Previous work

Search for low surface brightness dwarf galaxies around nearby host galaxies

MSc thesis, Toni Paradell, Valencian International University (VIU), October 2021


Dark Energy Spectroscopic Instrument (DESI) Legacy Imaging Surveys:
Legacy Surveys (LS)

The search for satellite galaxies beyond the Local Group requires accessing data of very deep astronomical surveys, in which very faint objects can be found, with apparent magnitudes of up to magnitude 25.

The survey on which this work is based is the Dark Energy Spectroscopic Instrument (DESI) Legacy Imaging Surveys (Dey et al. 2019), hereinafter the Legacy Surveys (LS). The LS provides very deep images in the g, r and z bands, and with a very large coverage.

LS coverage

The LS primary objective is to identify sources for the DESI instrument. It is actually composed of three projects in three different telescopes: Mayall z-band Legacy Survey (MzLS), Beijing-Arizona Sky Survey (BASS), and Dark Energy Camera Legacy Survey (DECaLS).

With this configuration, the LS reaches in z band the coverage shown in the figure above, a coverage of 14,000 deg2, from which 9,900 deg2 in the Northern galactic hemisphere, and 4,400 deg2 in the Southern galactic hemisphere. This is a similar coverage to the Sloan Digital Sky Survey (SDSS).

Along with these three projects, the LS includes also data from DECam grz surveys, the most important of which is the Dark Energy Survey (DES) in the Southern galactic hemisphere (The Dark Energy Survey Collaboration, 2005). The total coverage of the LS including DES reaches 20,000 deg2.

LS depth

Comparison of the photometric depth of SDSS, BASS and DECaLS in the r band. The 5-sigma depth estimated in SDSS is g = 23.13, r = 22.70, z = 20.71, while in the Legacy Surveys they are g = 24.7, r = 23.9, z = 23.0.

Selected host galaxies

The first decision of this study consists in the selection of the host galaxies around which satellite galaxies are searched on the LS images. The criteria applied to select host galaxies have been:

  • Coverage in the Legacy Surveys

  • Morphology: spiral galaxies, Milky Way type

  • Point of view: “edge-on”

  • Host galaxy distance: 7-26 Mpc

  • Galactic latitude: |b| > 30°

  • Already referenced in previous studies

(*) Distance to each galaxy was taken as the average of distances provided in studies which have been referenced by the NASA/IPAC Extragalactic Database
Distribution of the 36 identified candidates to be
a satellite dwarf galaxy of NGC 3669

Identified candidates

213 candidates have been identified around the 10 host galaxies (click on the galaxy name to get the distribution of candidates around the galaxy, as shown in the example on left for NGC 3669):

Photometry

The photometric characterization of the candidates was made by using IRAF (Image Reduction and Analysis Facility), and more specifically the Space Telescope Science Data Analysis System (STSDAS) package developed by the Space Telescope Science Institute (STScI). This package includes tasks for the photometric analysis of extended galactic-type sources, being ellipse the main task. Executing the ellipse task in several steps, we get the main geometric parameters of the candidate.

Examples of profiles from different candidates

Ideally, with the geometric parameters and Sérsic profile that we get after these steps, we could make the photometric measurements for each candidate image. However, this would hold true only if the object were completely isolated in the image, but in practice there are many other objects in the images, usually background galaxies but also sometimes image artifacts, which may be very close or even inside the object that we are analyzing, thus altering the measurements.

In order to solve this problem, the adopted strategy consists in the generation of a mask for each image, which sets the pixels that must be disregarded in the photometric model generation. The masks are generated by means of an iterative process which improves the mask, and hence also the model, in each iteration. The following schema summarizes the adopted strategy.

Example of the phometric model after iteration 1 (left) and iteration 4 (right) for candidate NGC2549-07 in r band

This methodology has produced, as a result, a full set of parameters for each candidate:

Geometric parameters

  • Coordinates of the center of the object

  • Distance to the massive galaxy

  • Ellipticity

  • Position angle

  • Effective radius

  • Sersic index

Photometric parameters

  • Apparent magnitude in g, r, z bands

  • Absolute magnitude in V band

  • Medium surface brightness at effective radius in g, r, z bands and in V band

Other

  • Papers, if any, in which the candidate has already been identified as a dwarf satellite galaxy

Example of the geometric and photometric data generated for each candidate, in this example for candidate NGC7814-08

Data for all candidates is available upon request, just simply drop a message to info@dgsat.org with your name, affiliation and motivation.

Comparison with the Local Group

Since we cannot measure the distance to the candidates, in order to have a higher degree of certainty that they really are satellites of the host galaxy, we compare their photometric characteristics with those of known and well-characterized dwarf galaxies of the Local Group. The comparison allows us to rule out those candidates whose characteristics differ significantly from those of the Local Group, since they could be background galaxies unrelated to the host galaxy, galactic cirrus or simply noise or artifacts from the Legacy Surveys images. For this purpose, we use the Milky Way and Andromeda satellite galaxies data compiled by McConnachie (2012) survey, data that the author keeps updated in his website.

Relationship in V band between absolute magnitude and mean surface brightness at effective radius (in mag arcsec−2) of known dwarf galaxies of the Local Group and of the candidates identified in this survey

Candidates characteristics

Brightness histograms

The histograms show that, based on their mode value (bin -10,-11), the most distant galaxies present a larger number of candidates on the side of the brightest than on the side of the least bright, with the exception of ESO146-14 in which the opposite happens. On the other hand, in the closest galaxies the trend seems the opposite, with a larger number of candidates of lower brightness.

From these observations it seems that the brightest candidates show proportions close to the real ones, while the less bright candidate observations may be limited by the instruments and telescopes used by the Legacy Surveys.

The cumulative total for all galaxies reflects the distribution that, on average, it was expected to find in the galaxies, somehow implying a statistical validation of the results.

Cumulative distribution of candidates by absolute magnitude in V band for the 10 host galaxies
Distribution of the number of candidates by absolute magnitude in V band for each of the 10 host galaxies, in the left column the closest, in the right column the furthest
Distribution of the candidates according to their direction, grouped in bins of 45°, for each of the 10 host galaxies, and the accumulated total for all of them

Spacial distribution

The distribution diagrams show that in some host galaxies their candidates have a preferred direction, while others exhibit a more isotropic distribution.

However, three of the host galaxies are accompanied by another galaxy at a relatively short angular distance. This is the case of NGC3669, with NGC3625 as a close companion; NGC4517 with NGC4517A; and NGC4631 with NGC4656. Such companion may have disrupted candidates distribution, and in fact in these three cases there is a significant reduction in the number of satellites in the direction of the companion.

From the remaining 7 galaxies, the one with a more isotropic candidates distribution is NGC100, which also happens to be the one with the largest number of identified candidates, and hence the most statistically significant. The other galaxies show more irregular distributions, and although the number of candidates is smaller and therefore they are less representative, some of them show preferred directions, as it is the case of NGC7814.

In general, it seems to be a tendency to get more isotropic distributions in the host galaxies in which a larger number of candidates have been identified. However, the figures are still not significant enough to extract conclusions, both in terms of number of host galaxies and in terms of number of identified candidates. More research work is still required.

Candidates already identified in previous works

A deep research in the literature shows that out of 198 candidates, 29 of them had been previously identified and referenced, which means that the remaining 169 candidates have not been found in any previous paper or survey, and hence this study has been the first one in which they have been referenced.

Identified candidates whose bibliographic references have been found, and the identifier of the candidate used in said reference. Some of the candidates are referenced in several papers or surveys