Disk Illumination and Jet Variability of the Herbig Ae Star HD 163296 Using Multi-epoch HST/STIS Optical, Near-IR, and Radio Imagery and Spectroscopy

Optical observations of HD 163296 were taken from 2018 March 18 to 2018 November 7 in the B, V, and I bands with an observing cadence of ∼1 observation day (Kafka 2018). Data were obtained through a request via AAVSO and taken with observers associated with Vereniging Voor Sterrenkunde, Werkgroep Veranderlijke Sterren (Belgium). The standard stars UCAC4 341-118672 and UCAC4 340-118109 were observed with the science target and utilized in the reduction. The data were reduced using the photometry reduction software LESVEPHOTOMETRY V1.2.0.90,¹⁰ including flat-fielding and bias subtraction. All of the optical photometric observations are plotted in Figure 2.

Zoom In Zoom Out Reset image size

On 2018 April 14 and 2018 July 18, HD 163296 was observed with HST/STIS for four sequential orbits at each of these epochs. It was observed using the STIS occulting bar positions A0.6 (06 width) and A1.0 (10 width) for orbits 1, 2, and 4, with each orbit offset in absolute roll angle by 15°. In coronagraphic mode, HST/STIS is filterless, with a CCD sensitive from 2000 to 10,300 Å. The (B–V) color-matched point-spread function (PSF) star HD 145570 was observed in orbit 3. The total exposure times for the July epoch were slightly shorter to allow for the entire CCD to be readout, whereas the April epoch only had the bottom half of the CCD readout. A summary of these observations is presented in Table 1. We note that while the A0.6 wedge data were taken, reduced, and analyzed, they did not reveal any information that was not seen in the A1.0 wedge, and thus we will not discuss these data any further in this paper.

Table 1.  HST/STIS Observations

Target Name	Date	Total	Notes
		Exposure (s)
HD 163296	1998 Sep 2	864	Wedge A1.0
HD 141653	1999 Jun 28	782	Wedge A1.0; PSF star
HD 163296	2018 Apr 14	4050	Wedge A1.0
HD 145570	2018 Apr 14	896	Wedge A1.0; PSF star
HD 163296	2018 Jul 18	3780	Wedge A1.0
HD 145570	2018 Jul 18	896	Wedge A1.0; PSF star
HD 163296	2017 Jul 23	1792	52' × 005 slit; PA = 85°
HD 163296	2017 Aug 10	2021	52' × 005 slit; PA = 453

Note. The HST/STIS coronagraphic and UV spectroscopic observing log. The total exposure is the sum total of individual exposures at each epoch.

Download table as:  ASCIITypeset image

We followed the basic PSF subtraction techniques outlined in Grady et al. (2000). First, we found the location of the star in each frame using the "x" marks the spot method developed by Schneider et al. (2009). Next, we performed PSF matching, where we subtracted the individual PSF frames from the science images in turn to see which PSF frame leaves the smallest residual. This PSF matching technique can help compensate for the "breathing"-induced PSF changes HST experiences during a given orbit. Once we found the best PSF match for each science frame, we subtracted the PSF star from the science star, scaling the PSF star flux to compensate for the different apparent brightness. Examples of the PSF-subtracted frames are shown in Figure 3, where a red plus sign denotes the location of the star HD 163296.

Zoom In Zoom Out Reset image size

We explored how to scale the flux of the PSF star to optimally subtract the central star flux of HD 163296. The nominal V-band magnitude for HD 163296 and the PSF star HD 141653 is V = 6.93 ± 0.14 (Ellerbroek et al. 2014) and 4.928 ± 0.009 (Høg et al. 2000) mag, respectively, suggesting a nominal PSF scale factor of 0.158. However, HD 163296 is a known aperiodic variable star (Ellerbroek et al. 2014; see also Figure 2). Thus, we used our AAVSO photometry to help constrain the PSF scaling. The 2018 July epoch observations have contemporaneous AAVSO photometry taken on 2018 July 15 with measured V-band magnitudes of 6.797 ± 0.013 and 6.729 ± 0.013 mag (scale factors 0.179 and 0.190), and the closest photometry points for the 2018 April epoch had values of 6.714 ± 0.011 and 6.812 ± 0.01 mag taken on 2018 April 15 (scale factors 0.193 and 0.176). Since the STIS observations are clear filter observations, the scale factors are not necessarily perfect, so we explored the scale factor space by performing the PSF subtraction and looking where the most stellar flux has been removed while minimizing the oversubtraction (e.g., negative flux) of the disk. For the July epoch, we found that the best scaling factor was 0.1773, corresponding to a V-band magnitude for HD 163296 of 6.806 mag. As for the April epoch, we found that the best scaling factor is 0.1680, corresponding to a V-band magnitude for HD 163296 of 6.87 mag. These best PSF scaling factors are within 0.1 mag of the contemporaneously observed AAVSO photometry.

Finally, the wedge and spider arms were masked, and the frames were aligned to a common orientation, rotated north, and median-combined to create the final images as shown in Figure 4. We looked for variation in flux from HD 163296 within a given epoch (2018 April or 2018 July) and did not detect any significant brightness changes in the disk. Note that while the 2018 April and 2018 July epochs were taken as part of the same observing program, we choose not to stitch these observations together, as is traditional (see Schneider et al. 2009), in order to enable us to explore potential variability between these epochs, as discussed below in subsections 3.1 and 5.1.

Zoom In Zoom Out Reset image size

We also rereduced the 1998 September epoch observation of HD 163296, which was originally presented in Grady et al. (2000). As noted by Grady et al. (2000), the 1998 September epoch did not have contemporaneous observations of a PSF star. We tested several other PSF stars previously observed with HST/STIS and the A1.0 wedge having similar (B–V) colors as HD 163296 (HD 135298, HD 36546, HD 145570, and HD 141653) to try and reduce the number of PSF subtraction artifacts in these data. We concluded that the PSF star HD 141653 used in Grady et al. (2000) was still the best PSF match. While we did use the same PSF star and scale factor as Grady et al. (2000), we adopted a different centering routine ("x" marks the spot), and we used PSF matching when subtracting the best-matching individual PSF frame from each science image as described above. We found that the same scaling factor of 0.213 utilized by Grady et al. (2000; PSF V = 5.194 mag, science V = 6.873 mag) best removed the stellar light from the science images. We visually compared the new final image using our new reduction to that of Grady et al. (2000). We found that we reduced the number of residual speckles around the ansae region of the disk, or the "wagon wheel spoke" effect. We will utilize our new reduction of the 1998 September epoch data shown in Figure 4 for the rest of this paper.

We monitored the near-IR behavior of HD 163296 from 2016 to 2018 using NASA's Infrared Telescope Facility (IRTF) and the Apache Point Observatory (APO) 3.5 m. We obtained 11 observations from 2016 April to 2018 September using the SpeX spectrograph (Rayner et al. 2003) at IRTF in its short- (0.8–2.4 μm) and long-wavelength modes (2.3–5.5 μm; see Table 2) using a 08 wide slit. Additionally, we performed three observations in 2018 April and 2018 May with the TripleSpec spectrograph (Wilson et al. 2004) at the APO 3.5 m telescope, covering a spectral range of 0.95–2.46 μm (see Table 2). We observed the A0V star HD 163336 for telluric corrections and flux calibration for all observations. Observations noted in Table 2 had contemporaneous IRTF/SpeX prism spectra taken with a 30 wide slit, which was utilized to correct for absolute flux variations. Observations without prism spectra were scaled to match the observations with prism data based on the optical flux component of their spectra (0.8–0.9 μm). These observations were reduced and calibrated using the standard reduction packages Spextool and Triplespectool, respectively (Vacca et al. 2003; Cushing et al. 2004). Note that the near-IR observations on 2018 May 16 and 2018 June 24 were previously presented in Rich et al. (2019). Sample SpeX and TripleSpec spectra are plotted in Figure 5.

Zoom In Zoom Out Reset image size

The VLA observed HD 163296 using the C-band receiver six times between 2018 March and 2018 November, with three observations in A configuration and three observations in D configuration, as described in Table 2. The two 1 GHz bandpasses were centered at 4.76 and 7.36 GHz. It was observed that J1820–2528 calibrated the complex gain, and 3C 286 was observed to calibrate the bandpass and absolute flux density scale. Our observations were intended to be scheduled 1 month apart and coincide with our HST/STIS observations; however, due to infrastructure work at the VLA, observations were not possible between 2018 May and September. We utilized CASA version 5.1.2 (McMullin et al. 2007) to reduce and analyze these data. Fluxes were measured from cleaned images by fitting a Gaussian profile to the flux images using the imfit tool. We also preformed RFI flagging. The measured fluxes are listed in Table 3.

The object HD 163296 was observed with HST/STIS using the FUV-MAMA detector, the G140M grating centered at 1218 Å, and the 52' × 005 slit as part of program 14690 (PI: Güenther). Two observations occurred on 2017 July 23 and 2017 August 10 at slit position angles (PAs) of 85° (approximately perpendicular to the jet axis) and 453 (approximately aligned with the jet axis of 425) and exposure times of 1792 and 2021 s, respectively. Both observations were reduced using the standard STIS pipeline, with standard reduction techniques that included dark subtraction, flat-fielding, and wavelength calibration. Subtracting the 2017 July epoch from the 2017 August epoch, scaled to match exposure depth, allows us to remove background and star continuum flux without affecting the jet. The resultant 2D spectra are shown in Figure 6.

Zoom In Zoom Out Reset image size

Scattered light from HD 163296's disk is observed in all of the HST/STIS epochs, as shown in Figure 4, from the edge of our effective inner working angle of ∼05 (51 au) to 5(500 au). The disk major axis is approximately oriented at a PA of 132°, similar to what has previously been observed with ALMA (Isella et al. 2016), and inclined at 42°. In the rereduced 1998 September epoch data, we observe the two ansae (SE and NW ansae) originally presented by Grady et al. (2000), located at a projected distance of 33 (330 au) from the star. The disk ring flux distribution is brightest on the near side of the disk (NE side), and we do not see any flux on the far side of the disk (SW side). This flux distribution is similar to the broken ring structures seen at the first ring (065, 66 au) in near-IR observations (Garufi et al. 2017; Monnier et al. 2017; Muro-Arena et al. 2018; Rich et al. 2019). While we refer to these features as ansae, we are interpreting them as the fourth dust ring in the HD 163296 system. We note that both ansae were not observed simultaneously in our 2018 data set. We only observed the SE ansae in the 2018 April epoch data, and we only observed the NW ansae in the 2018 July epoch data. Previously, Wisniewski et al. (2008) also only observed one ansa being present around HD 163296. The location of the ansae had not changed from the 1998 and 2018 epochs.

Assuming that the broken ring is circular and has an inclination of 42° and PA of 132° (Isella et al. 2016), we plotted an ellipse on the 1998 September epoch image (Figure 7) and found that the broken ring disk is consistent with a minor axis offset of 07. The 2018 April and 2018 July observations are also consistent with the same minor axis offset. Assuming that the fourth ring is perfectly centered around the star, the minor axis offset of 07 would correspond to a dust scale height of 64 au at a radial distance of 330 au. We will discuss the minor axis offset for HD 163296 in Section 5.

Zoom In Zoom Out Reset image size

We observe additional nonazimuthal scattered light from the near side of the disk north of the NW ansa, increasing in intensity as it approaches the NE minor axis. This feature is observed in all three epoch images (Figure 4) and labeled as "disk periphery" in Figure 8. It has a larger radial extent than the prominent ansae. We will discuss the origin of this feature later in Section 5. Finally, in the middle panel of Figure 4, showing epoch 2018 April, there is a faint linear streak radiating from the top of the disk at a PA of 45° along the jet axis. While this could be a signature of the jet, the linear streak resembles other radial streaks observed in the 2018 April and July epochs. Thus, the feature is most likely a residual.

Zoom In Zoom Out Reset image size

We computed the azimuthal dependence of the disk surface brightness within the ansae region of the disk (fourth ring, 300-360 au; see Figures 7 and 8), binning the flux within these elliptical annuli using an assumed inclination of 42° and PA of 132° (Isella et al. 2016). The resultant azimuthal surface brightness profile (Figure 9) includes 3σ error bars computed from photon-, read-, and dark-noise contributions. Several azimuthal bins are contaminated by background stars causing these regions to have abnormally high flux, such as the region near the SE minor axis.

Zoom In Zoom Out Reset image size

Figure 9 exhibits a large variation in the surface brightness of the ansae region of the disk both azimuthally (PA) and in time. In the 1998 September epoch, the disk is brighter closer to the NE minor axis than the SW minor axis. The SE (19.65 ± 0.11 STMag arcsec⁻²; PA = 124°) and NW (19.29 ± 0.10 STMag arcsec⁻²; PA = 332°) ansae exhibit similar surface brightnesses. By contrast, during the 2018 April epoch, we only detect the SE ansa, and its surface brightness (20.40 ± 0.02 STMag arcsec⁻²; PA = 124°) is ∼0.75 STMag arcsec⁻² dimmer than in 1998 September. The surface brightness near the expected location of the NW ansa is flat in 2018 April and much dimmer (20.68 ± 0.02 STMag arcsec⁻²; PA = 332°) than the 1998 September epoch. Conversely, we only detect the NW ansa in the 2018 July epoch. The surface brightness of the NW ansa (20.08 ± 0.01 STMag arcsec⁻²; PA = 332°) is ∼0.79 STMag arcsec⁻² dimmer than then 1998 September epoch. The surface brightness near the expected location of the SE ansa in the 2018 July epoch is dimmer (20.78 ± 0.02 STMag arcsec⁻²; PA = 124°) than both the 1998 September and 2018 April epochs. We will discuss the implications of these results in subsection 5.1.

We similarly computed the azimuthal surface brightness of the disk periphery region (420–500 au; see Figures 7 and 8), as seen in Figure 10. We note that all epochs exhibit the same trend, the surface brightness increases toward the NE minor axis (i.e., the near side of the disk), and we observe little azimuthal variation between the two 2018 epochs. However, the 2018 April (20.54 ± 0.02 STMag arcsec⁻²) and 2018 July (20.17 ± 0.02 STMag arcsec⁻²) epochs are dimmer than the 1998 September epoch (19.85 ± 0.10 STMag arcsec⁻²) at a PA = 12°. Figure 10 exhibits a large jump in surface brightness along the SW minor axis that is caused by background stars within those bins. An additional ∼1 STMag arcsec⁻² discontinuity in the surface brightness is present in the 2018 epochs at PA = 330°. This is further evidence of nonazimuthal variations within the disk and possibly due to a transient shadowing effect that was not present during the 1998 September epoch.

Zoom In Zoom Out Reset image size

Devine et al. (2000)and Grady et al. (2000) identified three HH knots, labeled A, B, and C, along the minor axis of their 1998 September HST observations (see Figure 4). To date, there have been 11 HH knots associated with HD 163296 (A, A2, A3, B, B2, C, D, E, F, G, and H; Devine et al. 2000; Grady et al. 2000; Wassell et al. 2006; Ellerbroek et al. 2014). Given the measured proper motion of the HH knots (Ellerbroek et al. 2014; red jet: ${v}_{t,{\rm{r}}{\rm{e}}{\rm{d}}}=0.28\pm 0.01$ arcsec yr^–1, blue jet: ${v}_{t,{\rm{b}}{\rm{l}}{\rm{u}}{\rm{e}}}=0.49\pm 0.01$ arcsec yr^–1), we predict that five of these HH knots (A3, B2, B, C, and D) should fall within the field of view of our 2018 July epoch data. Utilizing the HH knot proper-motion analysis from Ellerbroek et al. (2014, and references therein), we annotate the predicted locations of the HH knots for the 1998 September and 2018 July HST data (Figure 11). Surprisingly, we do not see any HH knots in either our 2018 April or 2018 July observations. We inserted artificial HH knots into our data at the predicted locations of HH knots B and C, performed aperture photometry, and found that the knots could be 15 times dimmer and detected with a 3σ significance above the background. We will discuss the nondetection of these HH knots in Section 5.3.

Zoom In Zoom Out Reset image size

Figure 6 shows the clear presence of the jet associated with HD 163296 in the 2017 UV observation out to a distance of 50 but does not reveal any evidence of the counterjet. Since the PA of the long slit was not perfectly aligned with the jet axis, and the slit is four times smaller (005) than previous HST/STIS long-slit observations of the jet (02; Devine et al. 2000), we expect much of the jet flux at larger distances from the star to be located outside of the slit. For this reason, we do not compare the flux of the jet that we observe to previous studies (Devine et al. 2000; Wassell et al. 2006; Günther et al. 2013). We expect the counterjet to be masked by the forward portion of the protoplanetary disk that extends out along the minor axis of the disk to a projected distance of 29. Thus, a portion of the counterjet could have been detected. We measure that the jet has a velocity of −336 km s⁻¹, which matches what was previously observed by Devine et al. (2000; 380–330 km s⁻¹).

Our optical (B-, V-, and I-band) photometry of HD 163296 throughout 2018 (Figure 2) exhibits evidence of moderate variability. After excluding discrepant photometry using a 3σ clipping algorithm, we find the average magnitude of the system to be 6.85 ± 0.15, 6.75 ± 0.11, and 7.27 ± 0.09 mag in the B, V, and I filters, respectively. The typical error for an individual observation is 0.02, 0.01, and 0.02 mag, respectively, in these filters. The clipping algorithm is necessary, as there are discrepant photometry data points that could induced by nonphotometric skies. Without observing logs, we cannot strategically remove these points any other way. We note that this analysis could be removing high-amplitude variations in flux in the light curve.

We observe two potential dipper events on 2018 June 7 and 2018 August 7 that each appear in all three bandpasses. The first event, which occurred between the two epochs of our HST coronagraphic observations, is immediately proceeded by a gradual rise to the average brightness over a several-day time frame. The event had an amplitude of ∼0.5 mag, although this is only a lower limit due to the sparse temporal sampling of our data. The 2018 June 7 event appears similar to four previous events noted by Ellerbroek et al. (2014) and is both smaller and shorter in duration than the 2001 event that reduced the system flux by ∼1 mag and lasted several months (Ellerbroek et al. 2014). The second dipper event was only observed by a single night observation, albeit in three filters. Robustly quantifying both of these dipper events is hampered by the sparse temporal coverage of our data. Thus, we label the first event as a potential dipper event and the second as a suggestive dipper event. We note that further analysis of the color of dipper events in 2012 of this system is presented in the appendix of Pikhartova et al. (2020).

We convolved a K-band filter with our near-IR flux-calibrated spectra to extract K-band magnitudes and quantify variability in these data (see Table 4 and Figure 12). We have also included historical K-band magnitudes from Ellerbroek et al. (2014) and citations therein in Figure 12. The K-band light curve exhibits variations on timescales of both months and years, with a total amplitude of ∼0.94 mag. Although our temporal coverage between 2012 and 2016 was poor, the system clearly dimmed by ∼0.5 mag during this time frame. Since 2016, the system has exhibited a gradual ∼0.4 mag increase in Ks-band flux. We do not see any outburst events in these data, such as the 2002 outburst that possibly coincided with the launch of an HH knot (Ellerbroek et al. 2014; Pikhartova et al. 2020).

Zoom In Zoom Out Reset image size

Table 4.  Near-IR Accretion Measurements Using Brγ

HJD	Kobs	EW(Brγ)obs	EW(Brγ)cs	L(Brγ)	log Macc(Brγ)
	(mag)	Å	Å	$({10}^{-3}{L}_{\odot }$)	(${M}_{\odot }$ yr−1)
2002 Mar 23a	4.5 ± 0.05	−3.0	−7.2 ± 1.2	1.8 ± 0.3	−6.73 ± 0.35
2002 Jul 18b	4.5 ± 0.05	−3.1	−7.3 ± 1.2	1.8 ± 0.3	−6.72 ± 0.35
2004 Jun 9c	4.8 ± 0.05	−4.7	−10.2 ± 1.2	1.9 ± 0.2	−6.69 ± 0.34
2005 Jul 6b	4.6 ± 0.05	−4.2	−8.8 ± 1.2	2.0 ± 0.3	−6.67 ± 0.34
2008 Mard	4.8 ± 0.05	−4.3	−9.8 ± 1.2	1.8 ± 0.2	−6.71 ± 0.34
2008 May 13d	4.8 ± 0.05	−4.3	−9.8 ± 1.2	1.8 ± 0.2	−6.71 ± 0.34
2009 Jul 15e	4.8 ± 0.05	−3.2	−8.7 ± 1.2	1.6 ± 0.2	−6.77 ± 0.35
2011 Oct 12f	4.8 ± 0.05	−4.2	−9.7 ± 1.2	1.8 ± 0.2	−6.71 ± 0.34
2011 Oct 14f	4.5 ± 0.05	−3.3	−7.5 ± 1.2	1.8 ± 0.3	−6.70 ± 0.35
2011 Oct 16f	4.3 ± 0.05	−3.9	−7.4 ± 1.2	2.2 ± 0.4	−6.61 ± 0.34
2012 Mar 24f	4.3 ± 0.05	−3.7	−7.2 ± 1.2	2.1 ± 0.4	−6.62 ± 0.35
2012 May 17f	4.2 ± 0.05	−3.7	−6.9 ± 1.2	2.2 ± 0.4	−6.60 ± 0.35
2012 Jul 5g	4.7 ± 0.05	−4.3	−9.3 ± 1.2	1.9 ± 0.2	−6.68 ± 0.34
2016 Apr 4	5.07 ± 0.11	−4.22	−11.28 ± 1.2	1.6 ± 0.2	−6.77 ± 0.35
2016 May 4	5.11 ± 0.11	−5.33	−12.66 ± 1.2	1.8 ± 0.2	−6.72 ± 0.34
2016 Jun 9	5.05 ± 0.11	−4.37	−11.33 ± 1.2	1.7 ± 0.2	−6.76 ± 0.34
2016 Aug 10	5.0 ± 0.11	−3.56	−10.23 ± 1.2	1.6 ± 0.2	−6.79 ± 0.35
2016 Sep 7	5.14 ± 0.11	−4.18	−11.76 ± 1.2	1.6 ± 0.2	−6.78 ± 0.35
2017 May 25	5.13 ± 0.11	−4.49	−11.99 ± 1.2	1.6 ± 0.2	−6.77 ± 0.34
2017 Sep 12	4.96 ± 0.11	−4.21	−10.63 ± 1.2	1.7 ± 0.2	−6.75 ± 0.35
2018 Apr 8	4.94 ± 0.11	−5.76	−12.04 ± 1.2	2.0 ± 0.2	−6.67 ± 0.34
2018 Apr 16	5.01 ± 0.11	−5.23	−11.91 ± 1.2	1.8 ± 0.2	−6.71 ± 0.34
2018 May 16	4.76 ± 0.11	−3.97	−9.31 ± 1.2	1.8 ± 0.2	−6.72 ± 0.35
2018 Jun 24	4.77 ± 0.11	−5.17	−10.54 ± 1.2	2.0 ± 0.2	−6.65 ± 0.34
2018 Aug 11	4.95 ± 0.11	−3.35	−9.69 ± 1.2	1.6 ± 0.2	−6.79 ± 0.35
2018 Sep 22	4.85 ± 0.11	−4.58	−10.37 ± 1.2	1.8 ± 0.2	−6.70 ± 0.34

Notes. Sources of original magnitude and EW measurements are indicated in footnotes.

^aBrittain et al. (2007). ^bSitko et al. (2008). ^cGarcia Lopez et al. (2006). ^dDonehew & Brittain (2011). ^eEisner et al. (2010). ^fMendigutía et al. (2013). ^gEllerbroek et al. (2014).

Download table as:  ASCIITypeset image

We calculate the mass accretion rate outlined in Ellerbroek et al. (2014). Equivalent widths (EWs) of the Brγ line were measured using the IRAF splot tool using a Gaussian function (Table 4). To calculate the line luminosity, we first correct our measured EW_obs by removing the effect of the photospheric line, as shown in Equation (1). We adopt the same photospheric EW (EW_phot = −22 Å) as Mendigutía et al. (2013) and the same reddening (A_V = 0.5 mag) as Ellerbroek et al. (2014). We use these values to calculate the line luminosity (L_line; Equation (2)). We then convert the line luminosity into the accretion luminosity in Equation (3) utilizing the relation by Fairlamb et al. (2017). Finally, the mass accretion rate is calculated in Equation (4) assuming R_* = 1.7 ${R}_{\odot }$ (Pikhartova et al. 2020). The values are listed in Table 4 and plotted in Figure 12. We recalculate the accretion rates previously presented in Ellerbroek et al. (2014) and include the updated distance, R_* values, and line luminosity in the accretion luminosity relation. We find no substantial changes in accretion rate from 2003 to 2018, as shown in Figure 12:

$$\begin{eqnarray}&&{{EW}}_{\mathrm{cs}}={{EW}}_{\mathrm{obs}}-{{EW}}_{\mathrm{phot}}{10}^{-0.4| {\rm{\Delta }}K| },\end{eqnarray} \tag{ 1 }$$

$$\begin{eqnarray}&&{L}_{\mathrm{line}}=4\pi {d}^{2}{{EW}}_{{cs}}{F}_{K}{10}^{0.4{A}_{K}},\end{eqnarray} \tag{ 2 }$$

$$\begin{eqnarray}&&\mathrm{log}\left(\displaystyle \frac{{L}_{\mathrm{acc}}}{{L}_{\odot }}\right)=4.46(\pm 0.23)+1.30(\pm 0.09)\times \mathrm{log}\left(\displaystyle \frac{{{\rm{L}}}_{\mathrm{Br}\gamma }}{{{\rm{L}}}_{\odot }}\right),\end{eqnarray} \tag{ 3 }$$

$$\begin{eqnarray}&&{\dot{M}}_{\mathrm{acc}}^{* }=\displaystyle \frac{{L}_{\mathrm{acc}}{R}_{* }}{{{GM}}_{* }}.\end{eqnarray} \tag{ 4 }$$

Fluxes from the C-band VLA observations are listed in Table 3. During the launch of an HH knot, we expect to observe the radio flux increase by 50%, based on previous measurements of HH knot launches (Devine et al. 2000). Additionally, we should observe a change in the spectral index, with the value possibly becoming negative. We do not observe any significant changes in the C-band flux to suggest the launch of an HH knot. Additionally, our spectral index values are positive, which is consistent with flux originating as thermal free–free emission. Our radio observations are consistent with no new HH knots being launched during 2018.

The outer disk, as imaged with ALMA, is a series of three concentric rings at 066, 10, and 16 (67, 102, and 160 au; Isella et al. 2018; Dent et al. 2019) with gaps between each ring. We observed two ansae that form a fourth ring at 325 (330 au). Assuming this ring is circular, we measured a minor axis offset value of 07, which would correspond to a dust scale height of 64 au (07 offset). This latest dust scale height measurement joins a host of other minor axis offset measurements, such as those of the first dust ring in the near-IR (00432 ± 00016, Rich et al. 2019; 006, Garufi et al. 2014; 0105 ± 0045, Muro-Arena et al. 2018; 01, Monnier et al. 2017). As we add more and more measurements, we can better map the vertical dust distribution as a function of stellocentric separation, which will aid modeling of the system.

It has been previously noted that HD 163296 exhibits azimuthally asymmetric disk illumination on timescales <4 yr (Wisniewski et al. 2008; Rich et al. 2019). Our observations (Section 3.2) indicate that the timescale for such illumination variations is <3 months. Large-scale, azimuthally asymmetric illumination has been observed for the innermost disk ring (066, 67 au; Rich et al. 2019) and now the fourth disk ring (325, 330 au; Wisniewski et al. 2008 and this work), suggesting that a similar mechanism is responsible for these phenomena. As discussed in Rich et al. (2019), such mechanisms could include a warped inner disk structure shadowing the outer disk (Sitko et al. 2008) or dust ejected above the midplane of the disk that shadows the outer disk (Ellerbroek et al. 2014).

Assuming that the material causing the illumination variations is in a Keplerian orbit around the star, the observed <3 month variability timescale and a stellar mass of 2.14 ${M}_{\odot }$ (Pikhartova et al. 2020) imply a semimajor axis of <0.5 au. This is consistent with the nominal outer edge of the inner disk,  0.41 au (Setterholm et al. 2018). If the source of the shadowing resides in the inner disk, this material must reach a scale height of at least 0.08 au above the midplane in the inner disk (at 0.41 au) to shadow the fourth dust ring. Pikhartova et al.'s (2020) MCRT modeling suggested that dust located interior to 0.5 au is ejected above the disk midplane to a scale height of at least 0.08 au, and this could account for the broadband optical and near-IR photometric variability observed in the system's 2002 outburst. If this ejected dust is not azimuthally symmetric, it could also explain the azimuthally asymmetric disk shadowing that has been observed in the first and fourth dust rings.

We can also estimate the azimuthal size of the source of the disk shadowing. Based on Figure 9, we estimate that the ansae in the fourth ring are at least ∼30° in azimuthal extent. Assuming the source material that creates the shadowing resides <0.5 au from the star, this material must subtend an azimuthal size of 0.26 au to create an ∼30° shadow in the fourth ring.

The optical and near-IR light curves, Figures 2 and 12, respectively, do not exhibit any large dips or bursts indicative of the launch of any new HH knots within our observing windows. The optical light curve exhibits two dipper events: one likely event on 2018 June 7 and one possible event on 2018 August 7. The 2018 June 7 event has a depth of at least 0.52 mag in the V band. The light curves also exhibit variations larger than individual photometric uncertainties, suggesting that additional short-timescale variations could be present.

Dippers have been attributed to young stellar objects with inclinations that are nearly edge-on, with fluctuations of the inner disk scale height causing dimming of the star (Cody et al. 2014). More recently, it has been shown that dipper events can occur in systems where the outer disk has a low to moderate inclination (0° to ∼60°) and might even be occurring independently of inclination (Ansdell et al. 2019); HD 163296 is one such system that has a moderate outer disk inclination (42°) and experiences dipper events (see Figure 2 and Ellerbroek et al. 2014). The disk's inclination of 42° suggests that for the star to be occulted by disk material, dust would have to reach a height of 0.37 au above the disk at the nominal inner disk radius (0.41 au).

We can estimate the azimuthal size of the feature responsible for the dipper using its observed duration. Based on the 2018 June 7 event, the duration of the event cannot be longer than 10 days. This duration is similar to some of the dipper events observed in the system by Ellerbroek et al. (2014). If we assume that the dust causing the event is located around the inner disk (0.41 au) and moving at Keplerian speeds, we find that the diameter of the occulting material is 0.3 au. This corresponds to an azimuthal angle of 27°.

We remark that the physical sizes we have derived for inner disk (<0.5 au) features that could cause both dipper events and large azimuthally asymmetric shadowing of the first and fourth disk rings are broadly consistent with one another. These similarities suggest that the same (or similar) mechanisms could be responsible for both types of observational phenomena in the system.

Finally, we note there is possible long-term variability in our near-IR flux observations. Our observations at the beginning of 2016 are ∼0.5 mag dimmer, and the K-band flux increases by 0.3 mag from 2016 to 2018. Long-term variability in the IR light curves of disk-hosting stars has previously been found by Cody et al. (2014). We do not observe similar long-term variations in the optical light curve. Multiwavelength modeling of these light curves is required to assess the origin of the different observed behaviors.

We had not detected the launch of a new HH knot as of 2018 November 3. Ellerbroek et al. (2014) predicted that the system should launch a knot every 16 ± 0.7 yr, calculated from the proper motion and radial velocity of current HH knots. Since the last known launch of an HH knot was ∼2002, a new HH knot was expected in ∼2018. Ellerbroek et al. (2014) also predicted that there should be a decrease in optical flux from the reddening due to the disk wind and an increase in the near-IR flux from thermal radiation from dust in the disk wind. Instead, we see two minor optical dimming events in 2018 (see Figure 2) and a general decrease in near-IR flux from 2016 to 2018 as compared to Ks-band photometry before 2013.

We cannot definitively rule out the 16 yr periodicity of knot launches, as there is an ∼32% statistical probability that such an event could have happened after our final observation in November 2018. Imaging and/or spectroscopic observations beyond the 3σ uncertainty of the periodic behavior suggested by Ellerbroek et al. (2014; e.g., in 2020 or later) that are capable of identifying new HH knots are recommended to confirm our nondetection.

We do not detect any of the previously launched HH knots in our 2018 April and 2018 July HST/STIS imaging. Günther et al. (2013) noted that if an HH knot is not shock-heated, the knot itself can cool on a timescale of τ = 0.4 yr. Thus, it is possible that the HH knots have cooled sufficiently such that they are no longer visible. We note that the lack of detection of HH knots does not necessarily correlate with a change in the activity of the jet itself. The jet associated with HD 163296 has a decades-long history of activity. As the HH knots are primarily fed energy through shock heating and cool on relatively short timescales, the HH knots could be interacting with less material than previous HH knots, and they have cooled sufficiently to no longer be detectable with our 2018 April and July epoch imagery. Future observations to directly measure the current jet activity and confirm the dimming of older HH knots, e.g., with deeper ground-based slit spectroscopy, would help test our interpretation of current data.

The last detection of the jet in Lyα was 2017 August 10. We do not see any emission beyond 50; however, this is most likely due to the jet being located outside of the slit. We also do not see any evidence of any of the HH knots in these data. The time lag between this UV jet detection and the 2018 epoch coronagraphic observations is large enough for significant cooling to have occurred.

Pinte et al. (2018) inferred the presence of a 2 M_Jup planet located at 260 au based on analysis of gas velocities. We do not observe any excess flux in this region of our image, which is expected, as such a planet is too dim to be detected with our HST/STIS observations. We can look for the effects of the planet on the disk. In Figure 8, we have placed a diamond at the approximate separation and PA of the proposed planet. Interestingly, the planet is located between the third ring (160 au) and the ansae region (fourth ring; 330 au). Much like the first and second rings (Teague et al. 2018), the fourth ring could be formed through the dynamics of the Pinte et al. (2018) planet. New dynamic modeling of the disk that includes the Pinte et al. (2018) planet and the two proposed planets by Teague et al. (2018) in the two inner gaps and replicates the shadowing phenomenon presented in this work would help form a complete picture of this complex system.