The most comprehensive survey on H I content in cluster galaxies to date is the work by Solanes et al. 2001. These authors compiled H I data on 1900 spiral galaxies in 18 nearby clusters. The data are mostly obtained with the Arecibo telescope, a single pixel telescope, and give information about the total amount of neutral hydrogen within the Arecibo beam (3 arcmin) centered on optically selected galaxies. Galaxies are earmarked as belonging to a cluster when they fall within a projected distance of 5 Abell radii (RA), i.e. within 7.5 h -1 Mpc, from the cluster center and have a radial velocity that is less than 3 times the average velocity dispersion from the cluster mean. Only clusters are included, for which there are good H I data for at least ten galaxies within 1 RA of the cluster center. HI defiency is calculated according to the recipe of Haynes and Giovanelli (1984). It is the log10( MH I observed / MH I expected), where the expected H I mass is derived from a sample of isolated spirals of the same morphological type and optical diameter. To get significant statistics the cluster sample is then divided in two groups, the deficient cluster sample, and the non deficient cluster sample. The deficient sample contains all clusters for which the H I deficiency distribution over galaxies is significantly different within 1 RA from that of the galaxies outside 1 RA. One of the most remarkable results of the analysis of that data base is shown in Figure 1. It shows the H I deficient fraction, i.e. the fraction of galaxies with a deficiency greater than 0.3, in bins of projected radius from the cluster center for the superposition of all the H I deficient clusters. Galaxies with a deficiency greater than 0.3, are galaxies that are deficient in neutral hydrogen by a factor two or more as compared to isolated galaxies of the same morphological type and size in the field. The percentage of H I deficient spirals increases monotonically going inward. What is surprising is that this monotonic rise starts as far out as two RA. This suggests that the effect of the cluster environment can be felt out to two Abell radii, far beyond the reaches of the dense ICM.
 |
Figure 1. Top: H I deficient fraction in bins of projected radius from the cluster center for the superposition of all the H I deficient clusters. Bottom: H I deficiency versus projected radius from the cluster center. Small dots show the radial variation of H I deficiency for individual galaxies, while the arrows indentify non detections plotted at their estimated lower limit. Large dots are the medians of the binned number distribution. From Solanes et al. 2001. |
No correlation is found for the fraction of H I deficient
spirals (i.e. the number of spirals with an H I deficiency
DEF 
0.30
within 1 RA of the cluster center compared to all galaxies
of that type found in that region) with global cluster properties,
such as X-ray luminosity, X-ray temperature, and radial velocity dispersion.
As pointed out in the paper, this could be a selection bias.
If stripped spirals would lose all their gas they may be transformed into
S0's and they would be left out from the statistics. It is
somewhat plausible that this is indeed the case since the fraction of
spirals is clearly anti correlated with X-ray luminosity.
The most important result of the paper apart from the extent of
the occurrence of deficient galaxies is the correlation between
deficiency and orbital parameters. This is shown in
Figure 2.
It shows for the composite deficient cluster the radial run of the line of
sight velocity dispersion for the most deficient spirals, the non deficient
spirals, for all spirals and for ellipticals and lenticulars.
If galaxies are on radial orbits the measured velocity dispersion should
decrease at large distances from the cluster center. Although all spirals
show a decrease in velocity dispersion at large distances, this effect
is by far the most pronounced in the H I deficient spirals. The
ellipticals and S0's have a constant velocity dispersion with
radius. This confirms the result by
Dressler (1986).
These results suggest that deficiency
is most pronounced when galaxies go through the dense cluster center at
high velocities and as such support the idea that ram pressure
stripping causes the deficiency.
 |
Figure 2. Radial run of normalized line of sight velocity dispersion for the composite H I deficient cluster. From Solanes et al. 2001. |