View Full Version : bending of light (and x-rays, radio, ..) by mass - how do we know it's real?

2006-Aug-06, 01:06 PM
In another Q&A thread (http://www.bautforum.com/showthread.php?t=45314), on gravitational lensing, the following question was raised:
How do we know that these four images in the links given by Nereid are actually lensed images of something behind the galaxy and not just separate objects nearby?The "four images" are the famous (http://antwrp.gsfc.nasa.gov/apod/ap050327.html) Einstein (http://www.astr.ua.edu/keel/agn/qso2237.html) cross (http://geogdata.csun.edu/~voltaire/roland/cross.html) (QSO 2237+0305 (http://www.journals.uchicago.edu/cgi-bin/resolve?id=doi:10.1086/374548); one more (http://adsabs.harvard.edu/abs/1999AAS...195.1708A)).

This is both a specific question, about this particular object (which, if you follow the links, you'll learn has been imaged in the x-ray, optical, and IR wavebands), and a general one, about how to determine whether gravitational lensing can account for an observation.

(More generally still, we could ask about the tests that one could do to ascertain the extent to which the deflection ("bending") of em radiation ("light") by mass is observed. This would include the apparent position changes of distant objects as they get near the Sun (and Jupiter), on the sky, weak lensing, shear, and the hundreds of events detected by searches such as MACHO and OGLE.)

How confidently can explanations other than gravitational lensing (strong or weak) be ruled out, for the now hundreds (hundreds of thousands?) of astronomical observations?

Tim Thompson
2006-Aug-06, 07:07 PM
How confidently can explanations other than gravitational lensing (strong or weak) be ruled out, for the now hundreds (hundreds of thousands?) of astronomical observations?
It's hard to separate this question entirely from that other Q&A thread (http://www.nsf.gov/od/lpa/news/press/00/pr0029.htm), because the answer depends on the theory of gravitational lensing, which is the topic of that other thread. Simply put, we observe something we think of as lensing, and then compare the details of the observsation to the details of theoretical predictions. If they are close, we say "lensing", and if they are not close we say "not lensing", and if they are almost close, we say "maybe lensing, maybe not lensing". Yee, 1988 (http://adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode=1988AJ.....95.1331Y&db_key=AST&d ata_type=HTML&format=) calls the Einstein Cross only a "candidate" for gravitational lensing, because they did not yet have enough data to be sure. On the other hand, Schneider, et al., 1988 (http://adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode=1988AJ.....95.1619S&db_key=AST&d ata_type=HTML&format=), are more confident, and conclude from their observtions that QSO 2237+0305 must be a gravitational lens.

One way to be sure you have a lens, perhaps independently from the model, is the case where the several lensed images of the same object all have the same detailed spectrum, aside from the amazing spatial coincidence of having random quasars "cross" a forground galaxy. But the only way to provide a real answer to a question like this is to examine the details of observation & analysis, as reported in the literature, for each instance. Bolton, et al., 2006 (http://adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode=2006ApJ...646L..45B&db_key=AST&d ata_type=HTML&format=&high=4366fa465112410) report the discovery of a new "Einstein cross" type lens, of unusually wide separation. In the case of the Einstein Cross, see for instance, Chae, Turnshek & Kheronsky, 1998 (http://adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode=1998ApJ...495..609C&db_key=AST&d ata_type=HTML&format=&high=4366fa465112410) (evidently not available on the web). But I think that the several references I posted (http://www.bautforum.com/showpost.php?p=799762&postcount=10) in the other thread must also shed some light on this question.

2006-Aug-07, 06:39 PM
There is also the case in gravitational lensing where a change in one of the images is also seen in the other images. This would only work if the path length differences are only on the order of ly.