Like restricting the sample to ScI galaxies? Like correcting the Tully-Fisher relation for the type effect? ( Type Dependence
Originally Posted by Cougar
Two things here. First, most of the analysis (the stuff not even discussed yet) is restricted to ScI group galaxies. In fact SO galaxies were not used. Second, the observed scatter of the TFR is very small for the local samples we're talking about (where "local" is within 200 Mpc). Galaxies at z=1 simply are not relevant to the results we're discussing. The results on the high z TF samples are inconsistent in their results. Some studies show very little luminosity evolution while others show a large luminosity evolution.
- Several recent papers have investigated the TF and found that it varies considerably between galaxy types and over time. For example in this paper http://arxiv.org/pdf/astro-ph/0603042 looking at the TF at z=1 they find that the TF is shifted upwards in magnitude, reflecting the extra star formation ongoing at that point . Another recent paper http://arxiv.org/pdf/astro-ph/0609076 investigating the TF for S0 galaxies found that the TF relation is offset to lower Magnitudes, because S0s are not forming stars and hence have faded since they were.
When you have redshift independent distances for calibration, it turns out that there is very little scatter in the TFR. In addition it also depends upon what wavelenth you look at. In the B-band star formation effects are larger (yet the Type dependent TFR significantly corrects for this because the narrow arms of ScI galaxies are well defined by current star formation), but in the K-band the type effect is much smaller because the underlying older population dominates the light.
The important point is that the TF depends on the star formation rate in a galaxy, if you select galaxies that are not forming stars at a similar rate you can get into trouble.
Irrelevant for several reasons. First, if you look at the distance distribution of the spirals you'll see that the ScI's match the distance distribution of the other galaxy types. No "assumption" was made. The distances were calculated with the TFR. In fact, that was the whole point of the TFR analysis. Arp and others had previously made assumptions that they were members or that they were on the backside of the cluster (based upon the large redshifts). But the Cepheid and TFR distances show that they are at the distance of the cluster.
Also star formation is related to position in a cluster (see below), if you aren't extremely careful about how you select your galaxies you can get really screwed up by picking galaxies that are not comparable.
[There are] very few spirals in clusters (such as Virgo), this is because the gas in them is stripped out during the fall into the cluster. These spirals become S0s as they fall into the cluster. This is very important because it means that the spirals that Russell assumes are in Virgo are almost certainly not, they form a shell around the main cluster.
Second, as for the "shell" concept. Can anybody explain why the ScI's are on the receding side of the shell while the Sab/Sb's are on the approaching side of the shell? Other than saying its an "accident" there is no reason to expect the spirals to be distributed by morphology in that fashion.
Nobody made that assumption. Distance were calculated ... and in the mean they fall at the cluster center. And most of them fall pretty close to the core distance. But the ScI's are not in filaments. If you look at their locations, they're scattered around the cluster both by coordinate and by distance. These concepts of shells and infalling filaments don't explain it.
Hence because Virgo has a fairly large virial radius > Mpc, it is not fair to assume they are all located in the cluster.... But it gets even worse, there are large filaments feeding into Virgo, http://www.ifa.hawaii.edu/~tully/
, in these filaments are galaxies that are streaming into Virgo at a significant speed. So the assumption that you make that if you look at enough spirals around Virgo it will even out so it seems like they are all in the centre wont hold.
There’s also a key point that Nereid made
summed it up best –
“It is strange that the supposedly strong correlations seem to get weaker with more data, exactly as you would expect if they were due to statistical noise.”
Pretty good criticism??? I think ToSeek summed it up best.
Are we talking about my papers or quasars now? If we were to accept Nereid's point as correct (which it likely isn't because incorrect assumptions are often made with these studies that use the larger data sets), the point has nothing to do with spiral galaxies - for which increased data has only led to a stronger case.