PI: Marjorie Decleir, postdoc in the ISM* group at STScI - mdecleir@stsci.edu
Group Website: ISM*@ST
Project Duration: 3 separate one year rotation projects for graduate students
Background and Context
Interstellar dust absorbs and scatters a large fraction of the starlight, hence influencing observations of many astrophysical objects. Understanding the amount of dust, the properties of the grains, and the interplay between dust and radiation, is crucial to derive precise knowledge of any object in the Universe that is obscured by dust. In addition, interstellar dust is fundamental to the star formation process and galaxy evolution.
In the Milky Way, we can investigate dust along individual lines of sight through its wavelength-dependent extinction effect on the light of a background star. Extinction curves are not only required to account for the dust in observations of a range of astrophysical objects, but also provide profound insights into the properties of interstellar dust, imposing fundamental constraints on models of dust grains. For example, the slope of an extinction curve gives an estimate of the average dust grain size along the line of sight, while extinction features provide a direct measurement of the composition of the dust grains in the interstellar medium.
Available student projects
I have three available projects related to the study of interstellar dust. All of these projects are suitable for a student-led publication.
1. Reconciling spectral multi-wavelength extinction curves
Fig. 1: Left: NIR extinction curve in black with a powerlaw fit in red (adapted from Decleir et al. 2022). Right: MIR extinction curve in blue with a functional fit in black (Gordon et al. 2021)
Extinction curves are commonly measured by comparing the observed spectrum of a dust-extinguished background star to the intrinsic spectrum of a similar star or stellar model (not affected by dust). These measurements are often limited to a specific wavelength range (e.g. the ultraviolet or the optical). Recently, we measured extinction curves in the near-infrared (NIR) using IRTF/SpeX spectra (Decleir et al. 2022), and mid-infrared (MIR) using Spitzer photometry and spectra (Gordon et al. 2021). The main reasons for this wavelength “split” are often the available data sets, taken with different telescopes and instruments, different astronomers being interested in different wavelength regimes, and different stars being suitable only for observations in part of the electromagnetic spectrum. In the NIR and MIR, we found that the extinction curve can be represented by a power law (see Fig. 1). However, for the same targets, we obtained different power law indices in the NIR (<5 micron) and in the MIR (> 5micron), and both pieces of the extinction curve do not always line up. Other studies have also shown that the extinction curve becomes flatter at longer wavelengths, but the transition wavelength varies between different studies. This heterogeneous set of measured extinction curves makes it very challenging for dust modelers to constrain the dust properties using all these separate pieces of constraints.
The goal of this project is to measure and fit extinction curves over a wavelength range as large as possible, for the same stars, and fit all wavelengths simultaneously. We will start from the overlap targets between the NIR and MIR samples, and measure the full NIR-MIR extinction curve using a stellar atmosphere model. We will investigate if the entire wavelength range can be fitted with one single powerlaw, or if multiple powerlaws or other functions are needed. This result will help to constrain dust grain models. If time permits, this project can be expanded to shorter wavelength regions (UV-optical). This project will be in collaboration with Karl Gordon, who is also an expert on dust extinction.
Planned student work:
- Obtain the NIR and MIR spectra for the overlap sample.
- Measure the full NIR-MIR extinction curves using stellar atmosphere models.
- Fit the full NIR-MIR extinction curves.
- Compare the results with the individual studies and with dust grain models.
- Write a short paper with the results.
Technical skills the student will learn:
- Use existing Python code to measure and fit extinction curves (based on Astropy fitting).
- Expand the code to simultaneously fit NIR and MIR spectra.
- Use Git and GitHub for version control and collaborative coding.
- Academic and technical writing.
2. Interstellar dust features at optical wavelengths
Recently, intermediate scale structure (ISS) features were discovered in a sample of extinction curves at optical wavelengths (see Fig. 2 and Massa et al., 2020), but these have not been identified so far. As part of the WISCI (Webb Investigation of Silicates, Carbons and Ices) collaboration, we obtained optical Hubble Space Telescope (HST) spectra for a sample of stars in the Milky Way, in order to study these ISS features in more detail.
Fig. 2: Optical extinction curve, showing the three weak ISS features.
The goal of this project is to study the ISS features in these lines of sight. We will use the new HST spectra and stellar atmosphere models to measure optical extinction curves. We will fit the continuum extinction and the ISS features simultaneously with models. The obtained properties (e.g. central wavelength, strength, width) of these features can then be compared to other sightline properties, such as the V-band extinction A(V) and the total-to-selective extinction R(V), as well as to dust features measured at other wavelengths. This will help to identify the dust grains that cause these ISS features. If time permits, this project can be expanded to also include Gaia optical spectra.
This project is part of the larger WISCI project, and will give the student the opportunity to meet and interact with other WISCI-members.
Planned student work:
- Obtain the optical HST spectra for the WISCI sample.
- Measure the optical extinction curve using stellar atmosphere models.
- Fit the continuum and ISS features in the optical extinction curve.
- Compare the obtained ISS properties with other sightline properties.
- Write a short paper with the results.
Technical skills the student will learn:
- Use existing Python code to measure and fit extinction curves (based on Astropy fitting).
- Expand the code to fit the ISS features.
- Use Git and GitHub for version control and collaborative coding.
- Academic and technical writing.
3. Background corrections in Swift UV images
The NASA Swift UVOT instrument has three near-UV filters that straddle the interesting dust absorption feature at 2175 Å (see Fig. 3 left). This UV bump has been identified as carbonaceous in nature, but the exact composition of the dust carrier of this feature is still under discussion. While the Swift filters are uniquely suited to study this dust bump in nearby galaxies, obtaining accurate photometry is challenging due to the high and variable background fluxes in the Swift images. These background variations are caused by scattered light from the Sun, Earth and Moon. The existing data reduction pipeline DRESSCode (Data Reduction for Extended Swift Sources Code, Decleir et al. in prep.) currently does not account for the variable background in the images.
The goal of this project is to implement a correction for these background variations into the DRESSCode pipeline. This project is more technical and coding focused. The student will run the pipeline for several galaxy images and match the background levels before the individual images are being coadded. We will then compare the new mosaiced images with the previous ones, and quantitatively assess the improvement. This will result in lower uncertainties on the photometry, which will benefit the study of the UV bump. If time permits, this project can be expanded to measure the UV bump strength for a number of galaxies and compare the results with previous literature studies.
This project is in collaboration with the Swift UVOT team at Penn State University, and the student will have a chance to meet and interact with the other team members.
Fig. 3: Upper left: Milky Way UV extinction curve showing the dust feature at 2175 Å. Bottom left: Transmission curves of Swift and GALEX filters. Right: Mosaic of Swift UVOT images, showing different background levels in every tile.
Planned student work:
- Obtain archival UVOT data for several nearby galaxies.
- Write code to measure the background levels in the different images and match them.
- Implement this code into the existing DRESSCode.
- Run tests on existing data sets and compare the results with previous results for which the background was not matched.
- Write a short paper with the results.
Technical skills the student will learn:
- Use and understand the DRESSCode pipeline (written in Python, based on the HEASoft package).
- Expand the pipeline to correct for the background variations.
- Use Git and GitHub for version control and collaborative coding.
- Academic and technical writing.
