SIMULATING THE GROUP FINDING IN THE DEEP2 SURVEY WITH PHOTOMETRIC REDSHIFT INFORMATION H.-Y. Baobab Liu1, B.-C. Paul Hsieh2, L.-H. Lin3 1. NTU; 2. ASIAA; 3. UC Santa Cruz

1
SIMULATING THE GROUP FINDING IN THE DEEP2 SURVEY
WITH PHOTOMETRIC REDSHIFT INFORMATIONH.-Y.
Baobab Liu1, B.-C. Paul Hsieh2, L.-H. Lin31.
NTU 2. ASIAA 3. UC Santa Cruz
ABSTRACT We developed a modified friend-of-friend
method for galaxy group/cluster finding by
considering the probability of two objects to be
physically associated based on their redshift
information and projected position on the sky.
This method is currently tested using the DEEP2
mock catalog. In the future work, we will use it
to identify galaxy groups in DEEP2 survey, which
has an averaged redshift 1.
INTRODUCTION In the previous work of DEEP2 group
(Gerke et at., 2005), they identified galaxy
groups using only galaxies with successful
spectroscopic redshift. From DEEP2 mock catalog
(Yan et al. 2004), we can see the total number of
groups drops by an order of magnitude if we
neglect those galaxies without spectroscopic
redshift measurement ( After mask-making in
FIGURE 1.) . Now we want to combine the
spectroscopic catalog with photometric redshift
information( Magnitude-limited in FIGURE 1.) to
obtain a group catalog with richness that is less
affected by the selection effect and
incompleteness problem. The improved richness can
reduce the uncertainty for the study of the
dependence of galaxy and galaxy group aproperties
on the environment.
FIGURE 1.
THE MODIFIED FRIEND OF FRIEND(FOF) METHOD BASIC
IDEA Instead of comparing the absolute
separation of two galaxies to the linking length,
whether the two galaxies belong to the same
galaxy group is determined by their PROBABILITY
OF BEING WITHIN LINKING LENGTH. FIGURE 2.
illustrates our modification. PDF1 and PDF2 are
the probability distributions of position of
object1 and object2 respectively. ADVANTAGE I.
Galaxies with and without spectroscopic redshift
have EQUAL FOOTING IN THIS ALGORITHM. II.
Spectroscopic data help to CONSTRAINT THE
POSITION OF GROUP DISADVANTAGE The
probability threshold has weak physical meaning,
and is optimized empirically.
FIGURE 2.
RESULT FIGURE 3. is our group finding result for
DEEP2 mock catalog, projected to the RA and DEC
coordinates on the sky. The reconstructed group
and real group are correlated. And some
reconstructed groups are found to be false
detection. TALBE 1. shows our preliminary
results of group finding in mock catalog with
different parameters used in algorithm.
Probability threshold is the Pth as in FIGURE 2.
. of mock group is the total number of groups
given in the DEEP2 mock catalog, and of
reconstructed group is the number of groups
identified by our algorithm. of Group not
identified are the number of real group we didnt
find while extra identification are the of
reconstruct groups which cant be identified with
real groups. Identical groups are those
reconstructed groups with same number of members
and members with a real group.
CONCLUSION We have developed a modified FOF
algorithm which can be used to identify galaxy
group/clusters using the photometric redshift
and/or spectroscopic redshift information
together with their projected coordinates on the
sky. This group-finding method has been tested
using the DEEP2 mock catalog. Our preliminary
results , however, suggest that the false
detection rate can be as high as 60 while the
completeness is around 40, which can be a result
of wrong choices of the linking length and the
probability threshold. We plan to address this
issue in the near future.
TABLE 1.
FUTURE WORK Those parameters in our algorithm
need further optimization. In addition, the
weighting of physical quantities of each members
in the groups, such as its position and and
velocity, need to be modified to obtain the
unbiased determination of center positions and
velocity dispersions of galaxy groups/clusters.
Linking Length plan-of-sky 100 kpc line-of-sight 3Mpc Linking Length plan-of-sky 100 kpc line-of-sight 3Mpc Linking Length plan-of-sky 100 kpc line-of-sight 3Mpc Linking Length plan-of-sky 100 kpc line-of-sight 3Mpc Linking Length plan-of-sky 100 kpc line-of-sight 3Mpc
Probability threshold 1 0.1 0.01 0.001
of mock group 159 159 159 159
of reconstructed group 13 93 166 166
of group not idenfied 150 126 95 95
of extra identification 4 56 95 95
of identical group 4 10 12 12
Reference Gerke, et al.,2005, ApJ, 6256-22 Yan,
R., White, M., Coil, A. C. 2004, ApJ, 607, 739