Extragalactic Data Challenge
To prepare for the launch of JWST, the JADES team has been developing the analysis tools and methods that will be used to explore and understand the galaxies that will be observed in this exciting dataset. To that end, we are undertaking a data challenge to test our ability to recover the properties for a sample of simulated galaxies in order to prepare for what we will be doing with the real galaxies. How well can we measure galaxy masses and rates of star formation? How can we recover galaxy distances using the JWST data? What is our plan for measuring the properties of gas and stars in the most distant galaxies ever observed? This challenge utilizes a simulated catalog of galaxies developed by members of the JADES team (the JAGUAR catalog, Williams et al. 2018), from which we generate fake JWST images and spectroscopy which we can use to test our methodology. JWST will let us observe thousands of faint objects for the first time ever, and so this data challenge is critical for helping us explore how well we will be able to understand the distant universe.
Estimating the distances to galaxies in our survey from their images alone is a difficult prospect. The JADES survey encompasses a series of deep images of a patch of sky taken in multiple filters that span the wavelength range covered by NIRCam, from about 1 micron to 5 microns in the near-infrared. Because the universe is expanding, distant galaxies are observed to be moving away from us, and the farther that they are, the faster their rate of recession. We can detect how fast a galaxy is receding because the galaxy’s movement away from us causes light we observe from them to be “redshifted” to longer and longer wavelengths: the light that a galaxy emits in the optical portion of the spectrum gets pushed into the infrared for us observing it at a great distance. As a result, and this is key, if we can figure out a galaxy’s redshift, we can tell the distance to the galaxy.
The easiest way to determine the redshift of a galaxy is by observing the galaxy’s spectrum and comparing the wavelength we observe important features to the wavelength they would be observed at if the galaxy were at rest. However, spectroscopy can be time consuming for a large number of objects, and so we want to explore techniques that we can use for our NIRCam images alone. What we’d like to do is look at how a galaxy looks at various wavelengths in the images to discover any broad spectroscopic features that we can use to tell the galaxy’s redshift. One of the primary ways that this has been done for star-forming galaxies for over twenty years is by using the "Lyman break” in a galaxy’s spectrum. Intergalactic hydrogen gas around star-forming galaxies will preferentially absorb all of the light at very short wavelengths where the galaxy would otherwise produce a lot of ultraviolet radiation from young stars. As a result, if we take an image on either side of the “break,” the galaxy will appear in the longer wavelength image, but not in the shorter wavelength image.
At larger and larger redshifts, this break is pushed to longer wavelengths, and we can use our deep JADES imaging in the F090W band to target galaxies where the break sits at around 1 micron, which corresponds to galaxies at z > 7, when the universe was only 700 Million years old. Below, we show a plot with one galaxy from the mock JAGUAR catalog in three filters, F090W (purple), F115W (blue), and F150W (dark red), where the Lyman break at this redshift means that we don’t detect the galaxy in the F090W filter. If we look for galaxies that satisfy this criteria across the image, we can select large subsamples of objects at high redshifts for further study.
The below image is a simulation of a region of the sky probed by JWST/NIRCam with the JADES deep survey. It was generated using the JAGUAR and Guitarra software developed at the University of Arizona, and was designed to help simulate our ability to detect extremely distant, faint galaxies. The mosaic combines images in three near-infrared filters, which we color blue, green, and red, although these images are simulated at wavelengths that are redder than optical wavelengths. In the insets, we focus on a selection of galaxies that highlight the versatility of JWST in finding extremely distant galaxies as well as faint, red galaxies that are much closer.
Relevant Publications
Williams et al (2018), “The JWST Extragalactic Mock Catalog: Modeling Galaxy Populations from the UV through the Near-IR over 13 Billion Years of Cosmic History,” The Astrophysical Journal Supplement Series: NASA ADS
