Detailed Studies of IPHAS sources. II. Sab 19, a true planetary nebula and its mimic crossing the Perseus Arm

The INT Photometric H$\alpha$ Survey (IPHAS) has provided us with a number of new-emission line sources, among which planetary nebulae (PNe) constitute an important fraction. Here we present a detailed analysis of the IPHAS nebula Sab\,19 (IPHASX\,J055242.8+262116) based on radio, infrared, and optical images and intermediate- and high-dispersion longslit spectra. Sab\,19 consists of a roundish 0.10 pc in radius double-shell nebula surrounded by a much larger 2.8 pc in radius external shell with a prominent H-shaped filament. We confirm the nature of the main nebula as a PN whose sub-solar N/O ratio abundances, low ionized mass, peculiar radial velocity, and low-mass central star allow us to catalog it as a type III PN. Apparently, the progenitor star of Sab\,19 became a PN when crossing the Perseus Arm during a brief visit of a few Myr. The higher N/O ratio and velocity shift $\simeq$40 \kms\ of the external shell with respect to the main nebula and its large ionized mass suggest that it is not truly associated with Sab\,19, but it is rather dominated by a Str\"omgren zone in the interstellar medium ionized by the PN central star.


INTRODUCTION
The INT Photometric H Survey (IPHAS: Drew et al. 2005; Barentsen et al. 2014) has mapped the Northern Galactic Plane within the latitude range |5 • |, discovering hundreds of new emission-line sources. Among those, many can be expected to be planetary nebulae (PNe), and indeed follow up spectroscopic observations have unveiled a large sample of new PNe. The first release of extended PNe based on the IPHAS catalogue identified 159 true, likely and possible PNe (Sabin et al. 2014).
We have started a series of detailed analyses of individual IPHAS objects. Sabin et al. (2020) and Rodríguez-González et al. (2020) described an evolved bipolar PN and a highly extincted bipolar PN, respectively. These two sources at an advanced evolutionary stage and found at large distances and affected by large amounts of extinction can be typically expected among IPHAS PNe. Here we have focused our attention in IPHASX J055242.8+262116, the source number #19 in Sabin et al. (2014)'s list that will be referred hereafter as Sab 19. This source, classified originally as a likely PN, is located on the Galactic plane along the Galactic anticenter ( =183.0219 • , =+0.0176 • ) and presents an intriguing triple-shell morphology.
We have obtained new images and spectroscopic information for this source and combined this information with archival radio and infrared observations. Sab 19 is confirmed to be a true small-size PN surrounded by a much larger Strömgren zone in the interstellar ★ E-mail:mar@iaa.es medium (ISM), which mimics a PN halo. The article is organised as follows. The imaging and spectroscopic observations are listed in 2. The morpho-kinematics as well as the nebular and stellar properties of Sab 19 are discussed in 3. Finally our discussion on the properties of the PN and its central star and our conclusions are presented in 4 and 5, respectively.

Optical Narrowband Imaging
Narrowband optical images of Sab 19 in the H , [N ] 6584 Å, and [O ] 5007 Å emission lines were obtained with the ALhambra Faint Object Spectrograph and Camera (ALFOSC) on the 2.5m Nordic Optical Telescope (NOT) at the Roque de los Muchachos Observatory (ORM, La Palma, Spain) on January 26, 2020. The detector was an E2V 231-42 2k×2k CCD with pixel size 15 m, providing a plate scale of 0. ′′ 211 pixel −1 and a field of view of 6. ′ 3×6. Three 600 s exposures were obtained for each filter, with a small dithering of a few arcsec between them to improve the image quality and remove cosmic rays. The observing conditions were excellent with a stable seeing of 1. ′′ 0 for all the images, as inferred from the FWHM of stars in the field of view. All the images were reduced using standard routines. After accounting for vignetting caused by the filters, the net field of view is ≈5. ′ 2.

Intermediate-dispersion Spectroscopy
Intermediate resolution spectra were obtained with the 10.4m Gran Telescopio Canarias (GTC) of the ORM with the Optical System for imaging and low-Intermediate-Resolution Integrated Spectroscopy (OSIRIS, Cepa et al. 2000) on January 2, 2018. OSIRIS was used with two Marconi 2048×4096 pixels CCD detectors with a 2×2 binning, leading to a spatial scale of 0. ′′ 254 pixel −1 . The R1000B grism was used, providing a spectral coverage from 3630 Å to 7500 Å and a dispersion of 2.12 Å pixel −1 .
The observations consisted of four exposures of 450 s and 750 s for a total exposure time of 4800 s. The slit was set at a position angle (PA) of 47 • with a length of 7. ′ 4. The slit width of 0. ′′ 8 resulted in a spectral resolution of 900. The data reduction, which includes wavelength calibration with HgAr and Ne lamps and flux calibration with the spectroscopic standard star Feige 110, was performed using standard routines.

High-dispersion Spectroscopy
A longslit high dispersion optical spectrum of Sab 19 was obtained with the Manchester Echelle Spectrometer (MES, Meaburn et al. 2003) mounted on the 2.12m telescope at the Observatorio Astronómico Nacional, San Pedro Mártir (OAN-SPM, Mexico). The observations were obtained on April 17, 2016 with a 2048×2048 pixels E2V CCD Marconi detector with a pixel size of 13.5 m pixel −1 . An on-chip 4×4 binning was applied, resulting in a spatial scale of 0. ′′ 702 pixel −1 and a spectral scale 0.11 Å pixel −1 . MES provides a slit length of 6. ′ 5 and the slit width was set to 150 m, corresponding to 1. ′′ 9. The slit width and spatial scale were suitable for the non optimal weather conditions, with a seeing ∼1. ′′ 8, providing a spectral resolution ≃12 km s −1 . The slit was arranged at PA 0 • and one exposure of 1800 s was obtained with an H filter with Δ = 90 Å to isolate the 87th echelle order. This order also contains the [N ] 6548,6584 Å emission lines. A calibration frame of a Th-Ar lamp was obtained immediately after the science exposure to perform the wavelength calibration. The data were reduced using standard routines.

Spitzer IRAC images
Sab 19 was observed on April 2010 by the Spitzer Space Telescope under the program GLIMPSE360: Completing the Spitzer Galactic Plane Survey (PI: Barbara A. Whitney). Images were obtained in the 3.6 and 4.5 m channels of the InfraRed Array Camera (IRAC, Fazio et al. 2004), a four-channel camera that provides simultaneous imaging at 3.6, 4.5, 5.8, and 8 m with similar 5. ′ 2×5. ′ 2 field of view (FoV). The two short wavelength channels use InSb detector arrays consisting of 256×256 pixels and a nearly same pixel scale of 1. ′′ 2 pixel −1 . Images were retrieved from the Spitzer Heritage Archive.

Morphology
The optical images in Figure 1 reveal an [O ] and H bright round main nebula with a double-shell morphology and an H-shaped fea-ture of diffuse emission brighter in [N ] and H . The main nebula consists of a 6. ′′ 4×6. ′′ 0 roundish inner shell and a concentric 16. ′′ 4×15. ′′ 6 outer shell. The outer shell has a notable brightness enhancement towards the North-Northeast direction with its apex along PA≈25 • , whereas the nebular emission along the opposite direction fades smoothly. The morphology of this outer shell is reminiscent of a bow-shock as caused by the motion of the nebula through the ISM. This suggestion is reinforced by the offset of the central star (CSPN) with respect to the centre of the outer shell by ∼0. ′′ 6 along the direction of the apex of the bow-shock.
The bright main optical nebula is surrounded by low surfacebrightness emission with an H-shaped morphology. This emission is not centred at the main nebula, but displaced towards the South-Southeast with a filamentary and fuzzy appearance. The emission is brighter in H , with some bright filaments in [N ] particularly towards the Southwest directon. The emission in these different emission lines is clearly stratified, with the lower ionization [N ] emission further away from the main nebula.
A comparison with Spitzer IRAC images in the available 3.6 and 4.5 m bands in Figure 2 reveals that the optical H-shaped feature is not only particularly prominent in these IR bands, but it is the brightest feature of a larger structure, a roundish shell with a diameter ≈7. ′ 5. This external shell is quite filamentary and, as the outer shell of the main nebula, it is brighter towards the Northeast direction and fainter and smoother towards the opposite direction. This might be indicative of the interaction with the ISM as the nebula moves through it, and actually the noticeable displacement of the main nebula towards the brightest Northeast rim of this external shell lends support to this idea. The IR emission in the H-shaped feature also follows the excitation structure revealed in the optical emission lines, thus suggesting that the emission from this H-shaped feature arises from an atomic, molecular or dusty component at lower excitation than the material responsible for the emission of the optical [N ] lines.  −0.7 kpc. Accordingly, the main nebula of Sab 19 has a radius of 0.10 +0.05 −0.03 pc, whereas its external shell has a radius of 2.8 +1.4 −0.8 pc.

Physical Conditions and Chemical Abundances
The GTC OSIRIS longslit spectra along PA 47 • have been used to extract one-dimensional spectra of the different structural compo-nents of Sab 19, namely the CSPN, the inner and outer shells of the main nebula, and the external shell. The spectrum of the latter corresponds to the only region detected, the bright filament of the H-shaped feature ≃70 ′′ Southwest of the main nebula (marked by the label H in Figure 2-right). The one-dimensional spectra presented in Figure 3 show generally a small number of emission lines, with very faint or even absent emission lines of low ionization species in the main nebula (e.g., [S ]) that become brighter in the external shell. We notice that the He 4686 emission line is not detected throughout the nebula, neither in the higher excitation main nebula nor in the lower ionization external shell. The intensity of the emission lines in these spectra relative to H and their 1-uncertainties was determined and dereddened using the extinction law by Fitzpatrick & Massa (2007) for =3.1 described by the coefficients listed in Table 1 and the logarithmic extinction coefficient (H ) also listed in Table 1 derived from the observed H to H line ratio adopting case B recombination. The 1-uncertainties of these logarithmic extinction coefficient imply they have a similar value through the different nebular components, with marginal evidence for a higher extinction in the external shell.
These relative intensities were analyzed using the nebular analysis tool (Olguín et al. 2011) based on the nebular package (Shaw & Dufour 1995). Unfortunately, the spectra display very few temperature or density diagnostic emission lines; only a temperature ≃11200 K can be determined from the relative intensities of the auroral-to-nebular [O ] optical emission lines in the spectrum extracted at the CSPN, and a density ≃300 cm −3 from the relative intensities of the [S ] doublet in the spectrum of the brightest feature of the external shell.
The determination of the ionic and elemental abundances listed in Table 2 is thus affected by the small number of emission lines in the spectra, but also by the uncertain determination of the physical conditions. For the main nebula, we have assumed a density of 4000 cm −3 as derived from radio observations (see Section 3.4) and adopted a temperature of 11200 K, whereas for the external shell we have adopted a density of 300 cm −3 and assumed a typical value of 10000 K for the temperature. Besides H + and He + , the main species in the main nebula is O ++ , which is almost 90 times more abundant than O + . Yet the He 4686 emission line is not detected, indicating the lack of species of high excitation including O 3+ and above. The nebula has thus a peculiar ionization balance, with most of the oxygen atoms as O ++ ions. The He and O abundances are typical of Type III and IV PNe, with sub-solar N/O ratios ((N/O) ⊙ =0.25±0.04, Asplund et al. 2009). Although the N/O ratio relies on the determination of the O + ionic abundances based on the [O ] 7320,7330 doublet, which is more sensitive to than the determination of the N + ionic abundances, a solar value for this ratio would require an electronic temperature 8000 K, which seems unlikely for the observed nebular excitation. On the other hand, the N/O ratio of the external shell is higher and marginally consistent with the solar value.

Kinematics
The MES echelle data provide kinematic information of the main nebular shell and the bright Southern region and Norhtern filament of the H-shaped feature intersected by the slit with PA=0 • (regions  Figure 4 reveals a lenticular shape for the main nebula of Sab 19 with brighter emission in the innermost region associated with its inner shell. Meanwhile, the line profile of the H emission associated with the regions of the H-shaped feature to the South of the main nebula (the Northern filamente is much fainter) is narrower than that of the main nebula and is basically consistent with a unique velocity, i.e., it does not show any velocity structure along the spatial extent ∼60 ′′ of the line.
The H line profile extracted from the main shell ( Fig. 5-top) is not resolved, but it is broader than the instrumental spectral resolution. The line can be fitted with two Gaussian components with radial velocities in the Local Standard of Rest (LSR) +81.8±2.1 km s −1 and +101.0±2.1 km s −1 , implying a LSR radial velocity LSR = +91.4± 1.3 km s −1 and an expansion velocity exp ≃ 9.6 ± 3.0 km s −1 . Adopting this expansion velocity for the outer shell, its 0.10 +0.05 −0.03 pc radius implies a kinematic age 10400 +5000 −3000 yr. The line profile of the H emission associated with the H-shaped feature is shown in the middle and bottom panels of Figure 5. As described above, it is narrow and it can indeed be fitted with a single Gaussian component with radial velocity LSR = +51.1±1.5 km s −1 for the Southern region H S and +49.4 ± 3.2 km s −1 for the Northern region H N . As illustrated in Figures 4 and 5, there is a remarkable ∼40 km s −1 shift between the radial velocities of the H-shaped feature and the main nebula of Sab 19.

Nebular Masses
The ionized mass ion of a PN can be estimated from its total H flux using, for instance, the relationship where ion is given in solar masses, (H ) is the extinctioncorrected H flux in units of 10 −11 erg cm −2 s −1 , is the distance in kpc, is the electron temperature in units of 10000 K, and is the electron density (Pottasch 1983).
The H flux can be derived from the measured H flux and the extinction derived from the nebular spectra (H ) of 1.9, as listed in Table 1. The H flux from the main nebula and the external shell have been derived from the NOT ALFOSC H image, computing the starsubtracted photon count rate within apertures encompassing the main nebular shell and the external shell (after excising the main nebular shell) and using the GTC OSIRIS spectra to flux calibrate them by comparing the image count rate within the nebular area covered by the GTC OSIRIS slit and the H flux derived from this spectrum. The observed flux of the main and external nebulae are 1.6 × 10 −12 and 3.6 × 10 −12 erg cm −2 s −1 , respectively. The extinction-corrected intrinsic H fluxes would be 2.9 × 10 −11 erg cm −2 s −1 for the main nebula and 6.4 × 10 −11 erg cm −2 s −1 for the external shell. The corresponding intrinsic H fluxes would be 1.0 × 10 −11 erg cm −2 s −1 for the main nebula and 2.2 × 10 −11 erg cm −2 s −1 for the external shell 1 . Therefore, the external shell intrinsic fluxes in these emission lines is approximately twice that of the main nebula, although its average surface brightness is ≈100 times lower. We note that the H intrinsic flux can also be derived from the radio flux at 5 GHz using the relationship given in Pottasch (1983). The radio flux at this frequency of 44±4 mJy (Gregory & Taylor 1986) implies an H flux of (1.3 ± 0.1) × 10 −11 erg cm −2 s −1 , in general agreement with the previous estimate for the main nebula. The density of the bright filament of the H-shaped feature of the external shell SW of the main nebula has been estimated to be ≃300 cm −3 (Table 1), which is certainly an upper limit to its average density. If this density is assumed to be the same for all ionized material in this shell, then Eq. 1 implies it has an ionized mass ≃0.5 ⊙ , that would be a lower limit for the ionized mass according to Eq. 1. On the other hand, the density of the main nebula is unknown. Since its H flux is known, we can use the relationship for the rms density 1 2 = 2.74 × 10 4 F(H ) 0.88 where is the filling factor and is the angular radius in arcsec (Pottasch 1983). Then, for the angular size and H flux of the main 1 Accounting for the uncertainties in image count rate, cross-calibration of the NOT images with the OSIRIS spectra, and extinction, the total uncertainties for these intrinsic H fluxes are <10% for the main nebula and <15% for the external shell.
nebula, an rms density of 800 − 1 2 cm −3 is derived. Therefore, according to Eq. 1, the ionized mass of the main nebula would be 0.10 1 2 ⊙ . The densities of the main nebula and external shell can also be derived from their radio emission. Sab 19 has been detected by several radio surveys carried out at different frequencies. The NVSS survey (Condon et al. 1998) detected it as NVSS J055242+262109 with an integral flux at 1.4 GHz of 45.5 ± 2.1 mJy. The source is reported to have an angular size 46 ′′ ×28 ′′ , even though the NVSS images have a 45 ′′ FWHM resolution. The Radio Patrol of the Northern Milky-Way (Gregory & Taylor 1986) detected an unresolved 5 GHz source at 17 ′′ from the central star of Sab 19, showing a flux of 44±4 mJy. More recently, the AMI Galactic Plane Survey (AMIGPS, Perrott et al. 2015) detected a 15.7 GHz counterpart with an (integrated) intensity of 33.0 mJy. The poor spatial resolution of this survey (3 ′ ) did not allow an estimation of its apparent size, and thus it was classified as a point source.
Assuming that the radio source associated with Sab 19 has the same dimensions at 1.4, 5, and 15.7 GHz, we can estimate the spectral index , considering ( ) ∝ . The spectral index between 1.4 and 5 GHz is −0.027 and that between 5 and 15.7 GHz is = −0.25. Therefore, at 1.4 GHz the radio continuum is near the turnover point between the optically thin and thick regimes, and the optical depth equals unity at this frequency (Pottasch 1983 where is the optical depth at the frequency , is the absorption coefficient, is the electron temperature (K), and and are the proton and electron density (in cm −3 ), respectively, integrated along the line of sight (in pc). Assuming that equals the unity at 1.4 GHz, = 10000 K and = , we obtain the emission measure ( ): where EM is given in cm −6 parsec −1 . An inspection of the radio emission in Figure 6 reveals that the 3 contour levels of the emission at 1.4 GHz spreads over an area about 72 ′′ ×54 ′′ . The linear size of this emission, 0.91 pc by 0.68 pc, would correspond to electron densities 2600-3000 cm −3 . If we assume instead that this radio emission uniquely arises from the main nebula of Sab 19 2 , with a diameter of 16 ′′ or 0.20 pc at 2.6 kpc, then the emission measure in Eq. 5 implies an average electron density ≃ 5600 cm −3 . Alternatively, if the emission were uniquely associated with the external shell, with a diameter of 7. ′ 5 or 5.7 pc at 2.6 kpc, it would imply an average electron density ≃ 1000 cm −3 , which is unphysical because it is notably larger than the value derived from the [S ] line ratio. Most likely, the emission measure in Eq. 5 is a combination of those of the main nebula and external shell. The density of the main nebula would be in the range from 2600 to 5600 cm −3 , which is consistent with the estimate of the rms density for values of the filling factor in the range 0.02-0.10 that would result in ionized masses in the range 0.016-0.035 ⊙ .

Properties of the Central Star
The excitation of the main nebula of Sab 19 is quite intriguing and may shed some light on the properties of its central star. The He 4686 emission line is not detected with an upper limit at 3 of 1.8% the intensity of the H line. The He 4686 to He 5867 line ratio 0.12 thus indicates an effective temperature 60,000 K, regardless of the optical thickness of the nebula (Gruenwald & Viegas 2000). The Stoy and Zanstra H temperatures ∼57300±3400 K and ∼53000±10000 K, respectively, derived from the intensity of the [O ] 5007 using Kaler & Jacoby (1991)'s prescriptions are consistent with the above upper limit. The stellar spectrum shows the broad spectral feature of C 5801,5812 (Fig. 3), but neither the C 5696 nor the O 5290 features. The range of effective temperatures proposed above for the central star of Sab 19 is consistent with these WR features (Acker & Neiner 2003).
The Lyman luminosity of the central star ( Ly ) can be used to constrain the stellar radius and luminosity. The nebular recombination rate can be obtained from the luminosity in the H line by comparing the total recombination rate with the bound-free transitions down to the = 2 level (Osterbrock. & Ferland 2006): where ,H ( ) are the case B and the n=2 recombination coefficients, and is the minimum ionization frequency of the hydrogen. Adopting the values of = 2.59 × 10 −13 and H = 7.69 × 10 −14 cm 3 s −1 for a temperature of 10000 K (Osterbrock. & Ferland 2006), the intrinsic H luminosities in the main nebula and external shell of 2.4 × 10 34 erg s −1 and 5.2 × 10 34 erg s −1 , respectively, derived from the extinction-corrected fluxes given in the section above, imply a total of 8.4 × 10 46 recombinations s −1 .
At frequencies where the nebula is optically thick to the stellar Lyman continuum the rate of incoming ionizing photons over the entire nebula ( Ly ) can be related to the free-free emission at the frequency by the following relationship (Rubin 1968): Ly = 5.59 × 10 48 1 + He + /(H + + He + ) 5 GHz where is the distance in kpc, is the total intensity in Jansky, and He + /(H + + He + ) is the fraction of He-recombination photons energetic enough to ionize the H. Since the He lines are absent in the optical spectrum, we can assume the latter to be null. Assuming e = 10000 K, we obtain at a distance of 2.6 kpc Lyman recombination numbers of 2.4×10 46 , 2.6×10 46 , and 2.2×10 46 s −1 at 1.4, 5 GHz and 15.7 GHz, respectively.
We note that the radio-derived recombinations are about 3.5 times lower than that derived using the H line. Several factors might be responsible for this discrepancy. For example, the recombination rate obtained from the H image is strongly dependent on the interstellar extinction, whereas the radio value, which is not affected by extinction, assumes that Sab 19 is matter-bounded. If the latter hypothesis is incorrect, then Ly should be considered as a lower limit.
These observed are compared to those computed assuming that the CSPN of Sab 19 emitted like a blackbody for various radii and temperatures in Figure 7. The gray area in the plot delimits the upper and lower limits for the effective temperature. The plot shows also the Lyman luminosity of evolving CSPNe of 0.546 ⊙ and 0.565 ⊙ calculated from the stellar temperatures and luminosities of the theoretical tracks by Schoenberner (1983). According to this figure, the observed Ly is compatible with a very low-mass CSPN, with CSPN 0.546 ⊙ . The theoretical tracks indicate that CSPN 1.2 × 10 3 ⊙ , and the stellar radius is CSPN / ⊙ = 0.3 − 0.4, as obtained from the H flux, or ∼ 0.2, as obtained from the radio data. Zhang & Kwok (1993) obtained various distance-independent parameters of PNe and analised how they vary as the nebula evolves. Their results were based on the theoretical tracks of the CSPN calculated by Schoenberner (1983), but they noted that the evolutionary tracks of the lowest-mass CSPNe had to be sped up in order to match the observations, which is in line with the latest evolutionary mod-  Zhang & Kwok (1993), based on the theoretical models by Schoenberner (1983). CSPNe evolve from the left to the right. The box marks the lower and upper limits of the temperature of the CSPN ( eff ) and the brightness temperature of the PN ( ) at 5 GHz, whereas the crosses correspond to the location in this plot of a sample of PNe with measured extracted from Zhang & Kwok (1993). els of Miller-Bertolami (2016). Figure 8 illustrates the evolutionary tracks of a 0.546, 0.565, and 0.598 ⊙ CSPNe by Zhang & Kwok (1993), corrected for the "speed up" effect. During a rapid evolutionary phase the brightness temperature ( ) of the nebula reaches its maximum, from which it slowly decreases as the PN evolves. The maximum at 5 GHz reached by a 0.598, 0.565 and 0.546 ⊙ nucleus is 200, 120 and 3K, approximately. This diagram allows the determination of the mass of the CSPN from two quantities: the nebular brightness temperature and the stellar effective temperature.
The average brightness temperature of Sab19 can be calculated using equation Assuming the dimensions of the 3 contour levels at 1.4 GHz shown in Figure 6 (72 ′′ × 54 ′′ ), we obtain a lower limit for of 0.80 K, whereas an upper limit of 12.1 K is obtained assuming that all radiation emitted at 5 GHz originates in the main nebula only. When plotted in Figure 8, these results indicate CSPN ≃ 0.56 ⊙ . Considering the uncertainties, this result agrees with the previous estimates obtained using the H and radio nebular fluxes shown in Figure 7.

The True and the Mimic
The analysis of the observations presented in previous sections has revealed a multi-component structure for Sab 19 consisting of a main double-shell nebula and a larger and fainter external shell with a prominent H-shaped feature. Their distinct properties suggest a distinct nature for each morphological component.
The main nebula of Sab 19 has a typical double-shell PN morphology. Its spectrum and the physical conditions and chemical abun-dances derived from it are also typical of PNe. It can be concluded that the main nebula of Sab 19 is a true PN. The nebula has subsolar N/O abundances ratio, which is suggestive of a type III or IV PN (Maciel & Koppen 1994), although it must be noticed that uncertainties in the values of O/H and N/O are large. The nebula has a small ionized mass, in the range from 0.016-0.035 ⊙ , and its central star is relatively cold with a low mass, 0.546-0.565 ⊙ . Apparently, the main nebula of Sab 19 is a type III PN descendant from a low-mass progenitor.
The external shell might then be interpreted as a halo resulting from an enhanced mass-loss episode associated with a thermal pulse in the last phases of the AGB, but this does not seem to be the case. First, the total ionized mass of the external shell, which has been estimated to be 0.5 ⊙ , is in sharp contrast with the low mass of its progenitor star. Most notably, the large radial velocity shift between this external shell and the main nebula, ≃40 km s −1 , casts serious doubts on its nature as a PN halo, as true haloes of PNe do not exhibit such large velocity shifts with their PNe (Guerrero et al. 1998). Possible mimics of PNe include a long list of sources, as discussed in detail by . Given the velocity discrepancy between the external shell and the main nebula of Sab 19 (as is also the case of the nebula PHL 932, ) and its location inside a much larger mid-IR patchy structure (Fig. 6), the most likely candidate for the external shell of Sab 19 among the usual suspects for PN mimics is a Strömgren zone in the ISM. To further reassure this possibility, the relationship between Sab 19 and the local ISM is investigated in the next section.

Sab 19 and Its Place in the Galaxy
At a distance of 2.6 +1.3 −0.7 kpc along galactic longitude ≃183 • (i.e., mostly at the Galactic anticenter), Sab 19 is located in the outskirts of the Perseus Arm, which extends from 1.4 to 2.9 kpc along this direction, as traced by very young high-mass stars (Reid et al. 2019). This description is consistent with the image in Figure 6, which shows large-scale patchy emission in the WISE W3 band at 12 m revealing a complex ISM along the line of sight of Sab 19. Since the mid-IR emission from the external shell of Sab 19 peaks in the Spitzer 3.6 m and WISE W2 4.6 m bands, whereas that of the ISM around it is brighter in the WISE W3 12 m band (Figs. 2 and 6), it can be concluded that the mid-IR emission from the external shell of Sab 19 includes line emission or is indicative of warmer dust than that in the ISM.
The positional coincidence of Sab 19 with the Perseus arm is, however, in sharp contrast with their respective radial velocities. The LSR velocity at the Galactic anticenter is expected to be null. Actually, LSR velocities in the range from −20 km s −1 to +10 km s −1 are measured in the Galactic H emission and giant molecular clouds along this direction (see Figure 3 in Reid et al. 2019). For instance, WB717 is a CO (J=1−0) source detected towards Sab 19 at LSR = −0.2 km s −1 , whereas other neighbouring CO sources detected within Galactic longitude ±1 • from Sab 19 exhibit a range of velocities −13 km s −1 < LSR < +9 km s −1 (Wouterloot & Brand 1989). The radial velocity of the main nebula of Sab 19, however, is notably different, LSR = +91 km s −1 . The discrepancy in radial velocity is also notorious on the tangential component of the velocity, i.e., the velocity on the plane of the sky. The proper motion components of Sab 19 measured by are RA = 1.234±0.205 mas yr −1 and DEC = 0.848±0.161 mas yr −1 , implying a large angle with the Galactic plane. On the other hand, the proper motion module of 1.50 mas yr −1 corresponds to a linear velocity on the plane of the sky ≃19 km s −1 for the distance given by Bailer-Jones et al. (2018). This confirms that the motion of Sab 19 is dominated by its radial velocity and that it does not corrotates with the Perseus Arm, reinforcing the idea that Sab 19 is not actually associated with it, but it is crossing it at a relative velocity 80 km s −1 . Actually, the distance of 2.6 kpc of Sab 19 would place it in the outer border of the Perseus Arm, which it would be "leaving" after a "short visit" of a few Myr (15 Myr if a thickness of 1.5 kpc is adopted for the Perseus arm according to Figure 3 in Reid et al. 2019), but we reckon that the distance error bar towards Sab 19 makes this claim uncertain.
The complete information on the position of Sab 19 in the phase space has been used to analyse its dynamic properties using the package (Bovy 2015). The initial values of the orbit are given by the coordinates, distance, proper motion and radial velocity of Sab 19, thought we reckon the large uncertainties in distance and proper motions. The nebula is then assumed to move under the gravitational potential MWPotential2014 included in , whose structure and physical properties are discussed in detail by Bovy (2015). As for our position and velocity in the Galaxy, the solar motion [-9.4, 12.6, 6.3] in km s −1 , height over the Galactic plane of ℎ = 16 pc, distance to the Galactic centre of 8.15 kpc and circular rotation speed at the Sun's position of 0 of 236 km s −1 have been adopted Reid et al. 2019). According to this model and initial conditions, Sab 19 has a highly-elongated orbit, with an apogee at almost 15 kpc from the Galactic centre and its perigee at 9 kpc. The orbit is contained in the Galactic disk, although it is able to reach a maximum height over the Galactic plane ≃700 pc, and the rotational velocity is 250 km s −1 . We note that, for the whole distance and proper motion error intervals, the orbit remains confined to the Galactic disk.

Just Passing By, but Leaving a Mark
The notable differences between the radial velocity of the main nebula of Sab 19 and the external shell make very unlikely the latter to be a halo ejected in late phases of the stellar evolution. Interestingly, the radial velocity of the external shell ( LSR ≃ +51 km s −1 ) is half way between that of the main nebula of Sab 19 ( LSR ≃ +91 km s −1 ) and that expected for the ISM (−13 km s −1 < LSR < +9 km s −1 ). This seems to imply that the material in this external shell has experienced an interaction with the moving PN.
To investigate into more detail the possible interactions caused by the motion of Sab 19 through the ISM, the direction of its motion according to the proper motions of RA = 1.234±0.205 mas yr −1 and DEC = 0.848±0.161 mas yr −1 has been overplotted on Figures 1, 2, and 6. The coincidence of the orientation of morphological asymmetries in the different shells of Sab 19 and the direction of its motion on the plane of the sky lends support to this interaction: the central star is offset towards the SW from the main nebula, which shows a bow-shock-like brightness enhancement at the NE direction (Fig. 1), the infrared shell is compressed along this direction and shows a more diffuse and fainter emission along the opposite SW trailing direction (Figs. 2 and 6), and the radio emission shows a bow-shock-like morphology towards the NE and a smooth decline in brightness towards the SW along the trailing direction.
The interactions of the outer shells and haloes of PNe with the surrounding ISM have been largely reported and described (e.g., Tweedy & Kwitter 1996;Wareing, Zĳlstra & O'Brien 2007). Indeed, the location of Sab 19 in the Perseus Arm and the large-scale IR emission around it indicates that the local ISM is relatively dense. The external shell of Sab 19 is not a halo, however, as most (if not all) the material in this shell belongs to a Strömgren zone in the ISM which is being ionized by the CSPN of Sab 19. Still, the H-shaped filaments in this external shell may arise as the result of Rayleigh-Taylor (RT) instabilities formed by the interaction of a fast-moving ionized shell with a cold, dense, perhaps magnetized ISM. The shape of the structures formed depend on many parameters, such as the intensity of the magnetic field, the pitch angle between the velocity of the PN and the local magnetic field, etc. For a fast-moving PN, the shock formed between the PN and the local ISM is isothermal, and the ISM magnetic field can contribute to form RT instabilities just behind the shock wave (Soker & Dgani 1997). At any rate, RT structures normally do not penetrate the ionized main PN shell, but they are restricted to the external, fragmented external shell (Soker & Dgani 1997;Dgani & Soker 1998), as is the case of Sab 19.
It is interesting to note that the module of the velocity vector of Sab 19 on the plane of the sky, ≃19 km s −1 , is about 5 times smaller than its radial component along the line of sight, ≃91 km s −1 . We can thus expect that the external shell of Sab 19 would be much thicker along the line of sight than its projection on the plane of the sky. If we keep in mind that the velocity component of Mira on the plane of the sky is ≃130 km s −1 , quite similar to the radial velocity of Sab 19, it could be envisaged the external shell of Sab 19 as the long tail left behind by Mira (Martin et al. 2007), but seen face on. A similar structure also on the plane of the sky might be observed in the PN HFG 1 (Boumis et al. 2009;Chiotellis et al. 2016).

The Nature of Sab 19
The peculiar velocity of Sab 19 and the maximum height of its orbit over the Galactic Plane can be used to further investigate its past evolution. It has been noted that the average peculiar velocity increases among the different Peimbert's types of PNe (Peimbert 1978), increasing from 20±14 km s −1 for the N-and He-rich type I PNe descending from more massive progenitors up to 170±80 km s −1 for type IV PNe of the halo evolving from low-mass progenitors (Maciel & Dutra 1992). The maximum height over the Galactic Plane of Sab 19 implies that it is not a halo type IV PNe, but it can be rather classified as a type III PN. The peculiar velocities of type III PNe are indeed typically high, ∼60 km s −1 , with many of them showing higher peculiar velocities such as Me 2-2 (−140 km s −1 ), IC 5217 (87 km s −1 ), or K 3-67 (−84 km s −1 ). Ortiz & Maciel (1994) observed that AGB stars can also be classified into types similar to the Peimbert scheme according to their kinematics. The theoretical models of Vassiliadis & Wood (1994) and Miller-Bertolami (2016) can be used to estimate the mass of the precursor ZAMS stars for a set of metallicities. The 0.546-0.565 ⊙ mass of the CSPN of Sab 19 is consistent with the 0.528-0.652 ⊙ final mass of a solar metallicity (Z=0.02) ZAMS star with an initial mass 1.00-1.25 ⊙ . Adopting a lower metallicity of Z=0.001, the final mass of 0.534-0.552 ⊙ implies initial masses of ZAMS stars in the range 0.90-1.00 ⊙ . Therefore, according to these theoretical models and the mass of the CSPN determined in Sect. 3.5, Sab 19 descends from a nearly solar-mass main-sequence star. Its maximum rotational velocity of 250 km s −1 makes it very likely a member of the thin disk population.

SUMMARY AND CONCLUSION
We have presented new and archival multi-wavelength images and new intermediate-and high-dispersion spectroscopic observations of Sab 19, aka IPHASX J055242.8+262116. These observations reveal that Sab 19 consists of two morphological components, a doubleshell main nebula and an external shell dominated by a prominent set of H-shaped filaments. At a distance of 2.6 kpc, as derived from G , the size of these shells is 0.10 and 2.8 pc, respectively. The origin of these two components is different.
The double-shell main nebula is a type III PN descending from a low-mass 0.90 − 1.25 ⊙ progenitor star, according to the small nebular ionized mass, the 0.56 M ⊙ low-mass CSPN, the peculiar velocity of the nebula, its small expansion velocity, and the nebular sub-solar N/O ratio. On the other hand, the higher N/O ratio, large ionized mass and radial velocity shift with the main nebula of the external shell suggested by the present data make it very likely to be a Strömgren zone in the ISM ionized by the CSPN of Sab 19. A complete spatio-kinematic study and investigation of the physical conditions in the external shell will certainly help to confirm it.
Sab 19 is located in the Perseus arm, but its peculiar radial velocity implies that it is not associated with it, but it is actually crossing it as it moves towards the apogee of its galactic orbit at about 15 kpc from the Galactic centre. Apparently, the progenitor of Sab 19 is a low-mass star of the thin disk on a very eccentric orbit that happened to become a double-shell PN as it was crossing the Perseus Arm. The interactions of the PN and its associated Strömgren zone with the local ISM are quite notorious.