1: \documentclass[11pt,twoside]{article}
2: % version 10/1/07
3:
4:
5: \usepackage{asp2006}
6: \usepackage{epsf}
7: %\usepackage{psfig}
8: \usepackage{lscape}
9: \usepackage{graphicx}
10:
11: \markboth{Isaac Shlosman}{Disk-Halo Interplay}
12:
13: \pagestyle{myheadings}
14: \setcounter{equation}{0}
15: \setcounter{figure}{0}
16: \setcounter{footnote}{0}
17: \setcounter{section}{0}
18: \setcounter{table}{0}
19:
20:
21:
22: \begin{document}
23: \title{Disk-Halo Interplay in Galaxy Evolution}
24: \author{Isaac Shlosman}
25: \affil{Department of Physics \& Astronomy, University of Kentucky,
26: Lexington, KY 40506, USA; email: shlosman@pa.uky.edu}
27:
28: \begin{abstract}
29: Some aspects of disk-halo interactions for models of in and out of equilibrium disk galaxies
30: are reviewed. Specifically, we focus on disk-halo resonant interaction without and in the presence
31: of a gas component. Another issue is the disk growth within an assembling triaxial dark matter
32: halo. We argue that while the triaxiality is the result of the merger process and the radial
33: orbit instability, it is the developing chaos that damps the first generation of bars and
34: washes out the halo prolateness. This chaos is triggered by the gravitational quadrupole
35: interaction(s) in the system and supported by a number of other processes which are characteristic
36: of baryons.
37: \end{abstract}
38:
39:
40:
41: \section{Introduction}
42:
43: The current paradigm of galaxy formation necessitates that galactic disks grow and evolve
44: within dark matter (DM) halos. As such, the baryonic disks serve as a test bed
45: for studying the DM properties, its dynamics and morphology, as well as mass,
46: momentum and energy exchange within the disk-halo system. Interaction between the galactic
47: disks and DM halos can in principle involve an exchange of these quantities.
48: For example, observations of the low angular momentum H\,I gas, deep in the halo of NGC~891,
49: require
50: cold extragalactic gas influx (e.g., Fraternali et al. 2007), maybe through filaments
51: (Dekel \& Birnboim 2006). On the
52: other hand, the X-ray halos of starbursts and some normal galaxies can be explained
53: by the supernovae-heated gas driven from the disk or by the shock-compressed gas in the
54: halo (e.g., Strickland et al. 2004).
55:
56: The situation is much more straightforward with the angular momentum ($J$) flow in the
57: system. Galactic disks are rotationally supported, while the DM halos have low $J$, and,
58: therefore, a low spin parameter $\lambda$ (e.g., Barnes \& Efstathiou 1987; Frenk, White
59: \& Efstathiou 1988). Most of the disk galaxies are barred,
60: especially in the NIR (e.g., Knapen, Shlosman \& Peletier 2000), with bar properties which
61: remain steady
62: up to the redshift of $\sim 1$ at least (Jogee et al. 2004; Elmegreen, Elmegreen \& Hirst
63: 2004), and
64: except for the very early Hubble types, they exhibit a spiral structure. This prevailing
65: {\it disk asymmetry} is extremely important for enhancing the $J$ transfer
66: between different morphological components --- the alternative way can be achieved, e.g., by
67: dynamical friction of baryons against the DM. The current understanding of galactic bar
68: formation relies on the spontaneous breakup of the axial symmetry in the disk, the so-called
69: classical bar instability (e.g., Hohl 1971), originally applied to isolated disks.
70: Paradoxically, while Lynden-Bell \& Kalnajs (1972) have shown that bars and
71: spirals facilitate the $J$ transfer, DM halos have been considered
72: as the main stabilizers against the bar instability for years (Ostriker \& Peebles 1973),
73: though they appear rather to be sinks of the disk momentum (e.g., Athanassoula \&
74: Misiriotis 2002).
75:
76: Within the framework of the classical bar instability, a stellar bar can develop only if
77: the unstable region loses its $J$. Gravitational torques serve as the mechanism that
78: drives this
79: process (Lynden-Bell \& Pringle 1974). Their action can be described in terms of a non-local
80: viscosity (Lin \& Pringle 1987; Shlosman 1991) which shortens the characteristic timescale
81: of the $J$
82: transfer dramatically, making this a dynamical rather then secular process. In principle,
83: $J$ can flow across the corotation radius (CR) to the outer disk, to the DM halo, or to
84: a flyby galaxy in case of a tidal galaxy interaction. The first option limits the $J$
85: flow because the mass of the outer disk is typically $\sim 20\%$ of the disk mass and its
86: ability to absorb $J$ quickly saturates (Fig.~1, left). On the other hand, the halo is
87: massive, its inner part has a comparable mass to the inner disk, and its overall $J$ is
88: small --- the halo is not supported by rotation. Fig.~1 shows also that the outer
89: halo, beyond the disk radius, is fully susceptible to the $J$ transfer, after the inner
90: halo efficiency has decreased.
91:
92: \begin{figure}[!t]
93: \plotfiddle{fig01a.ps}{3.5cm}{-90}{45}{45}{-170}{140}
94: \plotfiddle{fig01b.ps}{3.5cm}{0}{31}{31}{-10}{-16.5}
95: \caption{
96: \underline{\it Left:} Evolution of angular momentum, $J$, in the disk (upper panel) and
97: the halo (lower panel). Note, $J$ saturates in the disk outside the bar CR (from
98: Martinez-Valpuesta, Shlosman \& Heller 2006). \underline{\it Right:} The $m=2$ amplitude,
99: $A_2$, of the stellar and DM ghost bars (upper panel) evolution. The fast dissolution of
100: the ghost bar (lower) after its stellar counterpart is axisymmetrized. Note the very
101: short time period for the lower frame (I. Berentzen \& I. Shlosman, unpublished)}
102: \label{fig1}
103: \end{figure}
104:
105: The relatively local density response in the DM halo to the growing bar in the disk is
106: to generate
107: a shadow `ghost' bar in the DM (Athanassoula 2006, 2007; Berentzen \& Shlosman 2006).
108: The ghost has the pattern speed of the stellar bar but its mass distribution differs
109: and is more centrally concentrated. The particle orbits in the DM ghost, therefore, are
110: in resonance with the stellar orbits. The principal difference between these two bars
111: is that the DM bar is not strictly speaking a bar at all --- it is not self-gravitating
112: {\it per se} and represents a gravitational wake induced in the DM by its stellar
113: counterpart. When the stellar bar is axisymmetrized, the DM bar dissolves in a fraction
114: of a crossing time (Fig.~1, right).
115:
116: \section{Angular momentum transfer in collisionless disk-halo systems}
117:
118: The drain of $J$ from the inner disk appears to proceed selectively --- some of the orbits
119: lose much
120: more than their neighbors (Athanassoula 2004; Martinez-Valpuesta, Shlosman \& Heller 2006). Fig.~2
121: displays this phenomenon --- stellar/DM particles populating
122: orbits with specific
123: frequencies $\nu\equiv (\Omega-\Omega_{\rm b})/\kappa$ which correspond to lower resonances,
124: 0, $\pm$ 1:2, 1:3, 1:4, etc. dominate the $J$ exchange. Here
125: $\Omega$ is the angular velocity, $\Omega_{\rm b}$ is the bar pattern speed, and $\kappa$ is
126: the radial epicyclic frequency. More precisely, the particles which are more affected are
127: those ultimately trapped by the above resonances. The dominant resonance in the disk
128: which emits $J$ is the inner Lindblad resonance (ILR), and the dominant resonance in the
129: halo which absorbs it is the CR. So the direction of the $J$ transfer is confirmed to be
130: from the inner disk to the halo. Fig.~10 of Martinez-Valpuesta et al.
131: (2006) displays a `forest' of lower resonances in the halo outside the CR which
132: appear to be active in absorbing $J$ from the disk.
133:
134: \begin{figure}[!ht]
135: \begin{center}
136: \includegraphics[angle=0,scale=0.65]{fig02.ps}
137: \end{center}
138: \caption{Resonances and $J$ exchange. {\it Upper panels:} disk or halo mass per unit
139: normalized frequency $\nu\equiv (\Omega-\Omega_{\rm b})/\kappa$ as a function of $\nu$.
140: {\it Lower panels:} $\Delta J$
141: by the particles as a function of $\nu$. {\it Left panels:} disk, right panels: DM halo.
142: The vertical dash-dotted lines give the positions of the main resonances (ILR $\rightarrow$
143: 0.5, CR $\rightarrow$ 0, OLR $\rightarrow$ -0.5, etc). From Athanassoula (2004).
144: }\label{fig2}
145: \end{figure}
146:
147: The bar growth is closely related to the $J$ loss from the disk within the CR. All
148: orbits within the bar are near-resonant, i.e., trapped by the resonances, mostly by the
149: ILR. As the resonances sweep across the phase and configuration spaces, because the bar
150: slows down, disk particles can lose their $J$ abruptly, being added (mostly) to the outer bar
151: in this process (Martinez-Valpuesta 2006). Most of the untrapped particles by the bar
152: can be found between its end and the CR. Bars that do not slow down, also do not grow
153: in length. For the trapped particles, much of the $J$ evolution has ceased.
154:
155: While Athanassoula (2004) and Martinez-Valpuesta et al. (2006) have frozen the
156: gravitational potential in order to integrate the particle orbits, Ceverino \& Klypin (2007)
157: used live potential for the relatively short time period of 1~Gyr. It is
158: not clear whether the latter method provides any advantage because the short integration
159: time results in broader and less defined resonances.
160:
161: Different halos are expected to exhibit various efficiencies in their resonance interaction
162: with the disk. So far, we are aware of two factors which govern this process --- the
163: velocity dispersion in the halo and its mass density near the lower resonances (Athanassoula
164: 2003), however, additional dependencies are expected.
165:
166: Cosmological simulations of DM structure show that the halos tend to acquire a universal
167: density profile, $\rho(r)$, which is characterized by a central cusp with $\rho(r) \sim r^{-1}$
168: (e.g., Navarro, Frenk \& White 1997, hereafter NFW). This trend is challenged by
169: observations of galactic rotation curves that typically indicate the presence of a central
170: flattening, i.e.,
171: core, instead of the cusp (e.g., Bosma 2004), and by a theoretical estimate of dynamical friction
172: of a stellar bar against the DM background, which requires a density core to explain why
173: bars are still spinning (e.g., Sellwood 2006, 2007). But what is the possible cause of the DM core?
174:
175: A number of alternatives have been explored for this purpose. Weinberg \& Katz (2007)
176: have advanced the idea that stellar bars transfer $J$ to the DM cusp and flatten it.
177: However, the usage of an analytical bar and other conditions which prevent the system to
178: respond in a self-consistent way, make their arguments unconvincing, especially the need
179: for a large number of particles, $N > 10^8$, to resolve the bar-cusp interaction (e.g.,
180: Sellwood 2007). Live bars do not dissolve the DM cusps in the axisymmetric halos even with
181: effectively $N\sim 10^{10}$ (Dubinski,
182: Shlosman \& Berentzen, in preparation), but see Holley-Bockelman, Weinberg \& Katz (2005).
183: A possibility that the supernova feedback injected energy flattens the
184: cusp has been proposed (Mashchenko, Couchman \& Wadsley 2006).
185: The other option is that the DM cusp has
186: been flattened prior or during the disk formation, when massive baryonic clumps descended
187: to the center via dynamical friction (El-Zant, Shlosman \& Hoffman 2001; Tonini, Lapi \&
188: Salucci 2006). It should be
189: pointed out that the cusp existence requires a `temperature' inversion, when solving
190: the Jeans equations without the central velocity anisotropy. Such an anisotropy was
191: analyzed by Dehnen \& McLaughlin (2005). In the absence of the DM velocity anisotropy,
192: the cusp is thermodynamically unstable, and only the lack of an
193: energy flow in the collisionless matter allows for its existence --- a condition which
194: is resolved by the dynamical friction. Similarly, the DM cusps in galaxy clusters can be
195: erased by massive baryonic clumps --- those are played out by individual galaxies (El-Zant
196: et al. 2004).
197:
198: \section{Angular momentum transfer in the presence of gas}
199:
200: While the $J$ redistribution in the collisionless disk-halo system has many additional caveats,
201: it is interesting to explore the effects of an additional component --- the gas. Currently,
202: the gas fraction in disk galaxies probably does not exceed 10\%, but the gas is known
203: to influence the galaxy evolution well beyond its fractional mass. Moreover, the gas fraction
204: in disks at redshifts above unity can be substantially higher, left alone without the
205: dissipative effects in the gas the disk would not form in the first place.
206:
207: \begin{figure}[!t]
208: \plotfiddle{fig03a.ps}{1.7cm}{0}{106}{106}{-245}{-393}
209: \plotfiddle{fig03b.ps}{1.7cm}{-90}{33}{33}{-20}{123}
210: \caption{\underline{\it Left:} Evolution of the stellar bar ($m=2$ mode amplitude $A_2$) in
211: models with $f_{\rm g}=0\%$ (solid line) and 8\% (dashed) over the first 5~Gyr.
212: \underline{\it Right:} Evolution of $f_{\rm g}=8\%$ model (solid lane) and the test models with
213: gradually subtracted gravitational torques from the gas onto the stars --- 25\% subtracted
214: (dotted), 50\% (dash-dotted), 75\% (dashed) and 100\% (dash-dot-dot-dotted). From Berentzen
215: et al. (2007).
216: }\label{fig3}
217: \end{figure}
218:
219: Fig.~3 (left panel) displays the disk-halo evolution in the presence of various gas fractions,
220: $f_{\rm g}\sim 0\%-8\%$ (Berentzen et al. 2007). Similar bars develop in all models and experience a
221: vertical buckling that `heats' up the stellar `fluid' along the rotation axis, weakening
222: the bar. Although $f_{\rm g}$ differs dramatically among these models, the
223: evolution of the bar strength, $A_2$, shows the same
224: maximum and nearly identical decrease in $A_2$. This decrease has been mistakenly
225: attributed to the $J$ transfer from the gas to the stars in the
226: bar by Bournaud, Combes \& Semelin (2005). On the other hand, the extent of the plateau
227: in $A_2$ depends strongly on $f_{\rm g}$ and clearly anti-correlates with it.
228:
229: \begin{figure}[!t]
230: \plotfiddle{fig04a.ps}{1.7cm}{-90}{32}{26}{-200}{81}
231: \plotfiddle{fig04b.ps}{1.7cm}{0}{28}{23}{15}{-46}
232: \caption{\underline{\it Left:} Change in the angular momentum, $\Delta J$, within
233: the bar CR of the 8\% gas model shown in the left frame of Fig.~3. The solid line shows
234: $J$ lost by the stars, calculated directly from the model evolution. The
235: thin dashed line shows $J$ acquired by the stars from the gas, calculated from
236: integration of the corresponding gravitational torques in the simulation (it was shifted
237: down in $y$ to match the solid line), and vice versa, from the stars to the gas (dotted
238: line). The thick dashed
239: line provides the $J$ flow in the gas component, obtained from the model evolution
240: (from Berentzen et al. 2007). \underline{\it Right:} Position angle (in radians) of the
241: major axis of the DM halo in a cosmological simulation of a disk/halo formation in the
242: $\Lambda$CDM Universe with the WMAP3 parameters (from Romano-Diaz et al., in prep.).
243: }\label{fig3}
244: \end{figure}
245:
246: To show explicitly that the $J$ transfer from the gas does not alter the bar evolution
247: over the first 5~Gyr, Berentzen et al. (2007) have run a number of test models,
248: specifically removing the gravitational torques from the gas onto the stars and DM
249: (Fig.~3, right panel). The resulting $A_2$ curves are very similar. This test has been
250: supplemented by the direct analysis of $J$ redistribution in the disk halo system.
251: The inner (inside the CR) disk loses its $J$ to the outer disk and the DM halo.
252: The inner halo absorbs the disk $J$ and forms the ghost bar.
253: Fig.~4 (left frame) shows that the gas inputs into the stars $\sim 13\%$ of $J$ lost
254: by the disk stars to the halo. The overall $J$ balance in the system is hardly affected
255: by the gas presence.
256: Surprisingly, {\it the halo is capable of absorbing $J$ from the disk on a relatively
257: short timescale}. This timescale is defined by the gas, while the bar evolution
258: is not.
259: Bournaud et al. (2005) concluded that the input of $J$ by the gas torques contributes
260: to the bar dissolution --- the halo used in this work was frozen and, therefore, not
261: allowed to participate in the overall $J$ exchange. Note that the sharp decrease in
262: $A_2$ of the bar is not related to the angular momentum at all --- it results from
263: the buckling instability.
264:
265: \section{DM halo shapes and disk morphology: the source of chaotic dynamics}
266:
267: The angular momentum redistribution in the disk-halo system is facilitated by the
268: departure from axial symmetry. It is of paramount importance, therefore, that
269: cosmological simulations universally lead to the formation of triaxial DM halos, i.e.,
270: both prolate and flattened (e.g., Bullock 2002; Allgood et al. 2006). Hence we
271: can expect high-$z$ halos to follow this trend. On the other hand, observations
272: indicate that halos in the nearby Universe are rather oblate and somewhat flattened,
273: based on a number of arguments, e.g., the coherence of the Sagittarius tidal stream
274: (Ibata et al. 2001), polar rings (Sparke 1986; although their shape can be affected
275: by their self-gravity), diffuse X-ray emission around ellipticals (Buote et al. 2002),
276: H\,I isophotes of galactic disks (Merrifield 2002), weak lensing (Hoekstra, Yee \& Gladders
277: 2004), etc. The main issue, therefore, is {\it how the prolate DM halos at high redshifts
278: have been transformed into oblate ones in the local galaxies}. Both mass, angular
279: momentum and energy exchange between the disk and the halo are expected to contribute
280: to this process.
281:
282: While triaxial collisionless systems are not exactly in equilibrium and are expected
283: to evolve secularly, the
284: addition of baryons accelerates this process dramatically and reduces the asymmetry
285: (Dubinski 1994; Kazantzidis et al. 2004). However, these works did not
286: address the reason(s) for the shape evolution.
287:
288: Why do baryons reduce the halo triaxiality so dramatically and how is this effect related
289: to the internal dynamics of the system? Even in
290: pure DM halos, within the hierarchical merging framework, rich substructure forms
291: in the $\Lambda$CDM cosmology. These DM clumps join the main halo mostly along filaments.
292: The `thermalization' of such streamers and the accompanied dynamical friction that the clumps
293: experience when deep inside the main halo are themselves
294: irreversible processes that randomize the DM trajectories. The clumps also transfer $J$ to
295: the smooth DM component --- again changing the DM orbit shapes (El-Zant et al. 2001, 2004;
296: Tonini et al. 2006). The resulting inhomogeneities
297: inside the halo serve as multiple isotropic scattering centers. This process, while being
298: inherently related to the disk formation, in fact, precedes it. In elliptical galaxies it
299: may dominate the reshaping of the DM figure. Baryons only enhance this
300: trend. Dissipation leads to their accumulation in the inner halo, thus increasing the
301: baryon-to-DM mass ratio there and making them more dynamically important. If the baryons
302: are allowed to collapse to a disk, they become rotationally supported. But within
303: the prolate background potential, the disk acquires a substantial quadrupole moment,
304: as axisymmetric trajectories do not exist under these conditions.
305:
306: A new and interesting phenomenon occurs because the halo figure tumbles slowly while
307: the baryons in the disk are rotationally supported (Shlosman 2007; Heller, Shlosman \&
308: Athanassoula 2007b; Romano-Diaz et al., in prep.). Under these conditions, the inner
309: ILR (if it exists) approaches the center, while the outer ILR moves to large radii outside
310: the halo. This assures that the baryons in the disk respond out of phase with the DM
311: potential --- the required response of trajectories between the ILRs.
312: Support for the halo potential shape weakens and its prolateness is diluted.
313: {\it The near absence of halo figure rotation has important consequences
314: for the dynamics of galactic disks.}
315:
316: \begin{figure}[!t]
317: \begin{center}
318: \includegraphics[angle=0,scale=0.5]{fig05a.ps}
319: \includegraphics[angle=0,scale=0.503]{fig05b.ps}
320: \includegraphics[angle=0,scale=0.53]{fig05c.ps}
321: \end{center}
322: \caption{Comparison of developing chaos in various systems given by gray-scale
323: diagrams showing the maximal Liapunov exponents at 50~Gyr. The x-axis gives the
324: cylindrical distance from the rotation axis and extends to 6~kpc. The y-axis gives
325: the particle velocities normalized by the local circular velocity, $v/v_{\rm c}$, and
326: extends to $v/v_{\rm c}=1.2$. The frames display
327: the Liapunov exponents for the particle
328: trajectories in the system of \underline{\it Left:} a pure triaxial DM halo with a
329: core radius of 5~kpc and an equatorial potential ellipticity of 0.1.
330: \underline{\it Middle:} an axisymmetric halo with a core radius of 5~kpc, and with
331: a stellar bar of $\sim 5$~kpc. \underline{\it Right:} triaxial halo with a core radius
332: of 5~kpc and the equatorial potential ellipticity of 0.1, and with a 5~kpc stellar bar.
333: The gray-scale is logarithmic with the white background corresponding to a characteristic
334: time of $10^6$~yr or less, and the darkest background --- to $10^{10}$~yr or more
335: (from El-Zant \& Shlosman 2002).
336: }\label{fig5}
337: \end{figure}
338:
339: Even more important is the specifics of the baryon response to the halo prolateness ---
340: large-scale shocks induced in the growing and gas-rich disk by the gravitational torques
341: --- which act as a finite perturbation from the halo onto the disk. As a result, the
342: disk will be strongly asymmetric as long as the halo maintains its equatorial
343: prolateness within the disk region. The halo and disk quadrupoles will acquire different
344: pattern speeds (Heller, Shlosman \& Athanassoula 2007a,b). Their mutual interaction
345: will lead to chaos and dissolution of the least massive quadrupole, whether this is
346: a spontaneous or induced bar (El-Zant \& Shlosman 2002). The latter work found that the
347: presence of a large-scale
348: stellar bar in the disk immersed in a mildly prolate halo triggers a chaotic behavior in the least
349: massive component --- the bar. Using the Liapunov exponents as a measure of chaos in the
350: bar, El-Zant \& Shlosman found that the stellar trajectories quickly become chaotic, dissolving the
351: bar. When the slowly tumbling nonaxisymmetric perturbation in the {\it potential} has
352: reached 10\%, almost all trajectories integrated in time were found to be chaotic and have large
353: Liapunov exponents. More centrally concentrated halos appear to be more dominated by chaos.
354:
355: Fig.~5 quantifies the developing chaos in terms of the Liapunov exponents in three specific
356: cases of 5~kpc flat density core DM halos. Interestingly, a pure DM triaxial halo, with no
357: bar, is nearly
358: regular (Fig.~5, left panel), and no chaos is associated with this potential over the time period
359: exceeding the Hubble time. An axisymmetric halo which hosts a bar exhibits a healthy amount
360: of chaos around the stellar (and ghost) bar CR, as expected (Fig.~5, middle panel). It is the
361: addition of the
362: fast tumbling bar within a prolate halo that generates {\it continuous} chaos within the CR
363: and at most of the velocities (Fig.~5, right panel). The
364: characteristic timescale for the trajectories divergence is much less than the Hubble time.
365: The halo core potential can then be approximated as a quadratic form where the motion is separable
366: in Cartesian coordinates. Centrally concentrated
367: potentials do not have such symmetries --- the oscillations in the different degrees of freedom
368: are, in general, coupled, and no global integrals of motion exist. This makes plausible a
369: situation where most trajectories in the central regions conserve only energy, in which case
370: self-consistent solutions become impossible.
371:
372: Galaxies that are not
373: initially centrally concentrated may acquire a central mass concentration by accreting gas
374: via spontaneous or tidally-induced bars. The resulting coupling between the degrees
375: of freedom will destroy the integrals of motion for a large enough fraction of orbits and
376: dissolve the bar, which is replaced by e.g., a central mass concentration (CMC). This scenario
377: requires massive CMCs,
378: e.g.., supermassive black holes (SBHs) in excess of $10^9~{\rm M_\odot}$, which are not confirmed by observations
379: (El-Zant et al. 2003).
380:
381: Heller et al. (2007b) have compared the {\it shape} evolution
382: of an assembling DM halo with and without the baryons. Fig.~6 (left panel) shows the time evolution
383: of the pure DM shape from 2~kpc to 300~kpc. Fig.~6 (right panel) exhibits
384: the same evolution when the disk grows by converting the gas to stars during the collapse
385: of an isolated density perturbation embedded in the Hubble flow.
386: The halo triaxiality is reasonably stable with
387: time for the pure DM models at all radii --- only a mild secular decrease was found
388: over the Hubble time. Models with baryons show a different behavior within
389: and outside the disk radius (roughly 15~kpc). The inner halo prolateness is washed out
390: by the time a disk has formed, $\sim 2-3$~Gyr.
391: The outer halo prolateness and flatness go through an adjustment phase and then
392: stabilize at some value, remaining substantial.
393:
394: \begin{figure}[!t]
395: \plotfiddle{fig06a.ps}{1.9cm}{-90}{27}{27}{-215}{85}
396: \plotfiddle{fig06b.ps}{1.9cm}{-90}{27}{27}{-20}{150}
397: \caption{Evolution of DM halo intermediate-to-major axial ratio
398: $b/a$ for a pure DM model (left panel) and a DM with baryons model (right panel). The curves
399: correspond to 2~kpc (thick solid line), 5~kpc (dashed), 10~kpc (dot-dashed), 50~kpc
400: (thin solid), 100~kpc (dotted) and 300~kpc (dash-dot-dot-dotted). From Heller et al.
401: (2007b).
402: }\label{fig6}
403: \end{figure}
404:
405: Berentzen, Shlosman \& Jogee (2006) have verified in live $N$-body simulations that the interaction of
406: two or more quadrupoles leads to chaotic dynamics both in the disk and in the inner halo.
407: A typical decay time of stellar bars and disk ovalness was found
408: to lie in the range of a few Gyr, with a caveat that the halo itself suffers a reduction of its
409: quadrupole in the feedback, and, in extreme cases, the halo prolateness can be damped
410: ahead of that in the disk. More cuspy halos have been losing their prolateness within the radius
411: faster, allowing the bar to survive. This effect was further quantified in simulations with
412: a tenfold growing stellar
413: disk embedded in the halo which assembles under cosmological initial conditions
414: within a 4~Mpc computational box (Berentzen \& Shlosman 2006).
415: More massive disks have erased the halo prolateness faster and more efficiently ---
416: hence, maximal disks are particularly apt to axisymmetrizing the background halos.
417: Moreover, the loss of prolateness by the halo allowed the maximal disk to develop
418: a next generation bar via classical bar instability, albeit at much later times. On the other
419: hand, disks dominated by the DM at all radii will have a much smaller effect,
420: if at all, on the shapes of the inner halo.
421:
422: The emerging conclusion from the above works underlines the importance of self-consistent
423: models in understanding the
424: DM halo shape evolution. If the disk is not allowed to generate substantial
425: quadrupole(s), or, in simple terms, is prevented from responding to the halo prolateness,
426: the basic ingredients for its feedback onto the halo are virtually eliminated.
427: There are three basic ways the disk responds to the halo: (1) its orbits are out-of-phase
428: with the halo potential (the disk is elongated normal to the halo major axis),
429: (2) halo-induced and spontaneous stellar bars develop and tumble fast, and
430: (3) extensive stellar arms in the outer disk are triggered by the halo shape.
431: Recent interesting work by Debattista et al. (2007) has reached a different conclusion,
432: based on
433: simulations with a {\it frozen} disk which, therefore, cannot respond in any of the
434: above fashions. Furthermore, the disk is grown adiabatically, thus only reversible
435: processes are induced in the halo, making its orbits rounder. In fact, there is no basic
436: difference between slowly growing a rigid disk or a Plummer sphere within a responsive
437: halo --- the latter experiment was performed by Dubinski (1994).
438: It is natural that Debattista et al. find that the chaos plays no role in reshaping the
439: halo, and that only the (DM) substructure has an irreversible effect on the
440: halo shape.
441:
442: To summarize, {\it the origin of the halo triaxiality lies to a certain degree in the major
443: merger process and
444: in the radial orbit instability, while the reduction of the halo triaxiality is
445: governed by developing chaos and transfer of the angular momentum to the DM}.
446: These processes can be facilitated through
447: the basic mechanism --- the separation of baryons from the DM that is driven by dissipation.
448: Specifically, (1) the cooling baryons enhance the substructure in the overall mass
449: distribution during
450: the halo assembly. The resulting clumps deposit energy and $J$ in the background DM
451: via dynamical friction --- this includes minor mergers and quiescent accretion and
452: is an irreversible process. (2) The ability of the baryons
453: to cool allows them to become more centrally concentrated (in the global and not only in
454: the local sense) --- the centrifugal barrier delays this process, by converting it
455: from a dynamical to a secular one. However, even the presence of this barrier does not assure
456: the dynamical-to-secular transition because the baryonic disk forming in the center will not
457: be axisymmetric in the presence of a prolate DM potential, and even if it will form
458: axisymmetrically, the classical bar instability will spontaneously break this symmetry.
459: (3) The resulting interaction of two or more quadrupoles which produce incommensurable
460: perturbing forces both on baryons and the DM will generate chaos in the disk and the
461: triaxial halo and act to reduce the triaxiality. (4) The central cusp, whether baryonic
462: or DM, can induce chaos as well, although this is not assured in the most general
463: sense.
464:
465:
466: \section{Evolving Disk Morphology: Bars, Bulges and All That}
467:
468: The origin of galactic disks has emerged as one of the sticking issues in the overall
469: understanding of the galaxy formation process. Large gaps permeate our knowledge
470: of the origin of disk structural components, e.g., bars, bulges, SBHs, etc. The path to
471: the observed tight correlation between
472: the properties of the central SBH and the surrounding bulge is unknown as well,
473: although suggestions abound. Some of the components require a certain degree
474: of dissipation to accompany their formation, such as the disk itself, certain
475: types of bulges, and the SBH --- this erases the memory of the initial conditions
476: and the link to cosmology. In particular, the origin timescale for current disk morphology
477: and the amount of evolution that occured since is only known remotely. A number of controversies
478: have emerged: first, the angular momentum distribution of the observed (baryonic) disks
479: does not
480: match that of the disks obtained in cosmological numerical simulations --- the latter appear
481: to have radial scalelengths which are too small (e.g., Sommer-Larsen, Gotz \& Portinari
482: 2002) --- the so-called angular
483: momentum `catastrophe.' Moreover, the central bulges form too massive and there are
484: difficulties to explain the widespread bulgeless disks. Second, the numerical simulations
485: of DM halo formation show that they display a universal density profile characterized
486: by a steep central density cusp (e.g., NFW), while
487: observations point rather to a flat density core (e.g., de Blok \& Bosma 2002) ---
488: the core `catastrophe.' The central difficulty appears to be the conservation
489: of baryonic $J$ during the disk formation and major mergers events. The standard picture
490: of slow adiabatic collapse of a smooth baryonic component only aggravates the problem
491: by dragging the DM inward in an adiabatic compression.
492:
493: These and other discrepancies between observations and modeling cannot be fully
494: resolved by the current theory --- additional high-resolution numerical simulations
495: are required, focusing on the effect of various parameters of star formation, energy
496: feedback from the stellar evolution and growing SBHs. Indeed, recent numerical
497: modeling has
498: made progress in understanding the complexity of disk formation and evolution.
499: Models with {\it frozen} spherically-symmetric DM halos, e.g., by Samland \& Gerhard
500: (2003) and Immeli et al. (2004), have used a chemo-dynamical code to investigate the
501: bulge stellar population, as well as a possible bulge formation from massive
502: spiraling-in gas clumps, accompanied by a strong central starburst (see also Shlosman
503: \& Noguchi 1993). Okamoto et al. (2005) have assumed that the star formation proceeds
504: in two modes --- in a high density environment or shock-triggered, leading to
505: elliptical or disk galaxies, respectively. Hence, the star formation processes can
506: affect the final $J$ of the galaxy. Governato et al. (2004, 2007) have claimed to
507: produce disks with realistic exponential scalelengths, to fit the $I$-band and
508: baryonic Tully-Fisher relations, and to reduce the number of visible satellites
509: around the Milky Way-type halo, by combining the UV background and SN feedback.
510:
511: Heller et al. (2007a,b) have focused on the formation of the disk and its structural
512: components within the
513: context of the collapse of an isolated density perturbation embedded in the Hubble flow.
514: In agreement with Fig.~4b, $J$ was channeled into the internal circulation,
515: leaving the DM halo figure tumbling $\sim \pi$ radian in a Hubble time.
516: At $z=0$, a range of nearly bulgeless-to-bulge dominated disks were obtained, with
517: flat rotation curves which account for the observed disk/halo contributions.
518: A relatively large number of models was run to investigate the effect of star
519: formation and stellar evolution feedback onto the disk properties, normally unresolved
520: in cosmological simulations. Several single-varied
521: parameter evolutionary sequences have been constructed, such as varying
522: the energy thermalization parameter, which characterizes the feedback from stellar
523: evolution, the star formation threshold parameter, the collapse time for the star forming
524: clouds, and the gravitational softening in the gas. A number of correlations
525: along these sequences have been found, e.g., the thermalization parameter appears to
526: correlate with the bulge prominence, and a lower density threshold leads
527: to smaller, thicker disks.
528:
529: While the bar origin lies in the $J$ loss by the inner disk, exponential bulges appear
530: to be largely by-products of the bar evolution. The vertical buckling instability
531: in a bar leads to the formation of boxy/peanut-shaped bulges (e.g., Combes \& Sanders 1981;
532: Combes et al. 1990). Although the bar weakens as a result (Raha et al. 1991; Pfenniger
533: \& Friedli 1991) and shortens, in principle it cannot be destroyed (Martinez-Valpuesta
534: \& Shlosman 2004). Multiple bucklings of the bar add up to the growth of its bulge,
535: irrespective of the gas inflow (Martinez-Valpuesta et al. 2006).
536:
537: We have discussed earlier that the halo shape speeds up the $J$ exchange in the system.
538: This is especially pronounced in the early stages of disk growth, when its rotation
539: curve is probably dominated by the DM. The initial disk response to a finite perturbation
540: from the halo will naturally lead to a bar. The resulting mass cascade to the center,
541: e.g., in the form of nested bars (Shlosman, Frank \& Begelman 1989), can create the
542: environment for the formation and initial growth of a SBH (Begelman, Volonteri \&
543: Rees 2006).
544:
545: \section{Conclusions}
546:
547: We have summarized some aspects of disk-halo interactions in equilibrium and forming
548: disk galaxies. Because the overall problem is of a high complexity, progress in
549: this direction has been rather slow and dependent on numerical simulations. Nevertheless,
550: certain trends have emerged. The baryonic disk and DM halo exchange
551: mass, angular momentum and energy during the initial formation phase as well as
552: during the subsequent quiescent evolution. A number of processes, like star formation
553: as well as formation of the central SBH, require a much deeper understanding before
554: they can be modeled beyond the phenomenological approach. On the other hand, nonlinear
555: dynamics and the role of chaos in the evolution of disks and halos can be quantified
556: already at present.
557:
558: \acknowledgements %%% Text of acknowledgements runs on after this command.
559: I acknowledge illuminating discusssions with my colleagues, too numerous to list here.
560: I am grateful to my collaborators Ingo Berentzen, Clayton Heller, Yehuda Hoffman,
561: Inma Martinez-Valpuesta and Lia Athanassoula on these issues. The
562: relevant grants from NASA and NSF are acknowledged. I also thank the
563: conference organizers for generous support.
564:
565:
566: \begin{thebibliography}{}
567:
568: \bibitem[]{}Allgood, B., Flores, R.A., Primack, J.R., Kravtsov, A.V., Wechsler, R.H.,
569: Faltenbacher, A. \& Bullock, J.S. 2006, MNRAS, 367, 1781
570:
571: \bibitem[]{}Athanassoula, E. \& Misiriotis, A. 2002, MNRAS, 330, 35
572:
573: \bibitem[]{}Athanassoula, E. 2003, MNRAS, 341, 1179
574:
575: \bibitem[]{}Athanassoula, E. 2004, IAU Symp. 220, (eds.) S.D. Ryder, D.J. Pisano,
576: M.A. Walker, \& K.C. Freeman (San-Francisco: ASP), 255
577:
578: \bibitem[]{}Athanassoula, E. 2006, Nonlinear Dynamics in Astronomy and Physics,
579: eds. S. Gottesman et al., Annals New York Acad. Sci., 1045
580:
581: \bibitem[]{}Athanassoula, E. 2007, MNRAS, 377, 1569
582:
583: \bibitem[]{}Barnes, J. \& Efstathiou, G. 1987, ApJ, 319, 575
584:
585: \bibitem[]{}Begelman, M.C., Volonteri, M. \& Rees, M.J. 2006, MNRAS, 370, 289
586:
587: \bibitem[]{}Berentzen, I., Shlosman, I. \& Jogee, S. 2006, ApJ, 637, 582
588:
589: \bibitem[]{}Berentzen, I. \& Shlosman, I. 2006, ApJ, 648, 807
590:
591: \bibitem[]{}Berentzen, I., Shlosman, I., Martinez-Valpuesta, I. \& Heller, C.H. 2007,
592: ApJ, 667, September 20, in press (astroph/0703028)
593:
594: \bibitem[]{}Bosma, A. 2004, IAU Symp. 220 (eds.) S.D. Ryder, D.J. Pisano, M.A. Walker,
595: \& K.C. Freeman (San-Francisco: ASP), 39
596:
597: \bibitem[]{}Bournaud, F., Combes, F. \& Semelin, B. 2005, MNRAS, 364, L18
598:
599: \bibitem[]{}Bullock, J.S. 2002, The Shapes of Galaxies and Their Dark Halos, (ed.)
600: P. Natarajan (Singapore: World Scientific), 109
601:
602: \bibitem[]{}Buote, D.A., Jeltema, T.E., Canizares, C.R. \& Garmire, G.P. 2002, ApJ, 577, 183
603:
604: \bibitem[]{}Ceverino, D. \& Klypin, A. 2007, MNRAS, in press, astro-ph/0703544
605:
606: \bibitem[]{}Combes, F. \& Sanders, R.H. 1981, A\&A, 96, 164
607:
608: \bibitem[]{}Combes, F., Debbash, F., Friedli, D. \& Pfenniger, D. 1990, A\&A, 233, 92
609:
610: %\bibitem[]{}Combes, F. 2002, New Astron. Rev., 46, 755
611:
612:
613:
614: \bibitem[]{}Debattista, V.P., Moore, B., Quinn, T., Kazantzidis, S., Maas, R.,
615: Mayer, L., Read, J. \& Stadel, J. 2007, astro-ph/07070737
616:
617: \bibitem[]{}de Blok, W.J.G. \& Bosma, A. 2002, A\&A, 385, 816
618:
619: \bibitem[]{}Dehnen, W. \& McLaughlin, D. 2005, MNRAS, 363, 1057
620:
621: \bibitem[]{}Dekel, A. \& Birnboim, Y. 2006, MNRAS, 368, 2
622:
623: \bibitem[]{}Dubinski, J. 1994, ApJ, 431, 617
624:
625: \bibitem[]{}Elmegreen, B.G., Elmegreen, D.M. \& Hirst, A.C. 2004, ApJ, 612, 191
626:
627: \bibitem[]{}El-Zant, A., Shlosman, I. \& Hoffman, Y. 2001, ApJ, 560, 636
628:
629: \bibitem[]{}El-Zant, A. \& Shlosman, I. 2002, ApJ, 577, 626
630:
631: \bibitem[]{}El-Zant, A., Shlosman, I., Begelman, M.C. \& Frank, J. 2003, ApJ, 590, 641
632:
633: \bibitem[]{}El-Zant, A., Hoffman, Y., Primack, J., Combes, F. \& Shlosman, I. 2004, ApJ, 607,
634: L75
635:
636: \bibitem[]{}Fraternali, F., Binney, J., Oosterloo, T. \& Sancisi, R. 2007, New Astr., 51, 95
637:
638: \bibitem[]{}Frenk, C.S., White, S.D.M. \& Efstathiou, G. 1988, ApJ, 327, 507
639:
640: \bibitem[]{}Governato, F., Mayer, L., Wadsley, J., Gardner, J.P., Willman, B.,
641: Hayashi, E., Quinn, T., Stadel, J. \& Lake, G. 2004, ApJ, 607, 688
642:
643: \bibitem[]{}Governato, F., Willman, B., Mayer, L., Brooks, A., Stinson, G., Valenzuela, O.,
644: Wadsley, J. \& Quinn, T., 2007, MNRAS, 374, 1479
645:
646: \bibitem[]{}Heller, C.H., Shlosman, I. \& Athanassoula, E. 2007a, ApJ, 657, L65
647:
648: \bibitem[]{}Heller, C.H., Shlosman, I. \& Athanassoula, E. 2007b, ApJ, submitted
649: (astro-ph/07063895)
650:
651: \bibitem[]{}Hoekstra, H., Yee, H.K.C. \& Gladders, M.D. 2004, ApJ, 606, 67
652:
653: %\bibitem[]{}Hoffman, Y. \& Ribak, E. 1991, ApJ, 380, L5
654:
655: \bibitem[]{}Hohl, F. 1971, ApJ, 168, 343
656:
657: \bibitem[]{}Holley-Bockelmann, K., Weinberg, M. \& Katz, N. 2005, MNRAS, 363,991
658:
659: \bibitem[]{}Ibata, R., Irwin, M., Lewis, G.F. \& Stolte, A. 2001, ApJ, 547, L133
660:
661: \bibitem[]{}Immeli, A., Samland, M.O., Gerhard, O. \& Westera, O. 2004, A\&A, 413, 547
662:
663: \bibitem[]{}Jogee, S., Barazza, F.D., Rix, H.-W., Shlosman, I. et al. 2004, ApJ, 615, L105
664:
665: \bibitem[]{}Kazandzidis, S., Kravtsov, A.V., Zentner, A.R., Allgood, B., Nagai, D.
666: \& Moore, B. 2004, ApJ, 611, L73
667:
668: \bibitem[]{}Knapen, J.H., Shlosman, I. \& Peletier, R.F. 2000, ApJ, 529, 93
669:
670: \bibitem[]{}Lin, D.N.C. \& Pringle, J.E. 1987, MNRAS, 225, 607
671:
672: \bibitem[]{}Lynden-Bell, D. \& Kalnajs, A.G. 1972, MNRAS, 157, 1
673:
674: \bibitem[]{}Lynden-Bell, D. \& Pringle, J.E. 1974, MNRAS, 168, 603
675:
676: \bibitem[]{}Martinez-Valpuesta, I. \& Shlosman, I. 2004, ApJ, 613, L29
677:
678: \bibitem[]{}Matinez-Valpuesta, I. 2006, Ph.D. Thesis, University of Hertsfordshire
679:
680: \bibitem[]{}Martinez-Valpuesta, I., Shlosman, I. \& Heller, C.H. 2006, ApJ, 637, 214
681:
682: \bibitem[]{}Mashchenko, S., Couchman, H.M.P. \& Wadsley, J. 2006, Nature, 442, 7102
683:
684: \bibitem[]{}Merrifield, M.R. 2002, Shapes of Galaxies and Their Dark Matter Halos,
685: (ed.) P. Natarajan (Singapore: World Scientific), 170
686:
687: \bibitem[]{}Navarro, J.F., Frenk, C.S. \& White, S.D.M. 1997, ApJ, 490, 493 (NFW)
688:
689: \bibitem[]{}Okamoto, T., Eke, V.R., Frenk, C.S. \& Jenkins, A. 2005, MNRAS, 363, 1299
690:
691: \bibitem[]{}Ostriker, J.P. \& Peebles, P.J.E. 1973, ApJ, 186, 467
692:
693: %\bibitem[]{}Paczynski, B. 1978, Acta Astron., 28, 91
694:
695: \bibitem[]{}Pfenniger, D. \& Friedli, D. 1991, A\&A, 252, 75
696:
697: \bibitem[]{}Raha, N., Sellwood, J.A., James, R.A. \& Kahn, F.D. 1991, Nature, 352, 411
698:
699: %\bibitem[]{}Romano-Diaz, E., Hoffman, Y., Faltenbacher, A., Heller, C.H., Jones, D.
700: % \& Shlosman, I. 2006, ApJ, 637, L93
701:
702: %\bibitem[]{}Romano-Diaz, E., Hoffman, Y., Heller, C.H., Faltenbacher, A., Jones, D.
703: % \& Shlosman, I. 2007, ApJ, 657, 56
704:
705: \bibitem[]{}Samland, M. \& Gerhard, O. 2003, A\&A, 399, 961
706:
707: \bibitem[]{}Sellwood, J.A. 2006, astro-ph/0610468
708:
709: \bibitem[]{}Sellwood, J.A. 2007, astro-ph/07041047
710:
711: \bibitem[]{}Shlosman, I., Frank, J. \& Begelman, M.C. 1989, Nature 338, 45
712:
713: %\bibitem[]{}Shlosman, I., Begelman, M.C. \& Frank, J. 1990, Nature 345, 679
714:
715: \bibitem[]{}Shlosman, I. 1991, IAU Colloq. 124 on Paired and Interacting Galaxies,
716: J. Sulentic \& W. Keel, Eds. (Kluwer Acad. Publ.), 689
717:
718: \bibitem[]{}Shlosman, I. \& Noguchi, M. 1993, ApJ, 414, 473
719:
720: \bibitem[]{}Shlosman, I. 2007, IAU Symp. 235, Galaxy Evolution Across the Hubble
721: Time, F. Combes \& J. Palous (eds.), CUP, in press (astro-ph/0610727)
722:
723: \bibitem[]{}Sommer-Larsen, J., Gotz, M. \& Portinari, L. 2003, ApJ, 596, 47
724:
725: \bibitem[]{}Sparke, L.S. 1986, MNRAS, 219, 657
726:
727: \bibitem[]{}Strickland, D.K., Heckman, T.M., Colbert, E.J.M., Hoopes, C.G. \& Weaver, K.A.
728: 2004, ApJ, 606, 829
729:
730: \bibitem[]{}Tonini, C., Lapi, A. \& Salucci, P. 2006, ApJ, 649, 591
731:
732: %\bibitem[]{}Weinberg, M.D. \& Katz, N. 2002, ApJ, 580, 627
733:
734: \bibitem[]{}Weinberg, M.D. \& Katz, N. 2007, MNRAS, 375, 460
735:
736: \end{thebibliography}
737:
738: \end{document}
739: