4 The H I and the ICM in Coma

As mentioned earlier, physical mechanisms affecting galaxies in clusters are produced by interactions with one or more of the next three elements: the hot ICM, neighbor galaxies, and the cluster gravitational field. Two effects observed in H I in this work are explained on the basis of ICM-ISM interactions: the position of H I deficient galaxies relative to the ICM as drawn by the X-ray emission, and the fact that the most central galaxies detected in H I appear stripped in their outer regions, as predicted for ram-pressure stripping. If tidal interactions were predominant there should be evidence of optical signatures of the interaction, but none of the H I deficient galaxies in this work (except the merger NGC4922), neither detected nor undetected, display peculiarities in their optical morphology. It was also shown in Paper I that tidal interactions are unlikely to be the explanation for the disturbed H I disks in Coma or the triggering of starburst events, because there is no correlation between these effects and the number of close neighbors.

In order to confirm the role of ram pressure stripping in producing the observed H I distributions and in enhancing the star formation in Coma, it is fundamental to compare the H I imaging with 3D simulations of the ISM-ICM interaction. A thorough comparison will help to estimate the role played by different infalling orbits, inclination angles, and ICM densities, as well as to determine the stripping time scale and the regions of the disk where ram-pressure is more effective. To this aim we compare our H I observations with predictions made by 3D-simulations carried out by Abadi et al. (1999) and Vollmer et al. (2001b).

Simulations by Abadi et al. (1999) are a good point of departure even if the clumpiness of the ISM and different elapsed times after crossing the cluster core are not taken into account. They computed the expected radius of a ram pressure stripped gas disk ( $R_{\rm str}$) for a Coma cluster-like density and dispersion velocity. For instance, a typical galaxy in Coma with $v_{\rm rel}$ ${\sim}$ 1000 kms^-1 would have $R_{\rm str}$ ${\sim}$ 6kpc. In Table 1 we compare the observed H I radius (or upper limits if resolution is marginal) with the prediction of $R_{\rm str}$ made by Abadi et al. (1999). Columns 1 and 2 give the galaxy identification, Col. 3 gives the radial velocity relative to the cluster. In Col. 4 we give the observed H I radius in kpc estimated at a level of $3 \times 10^{19}$ $M_{\odot}$, taking the average between major and minor H I axis (none of the galaxies in Table 1 are very elongated). Column 5 gives the observed H I radii corrected for the beam; for galaxies marginally resolved (indicated with *) we give 0.5 times the beam size as an upper limit of the beam corrections, following Wild (1970). Column 6 gives the predicted value of $R_{\rm str}$ from Abadi et al. (1999), considering the corresponding velocity of the galaxy relative to the cluster.

 

 

Table 1: Comparison between the observed and predicted H I radius in kpc, for central galaxies in Coma.

CGCG	Other	$v_{\rm rel}$	Obs	Corr	Pred
	name	kms-1	$r_{{\rm HI}}$	$r_{{\rm HI}}$	$r_{{\rm HI}}$
(1)	(2)	(3)	(4)	(5)	(6)
160-055	NGC4848	41	6.8	5.0*	17.0
160-073	Mrk058	1575	7.9	5.8	4.5
160-086		481	6.0	5.6*	8.5
160-252	IC4040	758	10.2	8.0	7.5
160-257	NGC4907	1180	6.6	5.8*	5.5
160-260	NGC4911	997	12.7	12.2	6.5
160-095	NGC4921	1521	11.2	10.2	5.0
160-106	NGC4926-A	124	8.2	5.7	14.0

(*) Galaxies marginally resolved in H I.

The predicted H I radius is calculated assuming that the galaxy's distance to the center equals its projected distance and its velocity through the cluster equals its radial velocity. Considering those assumptions there is good agreement between observed and predicted values of the HI radius. The two galaxies (NGC4848 and 4926-A) that have significantly smaller radii than predicted must have a non negligible velocity in the plane of the sky, while the giants NGC 4911 and 4921 are probably at larger distance from the center and only in projection very close. For the three unresolved galaxies the H I distribution is limited to a central region of $\leq$6 kpc. Our results for H I deficient yet detected galaxies in Coma, confirm that most of the restoring force is coming from the central parts of the disk where the presence of the bulge is more important, as found by Abadi et al.'s simulation. They found that the gas is not completely removed by ram pressure, suggesting that other processes may be at work affecting those spirals which are not detected in H I in this survey.

Three dimensional simulations taking into account the clumpiness of the ISM were carried out for the Virgo cluster by Vollmer et al. (2001b). These authors found that stripping is very effective for galaxies that are on radial orbits through the cluster, in agreement with observational evidence provided by Dressler (1984) and Solanes et al. (2000). Vollmer et al. (2001b) found that time scales for ram pressure effects in Virgo may be as short as $3 \times 10^7$ yr, and in Coma they would likely be shorter because of its larger core size and more hostile ICM conditions. Interestingly, those authors found that galaxies showing important gas disruptions are not infalling but have already gone through the cluster core. If this is valid in Coma, it would confirm that all the central spirals (see Fig. 2 of Paper I) have already gone through the cluster core, while the H I regular galaxies in the outskirts of the X-ray emission are in the process of infalling (see Sect. 3). Some of the consequences in Coma are discussed in the next section.

 

 

Table 2: HI and radio continuum parameters of abnormal spectrum galaxies (top) and blue disks (bottom) in Coma.

ID	Other name	Morph.	$M_{\rm HI}$	$F_{{\rm cont}}$	rms noise	L1.4	SFR	SFR/$L_{\rm B}$
		type	108 $M_{\odot}$	mJy	mJy Beam-1	1020 WHz-1	$M_{\odot}$ yr-1	$M_{\odot}$(yr $L_{\odot})^{-1}$
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)a
D 77	Leda 83676	S0/a (SB)	<0.6	-	0.18	<3.17	<0.19	<0.23
D 94	Leda 83682	SA0  (PSB)	<0.5	-	0.18	<3.17	<0.19	<0.17
D 112	Leda 83684	SB0  (PSB)	<0.4	-	0.18	<3.17	<0.19	<0.03
D 21	MCG 5-31-037	SBa  (PSB)	<0.7	-	0.19	<3.35	<0.20	<0.09
D 73	RB 183	SA0  (PSB)	<0.9	-	0.17	<3.00	<0.18	<0.27
D 44	KUG1256+278A	S0    (SB)	<0.6	-	0.18	<3.17	<0.19	<0.18
D 43	NGC 4853	SA0  (SB)	<0.5	1.2	0.18	7.06	0.42	0.06
D 89	IC 3949	SA0  (PSB)	<0.3	2.1	0.18	12.35	0.73	0.18
D 127	RB 042	S0    (PSB)	<0.6	1.0	0.18	5.88	0.35	0.46
D 216	RB 160	Sa    (PSB)	<0.7	-	0.18	<3.17	<0.19	<0.28
D 99	Mrk 060	SB0  (PSB)	<0.7	1.2	0.19	7.06	0.42	0.23
D 146	RB 110	S0    (PSB)	<0.3	-	0.18	<3.17	<0.19	<0.16
D 61	CGCG 160-104	SA0  (PSB)	<0.7	-	0.10	<1.76	<0.10	<0.03
D 189	Leda 83763	S0    (PSB)	<1.5	-	0.17	<3.00	<0.18	<0.31
160-055	NGC 4848	Scd	4.3	16.6	0.18	97.37	5.75	0.38
160-073	Mrk 058	Sb	2.0	5.5	0.17	32.10	1.89	0.56
160-086		Sb	1.7	3.8	0.15	22.11	1.30	0.88
160-252	IC 4040	Sdm	3.3	15.0	0.18	88.20	5.20	1.00
160-098		Sbc	7.3	5.7	0.18	33.34	1.97	0.62
160-106	NGC 4926-A	Sa	6.0	3.1	0.15	18.17	1.07	0.49

^a In units of 10^-9.