View Full Version : Distant galaxies *generally* brighter due to lensing by foreground galaxies

2011-Jan-12, 10:07 PM
I noticed this news today:

In a paper published today in the science journal Nature, an international team of astronomers predicts that foreground galaxies will affect images of extremely far galaxies.... This means that a significant fraction of far background galaxies will appear on the sky near foreground galaxies. The good news is that the remote galaxies will appear brighter because of a phenomenon called gravitational lensing.

Of course, this kills a lot of old Arp arguments, but I thought further, "Hmm, could such significant lensing explain the data that has, up till now, been the reason we think the expansion is accelerating?"

But no, as Goldsmith says,

"The cosmological constant's claim to a nonzero value fundamentally rests on the finding that distant Type Ia supernovae reach maximum brightnesses approximately 25 percent fainter than the peak brightnesses they would attain in a universe with a cosmological constant equal to zero."

So, going the other way, perhaps the expansion is accelerating more than we thought since some of the Sne IA data may have been lensed, meaning they were actually more than 25% fainter, but the lensing made them appear more luminous.

However, I doubt the supernova search teams overlooked such an effect.

Ari Jokimaki
2011-Jan-13, 10:04 AM
Of course, this kills a lot of old Arp arguments,...

No, it doesn't. The fact that some pieces of evidence fit to a hypothesis doesn't take anything away from another hypothesis. It might be good for gravitational lensing hypothesis but it doesn't change anything in Arp's hypothesis. Arp's hypothesis still explains the data equally well it did before this study.

2011-Jan-13, 11:14 AM
It's worth noting that the effects of lensing only become significant when considering the extremely high redshift Universe (i.e., z>10).

Here is the paper on arxiv:


2011-Jan-13, 01:09 PM
Yes, the cosmological SN survey teams are aware of this possibility.

Note that one way (in theory) to quantify the effect is to compute the scatter around the mean brightness of SNe at particular redshifts; that is, the scatter of SNe at z=0.5, at z=1.0, at z=1.5, etc. If lensing is important, then one should see the scatter increase at high redshifts, since the likelihood that a SN is lensed should increase at high redshift.

Of course, in practice, there are a LOT of other factors that could also cause the scatter to increase systematically with redshift, such as measurement errors, changes in the metallicity of the progenitors, changes in the properties of dust in galaxies, changes in the mix of prompt and delayed SNe Ia, and so forth.