EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 684 (April 2024) A&A, 684 (2024) A73 Full HTML
Press Release
Open Access
Issue
A&A
Volume 684, April 2024
Article Number A73
Number of page(s) 22
Section Planets and planetary systems
DOI https://doi.org/10.1051/0004-6361/202348736
Published online 09 April 2024

© The Authors 2024

Licence Creative CommonsOpen Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This article is published in open access under the Subscribe to Open model. Subscribe to A&A to support open access publication.

1 Introduction

Over 5500 exoplanets1 have been discovered to date, and they are extremely diverse in terms of their masses, compositions, and distributions around their parent stars. Planet formation theories, such as the core-accretion (Pollack et al. 1996) or disk gravitational instability (Boss 1997) models, must be able to explain the resulting diverse planetary systems. To investigate the formation processes, we can study the circumstellar disks that shape the planet-forming environments. Disk sub-structures, such as rings or cavities, are expected byproducts of planet formation and are indeed associated with the protoplanet-hosting PDS 70 (Keppler et al. 2018, 2019; Haffert et al. 2019) and AB Aur systems (Currie et al. 2022), although the evidence for AB Aur b was recently disputed by Zhou et al. (2023). Multi-wavelength observations trace different disk regions, including the large, millimetre-sized dust grains near the midplane (e.g. ALMA Partnership 2015) at longer wavelengths. Scattered light can be captured from the upper surfaces of the disk at optical and near-infrared (NIR) wavelengths and provides information about the material through the measurements of phase functions and the degree of polarised light. Since central stars are observed close to the peak of blackbody emission, a high-contrast imaging technique is employed to reveal the faint structures in their immediate vicinities. Polarimetric differential imaging (PDI; Gledhill et al. 1991, 2001; Kuhn et al. 2001) is especially well suited to observing the optical and NIR scattered light of a circumstel-lar disk. Unpolarised stellar light becomes polarised after being scattered by circumstellar dust grains, and PDI can be used to remove the stellar component, revealing the fainter polarised light structures below the diffraction halo of the star.

Several instruments, including the Subaru High-Contrast Coronagraphic Imager for Adaptive Optics (HiCIAO; Hodapp et al. 2008; Suzuki et al. 2010), the Gemini South Gemini Planet Imager (GPI; Macintosh et al. 2006, 2014), the Very Large Telescope (VLT) Nasmyth Adaptive optics system COude (NACO) NIR camera (Lenzen et al. 2003; Rousset et al. 2003), and the VLT Spectro-Polarimetric High-contrast Exoplanet REsearch instrument (SPHERE; Beuzit et al. 2019), have exploited the PDI technique to observe a large number of young stellar objects (YSOs). These instruments utilise a polarised beam-splitter to separate the incoming light into two beams with orthogonal linear polarisations. The instrumental point spread function (PSF) is unchanged for both beams, as they are recorded simultaneously. The high contrast (~10−2–10−4; Avenhaus et al. 2018) between the faint scattered light disk and the bright stellar halo can be suppressed by subtracting measurements of the two orthogonal polarisation states. In particular, PDI is an effective imaging technique for circumstellar disks with low inclinations (e.g. HD 169142, i ≈ 13° and TW Hya, i ≈ 7°; Hales et al. 2006; Apai et al. 2004; van Boekel et al. 2017), for which angular differential imaging (Marois et al. 2006) leads to the self-subtraction of the face-on disk’s signal.

Observations that use PDI have revealed a large number of disks with different sizes, surface brightnesses, and morphologies in scattered light. Scattered light observations trace the upper layers of a circumstellar disk since the micron-sized dust grains found there are optically thick at optical and NIR wavelengths. Hence, circumstellar disks must be flared to intercept the stellar radiation at large distances (Chiang & Goldreich 1997; de Boer et al. 2016; Ginski et al. 2016). Transition disks with large dust-depleted inner cavities are frequently detected (e.g. Mayama et al. 2012; Canovas et al. 2013; Keppler et al. 2018; Maucó et al. 2020), and the observed circumstellar disks commonly show rings at varying radii (e.g. Quanz et al. 2013; Muro-Arena et al. 2018; Avenhaus et al. 2018). Additionally, spiral features are frequently detected in scattered light (see Fig. 9 of Benisty et al. 2023). The gas perturbations, coupled to the small grains that are traced in scattered light, are suggested to emerge from interactions with a companion or with the environment. Furthermore, the combination with sub-millimetre observations can reveal dust filtering at pressure maxima (e.g. Garufi et al. 2013; Maucó et al. 2020) and help identify fragmentation, possibly resulting from gravitational instability (Weber et al. 2023). In scattered light imaging, the misalignment of an (un-resolved) inner disk can cast a shadow onto the outer disk (Bohn et al. 2022). Depending on the magnitude of the misalignment, narrow shadow lanes (e.g. HD 100453; Benisty et al. 2017) or wide-angle obscurations can appear (e.g. HD 143006; Benisty et al. 2018). In the case of stellar multiplicity, the geometry of the circumstellar environment can be assessed further by interpreting which stellar component is responsible for the dust illumination (Weber et al. 2023; Zurlo et al. 2023). Depending on the size, composition, and porosity of the small dust grains, different scattering phase functions can be measured (Shen et al. 2009; Tazaki et al. 2016, 2019). By studying the dust properties in circumstellar disks, we can assess the efficiency of dust growth based on the size, composition, and porosity of the grains involved. PDI observations are not limited to Class II disks (Lada 1987): second-generation dust disks, or debris disks, are also observed with facilities such as VLT/SPHERE (e.g. HIP 79977; Engler et al. 2017, HR 4796A; Milli et al. 2019), Gemini South/GPI (e.g. HD 157587; Millar-Blanchaer et al. 2016), and Subaru/HiCIAO (e.g. HD 32297; Asensio-Torres et al. 2016).

For the most studied Class II disks (Lada 1987), the observed sub-structures are frequently explained by invoking the presence of planetary companions (e.g. ALMA Partnership 2015; van der Marel et al. 2019; Long et al. 2019; Asensio-Torres et al. 2021). The existence of sub-structures suggests that planet formation is already underway and began when the YSOs were still embedded in their natal envelopes, during the Class 0 or I phases (t < 106 yr; Garufi et al. 2022a). Furthermore, measurements of the dust masses of Class II disks appear incompatible with predicted planet formation efficiencies and the masses of exoplanetary systems (Manara et al. 2018; Mulders et al. 2021). The higher dust masses of Class 0 and I disks determined by Tychoniec et al. (2020) could indicate that giant planet formation commences before the protostellar envelope has dissipated (Cridland et al. 2022; Miotello et al. 2023). Alternatively, the accretion of material from the surrounding cloud can continually replenish the mass of the protoplanetary disk. The total mass budget available for planet formation therefore exceeds the disk mass at any given time (Manara et al. 2018; Garufi et al. 2022a). The two explanations put forward to solve the missing mass problem demonstrate the important role of embedded Class 0 and I objects in the formation of planets. However, the earliest YSOs are particularly difficult to observe at optical wavelengths due to their embedded nature. As a consequence, the optical wavefront sensors (WFSs) of most modern extreme adaptive optics (AO) systems do not allow for an adequate AO correction of deeply embedded YSOs. The NIR AO188 system, part of SCExAO on the Subaru telescope, is an exception as it provides AO for polarimetric imaging in the northern hemisphere. However, embedded protostars in the south were only observable to some older, ground-based instruments that were equipped with an infrared WFS. For completeness, the retired NICMOS instrument on the Hubble Space Telescope measured the polarised light of the earliest YSOs (e.g. Silber et al. 2000; Kóspál et al. 2008; Perrin et al. 2009), though JWST is not equipped with polarimetric capabilities.

In this work, we present a re-reduction of polarimetric archival data from NACO, the Nasmyth Adaptive Optics System (NAOS), which together with the COude Near-Infrared CAmera (CONICA) forms the NACO instrument at the VLT (Lenzen et al. 2003; Rousset et al. 2003). Initially installed at the Nas-myth B focus of UT4 in 2001, NACO was reinstalled at the Nasmyth A focus of UT1 from 2014 until its decommissioning in 2019. NACO operated at wavelengths between 1 and 5 µm, and NAOS was equipped with a visible (0.45–1.0 µm) and infrared (0.8–2.5 µm) WFS, enabling observations of embedded YSOs despite their faint optical magnitudes. NACO was equipped with a Wollaston prism (and also wire grids; see Sect. 2.2.4) to perform polarimetric observations, and a half-wave plate (HWP) was installed in 2003. In Sect. 2, we describe how our PDI PiPelIne for NACO data (PIPPIN)2 reduces the NACO polari-metric data with the PDI technique. Section 3 outlines a broad inspection of the PIPPIN-reduced data, and we present a novel method for assessing the detection significance of a polarised signal. Section 4 compares the reduced data with those of the SPHERE instrument. The conclusions are summarised in Sect. 5, and the reduced data archive has been published online3.

2 Reduction of NACO data

2.1 Selection of polarimetric observations

Since the polarimetric mode of NACO was not solely used to observe YSOs, we made a selection of observations of interest to this study. First, the European Southern Observatory (ESO) archive was searched for every polarimetric NACO observation classified as the SCIENCE data type. Using the object identifier and the astroquery Python package (Ginsburg et al. 2019), we searched the SIMBAD archive (Wenger et al. 2000) to select any object that was ever classified as one of the following categories: (candidate) Orion variable, (candidate) Herbig Ae/Be star, (candidate) T Tauri star, or a (candidate) YSO. However, a large number of observations have unclear object identifiers. In these instances, astroquery was utilised to locate the object closest to the target right ascension (RA) and declination (Dec) coordinates. In total, we find 57 candidate Class 0–III objects that potentially exhibit polarised light from circumstellar material. As these systems were observed in multiple filters, epochs, or with different instrument setups, we find a total of 243 datasets. Table A.1 lists the objects of interest and information on the observation setup for each dataset.

2.2 PDI PiPellne for NACO data (PIPPIN)

A general pipeline to reduce NACO data is provided by ESO4. However, this pipeline cannot reduce the polarimetric observations and thus previous works utilised custom, self-written pipelines (e.g. Apai et al. 2004; Quanz et al. 2011; Canovas et al. 2013). The different data reduction methods could lead to inconsistent scientific results. For instance, one of the rings of HD 97048 observed by Ginski et al. (2016) was not recovered from the same data in the earlier analysis of Quanz et al. (2012). Such discrepancies can be avoided by using a single, comprehensive pipeline. In this section, we describe the operation of our PIPPIN pipeline, which applies the PDI technique to polarimetric NACO observations. With the exception of an instrumental Mueller matrix model, PIPPIN largely follows the polarimetric data reduction outlined in de Boer et al. (2020). For a more detailed characterisation of the instrumental polarisation of NACO, we refer to de Boer et al. (2014) and Millar-Blanchaer et al. (2020).

2.2.1 FLATs, bad-pixel masks, and DARKs

To correct for any variations of the detector’s gain, PIPPIN performs a FLAT-fielding of the SCIENCE images. In general, internal lamp FLATs were taken for each filter and detector (i.e. S13, S27, L27, S54, and L54) that were used during the night. The polarimetric mask, which prevents the ordinary and extra-ordinary beams from overlapping, is also inserted when measuring the FLAT fields. The FLATs are DARK-subtracted and subsequently normalised by being divided by the median counts. The bad-pixel masks are generated by assessing which pixels had a non-linear response in the FLAT fields. The linearity of the pixel response is determined by comparing the FLATs observed with the internal lamp switched on (FLATon) to FLATs made with the lamp turned off (FLAToff). The factor by which the pixel-counts are expected to increase is computed by dividing the median of FLATon by the median of FLAToff. Pixels were flagged when their response deviated by more than 2σ from the expected increase. Similar to the FLAT fields, the bad-pixel masks are computed for each filter and detector used throughout the night.

2.2.2 Pre-processing

The PIPPIN pipeline can be described in two parts: the preprocessing and the application of PDI. The pre-processing commences by reading parameters from a configuration file that allows users to customise the data reduction. The configuration file must be located in the same directory as the SCIENCE observations; otherwise, the pipeline creates a default file. Table B.1 outlines the parameter keywords in the configuration file along with the recognised values, descriptions, and default values. After reading the configuration parameters, PIPPIN groups observations by the utilised detector, window-size, observing ID (if requested), filter, exposure time, HWP usage, and whether the Wollaston prism or wire grids were used. Each observation is DARK-subtracted and FLAT-normalised by division. The pixels flagged in the bad-pixel mask are replaced by the median counts of the surrounding square of 5 × 5 pixels.

To retrieve the approximate positions of the ordinary and extra-ordinary beams, PIPPIN applies a minimum-filter with a specific kernel-shape to the images. The filter consists of two squares of 3 × 3 pixels that are offset by the approximate separation of the beams, which in turn depends on the pixel scale of the utilised detector. The maximum in the filtered image yields the approximate location of the ordinary and extra-ordinary beams. This method avoids any persisting bad pixels or image artefacts such as the polarimetric mask. Subsequently, the initial guesses are used to retrieve more accurate PSF locations via a user-specified fitting method. For each beam, PIPPIN can employ a single 2D Moffat function or subtract two Moffat functions from each other to reproduce the flat top of a saturated PSF. Alternatively, the pipeline can use a maximum-counts method for the asymmetric PSFs that are encountered in the case of deeply embedded stars.

The sky-subtraction is performed by subtracting two dithering positions or by subtracting the median per row of pixels. To avoid contamination from the target, a region around the fitted beam centres is excluded in the median sky-subtraction method. This region is defined with the sky_subtraction_min_offset parameter in the configuration file. Moreover, this parameter ensures that the two dithering positions are sufficiently offset to perform a sky-subtraction. In addition, horizontal gradients are removed by a linear fit that excludes the region around the beams. The linear fit is applied to the average of five rows of pixels and a 2D Gaussian filter with σ = 5 pixels is applied to smooth out the resulting background approximation. Some observations show a distinct horizontal pattern, which can be removed by fitting each row of pixels individually and without applying a Gaussian filter. Next, the ordinary and extra-ordinary beams are cut out of the images by a user-specified crop-size. The maximum counts of the beams are evaluated with an iterative sigma-clipping to determine which observations suffered from a poor AO correction. Figure 1 shows an example of the open AO-loop analysis for observations of HD 135344B in the Ks band. The left panel shows the maximum counts of the ordinary and extra-ordinary beams for each observation in red and blue, respectively. The horizontal dashed lines show the 3σ-bounds used in the sigma-clip. The right panels show examples of the ordinary beams of two observations. In this work, the images presented with a blue colour map show PIPPIN-reduced data products. The upper-right panel of Fig. 1 shows an effective AO correction and the lower-right panel shows an example of an open AO loop. In the bottom panel, we notice that the point source is blurred, likely as a result of a tilting wavefront during the integration. The resulting maximum count of the (extra-)ordinary beam is measured lower than the 3σ-bound and this observation was removed. In this example, observations 3, 42, and 45 were ignored during the PDI application.

thumbnail Fig. 1

Open AO-loop assessment of HD 135344B Ks-band observations. Left panel: maximum counts of the ordinary (red) and extraordinary (blue) beams. The horizontal dashed blue and red lines are the 3σ bounds for the respective beam, indicating which observations have adequate AO corrections. The upper-right panel shows an example of an effective AO correction for the ordinary beam, and the lower-right panel shows the blurred result of an open AO loop.

2.2.3 Polarimetric differential imaging

Polarised light can be described with the Stokes formalism and the Stokes vector: S=(IQUV),$S = \left( {\matrix{ I \cr Q \cr U \cr V \cr } } \right),$(1)

where I is the total intensity, Q and U are intensities of the linear polarisation components and V describes the circular polarisation. As NACO was not primarily designed forpolarime-try, the observations suffer from instrumental polarisation (IP) and crosstalk effects. Reflections within the instrument can introduce polarised signal whose magnitude depends on the instrument configuration, altitude of the target object, etc. Furthermore, crosstalk between the linear and circular polarisation components reduces the polarimetric efficiency (Witzel et al. 2011). Hence, PIPPIN employs a multi-stage correction for these effects. A first-order correction for different transmission efficiencies is to impose that the stellar flux in the ordinary (Iord) and extra-ordinary (Iext) beams are the same, as described in Appendix C of Avenhaus et al. (2014a). Since the PSF core is often saturated in NACO observations, PIPPIN draws multiple, user-specified annuli and computes the total fluxes within them. For each annulus i, the ratio between the fluxes, Xord/ext ,i=∑pixels Iord ,i∑pixels Iext ,i,${X_{{\rm{ord/ext }},i}} = {{\sum\limits_{{\rm{pixels }}} {{I_{{\rm{ord }},i}}} } \over {\sum\limits_{{\rm{pixels }}} {{I_{{\rm{ext }},i}}} }},$(2)

is used to scale the ordinary and extra-ordinary images as Iord /Xord /ext,i and IextXord/ext ,i, ${I_{{\rm{ord }}}}/\sqrt {{X_{{\rm{ord }}/{\rm{ext}},i}}} {\rm{ and }}{I_{{\rm{ext}}}}\sqrt {{X_{{\rm{ord/ext }},i}}} {\rm{, }}$, respectively. This method implicitly assumes that the total flux in annulus i is unpolarised, thereby ignoring any polarisation induced by the interstellar medium or any intrinsic polarisation originating from an unresolved inner disk, for example. We note that this correction could overcompensate for a true disk signal if the disk is not axisym-metric and if its scattered light comprises a considerable fraction of the stellar signal.

If the HWP has a rotation angle of θ = 0°, the ordinary beam (Iord) measures light polarised in the +Q direction and the extraordinary beam (Iext) measures the perpendicularly polarised light in the −Q direction, both in the HWP reference frame. The Iq and Q components are found by addition and subtraction of the equalised beam intensities: IQ=Iord +Iext |θ=0°,${I_Q} = {I_{{\rm{ord }}}} + {\left. {{I_{{\rm{ext }}}}} \right|_{\theta = {0^^\circ }}},$(3) Q=Iord −Iext |θ=0°. $Q = {I_{{\rm{ord }}}} - {\left. {{I_{{\rm{ext }}}}} \right|_{\theta = {0^^\circ }}}{\rm{. }}$(4)

Measurements of the U component are made by rotating the incoming beam by 45°, which means that the HWP is rotated by θ = 22.5°. The IU and U components are calculated with IU=Iord +Iext |θ=22.5°,${I_U} = {I_{{\rm{ord }}}} + {\left. {{I_{{\rm{ext }}}}} \right|_{\theta = {{22.5}^^\circ }}},$(5) U=Iord −Iext |θ=22.5°$U = {I_{{\rm{ord }}}} - {\left. {{I_{{\rm{ext }}}}} \right|_{\theta = {{22.5}^^\circ }}}$(6)

The top panels of Fig. 2 show the resulting median Stokes Q and U images for HD 135344B. The position angle (PA) is −35°, so that the sky is rotated anti-clockwise to the axes of the detector as is indicated by the compasses in the figure. In the Q image, the positive signal aligns with the Y-axis and the negative signal aligns with the X-axis. The U image displays a similar butterfly pattern, but rotated by 45° since it measures different components of the disk.

Instrumental polarisation introduced downstream of the HWP can be removed by recording the −Q and −U parameters at θ = 45° and 67.5°, respectively. The instrumental QIP and UIP components are unaffected by this rotation of the HWP and contribute in the same manner as before: Q+=Q+QIP=Iord−Iext|θ=0°,${Q^ + } = Q + {Q_{{\rm{IP}}}} = {I_{{\rm{ord}}}} - {\left. {{I_{{\rm{ext}}}}} \right|_{\theta = {0^^\circ }}}$(7) Q−=−Q+QIP=Iord−Iext|θ=45°,${Q^ - } = - Q + {Q_{{\rm{IP}}}} = {I_{{\rm{ord}}}} - {\left. {{I_{{\rm{ext}}}}} \right|_{\theta = {{45}^^\circ }}},$(8) U+=U+UIP=Iord −Iext|θ=22.5°,${U^ + } = U + {U_{{\rm{IP}}}} = {I_{{\rm{ord }}}} - {\left. {{I_{{\rm{ext}}}}} \right|_{\theta = {{22.5}^^\circ }}},$(9) U−=−U+UIP=Iord −Iext|θ=67.5°.${U^ - } = - U + {U_{{\rm{IP}}}} = {I_{{\rm{ord }}}} - {\left. {{I_{{\rm{ext}}}}} \right|_{\theta = {{67.5}^^\circ }}}.$(10)

Using the double-difference method (Hinkley et al. 2009; Bagnulo et al. 2009), we can subtract the IP components: Q=12(Q+−Q−),$Q = {1 \over 2}\left( {{Q^ + } - {Q^ - }} \right)$(11) U=12(U+−U−).$U = {1 \over 2}\left( {{U^ + } - {U^ - }} \right)$(12)

Similarly, the IP-corrected intensities are found with the double-sum: IQ=12(IQ++IQ−),${I_Q} = {1 \over 2}\left( {{I_{{Q^ + }}} + {I_{{Q^ - }}}} \right)$(13) IU=12(IU++IU−).${I_U} = {1 \over 2}\left( {{I_{{U^ + }}} + {I_{{U^ - }}}} \right)$(14)

The total intensity is calculated as I=12(IQ+IU).$I = {1 \over 2}\left( {{I_Q} + {I_U}} \right)$(15)

The second row of Fig. 2 shows the median Q and U images resulting from the double-difference method. Due to the IP removal, the butterfly patterns show more distinct features than the Q+ and U+ images and the recorded noise outside of the disk is reduced.

An additional correction is made for the IP introduced upstream of the HWP, following the method outlined in Canovas et al. (2011) and de Boer et al. (2020). The correction is performed for each HWP cycle to mitigate temporal differences in the IP as a result of changing angles of reflection. As before, it is assumed that the stellar light is unpolarised and polarised signal near the star is ascribed to instrumental polarisation (Quanz et al. 2011). The median Q/I signal is computed over the same annulus i from Eq. (2) to obtain a scalar cQ. To obtain cU, we calculate the median U/I signal over the same annulus. Per annulus, the IP-subtracted linear Stokes components are found by subtracting the product of these scalars and the respective IQ or IU image: QIPS=Q−IQ⋅cQ,${Q_{{\rm{IPS}}}} = Q - {I_Q} \cdot {c_Q},$(16) UIPS=U−IU⋅cU${U_{{\rm{IPS}}}} = U - {I_U} \cdot {c_U}$(17)

By using multiple user-specified annuli, the pipeline retrieves various IP-subtracted results. The third row of panels in Fig. 2 displays the median QIPS and UIPS images where the annulus was drawn between a radius of 3 and 6 pixels. As expected from the correction, the QIPS measurement shows a decreased signal near the star compared to the Q image.

In Fig. 2, the UIPS signal is lower than QIPS as a result of crosstalk between the linear and circular Stokes components (Witzel et al. 2011). If a disk is unmistakably detected and approximately axisymmetric, this reduced efficiency of the Stokes U component relative to Q can be estimated following the method outlined by Avenhaus et al. (2014a). In an annulus with disk signal, the number of pixels where | QIPS| > | UIPS | is expected to be equal to the number of pixels where | UIPS| > |QIPS|. We can multiply the UIPS image by a factor of 1/eU so that the above assumption holds. The crosstalk-corrected components are then QCTC=QIPS,${Q_{{\rm{CTC}}}} = {Q_{{\rm{IPS}}}},$(18) UCTC=1eU⋅UIPS,${U_{{\rm{CTC}}}} = {1 \over {{e_U}}} \cdot {U_{{\rm{IPS}}}}$(19)

where we assume an efficiency of 100% for Stokes Q. By modelling the NACO IP with standard star observations, Millar-Blanchaer et al. (2020) conclude that the Stokes Q has an efficiency of ~90%. Since such a correction is not performed with PIPPIN, any quantitative polarimetry measurements on the reduced data products could be off by ~10%. The efficiency-correction should not be performed in instances with ambiguous signal and thus PIPPIN only makes the crosstalk-correction if requested.

Incomplete HWP cycles, with only measurements of Q± (or U±), are removed. If only the Stokes Q+ and U+ (or only Q− and U−) were recorded, PIPPIN will still be able to produce the final data products, but the double-difference method cannot be applied. At this point, the pipeline computes the median Q, U, IQ, Iu, and I over all observations. The final polarisation images (PI, Qϕ, and Uϕ) are described below in terms of Q and U, but we note that these data products are also calculated with QIPS/UIPS and QCTC/UCTC, if possible. The total polarised intensity is calculated as PI=Q2+U2.$PI = \sqrt {{Q^2} + {U^2}} .$(20)

This method of squaring Q and U can lead to the increase in noise in regions where the Q or U signal originating from the disk is low. A cleaner image can be found with the azimuthal Stokes parameters outlined in Monnier et al. (2019) and de Boer et al. (2020), analogous to Schmid et al. (2006), but with a flipped sign: Qϕ=−Qcos(2ϕ)−Usin(2ϕ), ${Q_\phi } = - Q\cos (2\phi ) - U\sin (2\phi ){\rm{, }}$(21) Uϕ=+Qsin(2ϕ)−Ucos(2ϕ), ${U_\phi } = + Q\sin (2\phi ) - U\cos (2\phi ){\rm{, }}$(22)

where ϕ is the azimuthal angle and is calculated for each pixel with ϕ=arctan(y−ystar xstar −x)+ϕ0,$\phi = \arctan \left( {{{y - {y_{{\rm{star }}}}} \over {{x_{{\rm{star }}}} - x}}} \right) + {\phi _0},$(23)

where (xstar, ystar) are the pixel-coordinates of the central star. If the disk has a low inclination and the scattered light emerges from single scattering events, the polarisation is oriented azimuthally with respect to the star. Consequently, the Qϕ image shows a positive signal as it measures polarisation angles of ±90°. Simultaneously, the Uϕ image is expected to show a negligible signal as it measures polarisation angles of ± 45°. However, a non-zero Uϕ signal can occur if there is crosstalk between Q and U, if the light is scattered multiple times (Canovas et al. 2015b), if the disk has a high inclination, and if an inadequate correction retains stellar or instrumental polarisation (Hunziker et al. 2021). If requested, PIPPIN can minimise the Uϕ signal in the same annulus used for the crosstalk-correction by fitting for the azimuth angle offset ϕ0, similar to Avenhaus et al. (2014a). Otherwise, the offset angle ϕ0 is set to 0°.

The median total intensity I, polarised intensity PI, and azimuthal Stokes parameters Qϕ and Uϕ with different levels of IP correction are shown in Fig. 3 for HD 135344B. Once PIPPIN has computed the final data products, these images are de-rotated using scipy.ndimage.rotate. Therefore, contrary to Fig. 2, the panels of Fig. 3 have north pointing up and east to the left. It is apparent from the total and polarised intensity images that the PDI technique applies an extremely effective suppression of the stellar signal, thus revealing the circumstel-lar disk and its spiral arms. In this example, we observe the Uϕ signal diminish as the IP corrections are performed. Since HD 135344B is observed at a low inclination and axisymmet-ric to a first order, we employed crosstalk-correction and Uϕ minimisation to produce the final Stokes images. For these Ks-band observations, we find a reduced efficiency of eU = 0.65, in agreement with Garufi et al. (2013) who find an efficiency of 0.61 (Avenhaus et al. 2014a). Similarly, the more extensive IP model presented by Millar-Blanchaer et al. (2020) results in an efficiency of eU = 0.7 ± 0.02 for Elia 2-25. Furthermore, we find an offset angle of ϕ0 = 5.3°, while 3.7° is derived in the previous analysis of these data (Garufi et al. 2013; Avenhaus et al. 2014a).

thumbnail Fig. 2

Median Stokes Q and U images with different levels of IP corrections for HD 135344B Ks-band observations. From top to bottom: Q+ and U+ components after equalising the ordinary and extra-ordinary fluxes, Q and U resulting from the double-difference method, QIPS and UIPS after subtracting the median IP within an annulus, and the crosstalk-corrected QCTC and UCTC components where the reduced Stokes U efficiency is accounted for. The characteristic butterfly pattern is visible in each panel, and the compasses show the orientation of the detector and the sky.

2.2.4 Non-HWP and wire-grid observations

Prior to August 8, 2003, NACO was not equipped with a HWP. Rather than rotating the HWP to modulate the direction of polarisation, observers would alter the PA by rotating the instrument on its rotator ring. PIPPIN automatically diagnoses whether the de-rotator flange of the telescope support structure was used (Lenzen et al. 2003). For these data, the HWP angles θ = 0, 22.5, 45, and 67.5° in Eqs. (7)–(10) are replaced by the PAs of the instrument: θPA = 0, 45, 90, and 135°. The Q+,U+,IQ±, and IU±${I_{{Q^ \pm }}}{\rm{, and }}{I_{{U^ \pm }}}$ images are also de-rotated to align the circumstellar structures before combining them with Eqs. (11)–(14). For the rotator observations, the IP subtraction of Eqs. (16) and (17) is also performed.

In the early stages of its operation, NACO was equipped with wire grids to carry out polarimetric observations, rather than the Wollaston prism. In our cross-validation of the ESO archive, we found four potentially young sources that were observed in this manner: V1647 Ori, NX Pup, Mon R2 IRS 3, and R Mon. PIPPIN adopts the Pol_00, Pol_45, Pol_90, and Pol_135 wire grids as measurements of the Stokes Q+, U−, Q−, and U+ components, respectively. The linear Stokes components are propagated in the presence of the HWP. The only beam that is present in the images is fit with a single Moffat function. Since the wire-grid observations are not limited by the height of the polarimetric mask, their final data products have a much larger field of view than those obtained with the Wollaston prism.

2.2.5 Supplemental data products

Since the disk is illuminated by the star, the scattered light brightness decreases by the inverse of the squared distance to the host star. To better visualise structures at larger separations from the star, PIPPIN also produces images that are multiplied by the squared, de-projected radius. The disk PA, PAdisk, is used to calculate the offsets along the major axis, Δxdisk, and minor axis, Δydisk, with Δxdisk =Δ(RA)⋅sinPAdisk +Δ(Dec)⋅cosPAdisk ,$\Delta {x_{{\rm{disk }}}} = \Delta ({\rm{RA}}) \cdot \sin P{A_{{\rm{disk }}}} + \Delta ({\rm{Dec}}) \cdot \cos P{A_{{\rm{disk }}}},$(24) Δydisk =Δ(Dec)⋅sinPAdisk −Δ(RA)⋅cosPAdisk ,$\Delta {y_{{\rm{disk }}}} = \Delta ({\rm{Dec}}) \cdot \sin P{A_{{\rm{disk }}}} - \Delta ({\rm{RA}}) \cdot \cos P{A_{{\rm{disk }}}},$(25)

where Δ(RA) and Δ(Dec) are the right ascension and declination offsets with respect to the star. Subsequently, the de-projected radius r is computed with r=Δxdisk 2+(Δydisk cosidisk )2, $r = \sqrt {\Delta x_{{\rm{disk }}}^2 + {{\left( {{{\Delta {y_{{\rm{disk }}}}} \over {\cos {i_{{\rm{disk }}}}}}} \right)}^2}} {\rm{, }}$(26)

where idisk is the disk inclination. As is shown in Table B.1, the disk PA, PAdisk and inclination idisk are specified in the configuration file for PIPPIN and are set to 0° by default. In cases where the disk inclination and PAs are unknown, the default values ensure that the images are scaled by the projected separation from the host star.

The height of the final data products is limited to ~3.0 arcsec (using the S27 detector) due to the polarimetric mask. Observations where the PA was rotated, rather than the HWP, cover a larger area of the sky. Since the sky rotates while the polarimetric mask remains stationary, the effective field of view is increased.

Figure C.1 depicts this increased sky coverage. An eight-pointed star emerges where at least one Q and one U component are covered and thus the polarised intensity can be computed within this shape. An inner octagon appears where every positive and negative Stokes component is observed. We note that the signal-to-noise decreases for areas outside of this octagon, due to the reduced number of observations. PIPPIN outputs the extended eight-pointed star images in addition to the data products resulting from the double-difference method, which are restricted to the inner octagon that has a complete coverage.

thumbnail Fig. 3

Final PIPPIN data products with different levels of IP correction. From left to right: median total intensity (I), polarised intensity (PI), and the azimuthal Stokes components Qϕ and Uϕ of HD 135344B observed in the Ks band. From top to bottom: equalised ordinary and extra-ordinary beams, IP-subtracted, crosstalk-corrected, and Uϕ-minimised results. The total intensity is shown with a logarithmic scale from 20 to 104 counts, whereas the other panels use a linear scale from −5 to +5 counts and a logarithmic scale up to ±90. The negative signal is depicted in orange, and in each image north points up and east to the left.

3 Inspection of NACO data

3.1 Identification of detections

The PIPPIN pipeline described above was used to reduce all observations listed in Table A.1. The table lists multi-epoch, multi-wavelength observations as well as different exposure times and whether the wire grids were used or the Wollaston prism, with the HWP or PA. For each set of observations, we indicate the (non-)detection of circumstellar material in the final data products. The detection significance of polarised signal was assessed via a template-matching method, akin to cross-correlation, applied to the Stokes Q and U images. In the case of a detection, we expect that the signal is present in multiple, adjacent pixels and forms a specific butterfly pattern. Synthetic Qsynth and Usynth templates of the expected butterfly patterns were constructed with Qsynth =−cos(2(ϕ−PA)), ${Q_{{\rm{synth }}}} = - \cos (2(\phi - PA)){\rm{, }}$(27) Usynth =−sin(2(ϕ−PA)), ${U_{{\rm{synth }}}} = - \sin (2(\phi - PA)){\rm{, }}$(28)

where ϕ is the azimuthal angle calculated with Eq. (23) and PA is the position angle of the observation, which is subtracted since PIPPIN de-rotates the final data products, including the QIPS and UIPS images. Subsequently, the Qsynth and Usynth templates were divided into multiple annuli with increasing radius and width of 2 pixels, roughly corresponding to one resolution element in the H (42 mas) and Ks band (56 mas) at a pixel scale of 27 mas pixel−1. Figure 4 shows an example of the Qsynth and Usynth templates and a single annulus for a PA of −35°, corresponding to the observations of HD 135344B. The values in the templates range from −1 to +1 and pixels outside of the annulus are set to 0, thus ensuring that they do not contribute when calculating the cross-correlation coefficient. In annulus i, a cross-correlation coefficient is calculated for the Q and U signals: CCQ,i=∑pixels QIPS,i⋅Qsynth ,i,$C{C_{Q,i}} = \sum\limits_{{\rm{pixels }}} {{Q_{{\rm{IPS}},i}}} \cdot {Q_{{\rm{synth }},i}},$(29) CCU,i=∑pixels UIPS,i⋅Usynth ,i,$C{C_{U,i}} = \sum\limits_{{\rm{pixels }}} {{U_{{\rm{IPS}},i}}} \cdot {U_{{\rm{synth }},i}},$(30)

where the sum is performed over every pixel within annulus i. In this manner, a positive pixel increases the coefficient if the respective quadrant expects a positive signal. A negative signal in the negative quadrants of the template also contributes, whereas a discrepant signal reduces the cross-correlation coefficient. A cross-correlation function (CCF) is constructed by computing a coefficient for each annulus. For the narrowband NB_1.64 observations of HD 135344B, the rightmost panel of Fig. 5 displays the CCFs for the QIPS and UIPS components in blue and red, respectively. The CCF was converted into a signal-to-noise (S/N) function by subtracting the mean coefficient between 35 and 50 pixels, and subsequently dividing by the standard deviation of coefficients within that same range, indicated by the grey shaded region in the right panel. The annulus-wise CCFs peak at a radius of 8 pixels with signal-to-noises of S/N ~ 13 and ~15, respectively for QIPS and UIPS. These maxima surpass our 5σ detection threshold, thereby identifying this observation as a detection. Although the template-matching method generally worked well, it failed to flag two observations of HR 4796 as detections, despite the polarised signal evident from a visual inspection. These non-detections can be ascribed to the high inclination and narrow features of HR 4796, while the outlined template-matching analysis works optimally for face-on disks.

In this reduction of the NACO data, many of the non-detections are likely the result of small or faint disks, or the absence of polarised light. Notably, IM Lup, GQ Lup, and EX Lup do not show polarised light in the NACO data, despite their prominent detections with SPHERE/IRDIS (Avenhaus et al. 2018; van Holstein et al. 2021; Rigliaco et al. 2020). The data of IM Lup and GQ Lup were not previously published, whereas Kóspál et al. (2011) also report a non-detection of polarised light in the EX Lup observations.

thumbnail Fig. 4

Templates for observations of HD 135344B with a PA of −35°. Top panels: complete Qsynth and Usynth templates. Values range from −1 to +1, and pixels outside of the image are set to 0. Bottom panels: example annuli used in computing the cross-correlation coefficients.

3.2 Analysis of detected polarised light

As demonstrated in Table A.1, in 22 out of the 57 observed systems, we find at least one set of observations with polarised signal originating from circumstellar material. Figure 6 presents a gallery of these detections and highlights a diverse collection of morphologies. As mentioned in Sect. 2, HD 135344B shows distinct spiral arms in its circumstellar disk while HD 142527 has spiral features in its eastern and western lobes. Furthermore, we detect rings in a large number of disks, including HD 169142, HD 163296, HD 97048, HR 4796, TW Hya, HD 142527, and Sz 91. HR 4796 is the only debris disk in our sample and its Qϕ image shows a narrow ring. The disks around HD 163296, HD 97048, and HD 100546 are offset from the central star, suggesting that the scattering surface is above the disk midplane as confirmed by Monnier et al. (2017), Ginski et al. (2016), and Sissa et al. (2018). The highly extended disk of HD 142527 shows a large inner cavity that is possibly cleared out by an inner companion (Biller et al. 2012; Close et al. 2014; Lacour et al. 2016; Claudi et al. 2019), undetected in polarised light. Moreover, we find narrow shadow lanes imprinted on the disks of HD 142527 and SU Aur, similar to Avenhaus et al. (2017) and Ginski et al. (2021). In these cases, misaligned inner disks prevent the stellar light from reaching certain areas of the outer disk. CR Cha, MP Mus, AK Sco, and Elia 2-25 show negligible structure due to their small sizes, but a significant butterfly pattern was detected, leading to their inclusion in Fig. 6. As a possible consequence of their small sizes, the polarimetric NACO data of CR Cha and MP Mus were previously unpublished.

Figure 6 displays a number of objects with extended features that appear inconsistent with circumstellar disks. SU Aur shows tail-like features, where Ginski et al. (2021) discover an inflow of material onto the disk by combining polarimetric SPHERE observations with ALMA line observations. The features of R CrA resemble the non-polarimetric SPHERE observations presented by Mesa et al. (2019) with scattered light towards the north-east, south-east, and south-west of the primary star. Although a brightness asymmetry is observed towards the east in the NACO Qϕ image, the companion inferred by Mesa et al. (2019) is not detected. To our knowledge, the detection of polarised light in the NACO observations of R CrA went unpublished before this work. Recently, Dong et al. (2022) reported that the binary star Z CMa experienced a close encounter with a nearby star (masked in the NACO observation), thereby ejecting the streamer structure that we also observe in the Qϕ image. As YLW 16A, Elia 2-29, Elia 2-21, and Parsamian 21 were observed with the rotator flange, Fig. 6 presents the extended data products described in Sect. 2. The polarised light of Elia 2-29 reveals three arcs to the east, north and north-west of the central star. The northern and north-western arcs have curved shapes that are reminiscent of spiral-like features (Huélamo et al. 2007). YLW 16A shows asymmetric polarised signal to the west and northwest of the binary components (Plavchan et al. 2013) that are still visible as intensity maxima. Parsamian 21 and Elia 2-21 appear to show bipolar outflows along the NW-SE and NE-SW axes, respectively. Both nebulae are distinctly asymmetric with the northern and north-eastern segments showing the largest emission surfaces, respectively for Parsamian 21 and Elia 2-21. At large separations, along the PAs of the Q± and U± measurements, one of the linear Stokes components dominates over the other. Hence, the majority of the signal in PI can be represented by the absolute values |Q±| or |U±|, which is shown with a grey colour map in Fig. C.1. The northern lobe of Parsamian 21 and the northern and eastern arms of Elia 2-21 are traced about ~2 arcsec further. To our knowledge, this work is the first publication of the polarimetric NACO observations of Elia 2-21 and YLW 16A. The polarised light of the reflection nebula NGC 2261, illuminated by R Mon, demonstrates distinct emission from the extended north-eastern and south-western components. The Stokes U component of the infrared source Mon R2 IRS 3, part of the Monoceros R2 molecular cloud complex, was not observed. Hence, Fig. 6 presents the median IQ, which also includes unpolarised (scattered) light. Despite this, filamentary structure is unambiguously detected. As discussed in Sect. 2.2.4, the two detections utilising polarimetric wire grids, R Mon and Mon R2 IRS 3, have much larger image sizes than the regular data products (i.e. compared to the upper rows of panels in Fig. 6).

thumbnail Fig. 5

Detection analysis of HD 135344B observed in the narrowband NB_1.64. Left panels: QIPS and UIPS images divided by the white lines into the four quadrants of the expected butterfly pattern. Right panel: annulus-wise CCFs, with the S/N shown against the annulus radius in pixels. The results for the QIPS and UIPS images are plotted in blue and red, respectively. The shaded region specifies the coefficients used in normalising and converting the CCF into a S/N function. The 5σ detection limit is indicated with a horizontal line.

3.3 Disk classification and brightness

Circumstellar polarised light is detected in nine out of the 14 Her-big stars in our sample. Grouping the Orion variables together with T Tauri stars, we detect polarised signal in eight out of 28 systems. For the YSOs, five of the 14 sources are flagged as detections with the template-matching analysis. The only high-proper motion star in our sample, HR 4796, also constitutes the only detection of a debris disk. However, our selection of young stars based on the SIMBAD object type could have missed some of the older Class III disks. Since NACO frequently observed disks known to be extended and thus potentially observable in scattered light, the gathered sample is certainly not unbiased. For that reason, a statistical analysis of the disk occurrence per object type is somewhat arbitrary.

Here we examine the disk brightness of the sample of cir-cumstellar disks detected by NACO. Since the disk inclination, disk extent, stellar brightness, and the distance to the source affect the total disk brightness, we made use of the polarised-to-stellar light contrast δpol (Garufi et al. 2017, 2022b; Benisty et al. 2023). The polarised flux per unit area, Fpol, was multiplied by the squared separation 4πr2 to account for the reduced stellar illumination. Subsequently, we normalised it by the stellar flux F* and averaged radially along the disk’s major axis. Thus, the polarised-to-stellar light contrast was computed as δpol=1rout −rin ⋅∫rin rout Fpol(r)F*⋅4πr2 dr,${\delta _{{\rm{pol}}}} = {1 \over {{r_{{\rm{out }}}} - {r_{{\rm{in }}}}}} \cdot \int_{{r_{{\rm{in }}}}}^{{r_{{\rm{out }}}}} {{{{F_{{\rm{pol}}}}(r)} \over {{F_*}}}} \cdot 4\pi {r^2}{\rm{d}}r,$(31)

where rin and rout are the inner- and outermost radii, respectively. This measurement expresses the fraction of stellar photons that became polarised as a result of scattering by the resolved disk. The contrast δpol is set by the geometry and composition of the circumstellar disk and is sometimes referred to as the geometrical albedo. Following the method outlined in Appendix B of Garufi et al. (2017), we performed a 2-pixel-wide cut (one resolution element) of the Qϕ image along the major axis of the disk. The photons measured along the major axis are scattered with angles close to 90°. This cut reduces the impact of the disk inclination on its brightness due to any asymmetry in the scattering phase function. The PA of the major axis varies for the sources observed by NACO, but was set to 0° when this axis could not be estimated. These ambiguous sources (e.g. CR Cha) were roughly azimuthally symmetric and therefore did not sig-niflcantly affect the derived contrast. The inner- and outermost radii (rin, rout) were determined by eye for each system detected in scattered light. The disk inclination and scale height were not taken into account when re-scaling the polarised flux by the projected separation. Similar to Garufi et al. (2017), we calculated the primary error of δpol by deriving the standard deviation in a resolution element of the Qϕ image. Subsequently, this noise estimate for each pixel was propagated through Eq. (31) to find δpol.

Figure 7 displays the polarised-to-stellar light contrast δpol against the apparent J-band magnitude mJ, measured as part of the 2 Micron All Sky Survey (2MASS; Cutri et al. 2003). The source names are shown along the top axes. Blue, orange, and red markers indicate whether the observation was performed in the H, Ks, or Ľ band. Diamonds (T Tau), circles (Her-big), and squares (debris) mark the object types. We note that HD 135344B is shown as a Herbig star (circle), in line with Garufi et al. (2014), but contrary to the SΓMBAD object type in Table A.1. Similarly, Parsamian 21 is depicted as an embedded YSO despite its Orion variable type reported in Table A.1. The δpol values of saturated PSFs are presented as upper limits in Fig. 7 (triangles; 99.75th percentile) because the stellar flux F* is underestimated. In some instances, the source was also observed with narrowband filters where the stellar PSF was not saturated due to the smaller filter width. Hence, we could estimate the broadband flux F*BB$F_*^{{\rm{BB}}}$, using the narrowband flux F*NB$F_*^{{\rm{NB}}}$, if the two filters had overlapping wavelength ranges. We calculated the stellar flux as F*BB=F*NB⋅texpBBtexpNB⋅∫TBB(λ)dλ∫TNB(λ)dλ,$F_*^{{\rm{BB}}} = F_*^{{\rm{NB}}} \cdot {{{t_{{{\exp }^{{\rm{BB}}}}}}} \over {t_{{\rm{exp}}}^{{\rm{NB}}}}} \cdot {{\int {{T^{{\rm{BB}}}}} (\lambda ){\rm{d}}\lambda } \over {\int {{T^{{\rm{NB}}}}} (\lambda ){\rm{d}}\lambda }},$$F_*^{{\rm{BB}}} = F_*^{{\rm{NB}}} \cdot {{{t_{{{\exp }^{{\rm{BB}}}}}}} \over {t_{{\rm{exp}}}^{{\rm{NB}}}}} \cdot {{\int {{T^{{\rm{BB}}}}} (\lambda ){\rm{d}}\lambda } \over {\int {{T^{{\rm{NB}}}}} (\lambda ){\rm{d}}\lambda }},$(32)

where texpBB and texpNB$t_{\exp }^{{\rm{BB}}}{\rm{ and }}t_{\exp }^{{\rm{NB}}}$ are the exposure times of the broadband and narrowband observations, respectively. We integrated over the corresponding transmission curves T(λ) to estimate how many photons should be detected for each photon in the narrowband filter. For the H band, we used the NB_1.64 and NB_1.75 narrowband filters. For the Ks band, NB_2.17, IB_2.18, NB_2.12, NB_2.15, and IB_2.21 were employed to compute the broadband flux and the NB_3.74 filter was used for saturated L′-band observations.

Since NACO was equipped with a NIR WFS, it could observe sources down to K ≈ 14 mag (Rousset et al. 2003). For comparison, SPHERE’s optical WFS has a magnitude limit of R ≈ 14 mag (Beuzit et al. 2019), GPI has a limit of I ≈ 10 mag (Macintosh et al. 2014), and SCExAO/CHARIS on the Subaru telescope is limited by R ≈ 13 mag5. For that reason, NACO could achieve a unique insight into embedded protostars, despite their faint optical magnitudes. The grey shaded region in Fig. 7 roughly indicates the sources that are inaccessible by the optical WFS installed on SPHERE. Since the estimated J-band magnitude limit of SPHERE depends on the spectral type of the observed source, the limit of mJ ~ 10 mag should be viewed as a crude assessment. From Fig. 7 we find that the four embedded protostars Parsamian 21, Elia 2-21, Elia 2-29, and YLW 16A, in addition to the low-mass star Sz 91, are likely not observable with modern PDI instruments.

The polarised-to-stellar light contrasts δpol are listed in Table A.1. We note that the δpol values are possibly underestimated by ~10% (see Sect. 2.2.3) due to the absence of a correction for the reduced Q efficiency resulting from crosstalk. For the sources with available mass estimates (also included in Table A.1), we fail to detect a trend between δpol and the stellar mass M*. Since the disk’s dust mass is related to the stellar mass (e.g. Pascucci et al. 2016), the absence of a distinct trend reveals that the disk brightness in polarised light is not strongly correlated with the abundance of dust in the system. Instead, the scattered light brightness is affected by the amount of light that is intercepted. The geometry of the system primarily influences the polarised-to-stellar light contrast δpol, with the dust composition acting as a secondary effect. Similarly, Garufi et al. (2022b) find no correlation between the disk brightness in scattered light and dust mass estimated from the 1.3 mm flux. We also find no apparent distinction in disk brightnesses between the T Tauri stars and Herbig stars. Comparing the obtained δpol results with those presented in Fig. 3 of Garufi et al. (2022b) and Table A.1 of Garufi et al. (2017), we find good agreement for HD 163296, HD 100546, HD 142527, HD 97048, HD 135344B, HD 169142, AK Sco, TW Hya, and CR Cha. For the extended systems SU Aur, R CrA, and Z CMa, we assessed the polarised-to-stellar light contrast ratio of a potential disk component at close separations, meaning that rout was limited to 25, 25, and 53 pixels, respectively. We find δpol ~ 1.5 × 10−3 for SU Aur, ≲ 3 × 10−2 for R CrA, and ≲ 6 × 10−3 for Z CMa. The brightest disk is found around HR 4796 with δpol ~ 0.3–0.4. This finding is somewhat surprising, given that it is a flat debris disk and therefore should not intercept much stellar light. However, the high brightness is also reported in previous works where it is argued that the scattering phase functions are consistent with large (~20 µm) aggregate dust particles composed of small monomers (Milli et al. 2017, 2019). For Sz 91, the lowest-mass star (M = 0.58 ± 0.07 M⊙; Maucó et al. 2020) where polarised light is detected, we determine upper limits of δpol ≲ 8 × 10−2 and ≲4 × 10−2 in the Ks and H band, respectively. The estimated contrasts of multi-wavelength observations do not show any clear discrepancies between H- and Ks-band observations, owing to relatively large uncertainties. Hence, it is difficult to draw any conclusions about the dust composition by evaluating the disk colour.

thumbnail Fig. 6

Gallery of young systems detected with NACO and reduced with PIPPIN. Each panel shows the polarised light on a logarithmic scale ranging between different values to highlight sub-structures. The highest degree of IP correction is used where possible. Scale bars in the lower-left corners of each panel indicate 100 AU at each object’s distance. HD 169142, R CrA, and Parsamian 21 are shown in the H band, MP Mus is shown in the IB_2.06 filter, and the other panels use Ks-band observations. Mon R2 IRS 3 shows the median Iq image because the Stokes U component was not observed. The images of YLW 16A and Elia 2-21 present the first polarised light detections in the NACO observations.
thumbnail Fig. 7

Polarised-to-stellar light contrast, δpol, plotted against the apparent J-band magnitude. The right panel shows a zoomed-in view of the bright m¡. The object names are listed along the top axes. The marker colours and symbols specify the observing filter and object type, respectively. Upper limits are shown when the stellar PSF was determined to be saturated. The error bars show the 3σ uncertainties. The grey shaded region shows the approximate magnitudes (mJ ≳ 10) inaccessible by the SPHERE AO system.
thumbnail Fig. 8

Comparison between PIPPIN-reduced NACO Qϕ observations (left panels) and the more recent SPHERE data (right panels). From top to bottom: SU Aur, HD 142527, and TW Hya observed in the H band with both instruments. The SPHERE observations were previously published by Ginski et al. (2021), Hunziker et al. (2021), and van Boekel et al. (2017) for SU Aur, HD 142527, and TW Hya, respectively.

4 Discussion

The morphologies shown in Fig. 6 are almost identical to the polarised intensity images presented in previous publications of these data (see Table A.1 for references). The data reduction performed by PIPPIN therefore appears to reproduce the results obtained by the custom pipelines in other works.

To study the performance between different instruments, Fig. 8 presents a comparison between the NACO and modern SPHERE observations of SU Aur (programme ID: 1104.C-0415(E), PI: Ginski), HD 142527 (programme ID: 099.C-0601(A), PI: Avenhaus), and TW Hya (programme ID: 095.C-0273(D), PI: Beuzit). In this comparison, we find the results of the different instrument characteristics. For instance, the NACO observations of TW Hya were made under better seeing conditions (~0.5 arcsec) than those made by SPHERE (~0.7 arcsec), but we find that the NACO polarised signal displays residual speckles over the circumstellar disk. The SPHERE Qϕ image does not show similar artefacts, likely due to the superior AO instrument. As part of the NACO instrument, NAOS had fewer actuators (185 active actuators for NAOS against 1377 for SAXO; Blanco et al. 2022) shaping the deformable mirror and its optical WFS operated at a lower frequency (1200 Hz versus 444 Hz; Fusco et al. 2006; Rousset et al. 2003), thus resulting in typical H-band Strehl ratios of ~ 10–35% as opposed to ~60– 80% for SPHERE observations (Fusco et al. 2014; van Boekel et al. 2017). Furthermore, the SPHERE NIR camera (IRDIS) has a pixel scale of ~12 mas pixel−1 (Maire et al. 2018) while the most-used S27 detector on CONICA had ~27 mas pixel−1. The NACO instrument was not equipped with a coronagraph in its polarimetric mode and thus short exposure times were utilised to avoid saturation by the central star. Each of the NACO observations presented in Fig. 8 employed considerably shorter single-frame integration times than the respective SPHERE observations (SU Aur: 0.35 versus 32 s, HD 142527: 0.3454 versus 16 s, TW Hya: 5 versus 16 s), thereby inevitably reducing the signal-to-noise. Ginski et al. (2021) trace the extended western structure of SU Aur out to ~6 arcsec, whereas the NACO data only confidently show signal out to ~4 arcsec. Moreover, the filamentary structure observed in the tails and disk of SU Aur (Ginski et al. 2021) are not resolved in the NACO data due to the reduced signal-to-noise. Lastly, the polarimetric mask of NACO limits the vertical extent of the final data products to ~3 arcsec. Hence, the north-western spiral structure of HD 142527 is eventually obscured in the NACO data.

5 Conclusions

We have presented a complete catalogue of polarimetric NACO images for YSOs, reduced in a homogeneous manner with a new pipeline that employs the PDI technique. Via a cross-examination with the object types reported on SIMBAD, 57 targets were identified as potentially young systems with polari-metric NACO observations. As a result of multi-epoch and multi-wavelength observations, a total of 243 datasets were reduced with the publicly available PIPPIN pipeline6. PIPPIN can handle observations made with NACO’s HWP as well as its de-rotator. In addition to the Wollaston prism, observations measured with wire grids can be reduced too. Various levels of corrections for instrumental polarisation are performed, depending on the type of observation.

The reduced data products were analysed with a template-matching method to evaluate the detection significance. This technique exploits the butterfly pattern in the Stokes Q and U images that should be present in the case of significant polarised light. We find that 22 out of the 57 observed systems revealed polarised light in at least one observation. These detections revealed a wide diversity of sub-structures, including rings, gaps, spirals, shadows, and in- or outflowing matter. Since NACO was equipped with a NIR WFS, unique polarimetric observations of embedded YSOs were made. To our knowledge, this is the first work to publish the reduced data products of the Class I proto-stars Elia 2-21 and YLW 16A. PIPPIN also revealed detections of polarised light in the L′ band for HD 100546 (Avenhaus et al. 2014b) and Elia 2-21. This long-wavelength filter (3.8 µm) is not available on current, state-of-the-art PDI instruments such as SPHERE/IRDIS, SCExAO/CHARIS, or GPI.

Alongside this article, we publish an archive of the reduced data products generated with PIPPIN on Zenodo7. As these observations were made in the past two decades, their combination with modern scattered light observations can be used to identify temporal changes in the sub-structures of planet-forming disks. In turn, such morphological changes can be used to infer the presence of a perturbing companion (Ren et al. 2020). Recent studies of the NACO data of HD 97048 and SU Aur (Ginski et al. 2016, 2021) have led to the discovery of previously unidentified features. With this work, we hope to have outlined the utility of NACO observations reduced with the PDI technique.

Acknowledgements

S.d.R. acknowledges funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program under grant agreement No. 694513. C.C. acknowledges support by ANID BASAL project FB210003 and ANID, – Millennium Science Initiative Program – NCN19_171. N.H. is funded by Spanish MCIN/AEI/10.13039/501100011033 grant PID2019-107061GB-C61. S.P. acknowledges support from FONDECYT grant 1231663 and funding from ANID – Millennium Science Initiative Program – Center Code NCN2021_080. M.R.S. was supported from FONDECYT (grant number 1221059) and ANID, – Millennium Science Initiative Program – NCN19_171. This work was also supported by the NKFIH excellence grant TKP2021-NKTA-64. The authors thank all PIs of proposals whose observations are used in this work. This includes Ageorges, Apai, Christiaens, de Kok, Feldt, Hardy, Jeffers, Milli, Mouil-let, Murakawa, Quanz, Schuetz, and Yudin. This research has made use of the SIMBAD database, operated at CDS, Strasbourg, France. SPHERE is an instrument designed and built by a consortium consisting of IPAG (Grenoble, France), MPIA (Heidelberg, Germany), LAM (Marseille, France), LESIA (Paris, France), Laboratoire Lagrange (Nice, France), INAF - Osservatorio di Padova (Italy), Observatoire de Genève (Switzerland), ETH Zurich (Switzerland), NOVA (Netherlands), ONERA (France) and ASTRON (Netherlands) in collaboration with ESO. SPHERE was funded by ESO, with additional contributions from CNRS (France), MPIA (Germany), INAF (Italy), FINES (Switzerland) and NOVA (Netherlands). SPHERE also received funding from the European Commission Sixth and Seventh Framework Programmes as part of the Optical Infrared Coordination Network for Astronomy (OPTICON) under grant number RII3-Ct-2004-001566 for FP6 (2004–2008), grant number 226604 for FP7 (2009–2012) and grant number 312430 for FP7 (2013–2016). This research was performed with the Python programming language. In particular, the SciPy (Virtanen et al. 2020), NumPy (Harris et al. 2020), Matplotlib (Hunter 2007), and astropy (Astropy Collaboration 2013, 2018) packages were utilised.

Appendix A Reduced systems and observations

Table A.1

Potential young systems observed by NACO in its polarimetric mode, sorted by right ascension.

Appendix B PIPPIN configuration keywords

Table B.1

Keywords and values recognised by PIPPIN in the configuration file.

Appendix C Extended data products of Parsamian 21 and Elia 2-21

thumbnail Fig. C.1

Polarised light for the embedded YSOs Parsamian 21 (left panels) and Elia 2-21 (right panels). The top panels show the polarised intensity, PI, with a blue colour map, while the grey colours display the absolute values of the linear Stokes components |Q±| and |U±|. In the bottom panels, these values are scaled by the squared separation from the centre. The dashed lines delineate the sections of the sky observed by one of the components. These sections overlap in the centre, resulting in an eight-pointed star, where the polarised intensity image can be computed as Q and U are both covered. The south-eastern region of the U− observation of Parsamian 21 is contaminated with signal from another dithering position, introduced during the sky-subtraction.

References

  1. Alcalá, J. M., Manara, C. F., Natta, A., et al. 2017, A&A, 600, A20 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  2. Alecian, E., Wade, G. A., Catala, C., et al. 2013, MNRAS, 429, 1001 [NASA ADS] [CrossRef] [Google Scholar]
  3. Alencar, S. H. P., Melo, C. H. F., Dullemond, C. P., et al. 2003, A&A, 409, 1037 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  4. ALMA Partnership (Brogan, C. L., et al.) 2015, ApJ, 808, L3 [Google Scholar]
  5. Apai, D., Pascucci, I., Brandner, W., et al. 2004, A&A, 415, 671 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  6. Asensio-Torres, R., Janson, M., Hashimoto, J., et al. 2016, A&A, 593, A73 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  7. Asensio-Torres, R., Henning, T., Cantalloube, F., et al. 2021, A&A, 652, A101 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  8. Astropy Collaboration (Robitaille, T. P., et al.) 2013, A&A, 558, A33 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  9. Astropy Collaboration (Price-Whelan, A. M., et al.) 2018, AJ, 156, 123 [Google Scholar]
  10. Avenhaus, H., Quanz, S. P., Schmid, H. M., et al. 2014a, ApJ, 781, 87 [Google Scholar]
  11. Avenhaus, H., Quanz, S. P., Meyer, M. R., et al. 2014b, ApJ, 790, 56 [Google Scholar]
  12. Avenhaus, H., Quanz, S. P., Schmid, H. M., et al. 2017, AJ, 154, 33 [Google Scholar]
  13. Avenhaus, H., Quanz, S. P., Garufi, A., et al. 2018, ApJ, 863, 44 [NASA ADS] [CrossRef] [Google Scholar]
  14. Bagnulo, S., Landolfi, M., Landstreet, J. D., et al. 2009, PASP, 121, 993 [Google Scholar]
  15. Benisty, M., Stolker, T., Pohl, A., et al. 2017, A&A, 597, A42 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  16. Benisty, M., Juhász, A., Facchini, S., et al. 2018, A&A, 619, A171 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  17. Benisty, M., Dominik, C., Follette, K., et al. 2023, Protostars and Planets VII, eds. S. Inutsuka, Y. Aikawa, T. Muto, K. Tomida, & M. Tamura, ASP Conf. Ser., 534, 605 [NASA ADS] [Google Scholar]
  18. Beuzit, J. L., Vigan, A., Mouillet, D., et al. 2019, A&A, 631, A155 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  19. Biller, B., Lacour, S., Juhász, A., et al. 2012, ApJ, 753, L38 [Google Scholar]
  20. Blanco, L., Behara, N., Valenzuela, J., & Kasper, M. 2022, SPIE Conf. Ser., 12185, 121856B [NASA ADS] [Google Scholar]
  21. Blondel, P. F. C., & Djie, H. R. E. T. A. 2006, A&A, 456, 1045 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  22. Bohn, A. J., Benisty, M., Perraut, K., et al. 2022, A&A, 658, A183 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  23. Boss, A. P. 1997, Science, 276, 1836 [Google Scholar]
  24. Bouvier, J., & Appenzeller, I. 1992, A&AS, 92, 481 [Google Scholar]
  25. Canovas, H., Rodenhuis, M., Jeffers, S. V., Min, M., & Keller, C. U. 2011, A&A, 531, A102 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  26. Canovas, H., Ménard, F., Hales, A., et al. 2013, A&A, 556, A123 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  27. Canovas, H., Perez, S., Dougados, C., et al. 2015a, A&A, 578, A1 [Google Scholar]
  28. Canovas, H., Ménard, F., de Boer, J., et al. 2015b, A&A, 582, L7 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  29. Chiang, E. I., & Goldreich, P. 1997, ApJ, 490, 368 [Google Scholar]
  30. Christiaens, V., Ubeira-Gabellini, M. G., Cánovas, H., et al. 2021, MNRAS, 502, 6117 [NASA ADS] [CrossRef] [Google Scholar]
  31. Claudi, R., Maire, A. L., Mesa, D., et al. 2019, A&A, 622, A96 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  32. Close, L. M., Follette, K. B., Males, J. R., et al. 2014, ApJ, 781, L30 [NASA ADS] [CrossRef] [Google Scholar]
  33. Coulson, I. M., & Walther, D. M. 1995, MNRAS, 274, 977 [NASA ADS] [Google Scholar]
  34. Covino, E., Terranegra, L., Vittone, A. A., & Russo, G. 1984, AJ, 89, 1868 [CrossRef] [Google Scholar]
  35. Cridland, A. J., Rosotti, G. P., Tabone, B., et al. 2022, A&A, 662, A90 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  36. Currie, T., Lawson, K., Schneider, G., et al. 2022, Nat. Astron., 6, 751 [NASA ADS] [CrossRef] [Google Scholar]
  37. Cutri, R. M., Skrutskie, M. F., van Dyk, S., et al. 2003, VizieR Online Data Catalog: II/246 [Google Scholar]
  38. de Boer, J., Girard, J. H., Mawet, D., et al. 2014, SPIE Conf. Ser., 9147, 914787 [NASA ADS] [Google Scholar]
  39. de Boer, J., Salter, G., Benisty, M., et al. 2016, A&A, 595, A114 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  40. de Boer, J., Langlois, M., van Holstein, R. G., et al. 2020, A&A, 633, A63 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  41. Dong, R., Liu, H. B., Cuello, N., et al. 2022, Nat. Astron., 6, 331 [NASA ADS] [CrossRef] [Google Scholar]
  42. Engler, N., Schmid, H. M., Thalmann, C., et al. 2017, A&A, 607, A90 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  43. Fairlamb, J. R., Oudmaijer, R. D., Mendigutía, I., Ilee, J. D., & van den Ancker, M. E. 2015, MNRAS, 453, 976 [Google Scholar]
  44. Fedele, D., van den Ancker, M. E., Petr-Gotzens, M. G., Ageorges, N., & Rafanelli, P. 2007, A&A, 472, 199 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  45. Fusco, T., Rousset, G., Sauvage, J. F., et al. 2006, Opt. Express, 14, 7515 [NASA ADS] [CrossRef] [Google Scholar]
  46. Fusco, T., Sauvage, J. F., Petit, C., et al. 2014, SPIE Conf. Ser., 9148, 91481U [NASA ADS] [Google Scholar]
  47. Garufi, A., Quanz, S. P., Avenhaus, H., et al. 2013, A&A, 560, A105 [CrossRef] [EDP Sciences] [Google Scholar]
  48. Garufi, A., Quanz, S. P., Schmid, H. M., et al. 2014, A&A, 568, A40 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  49. Garufi, A., Meeus, G., Benisty, M., et al. 2017, A&A, 603, A21 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  50. Garufi, A., Podio, L., Codella, C., et al. 2022a, A&A, 658, A104 [CrossRef] [EDP Sciences] [Google Scholar]
  51. Garufi, A., Dominik, C., Ginski, C., et al. 2022b, A&A, 658, A137 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  52. Ginsburg, A., Sipocz, B. M., Brasseur, C. E., et al. 2019, AJ, 157, 98 [NASA ADS] [CrossRef] [Google Scholar]
  53. Ginski, C., Stolker, T., Pinilla, P., et al. 2016, A&A, 595, A112 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  54. Ginski, C., Facchini, S., Huang, J., et al. 2021, ApJ, 908, L25 [NASA ADS] [CrossRef] [Google Scholar]
  55. Gledhill, T. M., Scarrott, S. M., & Wolstencroft, R. D. 1991, MNRAS, 252, 50P [CrossRef] [Google Scholar]
  56. Gledhill, T. M., Chrysostomou, A., Hough, J. H., & Yates, J. A. 2001, MNRAS, 322, 321 [Google Scholar]
  57. Gray, R. O., Corbally, C. J., Garrison, R. F., et al. 2006, AJ, 132, 161 [Google Scholar]
  58. Gray, R. O., Riggs, Q. S., Koen, C., et al. 2017, AJ, 154, 31 [NASA ADS] [CrossRef] [Google Scholar]
  59. Greene, T. P., & Lada, C. J. 2002, AJ, 124, 2185 [CrossRef] [Google Scholar]
  60. Haffert, S. Y., Bohn, A. J., de Boer, J., et al. 2019, Nat. Astron., 3, 749 [Google Scholar]
  61. Hales, A. S., Gledhill, T. M., Barlow, M. J., & Lowe, K. T. E. 2006, MNRAS, 365, 1348 [Google Scholar]
  62. Harris, C. R., Jarrod Millman, K., Stéfan J., et al. 2020, Nature, 585, 357 [NASA ADS] [CrossRef] [Google Scholar]
  63. Herczeg, G. J., & Hillenbrand, L. A. 2014, ApJ, 786, 97 [Google Scholar]
  64. Hinkley, S., Oppenheimer, B. R., Soummer, R., et al. 2009, ApJ, 701, 804 [NASA ADS] [CrossRef] [Google Scholar]
  65. Hodapp, K. W., Suzuki, R., Tamura, M., et al. 2008, SPIE Conf. Ser., 7014, 701419 [Google Scholar]
  66. Houk, N. 1982, Michigan Catalogue of Two-dimensional Spectral Types for the HD stars, 3. Declinations -40_f0 to -26_f0 (USA: Department of Astronomy, University of Michigan) [Google Scholar]
  67. Houk, N., & Smith-Moore, M. 1988, Michigan Catalogue of Two-dimensional Spectral Types for the HD Stars, 4, Declinations -26°.0 to -12°.0 (USA: Department of Astronomy, University of Michigan) [Google Scholar]
  68. Huélamo, N., Brandner, W., & Wolf, S. 2007, Rev. Mex. Astron. Astrofis. Conf. Ser., 29, 149 [Google Scholar]
  69. Hunter, J. D. 2007, Comput. Sci. Eng., 9, 90 [NASA ADS] [CrossRef] [Google Scholar]
  70. Hunziker, S., Schmid, H. M., Ma, J., et al. 2021, A&A, 648, A110 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  71. Hussain, G. A. J., Collier Cameron, A., Jardine, M. M., et al. 2009, MNRAS, 398, 189 [Google Scholar]
  72. Irvine, N. J., & Houk, N. 1977, PASP, 89, 347 [NASA ADS] [CrossRef] [Google Scholar]
  73. Keppler, M., Benisty, M., Müller, A., et al. 2018, A&A, 617, A44 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  74. Keppler, M., Teague, R., Bae, J., et al. 2019, A&A, 625, A118 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  75. Kóspál, Á., Ábrahám, P., Apai, D., et al. 2008, MNRAS, 383, 1015 [CrossRef] [Google Scholar]
  76. Kóspál, Á., Ábrahám, P., Goto, M., et al. 2011, ApJ, 736, 72 [CrossRef] [Google Scholar]
  77. Kóspál, Á., Ábrahám, P., Westhues, C., & Haas, M. 2017, A&A, 597, A10 [Google Scholar]
  78. Krautter, J., Wichmann, R., Schmitt, J. H. M. M., et al. 1997, A&AS, 123, 329 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  79. Kuhn, J. R., Potter, D., & Parise, B. 2001, ApJ, 553, L189 [NASA ADS] [CrossRef] [Google Scholar]
  80. Lacour, S., Biller, B., Cheetham, A., et al. 2016, A&A, 590, A90 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  81. Lada, C. J. 1987, in Star Forming Regions, eds. M. Peimbert, & J. Jugaku, IAU Symp., 115, 1 [Google Scholar]
  82. Lenzen, R., Hartung, M., Brandner, W., et al. 2003, SPIE Conf. Ser., 4841, 944 [Google Scholar]
  83. Levenhagen, R. S., & Leister, N. V. 2006, MNRAS, 371, 252 [Google Scholar]
  84. Long, F., Herczeg, G. J., Harsono, D., et al. 2019, ApJ, 882, 49 [Google Scholar]
  85. Macintosh, B., Graham, J., Palmer, D., et al. 2006, SPIE Conf. Ser., 6272, 62720L [Google Scholar]
  86. Macintosh, B., Graham, J. R., Ingraham, P., et al. 2014, PNAS, 111, 12661 [NASA ADS] [CrossRef] [Google Scholar]
  87. Maire, A. L., Rodet, L., Lazzoni, C., et al. 2018, A&A, 615, A177 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  88. Manara, C. F., Morbidelli, A., & Guillot, T. 2018, A&A, 618, A3 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  89. Marois, C., Lafreniere, D., Doyon, R., Macintosh, B., & Nadeau, D. 2006, ApJ, 641, 556 [NASA ADS] [CrossRef] [Google Scholar]
  90. Maucó, K., Olofsson, J., Canovas, H., et al. 2020, MNRAS, 492, 1531 [CrossRef] [Google Scholar]
  91. Mayama, S., Hashimoto, J., Muto, T., et al. 2012, ApJ, 760, L26 [NASA ADS] [CrossRef] [Google Scholar]
  92. Mesa, D., Bonnefoy, M., Gratton, R., et al. 2019, A&A, 624, A4 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  93. Millan-Gabet, R., & Monnier, J. D. 2002, ApJ, 580, L167 [NASA ADS] [CrossRef] [Google Scholar]
  94. Millar-Blanchaer, M. A., Wang, J. J., Kalas, P., et al. 2016, AJ, 152, 128 [NASA ADS] [CrossRef] [Google Scholar]
  95. Millar-Blanchaer, M. A., Girard, J. H., Karalidi, T., et al. 2020, ApJ, 894, 42 [NASA ADS] [CrossRef] [Google Scholar]
  96. Milli, J., Mawet, D., Pinte, C., et al. 2015, A&A, 577, A57 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  97. Milli, J., Vigan, A., Mouillet, D., et al. 2017, A&A, 599, A108 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  98. Milli, J., Engler, N., Schmid, H. M., et al. 2019, A&A, 626, A54 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  99. Miotello, A., Kamp, I., Birnstiel, T., Cleeves, L. C., & Kataoka, A. 2023, Pro-tostars and Planets VII, eds. S. Inutsuka, Y. Aikawa, T. Muto, K. Tomida, & M. Tamura, ASP Conf. Ser., 534, 501 [NASA ADS] [Google Scholar]
  100. Monnier, J. D., Harries, T. J., Aarnio, A., et al. 2017, ApJ, 838, 20 [Google Scholar]
  101. Monnier, J. D., Harries, T. J., Bae, J., et al. 2019, ApJ, 872, 122 [Google Scholar]
  102. Mora, A., Merín, B., Solano, E., et al. 2001, A&A, 378, 116 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  103. Mulders, G. D., Pascucci, I., Ciesla, F. J., & Fernandes, R. B. 2021, ApJ, 920, 66 [NASA ADS] [CrossRef] [Google Scholar]
  104. Müller, A., van den Ancker, M. E., Launhardt, R., et al. 2011, A&A, 530, A85 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  105. Murakawa, K., & Izumiura, H. 2012, A&A, 544, A58 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  106. Murakawa, K., Preibisch, T., Kraus, S., et al. 2008, A&A, 488, L75 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  107. Muro-Arena, G. A., Dominik, C., Waters, L. B. F. M., et al. 2018, A&A, 614, A24 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  108. Olofsson, J., Milli, J., Thébault, P., et al. 2019, A&A, 630, A142 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  109. Pascucci, I., Testi, L., Herczeg, G. J., et al. 2016, ApJ, 831, 125 [Google Scholar]
  110. Pecaut, M. J., & Mamajek, E. E. 2016, MNRAS, 461, 794 [Google Scholar]
  111. Perrin, M. D., Schneider, G., Duchene, G., et al. 2009, ApJ, 707, L132 [Google Scholar]
  112. Plavchan, P., Güth, T., Laohakunakorn, N., & Parks, J. R. 2013, A&A, 554, A110 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  113. Pollack, J. B., Hubickyj, O., Bodenheimer, P., et al. 1996, Icarus, 124, 62 [NASA ADS] [CrossRef] [Google Scholar]
  114. Quanz, S. P., Schmid, H. M., Geissler, K., et al. 2011, ApJ, 738, 23 [Google Scholar]
  115. Quanz, S. P., Birkmann, S. M., Apai, D., Wolf, S., & Henning, T. 2012, A&A, 538, A92 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  116. Quanz, S. P., Avenhaus, H., Buenzli, E., et al. 2013, ApJ, 766, L2 [Google Scholar]
  117. Reipurth, B., Hartmann, L., Kenyon, S. J., Smette, A., & Bouchet, P. 2002, AJ, 124, 2194 [CrossRef] [Google Scholar]
  118. Ren, B., Dong, R., van Holstein, R. G., et al. 2020, ApJ, 898, L38 [NASA ADS] [CrossRef] [Google Scholar]
  119. Rigliaco, E., Gratton, R., Kóspál, Á., et al. 2020, A&A, 641, A33 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  120. Rizzuto, A. C., Ireland, M. J., & Kraus, A. L. 2015, MNRAS, 448, 2737 [NASA ADS] [CrossRef] [Google Scholar]
  121. Romero, G. A., Schreiber, M. R., Cieza, L. A., et al. 2012, ApJ, 749, 79 [NASA ADS] [CrossRef] [Google Scholar]
  122. Rousset, G., Lacombe, F., Puget, P., et al. 2003, SPIE Conf. Ser., 4839, 140 [Google Scholar]
  123. Schmid, H. M., Joos, F., & Tschan, D. 2006, A&A, 452, 657 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  124. Shen, Y., Draine, B. T., & Johnson, E. T. 2009, ApJ, 696, 2126 [NASA ADS] [CrossRef] [Google Scholar]
  125. Silber, J., Gledhill, T., Duchêne, G., & Ménard, F. 2000, ApJ, 536, L89 [Google Scholar]
  126. Sissa, E., Gratton, R., Garufi, A., et al. 2018, A&A, 619, A160 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  127. Sissa, E., Gratton, R., Alcalà, J. M., et al. 2019, A&A, 630, A132 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  128. Skiff, B. A. 2014, VizieR Online Data Catalog: B/mk [Google Scholar]
  129. Staude, H. J., & Neckel, T. 1992, ApJ, 400, 556 [NASA ADS] [CrossRef] [Google Scholar]
  130. Suzuki, R., Kudo, T., Hashimoto, J., et al. 2010, SPIE Conf. Ser., 7735, 773530 [NASA ADS] [Google Scholar]
  131. Tazaki, R., Tanaka, H., Okuzumi, S., Kataoka, A., & Nomura, H. 2016, ApJ, 823, 70 [NASA ADS] [CrossRef] [Google Scholar]
  132. Tazaki, R., Tanaka, H., Muto, T., Kataoka, A., & Okuzumi, S. 2019, MNRAS, 485, 4951 [NASA ADS] [CrossRef] [Google Scholar]
  133. Torres, C. A. O., Quast, G. R., da Silva, L., et al. 2006, A&A, 460, 695 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  134. Tychoniec, L., Manara, C. F., Rosotti, G. P., et al. 2020, A&A, 640, A19 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  135. van Boekel, R., Henning, T., Menu, J., et al. 2017, ApJ, 837, 132 [Google Scholar]
  136. van den Ancker, M. E., de Winter, D., & Tjin A Djie, H. R. E. 1998, A&A, 330, 145 [Google Scholar]
  137. van der Marel, N., Dong, R., di Francesco, J., Williams, J. P., & Tobin, J. 2019, ApJ, 872, 112 [Google Scholar]
  138. van Holstein, R. G., Stolker, T., Jensen-Clem, R., et al. 2021, A&A, 647, A21 [EDP Sciences] [Google Scholar]
  139. Verhoeff, A. P., Min, M., Pantin, E., et al. 2011, A&A, 528, A91 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  140. Vieira, S. L. A., Corradi, W. J. B., Alencar, S. H. P., et al. 2003, AJ, 126, 2971 [Google Scholar]
  141. Virtanen, P., Gommers, R., Oliphant, T. E., et al. 2020, Nat. Methods, 17, 261 [Google Scholar]
  142. Weber, P., Pérez, S., Zurlo, A., et al. 2023, ApJ, 952, L17 [NASA ADS] [CrossRef] [Google Scholar]
  143. Wenger, M., Ochsenbein, F., Egret, D., et al. 2000, A&AS, 143, 9 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  144. Wilking, B. A., Meyer, M. R., Robinson, J. G., & Greene, T. P. 2005, AJ, 130, 1733 [NASA ADS] [CrossRef] [Google Scholar]
  145. Witzel, G., Eckart, A., Buchholz, R. M., et al. 2011, A&A, 525, A130 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  146. Zhou, Y., Bowler, B. P., Yang, H., et al. 2023, AJ, 166, 220 [NASA ADS] [CrossRef] [Google Scholar]
  147. Zurlo, A., Gratton, R., Pérez, S., & Cieza, L. 2023, Eur. Phys. J. Plus, 138, 411 [NASA ADS] [CrossRef] [Google Scholar]
2

PIPPIN is a publicly available Python package; see: https://pippin-naco.readthedocs.io for more information.

All Tables

Table A.1

Potential young systems observed by NACO in its polarimetric mode, sorted by right ascension.

In the text
Table B.1

Keywords and values recognised by PIPPIN in the configuration file.

In the text

All Figures

thumbnail Fig. 1

Open AO-loop assessment of HD 135344B Ks-band observations. Left panel: maximum counts of the ordinary (red) and extraordinary (blue) beams. The horizontal dashed blue and red lines are the 3σ bounds for the respective beam, indicating which observations have adequate AO corrections. The upper-right panel shows an example of an effective AO correction for the ordinary beam, and the lower-right panel shows the blurred result of an open AO loop.
In the text
thumbnail Fig. 2

Median Stokes Q and U images with different levels of IP corrections for HD 135344B Ks-band observations. From top to bottom: Q+ and U+ components after equalising the ordinary and extra-ordinary fluxes, Q and U resulting from the double-difference method, QIPS and UIPS after subtracting the median IP within an annulus, and the crosstalk-corrected QCTC and UCTC components where the reduced Stokes U efficiency is accounted for. The characteristic butterfly pattern is visible in each panel, and the compasses show the orientation of the detector and the sky.
In the text
thumbnail Fig. 3

Final PIPPIN data products with different levels of IP correction. From left to right: median total intensity (I), polarised intensity (PI), and the azimuthal Stokes components Qϕ and Uϕ of HD 135344B observed in the Ks band. From top to bottom: equalised ordinary and extra-ordinary beams, IP-subtracted, crosstalk-corrected, and Uϕ-minimised results. The total intensity is shown with a logarithmic scale from 20 to 104 counts, whereas the other panels use a linear scale from −5 to +5 counts and a logarithmic scale up to ±90. The negative signal is depicted in orange, and in each image north points up and east to the left.
In the text
thumbnail Fig. 4

Templates for observations of HD 135344B with a PA of −35°. Top panels: complete Qsynth and Usynth templates. Values range from −1 to +1, and pixels outside of the image are set to 0. Bottom panels: example annuli used in computing the cross-correlation coefficients.
In the text
thumbnail Fig. 5

Detection analysis of HD 135344B observed in the narrowband NB_1.64. Left panels: QIPS and UIPS images divided by the white lines into the four quadrants of the expected butterfly pattern. Right panel: annulus-wise CCFs, with the S/N shown against the annulus radius in pixels. The results for the QIPS and UIPS images are plotted in blue and red, respectively. The shaded region specifies the coefficients used in normalising and converting the CCF into a S/N function. The 5σ detection limit is indicated with a horizontal line.
In the text
thumbnail Fig. 6

Gallery of young systems detected with NACO and reduced with PIPPIN. Each panel shows the polarised light on a logarithmic scale ranging between different values to highlight sub-structures. The highest degree of IP correction is used where possible. Scale bars in the lower-left corners of each panel indicate 100 AU at each object’s distance. HD 169142, R CrA, and Parsamian 21 are shown in the H band, MP Mus is shown in the IB_2.06 filter, and the other panels use Ks-band observations. Mon R2 IRS 3 shows the median Iq image because the Stokes U component was not observed. The images of YLW 16A and Elia 2-21 present the first polarised light detections in the NACO observations.
In the text
thumbnail Fig. 7

Polarised-to-stellar light contrast, δpol, plotted against the apparent J-band magnitude. The right panel shows a zoomed-in view of the bright m¡. The object names are listed along the top axes. The marker colours and symbols specify the observing filter and object type, respectively. Upper limits are shown when the stellar PSF was determined to be saturated. The error bars show the 3σ uncertainties. The grey shaded region shows the approximate magnitudes (mJ ≳ 10) inaccessible by the SPHERE AO system.
In the text
thumbnail Fig. 8

Comparison between PIPPIN-reduced NACO Qϕ observations (left panels) and the more recent SPHERE data (right panels). From top to bottom: SU Aur, HD 142527, and TW Hya observed in the H band with both instruments. The SPHERE observations were previously published by Ginski et al. (2021), Hunziker et al. (2021), and van Boekel et al. (2017) for SU Aur, HD 142527, and TW Hya, respectively.
In the text
thumbnail Fig. C.1

Polarised light for the embedded YSOs Parsamian 21 (left panels) and Elia 2-21 (right panels). The top panels show the polarised intensity, PI, with a blue colour map, while the grey colours display the absolute values of the linear Stokes components |Q±| and |U±|. In the bottom panels, these values are scaled by the squared separation from the centre. The dashed lines delineate the sections of the sky observed by one of the components. These sections overlap in the centre, resulting in an eight-pointed star, where the polarised intensity image can be computed as Q and U are both covered. The south-eastern region of the U− observation of Parsamian 21 is contaminated with signal from another dithering position, introduced during the sky-subtraction.
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.