\documentstyle[12pt]{article} \begin{document} \centerline{\bf {STRUCTURE AND STAR FORMATION}} \centerline{\bf {IN CIRCUMNUCLEAR REGIONS OF SPIRAL GALAXIES}} \bigskip \bigskip \begin{abstract} Photometric and kinematic properties of inner regions of spiral galaxies are considered. It is shown that the fast rotation of nuclear gaseous disks makes them gravitationally stable or marginally unstable. Their star formation may continue however even in the absence of large-scale gravitational instability. \end{abstract} \bigskip {\bf Introduction} \bigskip Inner parts of galaxies within about 1 kpc from their nuclei in many cases appears to be peculiar regions by their properties even if we restrict our attention by normal galaxies without strong nuclear activity. First, circumnuclear regions often reveal rather complex - regular or irregular - photometric structure which may be connected with general structure of a main disk. It includes spiral arms (examples: M100, NGC 4314), short nuclear bars or nuclear rings of resonance nature (see Buta and Crocker 1993) or hot spots (Morgan 1958) which seems to be sites of active star formation. Most of these features are evidently related to large-scale bars in SB- galaxies, although in some galaxies a presence of bar is not evident from optical observations . Second, in many galaxies (especially of early types Sa-Sbc), including our own Galaxy, the central regions accumulate interstellar gas into dense molecular disk of about 1kpc- size. These central islands (or sometimes rings or bar-like features) of molecular gas are sooner a rule than an exception for massive spiral galaxies. In many cases a dense circumnuclear gas produces strongly enhanced star formation which is observed as starburst nucleus. Note that the presence of a large-scale bar is proved to be an efficient (but not the only one) mechanism for driving interstellar medium into the nuclear region, to feed starburst or mild star formation within a few hundred parsecs from the center, and- -under favorable conditions such as the formation of nuclear bar- to turn on the action of active nucleus (see for example Friedli and Benz, 1993, Telesco et al., 1993, Schlosman et.al.,1989). Third, as optical and CO - observations have shown, in many normal galaxies the nuclear gaseous disk is distinguished by rapid rotation of gas and stars, being often dynamically decoupled from the rest of the galaxy (Afanasiev et al., 1989, Zasov and Sil'chenko, 1994, Rubin et al.,.1997, Sofue, 1997 and references therein). A detailed comparison of spectra of nuclear HII- regions with the HII- regions of main disks which was carried out by Kennicutt et.al., 1993, and Luis et.al., 1997, showed that they are similar in many respects: their luminosities, masses of ionized gas and internal color excesses are within the same ranges, but some properties are systematically different: as a rule, HII nuclei emit stronger low-ionization forbidden lines compared to disk HII regions, they also have lower equivalent widths of Balmer emission lines, which gives evidence of bright background light of the stellar population. Nuclear emission regions also have unusually low filling factor (around $10^{-5}$ (Luis et al., 1997). It is worth mentioning also that the Hubble Space Telescope observations revealed the presence of unusually bright and compact young (globular?) stellar clusters in the regions of active circumnuclear star formation (Barth et.al., 1995). All these peculiarities show that the conditions of star formation there may differ from what we observe in spiral arms of galaxies. In this paper I describe briefly some results of investigations of nuclear regions of spiral galaxies, related to their structure and star formation. \bigskip \bigskip {\bf Rotation of inner discs.} \bigskip For a typical distance to galaxies considered here D = 10 - 20 Mpc a nuclear disk with diameter of about 1 kpc is seen at the angle 10'' -20''. It is very difficult to get the velocity field or velocity curve so close to the center. Nevertheless even the long slit observations gave strict evidences of fast rotation of nuclear regions of 0.5 - 1 kpc radius in many galaxies (Afanasiev et al., 1989, Rubin et al.,1997). For the most reliable estimates of velocity gradients it is necessary to use several cuts passing through a nucleus or two- dimensional velocity field assuming solid-body rotation. Without using two- dimensional data it is often impossible to verify rotational nature of gas velocities. As an example it is worth mentioning that from 17 galaxies where Rubin et.al. found dynamically detached nuclei only for a half of them circular rotation of circumnuclear gas fitted the data. The other method of seeking fast rotating dynamically decoupled nuclei is to measure line- of sight velocity gradient $dV_r / dR $ along the positional angle PA. This approach appears highly efficient. Observatons of more than a dosen of normal spiral galaxies obtained with Fabri-Perot and multipuple spectrograph observations at 6m telescope of SAO RAS (Laboratory of Spectroscopy and Photometry of Extragalactic objects, led by S.Dodonov) have shown that such nuclei may be revealed at distances 1 - 3 arcseconds fron the center (some examples are given in Zasov, Sil'chenko, 1996). Typical velocity gradients were found to be several hundreds km/s. They may be even higher if to take into account that the size of nuclei is comparable to the seeing during the observations. Note however that in the case of rigid-body rotation a finite angular resolutuion does not distort the slope of rotation curve (Fridman et.al., 1994). This method allows not only io determine velocity gradient in the nucleus, but also to find the position of dynamical axis (line- of- nodes) and hence to reveal its agreement or disagreement with the major axis of galaxies. Happily in most cases the divergence at R $\approx $ several hundreds parsecs is not too large which gives evidence of approximately circular rotation of gas (although bright exceptions were also found, f.e. NGC 895, NGC 972, NGC 7217). In principle, the turn of dynamical axis may be caused both by non-asymmetrical bar-like potential and by the inclination of nuclear disk to the plane of the main disk. To reveal its nature it is necessary to use both kinematic and photometric data. It is curious that unlike nuclear rings, dynamically revealed inner bars (mini-bars) do not strictly relate to large- scale optically visible bars: they were found not only in SB- galaxies, but also in SA-galaxies where the presence of bar is not evident (NGC 895, NGC 972, NGC 4100, NGC 7217). It means that these small features are of no resonance nature. Instead they may appear due to the instability of extended orbits of stars in slowly rotating stellar components (Lynden-Bell and Polyachenko bar- forming mechanism, see Polyachenko and Polyachenko, 1994). Radio observations of CO-rich nuclei also show fast rotation within inner several hundreds of parsecs with velocity gradients similar to those observed in the optically selected galaxies (velocity of rotation 200 - 250 km/s reaches within R = 100 - 200 pc (Sofue, 1997)), so optical and radio samples of nuclear regions are non-distinguishable by their gas kinematics. \bigskip \bigskip \centerline{\bf Nuclear disc stability and star formation} \bigskip As it was pointed in the Introduction, nuclear star formation takes place in galaxies of all types, especially those of late Hubble types. But the intensity of this process differs by several orders of magnitudes - from weak emission nuclei to bursting star formation nuclei and further up to ultraluminous far infrared galaxies galaxies which have both extremaly high luminosity in far infrared $(10^{11} - 10^{12}~ L_\odot)$ and large surface density of molecular gas ($10^3 - 10^4 ~M_\odot / pc^2$) in their nuclei (see f.e.Sanders et.al., 1991). Concerning HII nuclei of normal spiral galaxies, their properties are in general similar to properties of giant HII regions in the main disks, with the exception of higher concentration $n_e$ and very low filling factor $\approx 10^{-5}$ (Luis et.al., 1997). The latter may just be a sequence of large number density of star forming gas. Indeed, for a given excitation parameter of single ionizing star a volume filled by ionized gas is proportional to $n_e^2$, so it must be much lower than in the main disk and hence a less volume fraction is occupied by HII regions. One can expect that high resolution observations, if they were available, would show a presence of thousands of small HII regions related to O-stars, immersed into the dense non-homogeneous gas medium of nuclear disk. The other peculiarity of conditions of star formation in moleculal clouds of circumnuclear disks is high gas pressure caused by the high number density of gas (both molecular and HII). A good example is the nucleus of spiral galaxy NGC 1808 (Aalto et.al., 1994): the analysis of radio emission in different molecular lines led authors to conclusuion that $P/k~ \approx 10^8~ cm^{-3} \cdot K$), which is 3 - 4 order of magnitudes higher than in the disk of the Galaxy. A high pressure may be a factor which significantly enhances star formation efficiency in gas- rich nuclei. Fast rotation of inner parts of galaxies, especially of dynamically decoupled nuclei, is a factor, which, on the contrary, reduces star formation activity (even for rigid body rotation) due to conservation of angular momentum of collapsing parts of gaseous disks. For a flat disk a large-scale gravitational instability is absent if radial velocity dispersion $C_{gas}$ exceeds a critical value $$ C_{cr}\, =\,Q\,\pi\,G~\sigma_{gas} /\kappa (R). $$ Here$ \sigma$ is the surface density, $\kappa (R)$ is epicyclic frequency (for rigid-body rotation $\Omega~= const~\kappa\,= 2\Omega$). Stability parameter $Q$ = 1 for pure radial perturbations (Toomre' criterion). In galactic disks, where dispersion velocity of gas is about 10 km/s, parameter $Q~ = 1.5 - 2$ (Kennicutt, 1989, Zasov and Bizyaev, 1994). It agrees with non- WKB analysis of stability which shows that the threshold for instability reaches for $Q \approx 1.7$ for flat rotation curve but keeps close to 1 for$ \Omega \approx const$ (Polyachenko, 1990). Hence, as it follows from the equation, in the region of rigid-body rotation, which usually coincides with circumnuclear region, a lower value of $C_{cr}$ is necessary for the disc to be stable due to lower $Q$ and higher $|kappa/\Omega$. Note that the strict criterion for three -dimensional case has not obtained yet, and all usually adopted conditions for the instability are still approximate (see the discussion in the book Horuzhii, Fridman, 1994). Now the question arise - are circumnuclear disks gravitationally stable? The answer is definitely "yes", if their gas surface density is less or not much larger than at several kpc from the center in the main disk where the angular rotation is lower. But there are many galaxies (including our own) where radio observations in molecular lines show the presence of dense molecular disks in central kpc regions. For some of them $ \sigma_{gas}$ often exceeds $10^3~ M_\odot/pc^2$, which (for the usually adopted conversion factor $2\cdot10^{20} cm^{-2}(K\cdot km/c)^{-1}$ used here), which makes their self- gravity rather essential. Kinematically gas-rich nuclear discs (bars, rings) of galaxies are similar to the optically selected dynamically detached nuclei having velocity gradients 300 - 1000 km/s/kpc. Fig.2 shows histogram of distribution of $C_{cr}$ for galaxies where both rotation curve and CO intensity map or $\sigma_{gas}$ are available for the standard conversion factor $3\times 10^{20}~K\cdot cm^{-2}$ (for M33 the presence of HI was also taken into account). Radiuses and general shapes of molecular "islands" and references are given in the Table. Note that due to restricted angular observations $\Omega$ may be underestimated which allows to consider $C_{cr}$ as an upper limit. Observational estimates of velocity dispersion of molecular clouds in central parts of galaxies are very scarce and not too reliable. I will adopt 10 km/s as an lower limit of $C_{gas}$. For gas-rich nuclei it is definitely higher: direct measurements lead to values $C_{gas}\approx 12-50 km/s $ (see Kenney et.al. 1997 for NGC 3504, Lo et.al ,1984 for IC 342, Sofue et.al for NGC 4631, Ishizuki et.al.1990 for NGC 6946 and Garcia-Burillo, 1992 for NGC 891, Gusten, 1989 for our Galaxy). Hence general conclusion which may be done from histogram in Fig 2 is that a large parts of galaxies which are considered have nuclear discs which are on the threshold of gravitational stability or definitely stable. This is especially true for nuclear regions of galaxies poor of gas such NGC 7331 ( ,1997) or NGC 7217 (Zasov, Sil'chenko, 1997). The problem of lowering of conversion factor used for transition from the intensity $I_{co}$ to $\sigma_{gas}$ for warm and dense molecular gas is well known (see Schild, 1994 and references therein). Here I note that in the nuclear part of nearby actively star forming galaxy M 82 observations in different molecular lines (\raisebox{3pt}{12}CO, \raisebox{3pt}{12}CO) confirm that the conversion factor is lower with respect to usually adopted one (Wild et.al., 1992). It is worth noting that star formation still continues even in the gravitationally stable disks. A good illustration is NGC 7217, where the intensity of $H_\alpha$ in this gas-poor galaxy monotonously rises towards the center of inner disk(). In this galaxy, in spite of the low gas density, star forming regions do not only exist in the nuclear disk, but apparently have a surprisingly well-ordered spiral-like structure (Zasov, Sil'chenko, 1997). It is not visible in the original images of this galaxy because they are swamped by the emission of the substantially brighter bulge which have unusually high luminosity. Even if the nuclear disk of this galaxy was marginally unstable (independently on whether it has gaseous or stellar nature) the wavelength of perturbations corresponding to the maximum of instability growth increment would not correspond to the observed structure. For the observed velocity gradient $\approx 200 km/s/kpc$ unstable wavelength equals to several hundreds of parsecs, whereas the observed structure is a sort of "rippled surface" with a significantly smaller scale (the characteristic radial distance between neighboring arcs is about 1", or less than 0.1 kpc). Therefore, the observed pattern cannot be due to gravitational oscillations in either the stellar or the gaseous disk. A possible alternative is the presence of a hydrodynamical instability in the gaseous disk that does not require a high surface density to develop. One of the conditions for such an instability to occur is the presence of a local maximum in the rotation curve, followed by a region of rapidly decreasing angular velocity (see the discussion by Fridman in [17, 18]). In this galaxy such a region probably does exist at the boundary of the dynamically distinct nucleus, where velocity gradient rapidly changes (Zasov, Sil'chenko, 1994). Note that similar small-scale inner spiral-like situcture was found by HST observations of NGC () (Barth et al., 1995). \centerline{\bf Conclusions} 1.Inner (nuclear) regions of spiral galaxies within several arcseconds from the center often possess nuclear bars and/or inclined disks which may be revealed not only from photometric or gas distribution data, but also from the analysis of gas kinematics. Nuclear disks are usually distinguished by high angular velocity of gas. 2. In most cases one can epect that the inner molecular disks $\le 1 kpc$ size are gravitationally stable or marginally unstable. The dependence of conversion factor on the abundance strengthens this conclusion. 3. Star formation and some regular structure may be developed even in gravitationally stable disks and hence there is (are) some non-gravitational mechanism(s) which regulate(s) process of star formation there. Author is grateful to O.K.Sil'chenko for her key role in the observational programs and Yu.N. Efremov for valuable discussions. This work was partially supported by grant RFBI 97-02-17358 \centerline{Literature} Aalto, S., Booth, R.S., Black, J.H., Koribalski, B. and Wielebinski, R. (1994) Astron. Aph. 286, 365. Afanasiev V.L., Sil'chenko O.K., Zasov A.V. 1989, Astron.Aph., 213, L9. Barth, A.J., Ho, L.C., Filippenko, A.V., Sargent, W.L.W. (1995) Astron. J. 110, 1009. Benedict, G.F., Smith, B.J. and Kenney, J.D.P. (1996) Astron.J. 111, 1861. Buta R. and Crocker D.A.,1993. Astron.J., 105, 1344. Canzian, B., Mundy, L.G. and Scoville, N.Z. (1988) Ap.J. 333, 157.[253] Fridman, A.M., Lyakhovich, V.V. and Zasov, A.V. (1994) in "Physics of the Gaseous and Stellar Disks of the Galaxy", ASP Conference Series, vol.66, ed. by I. King, p.305 Friedly D., Benz W., 1993, Astron. Aph., 227, 27. Gerin,M., Nakai,N. and Combes,F. (1988) Astron.Aph 203. 44 [1097] Garman, L.E. and Young J.S. (1986) Astron.Aph. 154, 8. [4736] Handa,T., Nakai, N., Sofue, Y., Hayashi, M., and Fujimoto, M. (1990) PASJap, 42,1.[M83] Kennicutt, R.C. (1989). Astron.J., 97, 1022. Khoruzhii O.V. and Fridman, A.M. Fizika planetnuh kolets (in Russian), M., Nauka, 1994. Luis C.Ho, Filippenko A.V., Sargent W.L.W., 1997, astro-ph/9704101. Polyachenko, V.L. (1990), in "Dynamics of Astrophysical Discs", ed.by J.Sellwood, p. 199. Polyachenko, V.L. and Polyachenko (1994) Pis'ma v Astron.Zh. (Russian), 20, 494. Rubin V.C., Kenney J.D.P., Young J.A, 1997. Astron.J., 113, 1250. D.B.Sanders, N.Z.Scoville, B.T.Soifer, 1991, Ap.J. 370, 158 Schlosman, I., Frank, J. and Begelman M. (1989) Nature, 338, 45. Sofue, Y. (1997) PASJap. 49, 17. Sofue, Y., Doi, M., Krause, M., Nakai, N. and Handa, T. (1989) PASJap. 41, 113.[4258] Sofue, Y. and Handa, T. (1989) PASJap. 41, 937. [4631] Sofue, Y., Handa, T., Golla,G. and Wielebinski, R. (1990) PASJap 42, 745. [4631] Sofue, Y.and Irwin, J.A. (1992) PASJap 44, 353. [3079] Telesco C.M., Dressel L.L.and Wolstencropft R.D. (1993) Ap.J., 414, 120. Wild, W., Harris, A.I., Eckart, A., et.al (1992) Astron.Aph. 265, 447 [M82] Wilson, C.D. and Scoville, N. (1989) Ap.J. 347, 743 [M33] Wright, M.C.H., Ishizuki, S., Turner, J.L., Ho, P.T.P. and Lo, K.Y. (1993) ApJ. 406, 470 Zasov, A.V. and Bizyaev, D.V. (1994) in "Physics of the Gaseous and Stellar Disks of the Galaxy", ASP Conference Series, vol.66, ed. by I. King, p.305 Zasov,A.V. and Sil'chenko, O.K. (1996) in "Barred galxies". ASP Conference Series, vol.91, ed. by Buta R., Crocker D.A. and Elmegreen B.G. p. 207 Zasov,A.V. and Sil'chenko, O.K. (1997) Astron.J.(Ruaaian), Barred galxies. ASP Conference \end{document}