Se-Heon's e-page

THINGS (The HI Nearby Galaxy Survey)

HI survey program undertaken with the NRAO VLA for 34 nearby (10 < Mpc) galaxies. This covers most Hubble type galaxies and address a wide range of issues about the physical properties of galaxies such as the galaxy dynamics, dark matter distributions, star formation, and interstellar medimum etc. A detail description of the project including the data and science goals can be found in Walter et al. (2008). Following papers discussing the dynamics of THINGS galaxies have been submitted to the Astronomical Jounal (in the form of a special edition) accompanying the general THINGS data release.

High-resolution rotation curves and galaxy mass models from THINGS (AJ accepted )
W.J.G. de Blok, Fabian Walter, Elias Brinks, Clemens Trachternach, Se-Heon Oh and Robert C. Kennicutt, Jr.

High-resolution dark matter density profiles of THINGS dwarf galaxies: Correcting for non-circular motions (AJ accepted)
Se-Heon Oh, W.J.G. de Blok, Fabian Walter, Elias Brinks, and Robert C. Kennicutt, Jr.

The "cusp/core" crisis in the near-field cosmology

Considerable debates about the description of universe on the galaxy scales predicted by Λ Cold Dark Matter (ΛCDM) have been continued and the central density problem (so-called "cusp/core") lies at the very centre of them with other two crisis such as the missing satellite galaxies and the angular momentum catastrophe. The dark matter N-body simulations based on the CDM paradigm consistently predict the divergent density profiles towards the galaxy centre, which gives a "cusp" feature near the centre. According to the ΛCDM predition, this high concentration of dark matter in the galaxy centre is universal and should be found in every galaxy (from tiny to big galaxies). In contrast to the simulations, the followed observations have mostly shown that the dark matter particles in the galaxy centres are not such highly concentrated as predicted by ΛCDM but even cored.

To address this issue, we investigate the distribution of dark matter in dwarf galaxies using observations obtained as part of THINGS. Dwarf galaxies are dark matter dominated and the highly resolved THINGS observations provide an opportunity to trace in detail the overall dynamics and constrain the dark matter distribution. Especially, we develop a new method to remove the impact of random and small-scale non-circular motions from HI velocity fields in (dwarf) galaxies in order to better constrain the dark matter properties for these objects. This method extracts the circularly rotating velocity components from the HI data cube and condenses them into a so-called "bulk velocity field". We derive high-resolution (~0.2 kpc) rotation curves of IC 2574 and NGC 2366 based on bulk velocity fields derived from THINGS and find significant differences between the bulk veloicty field rotation curves and those derived from the traditional intensity-weighted mean velocity fields. The bulk velocity field rotation curves are significantly less affected by non-circular motions and constrain the dark matter distribution in our sample galaxies, allowing us to address the discrepancy between the inferred and predicted dark matter distribution in galaxies. Spitzer Infrared Nearby Galaxies Survey (SINGS) 3.6 micron data, which are largely unaffected by dust in these galaxies, as well as ancillary optical information, are used to separate the contribution of the baryons from the total matter content. Using stellar population synthesis models, assuming various sets of metallicity and star formation histories, we compare stellar mass-to-light ratios for the 3.6 and 4.5 micron bands. Using our predicted values for the 3.6 micron stellar mass-to-light ratio, we find that the observed dark matter distributions of IC 2574 and NGC 2366 are inconsistent with the cuspy dark matter halo predicted by ΛCDM models, even after corrections for non-circular motions. This result also holds for other assumptions about the stellar mass-to-light ratio. The distribution of dark matter within our sample galaxies is best described by models with a kpc-sized constant-density core. This has been submitted to the Astronomical Journal 2008.

Click images for high resolution.

TAROT (THINGS Automated ROtationcurve Task)

A rotation velocity of a galaxy reflects its mass distribution and thus it has been widely used for deriving the mass models of not only the baryonic matter but the dark matter in galaxies, which can be then compared to the theoritical prediction. Since the galaxy rotation was confirmed first in 1914 (Slipher) from the inclined absorption lines in the nucleus of M31 considerable efforts have been devoted on this area to measure the rotation velocity as accurate as possible.

The 'ROTCUR' task implemented in Groningen Image Processing SYstem (GIPSY) (Begeman 1989) is commonly used for deriving the rotation curves of a galaxy. This interactive task, however, requires standard inputs from the user for tilted-ring parameters such as galaxy center (XPOS, YPOS), inclination (INCL), and position angle (PA) etc and is time-consuming. The main reason of the time consuming work is that the input models of 'ROTCUR' is made manually and sometimes depends on subjective model choices. In this reason performing 'ROTCUR' manually is not suitable for a large number of samples. We therefore developed a pipeline so-calles TAROT to make an improvement on this critical part. With help of this pipeline we have made the procedures for deriving the rotation curves fully automatic without human intervention. TAROT consists of a number of C routines, C-shell scripts, and GIPSY tasks.

Bulk velocity fields

There are several ways to define a velocity field such as first-moment maps, single Gaussian, or hermite polynomial fits, as well as peak velocity fields. Some, such first-moment maps, Gaussian fits, and peak velocity fields, work well for single, symmetrical HI profiles, but fail to properly take into account the presence of multiple velocity components caused by non-circular motions. One can try and capture these non-circular motions by fitting multiple Gaussian components to the profiles. However, even if profiles are decomposed perfectly, we are still left with the fundamental problem that a decomposition into Gaussian components does not give much insight as to which component is more representative of underlying circular rotation of the disk and which one represents non-circular motion or additional velocity components along the line-of-sight.

In order to address this problem and extract the undisturbed "bulk motion" (i.e., the velocity most representative of the undisturbed rotation) from the HI data cube of a given galaxy, we devised an alternative Gaussian decomposition method. This method minimizes the effects of localized non-circular motions (such as those caused by star formation processes) and extracts the circularly rotating components from the HI data cube.

The intrinsic dynamics of galaxies

A tilted-ring model has been typically used for deriving the rotation curve of a galaxy (Begeman 1989). A set of concentric ellipse rings defined with their own centres, position angles, and inclinations are fitted to the velocity field of a galaxy, which provides the rotation velocity as a function of galaxy radius. Therefore, the uncertainties of the derived rotation velocity of a galaxy mainly come from the two sources: the used velocity field and the adopted geometry.

There are various types of velocity fields such as the IWM, single Gaussian, hermite, and bulk velocity fields. These are classified as a way by which the line-of-sight velocities of a galaxy are determined. Although perfect velocity field that correctly traces the line-of-sight velocities corresponding to the circular rotation of a galaxy is employed we are still left with another

uncertainty: the geometrical parameters. Fitting different geometrical rings to a same velocity field leads to different estimates of rotation velocities. It is therefore crucial to use not only an appropriate velocity field correctly tracing the circular rotation but also correct geometrical parameters to derive an accurate rotation velocity of a galaxy.

Unfortunately, these two sources (the velocity field and the geometry) are degenerated as the geometrical parameters are usually derived from the velocity field itself. Additional photometric data may break this degeneracy but no general guarantee that the photometric and kinematic morphologies of a galaxy are same exists.

In general, a tilted-ring model is fitted to a velocity field to derive not only the rotation velocity but also the geometrical parameters, such as position angle (PA), inclination (INCL), and the dynamical centre (x, y) of a galaxy. This method is basically dependent on the assumption that the circularly rotating components are dominant in the galaxy. Under this assumption a tilted-ring from a least squares fit is given as the best model at a certain galactic radius. However, significant fraction of non-circular motions in a galaxy, especially the large-scale ones induced by bars, spiral arms, or warps, makes the derived tilted-rings deviated from their intrinsic ones. As this tilted-ring model just provides the mathematically best fit the derived geometry of a galaxy is very dependent on the used velocity field. This is evident for a galaxy in a bar-like potential. The bar-like potential in particular results in a large variation of position angle as a function of galaxy radius with respect to those from the photometric morphology derived at the outskirt of the galaxy. We therefore suggest a new method based on the harmonic anaysis of a velocity field to derive the intrinsic geometrical parameters of a galaxy.
The details can be found here.

Simulations of non-circular motions in galaxies

Several possibilities have been proposed to explain the discrepancy between the predicted and observed dark matter density profiles in the galaxy centres. In a theoritical point of view, a modification of standard Cold Dark Matter (e.g., self-interacting dark matter), triaxial dark matter halos, and baryonic feedback have been suggested. It has been also argued that several observational systematic effects, such as beam smearing, pointing offsets, and non-circular motions, could affect the measured inner slope of the density profile, and may therefore "hide" the signatures of cusps in the central parts of galaxies (van den Bosch et al. 2000; Swaters et al. 2003; Hayashi & Navarro 2006). However, de Blok et al. (2003) found that no single systematic effect could explain the discrepancy between the observations and simulations, but still a

possibility of hiding cusps with multiple systematics combining the above effects remains. Kuzio de Naray et al. (2006) showed that the typical velocity difference required for reconciling the observed and predicted density profiles is about 20 km/s for their sample galaxies used. More recently, Oh et al. (2008) have shown that high-resolution dark matter density profiles of THINGS dwarf galaxies are inconsistent with the cuspy dark matter halo predicted by ΛCDM and are best decribed by pseudo-isothermal halo models, even after considering those systematic effects. If we can parameterize these systematic effects and measure them quantitatively then we would be able to relate the observational systematics with their effects on the dark matter distributions. We therefore design and simulate HI data cubes of model galaxies to examine the observational effects of the multiple systematics quantitatively. We include the kinematic information such as dark matter halos, baryonic contributions. In particular, we introduce the non-Gaussian pforiles of the data cube by adding small and large-scale non-circular motions into the single Gaussian profiles of the circular rotations. The basic parameters of galaxy properties such as HI velocity dispersion, HI surface density profile are constrained by THINGS observations. The velocity fields are extracted from the data cube and used to derive the rotation curves so as to measure the deviations of the recovered rotation curves from the input ones quantitatively. We finally perform the mass modelling using the rotation curves as observers normaly do and examine the possibility of hiding cusps from systematic effects. We find that the observed dark matter properties of THINGS dwarf galaxies are more consistent with the predictions about galaxies with the pseudo-isothermal halos rather than those about galaxies with the NFW halos in our simulations even after considering their systematic effects.
More details can be found here.

Last updated: 31st May 2008