School description > School description

Program

The planning of the school includes lectures (see below) and research projects that are both mandatoryto attend. The projects will be made in small groups with a tutor, and the last session (Monday 16th morning) will be for the presentation of the projects' results to the whole school. The social activities (dinner and guided tour) are optional, but the participants are warmly invited to attend (there is no additional cost; those not participating shall inform the organisers as early as possible in order to adjust the reservations). The off-school sessions, in the evenings are not mandatory. They will be self-organized by the participants to present their own research projects, or astronomical topics they are interested in. The access to the rooms to work on the projects will be possible at any time.

Lecturers and Lectures

1. Spectrographs and spectra -- observation and data-reduction techniques (Ranjan Gupta, IUCAA)
   • Basics of spectroscopy, including the description of spectrographs and of their characteristics, of observational aspects and data reduction.
     The lectures will cover basics of spectroscopy from the laboratory perspective and get into the astronomical aspects on spectroscopic instrumentation in particular the design of simple spectrographs and their limitations etc. All the three aspects i.e. low, medium and high resolution spectroscopy will be discussed in details. In the end observational preparation for actual telescope related aspects will also be described.
2. Methods of 3D spectroscopy – source detection and extraction (Johan Richard, CRAL)
   • Introduction to 3D spectroscopy. Detection of sources in data cube through emission lines (not seen in integrated light), extraction of 1D spectra from cubes, and characterization of the sources.
     Integral field spectroscopy, also called 3D spectroscopy, is a recent instrumental technique allowing observes to obtain the full set of spectra from astrophysical sources in a given sky region. By offering a contiguous spectral and spatial coverage of extended or multiple sources, integral field spectroscopy has transformed the field of distant galaxies studies, most notably by allowing us to measure their internal properties. I will introduce several concepts and instruments used for 3D spectroscopy and then present techniques typically used for the spectral analysis: in particular the automatic detection of sources in large 3D datacubes, optimal spectral extraction and decontamination, and the spectral analysis of extended sources.
3. Inverse methods for data fusion applied to 3D spectroscopie (Ferreol Soulez, CRAL)
   • Bayesian methods applied to inverse problems. Application to data fusion, for example to combine high spatial resolution images (HST) with spectroscopic cubes (MUSE), and extract the information.
     In many cases, observations from a single instrument is not informative enough for a precise comprehension of astrophysical phenomena. Only careful combination of observations from several different instruments may be able to provide sufficient information. For example, fusing HST and MUSE data to provide data cube with the spectral resolution of MUSE and the spatial resolution of HST. This lecture, will present methodological aspect of data fusion based on the 'inverses problems' framework. After a presentation of theoretical tools, I will focus on linear methods to provide a firm grasp on its behavior (in term of resolution, bias–variance tradeoff,…). Then I will present more state of the art methods (e.g. non linear method) and give insight on futur advance in this field.
4. Spectroscopic analysis and spectral classification (Philippe Prugniel, CRAL; Harinder Singh, Delhi University)
   • Extraction of information and astrophysical parameters from spectra: Feature measurements, model fitting. Automatic classification of spectra, deep learning.
     Spectroscopic analysis consists in extracting some physical information from spectra, and classification consists in assigning spectra to a given class of objects. This course will present various possible approaches to analyse spectra: This includes the 'measurements' of specific spectral features, like equivalent width, or spectroscopic indices that are afterward calibrated into physical quantities, or model fitting where non-linear models depending on some physical parameters are compared to an observation to constraint the values of these parameters. The course will also present automatic methods for classifying spectra. At a time when surveys are producing millions of spectra, such classification is vital and must allow one to discover new types of objects (data driven classification).
5. Planetary surfaces spectroscopy (Cathy Quantin-Nataf, LGL, UCBL, Lyon)
   • Observation of planetary surfaces. Spectral analysis, spectral unmixing. Mineralogic composition.
     Spectroscopy is the first exploration strategy to assess the composition of planetary worlds. Many spectrometers have been installed onboard spacecraft, lander or rovers. We will review the principles of spectroscopy of rocky and icy surfaces, the related laboratory measurements and the main results acquired the past 30 years of solar system exploration. A large part of the examples will be dedicated to Mars exploration from the first spectrometer to the last generation of spectro-imagers. We will review the processing pipeline, as well as the data meaning and data reduction pipeline.
6. Gnuastro (tutorial) and principles for preparing 'reproducible' papers (Mohammad Akhlaghi, CRAL, Lyon)
• Reproducible research and tutorial on estimating broad-band galaxy colors using GNU Astronomy Utilities.
  With the high speed processors and easy access to large volumes of data, data analysis and the scientific interpretations of the results has become a very complex process. The analysis usually involves many steps that cannot be adequately (in sufficient details), described in any paper. Therefore it is necessary to publish the scripts and lower-level software and their configuration files, that were used to generate the results in the paper/report. In this session, first we will follow a tutorial to derive broad-band colors of Hubble Space Telescope galaxies in a deep field using the large collection of tools available on the command-line in GNU Astronomy Utilities (Gnuastro, see https://www.gnu.org/software/gnuastro/manual/html_node/General-program-usage-tutorial.html). Then we will describe how the results of this small and simple research project can be presented as a paper/report in a reproducible format (see https://gitlab.com/makhlaghi/reproducible-paper). Please install Gnuastro prior to this tutorial, so you can follow the steps in practice (see https://www.gnu.org/software/gnuastro/manual/html_node/Quick-start.html).