2022 Sagan Summer Workshop

Exoplanet Science in the Gaia Era

Hands-On Sessions

Background References

Before the Workshop

The hands-on sessions will include activities to access the Gaia Catalog, visualize Gaia data, and carry out orbital fits using the open-source code Orvara (see below for details).

Most of the hands-on session activities will use Python notebooks. You can choose to either install Python and associated software packages on your computer and use Jupyter notebooks, or download and use Google Colaboratory (Colab) notebooks. We have both Colab notebooks and Jupyter notebooks available with the same exercises; the former run in your Google Drive using a virtual machine and do not require a Python installation, so they provide the easiest way to work on the activities. No prior Python experience is required to participate in the hands-on sessions.

In addition, for hands-on session on Gaia, you are asked to create a Gaia archive account and install the visualization tool Glue, as well as download a few catalog files.

Here are the instructions (including the above):

You only have to follow one set of instructions, not both. Please complete these steps before the workshop, so you are ready to start right away working on the hands-on activities. If you have any problems installing Python, we recommend using Google Colab instead. If you prefer to install Python, but are having problems, you can ask questions on the Slack channel #python-help.

During the Workshop

Monday, July 25: Accessing the Gaia Catalog

The Gaia Data Release 3 (DR3), on June 13, 2022, will be the largest and richest resource on fundamental properties of stars. These properties are essential in the characterization of exoplanets, thus being comfortable with extracting the relevant information from the Gaia catalog is important.

In this hands-on session you will learn how to access the Gaia catalog data through the Gaia archive's Astronomical Data Query Language (ADQL) query interface and programmatically in Python. For more information on ADQL and accessing the Gaia data archive, please see their Demos and Tutorials page. Topics covered:

  • Basic ADQL queries
  • Complex queries to combine data from several Gaia DR3 tables
  • Cross matching of an exoplanet host sample to Gaia DR3, using: tools available in the Gaia archive; Topcat; or Python
  • Accessing multidimensional data in Gaia DR3, such as the lightcurves of candidate transiting exoplanets

Here is some reference material for this hands-on session:

Tuesday, July 26: Visualizing Gaia Data

Exploring a large and complex data set such as Gaia DR3 is made easier through modern visualization tools that can quickly render plots of large amounts of data and that allow linking different views of the same data.

In this hands-on session you will explore the Gaia Catalog of Nearby Stars (GCNS) using Glue.

Topics covered:

  • Plot the observational Hertzprung-Russell diagram (HRD) for the GCNS
  • Do the same for different volume-limited or magnitude limited samples
  • Use the techniques learned in the first hands-on session to extract exoplanet host star data and overlay these on the observations HRDs
  • Make plots of different quantities for the same sample of stars and use Glue's data-linking capabilities to explore the relations between the plotted quantities; for example, what is the spatial distribution of a sample of stars selected from a color-magnitude diagram?

Here are the step-by-step instructions on how to use Glue: Glue_Gaia_and_The_Exoplanet_Archive.pdf

Wednesday, July 27: Orbital Fits using Orvara

You will learn how to use the open-source code Orvara to fit the mass and orbit of a known brown dwarf around a main-sequence star. Three complementary types of data will be available:

  1. absolute astrometry from Hipparcos and Gaia
  2. relative astrometry from direct imaging
  3. radial velocities

As mentioned above, you will either install the code or use a Google Colab notebook. The goals of this session are to:

  • obtain an orbital fit
  • estimate the convergence of that fit, and
  • to assess the reliability of the results using the chi squared values of the residuals.

During this session, you should be able to produce a chain from Markov Chain Monte Carlo, discard an appropriate amount from the beginning as burn-in, and plot all of the results. If there is time remaining, the fit can be repeated omitting one form of data, e.g. relative astrometry from imaging, and assessing the changes in the fit and the ability to predict the location of a companion in advance of imaging.

Here is the document with details on the Orvara hands-on activities, including the orbital fitting group projects (see Thursday): Orvara_Hands-on_Session.pdf

Thursday, July 28: Orbital Fitting Group Projects

This session will expand on orbit fitting and mass measurements and will be more open-ended. Data for several systems will be supplied: some will have only two of the three data types (absolute astrometry, relative astrometry, and radial velocities), some will have more than one companion, and some will not allow for a satisfactory orbital fit.

You will work in groups to explore these systems, including

  1. predicting the current locations of companions;
  2. assessing evidence for underestimated uncertainties or additional companions;
  3. implementing priors on masses where appropriate;
  4. diagnosing and achieving convergence;
  5. explaining degeneracies and covariances in the orbital parameters.

Besides the different group projects using Orvara, there is an additional notebook for orbital fitting of Gaia-like data (links are provided in the Python and Google Colab instructions documents). This activity requires some Python programming and is intended as a challenge, with valuable insights gained on fitting Gaia astrometric data. Here is cheat sheet for this activity.

Friday, July 29

In-person attendees will give short, informal presentations on their group projects on Friday afternoon of the workshop.

Our remote attendees can choose to do the hands-on session activities on their own time. We will post some notebooks with answers on this website after the workshop.


We will have Slack available for asking questions at any time during the workshop week (three channels for the hands-on sessions, #hands-on-session-gaia, #hands-on-session-orvara, and #python-help). Questions posted on Slack will be answered as timely as possible, but expect some delays.

In addition, there is the opportunity to also use the Gather space that we have set up for the workshop to meet and work on the hands-on activities. Gather is a virtual environment in which you move around with an avatar and can talk to colleagues. We have set up a “Hands-on session lab” where you can meet and (virtually) sit at tables and work together in groups.


Workshop Code of Conduct.

2022 Workshop home page

(last updated August 15th, 2022 13:44:13)