Max Planck Institute for Astronomy - Annual Report 2007
Max Planck Institute for Astronomy - Annual Report 2007
Max Planck Institute for Astronomy - Annual Report 2007
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
The Search <strong>for</strong> Needles in the Hay Stack<br />
– Four New Dwarf Galaxies<br />
In fact, four new dwarf galaxies were already discovered<br />
among the SDSS data in 2005 and 2006. They were, as<br />
is generally so, named after the constellations in which<br />
they were found. These were: Ursa Major I and II, Canes<br />
Venatici, and Bootes. An international astronomer team<br />
led by the MPIA and the University of Cambridge were<br />
able during the reporting year to discover four more dwarf<br />
galaxies. An additional faint object was added which was<br />
very likely a globular cluster.<br />
In the context of the SDSS, an area approximately the<br />
size of 8000 square degrees around the galactic north pole<br />
(corresponding to about one fifth of the entire celestial<br />
sphere) was filmed in five photometric spectral ranges.<br />
The dwarf galaxies discovered so far cannot be seen on<br />
these images with the naked eye, rather they come to<br />
light only after systematic, computer-supported searches<br />
<strong>for</strong> over-densities of stars within specific brightness and<br />
color ranges.<br />
An important criterion in searching <strong>for</strong> dwarf galaxies,<br />
and the later interpretation of data, is also the SDSS’s sensitivity.<br />
Bright stars in the red-giant branch can be detected<br />
at distances of up to three million light years; while very<br />
faint stellar systems which hardly contain red giants can<br />
only be detected at a distance of one million light years.<br />
Thus the SDSS reached farther than any survey to date.<br />
Under these boundary conditions, Sergej Koposov,<br />
Hans-Walter Rix and Eric Bell at MPIA with their colleagues<br />
developed a program to detect dwarf galaxies.<br />
A program of this type works as follows: One searches<br />
within an area of the sky <strong>for</strong> stars with selected colors<br />
and brightness and compares their numbers with expected<br />
values <strong>for</strong> background stars. A proven method in<br />
the search <strong>for</strong> overdensities in stars or <strong>for</strong> other deviations<br />
from average values within a particular region is<br />
the application of a spatial kernel: average values <strong>for</strong><br />
all data points within a certain area of the sky (in our<br />
case the colors and brightness of the stars) are sought.<br />
The comparison with these average values will reveal<br />
deviations such as the searched-<strong>for</strong> overdensities in the<br />
angular scale of the selected kernel.<br />
In order to test the efficiency of the program, the<br />
values of “artificial” dwarf galaxies and globular clusters<br />
were inserted. In this manner the search program’s<br />
detection limit was determined <strong>for</strong> seven distance ranges<br />
from 26 000 to 3.2 million light-years in dependence on<br />
the searched <strong>for</strong> object’s brightness (Fig. II.7.1). The<br />
size of the kernel played a significant role here: While a<br />
smaller kernel allowed <strong>for</strong> the better discovery of faint<br />
objects than a larger kernel did, this was at the expense<br />
of galaxy size. A large kernel adds up more stars from<br />
extended objects and there<strong>for</strong>e allows <strong>for</strong> the recognition<br />
of objects with lower surface brightness than a smaller<br />
kernel would have.<br />
II.7. Dwarf Galaxies – the “Missing Satellaties” of the Milky Way 51<br />
<br />
0.4<br />
0.2<br />
0<br />
–0.2<br />
–0.4<br />
0.4<br />
0.2<br />
0<br />
–0.2<br />
–0.4<br />
0.4<br />
0.2<br />
0<br />
–0.2<br />
–0.4<br />
0.4<br />
0.2<br />
0<br />
–0.2<br />
–0.4<br />
0.4<br />
0.2<br />
0<br />
–0.2<br />
–0.4<br />
0.4<br />
0.2 0 –0.2 –0.4 0.4<br />
<br />
0.2 0 –0.2 –0.4<br />
Fig. II.7.4: Left column: SDSS images of indicated objects as<br />
well as the positions of the su b a r u and INT fields. Right column:<br />
Images filtered with a 4 kernel.<br />
As a compromise, a kernel of 4 arc minutes was used<br />
where the red limiting magnitude lay around 22.0 mag.<br />
In the program’s trial runs, it turned out additionally that<br />
HII regions and clusters of galaxies were erroneously<br />
identified as dwarf galaxies (Fig. II.7.2). This problem<br />
could be eliminated through a comparison to existing<br />
catalogues of these objects.<br />
Through this method, MPIA astronomers and their<br />
colleagues were able to identify four as yet unknown<br />
dwarf galaxies: Coma Berenices (Com), Canes Venatici<br />
II (CVn II), Leo IV, and Hercules (Her). The fifth object,