Christina M. Bird

Detecting bimodality in astronomical datasets

We discuss statistical techniques for detecting and quantifying bimodality in astronomical datasets. We concentrate on the KMM algorithm, which estimates the statistical significance of bimodality in such datasets and objectively partitions data into sub-populations. By simulating bimodal distributions with a range of properties we investigate the sensitivity of KMM to datasets with varying characteristics. Our results facilitate the planning of optimal observing strategies for systems where bimodality is suspected. Mixture-modeling algorithms similar to the KMM algorithm have been used in previous studies to partition the stellar population of the Milky Way into subsystems. We illustrate the broad applicability of KMM by analysing published data on globular cluster metallicity distributions, velocity distributions of galaxies in clusters, and burst durations of gamma-ray sources. PostScript versions of the tables and figures, as well as FORTRAN code for KMM and instructions for its use, are available by anonymous ftp from kula.phsx.ukans.edu.

Substructure in clusters and central galaxy peculiar velocities

Formation theories for central dominant galaxies in clusters require them to be located at the minimum of the cluster gravitational potential. However, 32\% (8 out of 25) of the clusters with more than 50 measured redshifts have central galaxies with significant velocity offsets (with respect to other cluster members). By studying their velocity distributions and correlations between velocity and position, I show that the presence of a large peculiar velocity is strongly correlated with the presence of substructure in these massive systems. About 85\% (21 of 25) of all well-studied clusters show some evidence for substructure, in contrast to the 30-40\% found when using only galaxy or gas distributions. The correlation between substructure and central galaxy location verifies the hypothesis of Merritt (1985) and Tremaine (1990) that high peculiar velocities are indicative of recent merger events between less-massive systems of galaxies. Dynamical friction should act quickly to pull the central galaxy, the most massive discrete object in a cluster, to the minimum of the potential. The less-massive galaxies retain information about their primordial subclusters for a longer period of time. I use an objective partitioning algorithm to assign cluster galaxies to their host subclumps. When galaxies are allocated in this fashion to their subclusters, 75\% of the significant velocity offsets are eliminated. Only 2 out of the 25 clusters have central galaxies which are not centrally-located when substructure is considered in the analysis.

Astronomical applications of distribution shape estimators

We present alternative measures to the classical coefficients of skewness and kurtosis, called the asymmetry and tail indices, for comparing the shapes of one-dimensional velocity distributions. We then apply these estimators to the Dressler and Shectman catalogues of clusters of galaxies to compare the dynamical evolution of these systems

Kinematics and dynamics of the MKW/AWM poor clusters

We report 472 new redshifts for 416 galaxies in the regions of the 23 poor clusters of galaxies originally identified by Morgan, Kayser, and White (MKW), and Albert, White, and Morgan (AWM). Eighteen of the poor clusters now have 10 or more available redshifts within 1.5/h Mpc of the central galaxy; 11 clusters have at least 20 available redshifts. Based on the 21 clusters for which we have sufficient velocity information, the median velocity scale is 336 km/s, a factor of 2 smaller than found for rich clusters. Several of the poor clusters exhibit complex velocity distributions due to the presence of nearby clumps of galaxies. We check on the velocity of the dominant galaxy in each poor cluster relative to the remaining cluster members. Significantly high relative velocities of the dominant galaxy are found in only 4 of 21 poor clusters, 3 of which we suspect are due to contamination of the parent velocity distribution. Several statistical tests indicate that the D/cD galaxies are at the kinematic centers of the parent poor cluster velocity distributions. Mass-to-light ratios for 13 of the 15 poor clusters for which we have the required data are in the range 50 less than or = M/L(sub B(0)) less than or = 200 solar mass/solar luminosity. The complex nature of the regions surrounding many of the poor clusters suggests that these groupings may represent an early epoch of cluster formation. For example, the poor clusters MKW7 and MKWS are shown to be gravitationally bound and likely to merge to form a richer cluster within the next several Gyrs. Eight of the nine other poor clusters for which simple two-body dynamical models can be carried out are consistent with being bound to other clumps in their vicinity. Additional complex systems with more than two gravitationally bound clumps are observed among the poor clusters.

Globular cluster clustering in M31

We present an analysis of the positions and velocities of M31 globular clusters, which indicates the presence of significant substructure in the globular cluster system. We suggest this clustering is the surviving signature of gaseous clumps out of which the halo of M31 formed. A mixture-modeling analysis of the metallicity distribution of the M31 globulars indicates that, like the Milky Way, the metallicity distribution is bimodal. We show that the metallicity of genuine halo clusters in the Milky Way and M31 is similar to that of globulars around nearby dwarf galaxies. This suggests that such clusters formed from primordial material, whereas the disk globulars in the Galaxy and M31, as well as metal-rich globulars in elliptical galaxies, formed in some secondary event

Abell 548: An X-ray and optical analysis of substructure

X-ray observations of Abell 548 reveal that the hot gas in this cluster is at least as complex as the galaxy distribution. Our ROSAT Position Sensitive Proportional Counter (PSPC) image is used in conjunction with optical data from the Minnesota plate-scanning project and redshift data in the literature to investigate the degree of substructure in the intracluster medium and the galaxy distribution. A548 has several X-ray components: hot gas associated with clumps of galaxies, individual sources, and a newly discovered diffuse component. This diffuse component may be promordial gas falling into the cluster for the first time. The optical data suggest that this cluster consists of four major components, not three (as indicated when velocity data are ignored). Simple models of the orbital dynamics suggest that the cluster may not yet have reached its turnaround radius.

Subcluster mergers and galaxy infall in A2151

We have obtained a 12.5 ksec image of the Hercules Cluster, A2151, with the {\it ROSAT} PSPC. Comparison of the optical and X-ray data suggest the presence of at least three distinct subclusters in A2151. The brightest X-ray emission coincides with the highest-density peak in the galaxy distribution, and is bimodal. The northern subclump, distinct in position and velocity, has {\it no} detectable X-ray gas. The eastern subclump, apparent in the optical contour map, is indistinguishable from the main clump in velocity space but is clearly visible in the X-ray image. X-ray spectra derived from the central peak of emission yield a best-fit temperature of 1.6 keV. The emission coincident with the eastern clump of galaxies is cooler, 0.8 keV, and is outside the 90\% confidence intervals of the central peak temperature. We suggest that the eastern and central subclusters have recently undergone a merger event. The lack of X-ray emission to the north suggests that those galaxies do not form a physically-distinct structure (i.e. they are not located within a distinct gravitational potential), but rather that they are falling into the cluster core along the filament defined by the Hercules Supercluster.

The effects of substructure on galaxy cluster mass determinations

Although numerous studies of individual galaxy clusters have demonstrated the presence of significant substructure, previous studies of the distribution of masses of galaxy clusters determined from optical observations have failed to explicitly correct for substructure in those systems. In this {\it Letter} I present the distributions of velocity dispersion, mean separation, and dynamical masses of clusters when substructure is eliminated from the cluster datasets. I also discuss the changes in these distributions because of the substructure correction. Comparing the masses of clusters with central galaxies before and after correction for the presence of substructure reveals a significant change. This change is driven by reductions in the mean separation of galaxies, not by a decrease in the velocity dispersions as has generally been assumed. Correction for substructure reduces most significantly the masses of systems with cool X-ray temperatures, suggesting that the use of a constant linear radius (1.5

Dark matter and dynamics in the Hercules Cluster (A2151)

h _{100} ^{-1}

The velocity dispersion‐temperature correlation from a limited cluster sample

Mpc in this study) to determine cluster membership is inappropriate for clusters spanning a range of temperatures and/or morphologies.