Chris O’Dea (STScI) Stefi Baum, Anton Koekemoer, Jennifer Mack (STScI) Ari Laor (Technion)

1. HST FUV Imaging of the Luminous Emission Line Nebulae in the Central Dominant Galaxies in A1795 and A2597 Chris O’Dea (STScI) Stefi Baum, Anton Koekemoer, Jennifer Mack (STScI) Ari Laor (Technion)

2. Outline • Background on A1795 and A2597 • Previous Evidence for Star Formation • The New HST FUV Continuum and Lyα Observations • Star Formation • Ionization of the Nebula • Summary • A Scenario

3. The “Head-Tail” Nebula in A1795 • Luminous X-ray cluster, mass accretion rate ~35 M⊙/yr ( Ettori et al. 2002) • Bright radio source 4C26.42 • Very luminous emission line nebula; central region in which the gas follows the edges of the radio source (van Breugel et al 1984; Pinkney et al. 1996) • A system of bright filaments stretching 45 arcsec south (Cowie et al. 1983). • Chandra detects 40” X-ray filament along the emission line filament (Fabian et al. 2001). (Top) Adaptively smoothed Chandra X-ray image. (Bottom) Hα+[NII] image of the same field from Cowie et al. (1983). Figure from Fabian et al. (2001).

4. Blue Lobes, Dust & Gas in A1795 • Blue light associated with the radio lobes was detected in ground based U band images by McNamara & O’Connell (1983). • WFPC2 imaging reveals the blue light is composed of knots which lie along the radio source (McNamara et al. 1996). • The filaments wrap around the radio lobes. Central 30 kpc of A1795. Sum of V and R WFPC2 images with a galaxy model subtracted. Central insert is smoothed to 0.1” resolution and contours of VLA image are superposed. Lower left insert is unsmoothed. Upper left insert is V-R color map. (McNamara et al. 1996).

5. The Complex Center of A2597 • This is a luminous X-ray cluster (e.g., Crawford et al. 1989; Sarazin et al. 1995, 1997) • The X-ray emission is clumpy in the inner ~ 40 arcsec (McNamara et al. 2001). • Estimated temperature drops to ~1-2 keV at 10 kpc Radial variation of (a) surface brightness, (b) electron density, c) temperature, and (d) temperature. (McNamara et al. 2001) Smoothed Chandra 0.3-8.0 keV X-ray image of A2597. McNamara et al. (2001).

6. Filaments in A2597 • The optical emission line nebula is clumpy and filamentary. • The filaments tend to wrap around the radio lobes.

7. Previous Evidence for UV Light and/or Star Formation • Previous low angular resolution observations of A1795 and A2597 have shown that the FUV continuum has an integrated spectrum and/or colors which are consistent with those of a “young” stellar population • Spectra – Hu 1992; Crawford & Fabian 1993; Oegerle et al 2001 • Imaging – Smith et al. 1997; Mittaz et al. 2001 • Analyses of the optical properties of the nebula also concluded that young stars were present, e.g., Johnstone & Fabian 1988; Crawford & Fabian 1992, 1993; McNamara & O’Connell 1993; Allen 1995; Voit & Donahue 1997; McNamara et al. 1996,1999; Koekemoer et al. 1999. A1795. U (~3500Å) and UVW2 (~2000Å) images of the central 35”x48” with the cD galaxy located near the middle of the images. Note the UV blob in the filament 20 arcsec to the south of the cD. Mittaz et al. (2001).

8. Evidence for a Distributed Source of Ionization • The Hα surface brightness falls off much slower than expected for a central ionizing source (Conselice et al. 2001; see also Johnstone & Fabian 1988; Heckman et al. 1989). Observed mean Hα intensity profile of the ionized gas around NGC 1275, as measured in elliptical apertures, shown as a red line with error bars, plotted on the left intensity scale. The dashed line shows the predicted intensity with projected radius, R, for a central ionization source model with the only losses being r-2 dimming. Using the scale on the right, we show the upper limit to the projected ionized gas filling factor versus radius (solid blue line). The inner 3.8 h kpc is almost completely filled with overlapping filaments, while the outer parts of the system are very sparsely occupied. The green dotted line above the mean intensity profile at larger radii illustrates the brightness that would be observed at each radius if the ionized gas projected filling factor were unity.

9. The Blue Lobes in A2597 • Blue lobes detected in ground based optical data by McNamara & O’Connell (1993). • HST WFPC2 observations show blue light tends to wrap around the radio source (Koekemoer et al. 1999).

10. Our FUV and Lyα Observations • HST/STIS FUV MAMA ½ orbit in Lyα line and in adjacent redward FUV continuum • No geocoronal Lyα included • Archival WFPC2 and NICMOS observations re-analyzed

11. The FUV Spectrum of A2597 • Lyα is the only bright line in the FUV (in agreement with Hu 1992). • There is no significant emission line contamination of the FUV continuum measured in the F25QTZ filter. • The relative faintness of the other UV lines may place constraints on the contribution of shocks to the ionization. A2597. HST STIS FUV MAMA long slit spectrum with low dispersion G140L grism. Baum et al. (2003).

12. Lyα, X-ray, and Radio Emission in A1795 • The Lyα is distributed in a diffuse component which is aligned with the inner radio source axis, as well as in bright filaments and/or clumps along the radio source edge. • The peaks in the Lyα and the X-rays seem to coincide. • The X-ray peak is not centered on the nucleus of the galaxy. A1795. Blue contours are Chandra image from Fabian et al (2001). White contours are the VLA image from Ge & Owen (1993) courtesy of Greg Taylor and Jing Ping Ge. Orange is Lyα emission line. FOV is 19x19 arcsec. (O’Dea, Baum, Koekemoer, Mack & Laor 2003)

13. Lyα and FUV Continuum in A1795 The FUV continuum follows the bright Lyα emission. A1795. Color image of smoothed Lyα emission line with contours of FUV continuum. (O’Dea, Baum, Koekemoer, Mack, & Laor 2003)

14. A1795 • The UV continuum exhibits both a diffuse component as well as a set of bright knots which lie preferentially along the radio source edge. (Does this mean that there are two modes of star formation in the nebula?) • The Lyα and Hα are distributed in a diffuse component which is aligned with the inner radio source axis, as well as in filaments along the radio source edge. • ~60% of the Lyα and FUV continuum is in the diffuse component. The rest is in the compact knots. A1795. (J1348+2635). Montage of HST images and VLA 8.4 GHz image.The UV continuum and Lyα images are from our STIS observations. The other HST data are archival WFPC2 observations. The Lyα and Far-UV continuum images are smoothed to bring up low surface brightness emission. VLA image is from Ge & Owen (1993) courtesy of Greg Taylor and Jing Ping Ge. (O’Dea, Baum, Koekemoer, Mack, Laor 2003)

15. Lyα, X-ray, and Radio Emission in A2597 • The Lyα and the X-rays are both “clumpy”. • The central Lyα plateau coincides with X-ray peak, and the radio source. • The extended filaments of Lyα seem independent of the X-ray emission. • The radio source is oriented along the steepest gradients in the X-ray emission. A2597. Blue contours are Chandra image from McNamara et al (2001). White contours are VLA image from Sarazin et al. (1995). Orange is Lyα emission line. FOV is 20x20 arcsec. (O’Dea, Baum, Koekemoer, Mack & Laor 2003).

16. Lyα and FUV Continuum in A2597 • The Lyα is distributed in three components – • (1) a bright inner region coincident with the radio source • (2) a diffuse extended component • (3) large scale arcs • The FUV continuum follows the bright Lyα emission. A2597. Color image of smoothed Lyα emission line with contours of FUV continuum. (O’Dea, Baum, Koekemoer, Mack, & Laor 2003)

17. A2597 • The UV continuum exhibits both a diffuse component as well as a set of bright knots which lie preferentially along the radio source edge. • The Lyα and Hα are distributed in a diffuse component, as well as in filaments along the radio source edge. • ~60% of the Lyα and ~70% of the FUV continuum is in the diffuse component. The rest is in the compact knots. A2597. Montage of HST images and VLA 8.4 GHz image.The UV continuum and Lyα images are from our STIS observations. The other HST data are archival WFPC2 observations. The Lyα and Far-UV continuum images are smoothed to bring up low surface brightness emission. VLA image is courtesy of Craig Sarazin. (O’Dea, Baum, Koekemoer, Mack, Laor 2003)

18. Correlations in A1795 • The observed Lyα / Hα ratio is in the range 2-5. • There is a correlation between Lyα and Hα. • There is a correlation between Lyα and the FUV continuum. • There is a weaker correlation between Hα and FUV continuum. • There is no correlation between the Lyα / Hα ratio and the FUV continuum. • This is consistent with in situ ionization of the gas by the FUV continuum. A1795. Plots of the relationship between FUV continuum, Hα, Lyα, and the Lyα/ Hα ratio. The data have been smoothed and re-sampled onto pixels 4 times larger in order to increase the S/N. Pixels are included in the plot when the S/N is greater than 3 in both quantities. (O’Dea, Baum, Koekemoer, Mack, Laor 2003)

19. Spatial Coherence of the Correlations • Overall, there is a correlation between FUV continuum and Lyα flux. • Individual regions have coherent values and occupy smaller regions on the plane. (Top). A1795. Plot of the relationship between FUV continuum and Lyα for selected regions. The data have been smoothed and re-sampled onto pixels 4 times larger in order to increase the S/N. Pixels are included in the plot when the S/N is greater than 3 in both quantities. (Bottom) A2597. Plot of the relationship between FUV continuum and Lyα for selected regions. The data have been smoothed and re-sampled onto pixels 4 times larger in order to increase the S/N. Pixels are included in the plot when the S/N is greater than 3 in both quantities. (O’Dea, Baum, Koekemoer, Mack, Laor 2003)

20. Structure in Lyα / Hα • We do not see large amplitude fluctuations in the Lyα / Hα ratio. Instead, there are coherent features in the Lyα / Hα ratio. • This implies the factors which affect the ratio (e.g., ionization, dust) are not changing rapidly across the nebula. • A2597 has higher Lyα / Hα ratio at “hot spot” in northern radio lobe. (Top) A1795. Contour plot of radio emission superposed on color plot of the Lyα/ Hα ratio. Coherent features in the emission line and ratio are labeled. VLA image is from Ge & Owen (1993) courtesy of Greg Taylor and Jing Ping Ge. (Bottom) A2597. Contour plot of radio emission superposed on color plot of the Lyα/ Hα ratio. Coherent features in the emission line and ratio are labeled. VLA image is courtesy of Craig Sarazin. (O’Dea, Baum, Koekemoer, Mack, Laor 2003)

21. Hot Stars Can Power the Nebulae. • We use the observed Hα luminosity to estimate ionization rates assuming case B. A1795 ~ 6 x 1053 s-1 and A2597 ~ 7 x 1053 s-1. • We calculate the number of hot stars required to produce the ionization rates using Panagia (1973). • We calculate the expected counts in our STIS FUV-MAMA image using SYNPHOT, and the stellar SEDs. • We find that the observed FUV continuum accounts for about half of the hot stars (e.g., O5) necessary to power the nebulae. This is probably close enough given the uncertainties and the fact that we are probably not seeing all the FUV continuum. • Lifetimes of the hot stars are short implying star formation is either very recent (<5x 106 yr) or ongoing. Number of photons below 912Å as a function of time. (Top) Continuous star formation model, 1 M⊙ yr-1 . (Bottom). Instantaneous burst, 106M⊙. Solid line, α=2.35, Mup = 100 M⊙, long dashed line α=3.30, Mup = 100 M⊙, short dashed line α=2.35, Mup = 300 M⊙. Starburst99, Leitherer et al. (1999).

22. Properties of the Stars in A1795 • The colors of the FUV knots are consistent with those of young stars with ages ~107 yr. • Observed FUV emission is consistent with • Ongoing star formation: Star formation rate ~ 1 M⊙ yr-1. • Instantaneous burst: mass in stars ~ 7 x 106M⊙. • Individual knots are ~0.5% of the total FUV continuum, e.g., 0.005 M⊙ yr-1 for 107 yr gives a mass of 5x104M⊙ for a young star cluster. • Results are similar for A2597. A1795. We plot Bruzual-Charlot models for both constant star-formation and instantaneous burst scenarios, for a Scalo and Salpeter IMF, and metallicities ranging from solar to 0.02 solar.In all cases the ratios are defined such that positive values are "redder", negative values are ``bluer". The black dots along each curve in the bottom panel represent logarithmic increments of 1e6, 1e7, 1e8, 1e9 and 1e10 year (the curves always start off at the blue end, and gradually evolve toward redder ratios. (O’Dea, Baum, Koekemoer, Mack, Laor 2003)

23. The Temperature Problem • Hot stars are not capable of heating the nebula to the inferred temperatures (Voit & Donahue 1997). • This implies an additional source of “non-ionizing” heat for the gas. Comparison between model H II regions and temperatures measured from [S II] and [O II] line ratios. The solid (S II) and dotted lines (O II) give the temperatures inferred from the S II and O II line ratios given by CLOUDY (Ferland 1993) photoionization models of H II regions. The input spectra for the "Stellar" models are Kurucz model atmospheres with effective temperatures (Teff) ranging from 35,000 to 50,000 K. The lines labeled "Blackbody" give the results for blackbody input spectra with temperatures from 30,000 to 100,000 K. In the photoionized gas, the abundances are 0.5 solar with Galactic depletions, and the ionization parameter is log U = -4.0. All the models fail to reproduce the observed temperatures. (Voit & Donahue 1997). .

24. What Ionizes the Nebula? • Hot stars may provide the bulk of the global ionization. • The radio source may provide some additional ionization to gas along the lobes. • The AGN “ionization cone” may also ionize some gas. • Other mechanisms?

25. Comparison between Low-z Cooling Flows and High-z RGs • High-z RGs have Lyα halos which are much brighter than the low-z cooling flows (~2x1042 vs. ~5x1044 ergs/s) • High-z RGs have Lyα halos which are much larger than the low-z cooling flows (~15 kpc vs. ~100 kpc) • The rest-frame equivalent width of Lyα are similar (~130, 500 vs. ~100-500). • The Lyα/ Hα ratios are similar (2-6 vs. 4-5) • Thus the “physics” of these nebulae may be similar at low and high z. Grayscale Lyα image of 4C41.17 with a contour representation of the 4.9 GHz VLA map (Carilli, Owen, & Harris 1994) overlaid. The grayscale has been colorcycled to show the details of the high and low surface brightness simultaneously. The radio core is identified with a cross. (Reuland et al. 2003).

26. Summary I: The Results • The Lyα and FUV emission consist of a diffuse component ~60% and more compact features – knots/filaments - which lie preferentially along the radio source edges. • There are correlations between the FUV continuum flux and the emission line fluxes. • The observed Lyα/Hα ratios are in the range ~2-6. The ratios exhibit spatially coherent structure. • The colors of the continuum emission is consistent with those of young stars ~107 yr. • The observed FUV continuum flux is consistent with about half the number of hot stars which are required to ionize the nebula. (We may be missing some FUV flux). • Although smaller and less luminous than the Lyα nebulae in high-z RGs, some key properties of the cooling flow nebula are similar.

27. Summary Reloaded: The Scenario • The results from A1795 and A2597 are in good agreement, suggesting that these results are not atypical of cooling flow clusters. • The radio source expands into a nebula consisting of dense atomic/molecular clouds– influencing the morphology of the gas (snowplowing the gas into arcs and filaments which wrap around the radio lobes) - implies slow expansion of the radio lobes? • Star formation occurs through out the nebula. • The star formation is locally strongly enhanced by the radio source. • The hot young stars, e.g., O5, probably provide the bulk of the photons which ionize the nebula. • In addition, the central nucleus may ionize gas along its ‘ionization cone’. • Similar processes may occur in the Lyα nebulae in high-z RGs.

28. The End