004; Momose et al. 2014, 2016; Guo et al. 2020), LABs have a physical scale of $\sim$ 10 -100 kpc and a luminosity of 10${}^{43}-10^{44}$ erg s ^-1 (Steidel et al. 2000; Matsuda et al. 2004; Shibuya et al. 2018; Herenz et al. 2020), and ELANs are the most luminous ($>10^{44}$ erg s ^-1) and extend to a large-scale of several hundreds of kpc (larger than the virial radius of dark matter halo) around $z>2$ QSOs (Cantalupo et al. 2014; Hennawi et al. 2015; Cai et al. 2017; Arrigoni Battaia et al. 2018; Nowotka et al. 2022). Based on these classifications Ly$\alpha$ nebulae around two QSOs come under bright LABs with a Ly$\alpha$ luminosity of $\sim 3-6\times 10^{44}$ erg s ^-1 and physical size of $\sim$ 30 – 45 kpc (Fig. 5). These QSOs are the most luminous and have a similar range of luminosity as ELANs (e.g., Slug, Jackot, MAMMOTH-1 in Fig. 5). Most importantly the spatial extent of the Ly$\alpha$ nebulae strongly depends on the limiting surface brightness. Kimock et al. (2021) demonstrated Ly$\alpha$ luminosity and enclosed area as a function of limiting surface brightness through high-resolution cosmological zoom-in simulations. Comparing with their simulation results suggest that the current miniJPAS depth can only trace a small area near the center of nebulae and these two nebulae maybe two new ELANs. Much deeper NB imaging is needed to confirm whether these objects are new ELANs.

Here, we discuss several mechanisms, including resonant scattering (e.g., Hayes et al. 2011; Steidel et al. 2011), gas photo-ionization by central AGN (e.g., Haiman & Rees 2001; Geach et al. 2009; Hayes et al. 2011), shocks due to gas flows (e.g., Taniguchi & Shioya 2000), and cold gas accretion (e.g., Haiman et al. 2000; Fardal et al. 2001; Furlanetto et al. 2005; Dijkstra & Loeb 2009; Rosdahl & Blaizot 2012), which could be responsible for the extended Ly$\alpha$ emission around QSOs. A combination of all these mechanisms may also act together to power Ly$\alpha$ nebulae.

The morphology of rest-frame UV lines such as CIV $\lambda$1549 and HeII $\lambda$1640 along with Ly$\alpha$  provide additional information about different physical mechanisms explained in the previous section. For example, CIV/Ly$\alpha$ ratio provides a diagnostic for the halo gas metallicity and ionization state of halo gas in the CGM, while extended CIV morphology indicates the size of the metal-enriched halos from galactic outflows from central AGN (Arrigoni Battaia et al. 2015). Photoionization and galactic outflows are the two possible reasons for the Ly$\alpha$ emission if there is a detection of extended emission from both CIV and HeII lines. However, the UV line ratios (e.g., CIV/Ly$\alpha$ and HeII/Ly$\alpha$ ) are different in both scenarios. If the Ly$\alpha$ nebulae is powered by shocks due to galactic-scale outflows or a shell-like or filamentary morphology with large Ly$\alpha$  width of $\sim$ 1000 km/s and enormous Ly$\alpha$ ($\sim$ 100 kpc) size is expected (Taniguchi & Shioya 2000; Taniguchi et al. 2001; Ohyama et al. 2003; Wilman et al. 2005; Mori & Umemura 2006). Both QSOs have a large Ly$\alpha$  width of $>$ 1000 km/s, suggesting that shell or filamentary-like morphology could be expected for the extended Ly$\alpha$ emission surrounding them. However, the nominal depth of miniJPAS data cannot trace the extended size of these nebulae. Additionally, the NB covering the CIV emission line in both QSOs show extended morphology which suggests that the neutral gas expelled by outflows that are carbon (metal) enriched.

In gravitation cooling radiation, collisionally excited CGM gas emits Ly$\alpha$ photons and converts gravitational energy into thermal and kinetic energy as it falls into the dark matter (DM) halo potential. The luminosity of Ly$\alpha$ nebulae produced by cooling radiation (in a 10¹¹ DM halo) predicted to be $\leq$ 5$\times 10^{41}$ erg s ^-1 through the hydrodynamic simulation of Dijkstra & Loeb (2009) and Rosdahl & Blaizot (2012). If Ly$\alpha$ nebulae has a luminosity higher than this value, it suggests that the mechanism producing Ly$\alpha$ emission is not only due to gravitation cooling radiation. Additionally, the powering mechanism for Ly$\alpha$ emission due to gravitational cooling radiation is not expected to produce extended CIV emission, but HeII emission can be expected (Arrigoni Battaia et al. 2015). This implies that the increased Ly$\alpha$ luminosity and extended CIV emission in both QSOs exclude a significant contribution from gravitational cooling.

Resonant scattering does not contribute to the large scale SB of ($>$ 100kpc) of Ly$\alpha$ nebulae because resonantly scattered photons can escape the system effectively only in very small scales ($<$ 10kpc) (Cantalupo et al. 2014; Borisova et al. 2016). Since CIV is a resonant line, the extended CIV emission in both QSOs could also arise due to the resonant scattering by the same medium scattering Ly$\alpha$ with non-detection in the extended non-resonant HeII line. HeII does not have an extended structure in QSO2, but QSO1 has a faint extended emission signature in miniJPAS images. Therefore, the extended emission for HeII in QSO1 may be powered by a different scenario.

The double-peaked line profile in Ly$\alpha$ spectrum of QSOs at high redshift is common due to the resonant nature of Ly$\alpha$ (Dijkstra et al. 2006; Tapken et al. 2007; Yang et al. 2014; Cai et al. 2017). The Ly$\alpha$ line profile shows asymmetric due to the bulk motion of neutral hydrogen and Ly$\alpha$ photons escape in a double-peaked line due to the high optical depth at the line center and it is absorbed and re-emitted in another direction (Sanderson et al. 2021; Dijkstra et al. 2007; Matthee et al. 2018). Consequently, Ly$\alpha$ photons diffuse spatially and also in frequency space. The double-peak profile in the Ly$\alpha$ emission line seen in the SDSS spectra of both QSO1 and QSO2 indicates that the Ly$\alpha$ emission is powered by resonant scattering. Ly$\alpha$ line of both QSOs are redshifted with respect to other emission lines ($\Delta$v = 318.79 km/s for QSO1 and $\Delta$v = 1052.39 km/s for QSO2) suggests that the Ly$\alpha$ photons scatter through large-scale outflows (Verhamme et al. 2006; Steidel et al. 2011). The galactic outflows (inflows) would also lead to an enhanced red (blue) peak in the Ly$\alpha$ profile (Zheng & Miralda-Escudé 2002; Verhamme et al. 2006; Dijkstra et al. 2007; Laursen et al. 2011; Weinberger et al. 2018). Moreover, IGM resonant scattering causes the diminishing of the blue peak. The simulation results from Laursen et al. (2011) indicate that at z$=$3.3, IGM transmission in the blue side of Ly$\alpha$ line is 80%. The Ly$\alpha$ profile of both QSOs shows an enhanced red peak with a diminished blue peak. Such spectral features indicate gas outflows. Therefore, galactic outflows play a role in powering Ly$\alpha$ emission in both QSOs and show redshifted Ly$\alpha$ profile with enhanced red Ly$\alpha$ peak.

Here, we briefly outline the physics underlying the different scenarios that may explain double-core morphology in Ly$\alpha$ emission.

Double-cored Ly$\alpha$ has been seen in some high-$z$ radio galaxies, where it appears to be related to jet interactions (e.g., Overzier et al. 2001), but neither QSO1 or QSO2 are radio-loud galaxies. Gravitational lensing is one of the scenarios that can explain the unusual morphology in QSOs (Walsh et al. 1979; Meyer et al. 2019). However, the single-core QSO continuum component in miniJPAS BB (e.g., rSDSS) and NB images (e.g., J0490 in QSO1 and J0460 in QSO2) lacks evidence for lensed QSOs or the corresponding lensing galaxy, or even binary QSOs. The single-point structure in the HST F606W and F814W images of QSO1 also disfavor the QSO lensing hypothesis. Ionization echoes are often seen from type 2 AGN/QSO, as the echoes were enlightened by the previously active but now inactive SMBH. AGN ionization echoes are fossil records of the rapid shutdown of luminous QSOs, are uncovered in low-$z$ ($z=0.05$ to 0.35) characterized by high-ionization gas extending more than 10kpc from AGN and show strong [OIII]$\lambda$ 5007 emission powered by type 2 AGN (Schawinski et al. 2010; Keel et al. 2012, 2015; Schweizer et al. 2013). The two QSOs reported here are all type 1 QSOs. Since these two QSOs have broad emission lines in their spectra, and AGN counterparts, we can rule out the AGN ionization echoes hypothesis.

The morphology of Ly$\alpha$ emission due to outflows strongly depends on the orientation of the outflow with respect to the line of sight. Outflows emerging from QSOs show double-lobed (bipolar) structure in their morphology if viewed perpendicular to the cone axis (Weidinger et al. 2005; Borisova et al. 2016; Sanderson et al. 2021). The morphology of the miniJPAS Ly$\alpha$ image and line profile from SDSS spectra support the hypothesis that outflows in these two QSOs are contributing to the double-core Ly$\alpha$ morphology in the miniJPAS image. Alternatively, it is possible that high central HI column density or dust obscuration may be the reason for the asymmetric double-core structure in QSO2. Erb et al. (2019) explained the offset by a significant variation of the neutral hydrogen column density across the object. Chen et al. (2021) identified 2 Ly$\alpha$  nebulae spatially offset from the associated star-forming regions, suggesting large spatial fluctuations in the gas properties (see also Claeyssens et al. 2022). Borisova et al. (2016) found no evidence for ”bipolar ionization cone illumination” patterns in their study of Ly$\alpha$  nebulae around quasars, except for one quasar (see their Section. 4.2).