Tomography of the Unique Ongoing Jet in the Planetary Nebula NGC 2392

Hubble Space Telescope (HST) Wide Field and Planetary Camera 2 (WFPC2) images of NGC 2392 will be used through this work for different purposes. Images in the F502N, F656N, F658N, and F671N filters (program IDs 8499 and 8726) corresponding to the [O iii] λ5007, Hα λ6563, [N ii] λ6584, and [S ii] λ λ6716,6731 emission lines, respectively, were retrieved from the Hubble Legacy Archive. ⁶ The count rate of each image was divided by the total throughput efficiency of the filter at the line wavelength. ⁷

IFS observations of NGC 2392 were obtained on 2020 January 4 and 25 using the Multi-Espectrógrafo en GTC de Alta Resolución para Astronomía (MEGARA; Gil de Paz et al. 2018) at the 10.4 m Gran Telescopio de Canarias (GTC). Both observing runs were performed under clear conditions and good seeing conditions with values in the range 10–12. The high-resolution Volume-Phased Holographic (VPH) grism VPH665-HR was used, providing a spectral dispersion of 0.098 Å pix⁻¹ and a full-width at half-maximum (FWHM) spectral resolution ≈16 km s⁻¹, that is, R ≈ 18,700. The spectral range 6405.6–6797.1 Å covered by the VPH665-HR grism includes the key emission lines of [N ii] λ λ6548,6584, Hα, and [S ii] λ λ6716,6731.

The integral field unit (IFU) mode was used. It has 567 hexagonal spaxels with a diameter of 062 and a field of view (FoV) of 125 × 113. Since the nebular size of NGC 2392 (∼48'') is larger than the instrument FoV, observations in twenty different pointings were obtained as shown in Figure 1 to cover the inner shell and most of the outer nebular shell, with special coverage of the diameter along position angle (PA) 70° to map the outflow (see García-Díaz et al. 2012), which is detected in pointings "A," "B," "E" to "O," "S," and "T." The FoV of nineteen out of the twenty pointings overlap ∼2'' along the east–west direction and 05 along the north–south direction with adjacent fields. The twentieth pointing ("J" in Figure 1) was centered at the location of the central star. Three 5 minute exposures were obtained at each position to facilitate removal of cosmic rays.

Zoom In Zoom Out Reset image size

2.2.1. Data Reduction

2.2.2. Data Analysis: Mapping the Jet

As a first step in the data analysis, the MEGARA data cubes were "sliced" at a particular wavelength range to produce a pseudo-narrowband image at a specific velocity interval of an emission line. This provides only crude information of the spatial distribution and kinematics (radial velocity and velocity width) of the gas in the jet of NGC 2392. An inspection of the Hα, [N ii], and [S ii] emission line profiles at different locations reveals the presence of multiple kinematic components associated either with the nebula (bottom panel of Figure 2) or with both the nebula and the jet (middle and top panels of Figure 2). The emission of the high-velocity jet is generally well isolated from that of the nebular shells in the [N ii] and [S ii] emission lines. Therefore, the line intensities, centroids, and widths at each spaxel can be fitted using a single Gaussian function. On the other hand, the Hα jet emission cannot be resolved properly from the Hα nebular emission at all locations given its low surface brightness contrast with the Hα nebular emission. These difficulties are further aggravated by the broader thermal width of this line. Finally, we note that the [N ii] λ6548 line can be ignored relative to the three-times-brighter [N ii] λ6584 line.

Zoom In Zoom Out Reset image size

Taking into account all comments above, we applied a procedure to obtain the best single Gaussian fit of the [N ii] λ6584 and [S ii] emission line profiles on a spaxel-by-spaxel basis (see Figure 3 for examples of these fits). The fitting was performed to all spaxels with sufficient signal-to-noise ratio using the Levenberg–Marquardt least-squares fitting routine mpfitexpr (Markwardt 2009) within the Interactive Data Language ⁸ (IDL) environment for the [N ii] λ6584 and [S ii] emission lines separately (e.g., Cazzoli et al. 2020). We found that constraints to the line widths of FWHM < 1 Å (≈45 km s⁻¹ at the rest-frame wavelengths of these lines) and absolute radial velocities with respect to the systemic velocity ⁹ (∣v_r ∣) between 130 km s⁻¹ and 250 km s⁻¹ produced optimal results. At each spatial position, we additionally imposed that each [S ii] emission line shares the same kinematics (velocity and velocity dispersion).

Zoom In Zoom Out Reset image size

The output consists of fit parameters for each spectral feature including the central wavelength, width, and flux of the line along with their uncertainties. A heliocentric velocity correction of ∼4 km s⁻¹ was applied (the exact value depending on the observing time of each pointing). Finally, the observed line width was corrected for the effect of instrumental width (σ_ins ≈ 6.8 km s⁻¹) by subtracting it quadratically from the observed line width. The final maps of surface brightness, radial velocity with respect to the systemic velocity, and instrumental-width-corrected velocity width of the jet are presented in Figure 4. In these maps, regions for which the significance of the [N ii] λ6584 and [S ii] emission lines is lower than 5σ and 3σ, respectively, have been masked out. The kinematics in the [S ii] emission lines, which is basically consistent with that of the [N ii] emission line, is not presented.

Zoom In Zoom Out Reset image size

The jet in NGC 2392 consists mostly of two blobs of similar angular extent and a few fainter outer knots (Figure 4(a)). These three components will be referred to hereafter as the inner, mid, and outer jets, respectively. Despite the morphological differences between the Western and Eastern components, the three components of the jet appear to twist point-symmetrically from the central star in a subtle S-shaped morphology. The total flux in the [N ii] λ6584 emission line is 6.6 × 10⁻¹⁴ erg cm⁻² s⁻¹, i.e., approximately 1000 times fainter than the nebula emission in this line. The average (peak) surface brightness of the jet in this line is 1.5 × 10⁻¹⁶ (4.5 × 10⁻¹⁶) erg cm⁻² s⁻¹ arcsec⁻², with a jet-to-nebula surface brightness ratio peaking at ≈0.5 for the mid jet, but staying lower than 0.01 for the inner jet that is projected onto the bright inner shell.

The comparison of the jet morphology and spatial extent with the nebular emission shown in Figure 5 is clarifying. In spite of the much brighter nebular emission, the inner jet can be traced down to exactly to the location of the central star of NGC 2392 (Figure 5(a)). The inner jet brightens then at a location immediately interior to the inner shell rim and breaks through the walls of the inner shell reaching the nebular outer shell (Figure 5(a)). Meanwhile, the mid jet extends across the nebula outer shell, with its outermost tips being found outside the nebular edge (Figure 5(b)). The outer jet is definitely beyond the nebular outermost edge.

Zoom In Zoom Out Reset image size

The velocity map in Figure 4(b) shows the Western component of the jet to approach and the Eastern component to recede from us. The jet velocity varies steadily with radius, with a notable ≈20 km s⁻¹ increase from the inner to the mid jet and a small decrease between the mid and outer jets. The velocity breaks among the different components of the outflow are suggestive of episodic ejections, with small changes in the direction of the motion of the different components of the outflow. That said, the velocity dispersion in Figure 4(c) is notably small, close to the thermal width of the line at an electronic temperature (T_e) of 10,000 K. Therefore, although the varying morphology of the approaching and receding components of the jet hints at a complex dynamics of the outflow in its interaction with the nebula, the coherent kinematics rather implies a steady motion of the gas in each component of the jet with little interaction with the bulk of the nebular material.

The GTC MEGARA high-dispersion IFS observations presented in panels a, b, and c of Figure 4 provide the first clean and complete view of the morphology and kinematics of the jet of NGC 2392 in the [N ii] λ6584 emission line. The spatial extent of the inner jet could not be described in previous works based on Fabry–Perot (Reay et al. 1983) and systematic mappings using multi-long-slit echelle spectroscopic (Balick et al. 1987; García-Díaz et al. 2012) observations due to the extremely low emission contrast between the inner jet and the inner shell. That issue is further aggravated by the unexpectedly (Jacob et al. 2013) fast expansion velocity of this shell (≈120 km s⁻¹, García-Díaz et al. 2012), which makes difficult to deblend the emission of the jet from that of the inner shell. In addition, the outer jet could not be detected in previous observations due to its low surface brightness.

The morphology of the jet in the [S ii] λ λ6716,6731 emission lines (Figure 4(d)) is very similar to that in [N ii], although the total flux in the [S ii] lines is 1.3 × 10⁻¹⁴ erg cm⁻² s⁻¹, i.e., about 5 times fainter. The average (peak) surface brightness of the [S ii] emission lines is 5.7 × 10⁻¹⁷ (1.4 × 10⁻¹⁶) erg cm⁻² s⁻¹ arcsec⁻². The [S ii] λ6716 to [S ii] λ6731 line ratio free from nebular emission has been used to estimate the electronic density (N_e) of the material in the jet assuming a T_e of 10,000 K (Figure 4(a)). The density of the jet is lower than that of the nebula (Pottasch et al. 2008), with values of N_e in the range from the low-density regime ≈100 cm⁻³ up to a few thousands of cm⁻³, for a median N_e of 380 ± 160 cm⁻³. The integrated values of N_e for the different components of the jet reveal a density gradient, from ≃400 cm⁻³ for the inner jet, to ≃150 cm⁻³ for the mid jet, and, at the low-density regime, ≲100 cm⁻³ for the outer jet.

In retrospect, we have examined the HST WFPC2 images of NGC 2392 to search for the jet emission. The inner jet, projected onto the bright emission of the inner shell, cannot be detected, nor can the weak emission from the outer jet, but it is possible to identify in the [N ii] and [S ii] images and [N ii]/[O iii] and [S ii]/[O iii] ratio maps the emission from the mid jet projected onto the outer shell (Figure 6). The morphology of the mid jet shown by the GTC MEGARA data is confirmed in these HST images and ratio maps; the Eastern mid jet is revealed as a curving-blob, whereas the Western mid jet has a loop-like morphology with brighter emission at its tip. We remark that the identification of the emission from the mid jet in HST images and ratio maps has become possible only after a close comparison with the GTC MEGARA data, as the emission from the mid jet is basically indistinguishable from the numerous low-velocity patches of diffuse emission of the outer shell.

Zoom In Zoom Out Reset image size

The ionized mass of a nebula can be derived using the relationship:

$$\begin{eqnarray}&&{M}_{{\rm{i}}}=11.06\times F({\rm{H}}\beta )\times {d}^{2}\times {T}_{{\rm{e}}}^{0.88}\times {N}_{{\rm{e}}}^{-1}\ {M}_{\odot }\end{eqnarray} \tag{ 1 }$$

if N_e is derived from density diagnostic emission lines and F(Hβ) in units of 10⁻¹¹ erg cm⁻² s⁻¹, d in kpc, and T_e in 10,000 K (Pottasch 1984). It is possible to derive the intrinsic Hβ flux from the observed Hα flux adopting the recombination case B for an Hα to Hβ theoretical ratio of 2.86 (Osterbrock & Ferland 2006) and correcting the reddening using a value of 0.23 for c(Hβ); (Pottasch et al. 2008). Since it is not possible to build a complete map of the surface brightness in the Hα emission line, the [N ii] λ6584 to Hα emission line ratio has been measured in spectra extracted in a number of representative apertures and its average and dispersion values estimated to be 1.5 ± 0.2.

Accordingly, the intrinsic Hβ flux is derived to be (2.6 ± 0.4) × 10⁻¹⁴ erg cm⁻² s⁻¹. Adopting the distance of 1.84 ± 0.16 kpc derived from the Gaia EDR3 parallax (Gaia Collaboration et al. 2020, in preparation) and assuming a T_e of 10,000 K, an ionized mass M_i of (2.5 ± 1.1) × 10⁻⁴ M_⊙ is derived for the jet of NGC 2392. The error bar includes the 1σ uncertainties in flux, distance, and electronic density.

Besides the jet mass, its inclination with the line of sight is a key parameter that would allow the determination of its space velocity, age, and mass-loss rate. In general, it would not be possible to derive the inclination of a jet unless additional assumptions were adopted. As for the jet of NGC 2392, its inclination angle has been proposed to be 7° assuming it has the same age of the inner shell (Gieseking et al. 1985), and in the range 33–10° assuming it shares the same tilt with the line of sight of the inner shell (O'Dell & Ball 1985). The inner jet is found at an average PA on the plane of the sky of ∼55° and then it twists to ∼70° in the mid jet (Figure 4(a)). The jet is thus clearly misaligned with the inner shell whose symmetry axis PA has been estimated to be 20–25° (O'Dell et al. 1990; García-Díaz et al. 2012), questioning the above assumptions on the similarity between the jet and inner shell inclinations.

The S-shaped morphology of the jet and its correlated radial velocity variations (Figures 4(a) and (b)) are consistent with the spatio-kinematic behavior of an episodic precessing jet. Following Guerrero & Manchado (1998), a precessing jet can be described by the half-aperture angle of the precession cone (θ), inclination angle of its symmetry axis with the line of sight (ι), and space velocity (v). Since the approaching and receding components of the jet are found at different sides of the central star, it is inferred that the half-aperture angle of the precession cone θ is smaller than the inclination angle ι as assumed in Figure 7. The maximum and minimum systemic radial velocities of the jet (${v}_{\max }$ and ${v}_{\min }$) depend on the expansion velocity of the jet v and on θ and ι as

$$\begin{eqnarray}&&{v}_{\max }=v\cos (\iota -\theta )\end{eqnarray} \tag{ 2 }$$

$$\begin{eqnarray}&&{v}_{\min }=v\cos (\iota +\theta )\end{eqnarray} \tag{ 3 }$$

which can be expressed as

$$\begin{eqnarray}&&{v}_{\max }-{v}_{\min }={\rm{\Delta }}v=2\,v\sin \iota \sin \theta .\end{eqnarray} \tag{ 4 }$$

Since v is not known, though it can be substituted using Equation (2), after some calculations we derive the following expression

$$\begin{eqnarray}&&\tan \iota \tan \theta =\displaystyle \frac{{\rm{\Delta }}v}{2\,{v}_{\max }-{\rm{\Delta }}v}.\end{eqnarray} \tag{ 5 }$$

The radial dependence of the systemic radial velocity of the jet (v_r) has been obtained by computing its average value at each radius from Figure 4(b). The values of v_r for the approaching and receding jet and its average value are shown in Figure 8 (left). From this figure, values of 183.2 km s⁻¹ and 21.9 km s⁻¹ are obtained for ${v}_{\max }$ and Δv, respectively. The small value of the ratio ${\rm{\Delta }}v/{v}_{\max }$ hints at small values of ι and θ.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The projection of the half-aperture angle of the precession cone θ on the plane of the sky ($\theta ^{\prime}$) depends on θ and ι according to

$$\begin{eqnarray}&&\tan \theta ^{\prime} =\displaystyle \frac{\tan \theta }{\sin \iota }\end{eqnarray} \tag{ 6 }$$

as can be inferred from Figure 7. The averaged PA at each radius of the jet has been computed from the surface brightness map in Figure 4(a). The PA of the receding and approaching components of the jet and their average value are shown in Figure 8 (center), from which a value of 102° is derived for $\theta ^{\prime}$.

The values of ${v}_{\max }$ and Δv define the red curve in Figure 8 (right). Similarly, the value of $\theta ^{\prime}$ defines the blue curve in Figure 8 (right). An inclination angle of the jet ι of 330 and half-aperture angle of the precession cone θ of 56 are defined where these two curves cross. According to Equation (2), the space velocity of the jet is derived to be 206 km s⁻¹.

The S-shaped morphology of the jet and its velocity variations have been subsequently modeled assuming a continuous ejection of a precessing jet with the velocity v, inclination ι, and aperture angle θ given above. This simple analytical model indeed provides a reasonable fit to the observed morphology (Figure 9, left) and kinematics (Figure 9, right) of the jet of NGC 2392—taking into account their notable complexity and the episodic nature of the ejection—for a precession period of 3200 yr and an age of 2400 yr at the adopted Gaia distance of 1.84 kpc. The full linear span of the jet would be 1.0 pc, i.e., ≈2.3 times larger than the nebular diameter of 0.43 pc derived from its angular size. Its mass-loss rate would be 1.0 × 10⁻⁷ M_⊙ yr⁻¹ and the mechanical luminosity 0.4 L_⊙, a tiny fraction of the 3000–6000 L_⊙ stellar luminosity (Herald & Bianchi 2011).

Zoom In Zoom Out Reset image size

Figure 9 is additionally overlaid with variations of the different parameters that also produce reasonable fits to the observations. These variations define the uncertainty in jet parameters, but they can be rather envisaged as intrinsic velocity or direction fluctuations in the launch of the jet or variations caused by its interaction with the nebular material that are not considered in the simplistic model used here to describe its morphology and kinematics.

The brightening of the inner jet at the inner shell rim (Figure 5(a)) suggests the interaction of the jet with the inner shell. Projection effects can be important, though, as the inner jet has a full linear span of ≃0.4 pc, whereas the maximum size of the inner shell, as projected in the plane of the sky, is only ≃0.2 pc. The size of the inner shell along the line of sight is most likely considerably larger, as suggested by the model presented by García-Díaz et al. (2012), and may be comparable to the inner jet extent. The similarity between the expansion velocity of the inner shell, ≃120 km s⁻¹ (García-Díaz et al. 2012), and that of the inner jet at the location of its rim, ≃150 km s⁻¹, suggests that the inner shell is being dragged by the jet.

It is also worthwhile to note the correspondence between the mid jet morphology and some cometary knots of the outer shell. The loop-like feature of the Western mid jet seems to surround a cometary knot and the Eastern mid jet curves as it seems to reach a cometary knot. These morphological correspondences might be indicative of the interaction of the jet with nebular material, but we note that those cometary knots have been suggested to be close to the plane of the sky, whereas the jet is only 33° apart from the line of sight.

A comprehensive study of the detailed spatio-kinematic properties of the inner and outer shells of NGC 2392 and its cometary knots to investigate the possible jet–nebula interactions will be the goal of a subsequent work (J. S. Rechy-García et al. 2021, in preparation).

The close-up view of the jet emission in Figure 5(a) shows that, within the spatial resolution provided by the present observations, the jet of NGC 2392 emanates from its central star, confirming previous suggestions based on the emergence of the jet emission from the stellar continuum in high-dispersion echelle spectroscopic observations (Gieseking et al. 1985; García-Díaz et al. 2012). On these grounds, the high-ionization polar stellar wind proposed for its central star (Prinja & Urbaneja 2014) could be interpreted as the onset of the jet rather than an isotropic stellar wind. Indeed, the terminal velocity of such stellar wind ≈300 km s⁻¹ is abnormally low for central stars of PNs, but close to the velocity of the jet.

The launch mechanism of the jet of NGC 2392 is therefore presently active, i.e., it is an ongoing jet (Miszalski et al. 2019). This is also the case of the late AGB jets in BD+46°442 and IRAS 19135+3937 (Bollen et al. 2019). The mass-loss rates of those jets is (0.5–1) × 10⁻⁶ M_⊙ yr⁻¹ and (1.3–4) × 10⁻⁶ M_⊙ yr⁻¹, respectively (Bollen et al. 2020). Adopting a ratio of ejected mass-loss rate to accretion rate for the jet in the range 0.01–0.3 for NGC 2392, similar to that adopted by Bollen et al. (2020), the accretion high-regime would be in the range 4 × 10⁻⁷–10⁻⁵ M_⊙ yr⁻¹. The high rate of accretion of late AGB jets sustained by the heavy stellar mass loss at this phase is apparently not operating in NGC 2392, which is a mature PN.

This seems natural, as once the central star of a PN enters the post-AGB phase and its heavy mass loss ceases, the accretion onto a companion and the jet formation can be expected to come to an end, too. Indeed, jets are found to be mostly coeval with their PNs (Guerrero et al. 2020). Only the jets of the post-common envelope (post-CE) PNs Hb 4, NGC 6337, and NGC 6778 (García-Díaz et al. 2009; Tocknell et al. 2014; Derlopa et al. 2019) have been proposed to originate during the post-AGB phase. These post-CE jets would be fed by the re-accretion of nebular gas or material from the remains of a circumbinary disk onto the primary (Soker & Livio 1994). Besides the case of Hen 2-90 (Sahai & Nyman 2000; Guerrero et al. 2001), which might be a symbiotic star rather than a PN (Schmeja & Kimeswenger 2001), all of the jets in PNs, including those in post-CE systems, are spatially detached from their central stars, with notable spatial gaps between them and the innermost regions of the jets. It is then implied that the jet collimation in PNs ceased hundreds to thousands of years ago, making the jet in NGC 2392 the only ongoing one observed in a mature PN.

NGC 2392 has been proposed to be an analog to NGC 6543 and NGC 7009 with a pole-on orientation (García-Díaz et al. 2012). More naturally, the large jet-to-nebula size ratio of NGC 2392 derived above, ≈2.3, and the fast expansion velocity and S-shaped morphology of its jet, would make it rather an analog to Fleming 1, the archetype of PNs with bipolar, rotating, episodic jets (BRET, López et al. 1993; Palmer et al. 1996). The physical structure of NGC 2392, with its ring of high-density knots (O'Dell et al. 1990) and long precessing jet, is also reminiscent of IPHASX J194359.5+170901 (Corradi et al. 2011). Interestingly, both Fleming 1 and IPHASX J194359.5+170901 harbor post-CE binary systems (Boffin et al. 2012; Miszalski et al. 2013). Their jets have been suggested to pre-date their formation (Tocknell et al. 2014), which may be the case of NGC 2392 as well given the age of 1140/d yr kpc⁻¹ proposed for the inner shell (O'Dell & Ball 1985). NGC 2392 would then be a young twin of Fleming 1 and IPHASX J194359.5+170901 caught in the short time lapse between the PN formation and the jet extinction.

Finally, it is worth noting that the central star of NGC 2392 has recently been recognized to be in a binary system (Miszalski et al. 2019). Its orbital period is yet unsettled, as periodic radial velocity variations of 3 hr (Prinja & Urbaneja 2014) and 1.9 days (Miszalski et al. 2019) and a hard X-ray emission modulation of 6 hr (Guerrero et al. 2019) have been reported. The companion of the central star of NGC 2392 has been proposed to be a hot white dwarf, as its effective temperature of ≃43,000 K (Méndez et al. 2012) cannot explain its high nebular excitation, being notably discrepant from the higher He ii Zanstra temperature (Heap 1977; Pottasch et al. 2008; Miszalski et al. 2019). This would make the central star of NGC 2392 one of the few double-degenerate ones as Fleming 1 itself or Hen 2-428 (Boffin et al. 2012; Santander-García et al. 2015; Reindl et al. 2020). The present-day launch and collimation of a jet in NGC 2392 strongly supports the presence of a binary system and an accretion disk around a compact source at its heart. The low mass-loss rate of its central star, (3–5) × 10⁻⁸ M_⊙ yr⁻¹, is not capable, however, of feeding the material in the jet (Herald & Bianchi 2011), which would be rather accreted from a circumbinary disk remnant of a CE episode late in the AGB phase. Indeed, the peculiar chemical abundances of the central star of NGC 2392 have been suggested to originate in such a CE phase (Méndez et al. 2012). The central star of NGC 2392 would be in its final transition toward a double-degenerate binary system, when the second component of the system evolves into the white dwarf stage ejecting its envelope to form a PN-like shell. It is of the utmost importance to determine the real orbital period and masses of the double-degenerate binary system in NGC 2392 to assess its possible fate as a Type Ia supernova.