Proceedings of the Particle Physics and Early Universe Conference (PPEUC).
For this paper: postscript previous 2 Fitting procedures and ... up Abstract... next Acknowledgements...

3 Discussion and conclusions  

In order to discuss the cosmological consequences of our results, we converted our results into limits on intrinsic physical properties of the SSC (for details see Molnar & Birkinshaw (1996)). Figure 1 shows 95 % upper limits for the central electron number density as a function of electron temperature for truncated beta models with (Fig. 1a) and (Fig. 1b), which bracket the likely range of this parameter, and a core radius of suggested by optical observations, from our fits to COBE DMR, HEAO and ROSAT data. We also show the GINGA limits (Day et al. (1991)), and estimates of the limits that could be reached using moderate-duration XTE scans across the SSC (Jahoda et al. (1996)).

Figure 1:  95% upper limits on the central electron number density of intra-supercluster gas as a function of electron temperature from the COBE DMR (Galactic emission removed) and HEAO-1 A2 data. The solid line represents the limits that can be set based on the sensitivity of GINGA (see Day et al. (1991)). The 95% estimated sensitivity limits that could be achieved by a long XTE scan (Jahoda et al. (1996)) are also shown. The two panels correspond to truncated isothermal beta models with a core radius of , and (left) beta = 0.6; (right) beta = 1.2.

The HEAO-1 A2 limits are slightly less sensitive than the GINGA limits of Day et al. (1991), and more sensitive than those obtained using the COBE DMR data unless the gas temperature is unexpectedly high. The ROSAT All-Sky Survey results provide slightly better constraints than the COBE results, but the results from the other satellites are more stringent. The central electron density in the SSC is constrained to be less than about for most values of . XTE scans should achieve about a factor of three improvement on the GINGA limits. Two proposed missions, to study the CBR, MAP and PLANCK, have () sensitivities of at 90 GHz (the highest MAP frequency, where galactic contamination has the minimum antenna temperature), and at 143 GHz (a PLANCK frequency, where the SZ effect has its maximum in intensity units) with an angular resolution of 20 and 10 arc minutes respectively (Bennett et al. (1995), Tauber et al. (1996)). MAP and PLANCK should be able to detect gas in the SSC with central electron densities of about and , based on their sensitivity alone, but the practical detection limit is likely to be set by confusion with primordial fluctuations.

White & Fabian (1995) have shown that the average baryonic densities in a sample of 13 clusters are considerably higher () than the limit on the average baryon density of the Universe derived from standard nucleosynthesis theories (assuming , Walker et al. (1991), Smith et al. (1993)). It is important to find out if this baryon problem exists at a supercluster scale. Fabian (1991) estimated the baryon fraction in the SSC by extrapolating the gas mass and luminosity relation for clusters of galaxies (Jones & Forman (1984)), and concluded that . We derived limits on the average gas density over a core radius in the SSC >from the same two truncated beta models we used for Figure 1 ( 0.6 and 1.2) using GINGA sensitivity limits, the estimated sensitivity of proposed XTE scan, and the expected sensitivities of MAP, and PLANCK missions. We plot the 95 % upper limits on the ISC gas density in units of the critical density ( ) on Figure 2. We also plot average densities derived from the same isothermal models with a central peak (4 ) equal to the estimated primordial fluctuation on a scale at two frequencies, 90 GHz, and 143 GHz. These models represent limits below which single frequency observations are limited by primordial fluctuations.The lower lines represent limits using , the upper lines belong to .The area enclosed by these lines represent the upper limits over the full likely range of beta. The prediction for the average baryon density derived from standard nucleosynthesis theories is also shown with 2 error bars (two horizontal lines). The limits on based on GINGA sensitivities range from 0.19 to 0.08 with increasing temperature. Thus we can conclude that the SSC does not contain much more hot () large scale diffuse ISC gas than it is predicted by standard nucleosynthesis theories. Any reasonable value for would not change our conclusion.

The temperature and distribution of the intra-(super)cluster diffuse gas strongly depend on the dynamics of the clustering (Governato et al. (1996)). The lack of such diffuse gas could tell us about the efficiency of clustering (Frisch et al. (1995), West et al. (1995)). Therefore it is important to continue to search for ISC gas in superclusters with instruments with higher sensitivity. Assuming hot ISC gas (), we can conclude from Figure 2 that XTE would be able to detect gas in the SSC with . Single frequency observations with sensitivities are limited by primordial fluctuations.

The estimated 95 % detection limits of MAP and the PLANCK for extended objects at 90 and 143 GHz make it possible to detect ISC gas in the SSC at density and or better. These limits are much lower than the limits derived from standard nucleosynthesis theories, and the single frequency primordial limits. These observations are limited by primordial fluctuations on angular scale of the observed extended object, and the accuracy of doing spectral separation of the SZ effect from primordial fluctuations and Galactic contamination. A detailed modeling of galactic contamination and point sources will be clearly needed to analyze the data collected by the next generation of satellites, but the signature of ISC gas should be evident in their result.

Figure 2:   95 % upper limits on the average intra-cluster gas density, , as a function of electron temperature in the core () of the SSC in units of the critical density, that can be derived based on non-detections with GINGA (see Day et al. (1991)), and a long XTE scan (Jahoda et al. (1996)), and limits derived from models with a central peak (4 ) equal to the estimated primordial fluctuation on a scale (PRIM) using truncated isothermal models with a core radius of , and beta = 0.6 (lower line), or beta = 1.2 (upper line). The two horizontal lines (NS LIMIT) bound at the average baryon density required by standard nucleosynthesis theories.

PPEUC Proceedings
Thu Jun 19 14:50:07 BST 1997