"Remote Sensing" of Exo–Kuiper Belts
Like the debris disks in extrasolar systems, our own Asteroid Belt and Kuiper Belt are born from the mutual collison of minor objects in planetary systems. The existence of such exo–Kuiper belts not just informs the architecture of planetary systems, but also sheds light on the composition of them. Indeed, the Asteroid Belt and Kuiper Belt provide the water that is essential to life on Earth, with the fomer being the primary resevoir (Morbidelli et al. 2000).
Using both the Hubble Space Telescope (e.g., Schneider et al. 2014) and ground-based extreme-adaptive-optics-equipped coronagraphic instruments (e.g., the Gemini Planet Imager, see Esposito et al. 2020), the study of them at multiple wavelengths is at the horizon. This "remote sensing" makes it possible to inform the dust composition (water, volatiles, etc.) for extrasolar systems.
1. Foundation: authentic recovery of exo–Kuiper Belts
High-contrast imaging using space- and ground-based instruments requires both state-of-the-art data reduction methods (see also High Contrast Imaging) and ways to offset their potential biases. For the latter, it is necessary to not just recover the disks, but also quantify their uncertainties.
By looking at its Kuiper Belt, HD 191089 is a remote young cousin of our own Sun. Using a combination of radiative transfer modeling and Monte Carlo Markov Chain exploration, we studied the circumstellar disks using the MCFOST software (HD 191089, ApJ, arXiv, ADS). To efficiently sample the parameter space, I come up with the DebrisDiskFM framework that can perform parallel calculation across multiple nodes on a computer cluster.
Technical highlight: using probability integral transform, I demonstrated that current simple dust models are unable to consistently produce observations at different wavelengths.
Scientific highlight: the closest Kuiper Belt sibling was found. See NASA Astrobiology webpage, SciTechDaily webpage, CosmicDiary post, and the SETI institute tweet.
Technical implication: the framework for disk modeling is established for efficient paramter optimization, this is based on reducing the single exploration time from ~2 weeks on a server to ~1 day on a computer cluster.
Project timescale: 2018 January to 2019 July.
2. Application #1: face-on study of an M-star's debris disk
M stars are now the best places for planet search using current and upcoming telescope instruments, in the sense that the planet-to-star contrast is optimal for planet imaging, and that they are populus (more than 70% of the Milky Way; Muench et al. 2002) while likely hosting more planets (Howard et al. 2012). TWA 7, the only known M star that hosts a face-on disk, offers the best vantage point to complement the study of debris disk structures for M stars.
Technical highlight: using a sequential approach of disks modeling with DebrisDiskFM, and by masking out certain regions of interest, we can determine the disk structure and study its small varaition.
Scientific highlight: we recover a moving spiral in the system by combining Hubble and SPHERE observations, and it might be driven a hidden planet. In addition, the dust distribution of the tail of the system is unsual, with a powerlaw index of -0.7 rather than classical expectation of -1.5.
Technical implication: by masking out certain regions of interest, we can study the minute structure in debris disks.
Project timescale: 2020 March to 2021 May.
3. Application #2: completing mission impossible in extracting a circumstellar disk
Oct 2021 --- Nov 2023: Extraction of HD 53143 debris disk using archival Hubble Space Telescope (HST) STIS coroangraphic observations (Stark, Ren et al. 2023; AJ, arXiv, ADS)
Technical highlight: I managed to recover the HD 53143 debris disk using an ensemble of other stellar images, which was the only way to recover the disk then.
Scientific highlight:
Using the HST/STIS coronagraphic archive that I constructed (Ren et al. 2017), I recovered the disk (right) that cannot be achieved with classical reduction (left).
The forward scattering part is mysteriously missing, after disk modeling performed by me.
Based on my recovery and modeling of the HD 53143 system, I was originally expected to be the first author for the publication. Nevertheless, I yielded the first authorship to Dr. C. C. Stark (NASA Goddard) the original principal investigator of the proposal (HST GO-16202) due to extreme oversubscription of my time. I retained the corresponding authorship to answer questions on the reference sifting, preprocessing, post-processing, and forward modeling of the paper.
Sucessful recovery of the HD 53143 disk with HST/STIS coronagraphic archive (colored). In comparison, classical reduction cannot recover the disk well (gray).
4. Application #3: ensemble study of debris disk resolved at different wavelengths by Hubble coronagraphs
Oct 2019 --- Nov 2022: Multi-wavelength imaging of debris disks using Hubble Space Telescope (HST) coronagraphs (Ren et al. 2023b; A&A, arXiv, ADS).
Technical highlight: reduction, modeling, and analysis of debris disk reflectance in a systematic way.
Scientific highlight:
Debris disks in HST wavelenghts (0.6 µm, 1.1µm, and 1.6µm) are predominantly blue.
The more luminous the host star, the less blue the disk.
The abedo-color distribution may resemble that of solar system minor objects.
This confirms a simulation study (Thebault & Kral 2019), in which the authors found it hard to explain an observed "red" system: the system is blue in my work!
This is also a collaborative effort between debris disks imaging and solar system studies.
Non-science but management lessons learnt: while the results were already obtained by Aug 2021 (funding completion), the actual manuscript was not completed until Nov 2022. This is not due to further analysis or manuscript writing, but due to a waste of my 15 calendar months of free labor! The primary obstacles are (1) to convince one collaborator who made me to align my blue results with a few studies on that "red" system, yet these studies' limitations were already clearly identified by a later paper from that team. This year-long distrust from that person triggered my initial plan to leave debris disk imaging. (2) To persuade the same person -- who wanted to use Mie theory to explain the observations -- that Mie theory has been rigorously shown to be not appropriate in explaining disk observations. This waste of ~another year (for both me and another young scientist) consolidated my determination to leave debris disk imaging, at least to cut ties with that person. Were it not for the unwavering support of my official PhD and postdoc advisors, I would not have had the courage to confront that individual and to say no.
Personal suggestions to young scientists when encountering similar issues: if someone's behaviors make you uncomfortable, they are highly likely to be not appropriate. However, I was too timid to stand out until when I confirmed that individual's behavoirs on a few other scientists younger than me, then I decided to stand out for them. As a result, no means no, no matter in what environment and regardless of your gender/race/age. If you are in the United States, cite Title IX.