1: \documentstyle[12pt,aasms4]{article}
2:
3: \begin{document}
4:
5: \noindent
6: {\it Dissertation Summary}
7:
8: \begin{center}
9:
10: \title{\large \bf On the Formation and Evolution of Stellar Bars in Galaxies\footnote{See
11: http://www.astro.iag.usp.br/$\sim$dimitri/phdthesis/phdthesis.html where this work, originally written in
12: Portuguese, is available.} }
13:
14: \end{center}
15:
16: \author{ Dimitri Alexei Gadotti }
17:
18: \affil{ Astronomy Department - University of S\~ao Paulo, Rua do Mat\~ao, 1226, S\~ao Paulo/SP, Brasil,
19: 05508-900 }
20:
21: \begingroup
22:
23: \parindent=1cm
24:
25: \begin{center}
26:
27: Electronic mail: dimitri@astro.iag.usp.br
28:
29: Thesis work conducted at: Astronomy Department - University of S\~ao Paulo
30:
31: Ph.D. Thesis directed by: Ronaldo Eust\'aquio de Souza ; ~Ph.D. Degree awarded: 2003 December
32:
33: {\it Received \underline{\hskip 5cm}}
34:
35: \end{center}
36:
37: \endgroup
38:
39: \keywords{ galaxies: evolution -- galaxies: formation -- galaxies: fundamental parameters
40: -- galaxies: halos -- galaxies: kinematics and dynamics -- methods: n-body simulations }
41:
42: We have done a detailed study on the structural and kinematical properties of lenticular
43: and early- and late-type spiral galaxies with bars, aiming to explore the formation and evolution processes
44: of stellar bars in galaxies, and their implications on the global formation and evolution of galaxies.
45:
46: Firstly, using high signal-to-noise spectra obtained along the major and minor axes of the bars in a sample
47: of 14 face-on galaxies, we have determined the line of sight stellar velocity
48: distribution in the bars' vertical axis, in several points, reaching 20'' from the center. This was done with
49: an algorithm properly developed for this task, which parameterizes the velocity distribution using Gauss-Hermite series
50: (see, e.g., R. P. van der Marel \& M. Franx 1993, \apj, 407, 525),
51: allowing for an accurate determination of the kinematical parameters. These spectra were observed with the
52: 1.5 m ESO telescope at La Silla, Chile, and with the 2.3 m Steward Observatory telescope on Kitt Peak, Arizona.
53: With these data, it was possible to develop a diagnostic tool that allows one to estimate the ages of bars, and distinguish
54: between recently formed bars and evolved bars. Furthermore, we could evaluate the vertical structure of disks
55: and bars in galaxies. We were able to separate evolved bars from recently formed ones based on the
56: assumption that bars form within disks, thus having initially a thin vertical structure, which can be recognized
57: by low values for the vertical stellar velocity dispersion, $\sigma_z$. As bars age, processes like vertical resonances
58: (e.g., F. Combes \& R. H. Sanders 1981, \aap, 96, 164) and the hose instability (e.g., D. Merritt \&
59: J. A. Sellwood 1994, \apj, 425, 551) contribute to make $\sigma_z$ higher, i.e., turning them vertically thick
60: and possibly originating the characteristic boxy/peanut morphology (see Figure 1).
61:
62: Secondly, through realistic $N$-body simulations with {\sc nemo} (P. J. Teuben 1995, ASP Conf. Ser., 77, 398),
63: which represent a $2-3$ Gyr evolution of isolated galaxies with bulge, disk and a rigid dark matter
64: halo, we have studied the necessary conditions to the formation of bars in galaxies with different morphological
65: types (i.e., different bulge/disk ratios). In this way, we could also check the time scales involved in the processes
66: which give bars an important vertical structure. These simulations show that the current scenario for bar formation
67: (i.e., via the dynamical bar mode instability that originates globally in kinematically cold disks)
68: is not able to explain naturally the existence of bars in lenticular
69: galaxies, which are kinematically hot stellar systems and have prominent bulges. While bars develop
70: conspicuously in our $N$-body realizations of late-type spirals, the bar mode instability is suppressed in
71: the realizations of galaxies with important bulges and kinematically hot disks. We thus suggest a new
72: mechanism for bar formation in galaxies, which is based on the accommodation of the stellar orbits
73: within a triaxial, or prolate, dark matter halo, which is sufficiently eccentric (D. A. Gadotti \& R. E. de Souza
74: 2003, \apjl, 583, 75). As we showed with $N$-body simulations which consist of only a spheroidal within
75: such a halo, this mechanism rapidly produces perennial bars in hot stellar systems, and may be an
76: explanation for the existence of barred galaxies in which the disk component is almost absent,
77: a discovery which is also a product of this work (see below). Furthermore, this mechanism may
78: account for the serious drawbacks confronted in the current scenario when applied to barred early-type
79: spirals and lenticulars, as the triaxial halo may help in the onset of the bar mode instability in otherwise stable disks.
80: We have also verified that a typical timescale for the thickening of bars, which cause the boxy/peanut morphology,
81: is of the order of 1 Gyr. However, the values for the bar $\sigma_z$ in our simulations after 1 Gyr are too low ($<$ 50 Km/s)
82: when compared to the ones we observed in evolved bars ($\sim$ 100 Km/s). Thus, we suggest that another process
83: may be playing a relevant role after the vertical resonances and/or the hose instability to make $\sigma_z$ as
84: high as we observe in evolved bars. We have theoretically verified that the Spitzer-Schwarzschild mechanism,
85: originally proposed for the Galaxy (see L. Spitzer \& M. Schwarzschild 1951, \apj, 114, 385),
86: is quantitatively able to explain these observations if we assume that giant molecular clouds are twice as much
87: concentrated along the bar as in the remaining of the disk. The timescales involved in this process
88: should be $\sim 5-10$ Gyr, which is in agreement with the age differences between young and evolved
89: bars we have estimated with optical colors (see below).
90:
91: Thirdly, using images obtained in $B$, $V$, $R$, $I$ and $K\!s$ for a sample of 19 galaxies, and images obtained
92: in $R$ for a sample of 51 galaxies, including elliptical galaxies, we have performed a detailed structural analysis in galaxies
93: covering the whole Hubble morphological sequence. The optical imaging was done with the 0.6 m telescope at the Pico
94: dos Dias Observatory, Brazil, and with the 1.5 m Steward Observatory telescope on Mount Bigelow, Arizona.
95: The infrared imaging was done with the 2.3 m Steward Observatory telescope on Kitt Peak, Arizona.
96: We were then able to determine the structural parameters which better describe the bulge and
97: the disk in these galaxies, as well as obtaining residual images which reveal important sub-structures. In order to perform
98: this analysis we have developed a specific algorithm, and built an atlas of structural analysis in galaxies.
99: This algorithm, named {\sc budda} (which stands for BUlge/Disk Decomposition Analysis\footnote{See
100: http://www.astro.iag.usp.br/$\sim$dimitri/budda.html for the code's internet site.}) performs a 2D
101: bulge/disk decomposition on galaxy images, assuming a S\'ersic bulge and an exponential disk.
102: This part of the work revealed the existence of barred lenticular galaxies whose disk component is
103: negligible (NGC 4608, NGC 5701 and possibly NGC 2217), as well as a difference in the optical colors
104: of young and evolved bars which amounts to $B-V=0.4$ mag. According to our knowledge on the
105: evolution of stellar populations, this difference may be translated to a difference in age of the order of 10 Gyr.
106: This result mean that bars may be, at least in some cases, a long standing structure. We have also found
107: a correlation between the bulge S\'ersic index $n$ and the galaxy $B-I$ color gradient, in the sense that
108: bulges with a more concentrated distribution of mass (i.e., with higher values for $n$) have flatter color gradients.
109: As shown by D. A. Gadotti \& S. dos Anjos (2001, \aj, 122, 1298), galaxies with flat color gradients are likely
110: to be the ones in which the bar secular evolution processes related to bulge building are relevant. Thus, the
111: correlation found corroborates the scenario in which bulges in lenticular and in spiral galaxies are, at least
112: partially, formed through the secular evolutionary processes in bars. Moreover, our results showed that
113: longer bars appear exclusively in galaxies with faint disks, that there is a correlation between the color
114: of the bar and the central disk intensity (in the sense that redder bars are in galaxies with fainter disks), and
115: that evolved bars are longer than young bars, in average. Altogether, these results seem to indicate that,
116: during its evolution, a bar grows stronger by capturing stars from the disk, which becomes fainter, in agreement with
117: recent numerical and analytical results (E. Athanassoula 2003, \mnras, 341, 1179). Thus, another possibility to
118: explain the existence of barred galaxies with almost no disk is that their bars may have evolved
119: so strongly by consuming their disks that there is almost no disk left.
120:
121: Finally, our structural analysis also showed that around 1/3 of the elliptical galaxies harbor inner disks, which may be
122: the result of recent mergers, and that visual morphological classification of galaxies is wrong in nearly 1/3 of the
123: elliptical and lenticular galaxies, being equally easy to misclassify an elliptical as a lenticular and the other way around.
124:
125: This work was supported by FAPESP grant 99/07492-7.
126:
127: \figcaption[fg1.eps]{Measured values for $\sigma_z$ along the bar major axis of 4 galaxies in our sample. The
128: two panels at left are examples of recently formed bars while the ones at right are examples of evolved bars.
129: In NGC 5383, the high values of $\sigma_z$ in the inner region correspond to its bulge, while the drops at each
130: side of the center are caused by its inner spiral arms. In NGC 5850, an inner bar identified in the {\sc budda}
131: residual images is the likely reason for the noticeable drop in the center.}
132:
133: \newpage
134:
135: \plotone{fg1.ps}
136:
137: \end{document}
138: