APOD: Galaxy Cluster Magnifies Distant... (2014 May 05)

[geckzilla]

It's something about the way they get lensed. There are always often oddly red cores with oddly blue arms for lensed galaxies. . . .

I'm thinking that when we make our own lenses, we make them in a shape that is convenient, to produce good optical results. However, these natural gravitational lenses are not shaped for our purposes. So, I'm wondering if that produces the kind of "optics" you are reporting. Are most gravitational lenses of a similar "shape", perhaps like looking at something through a spherical crystal ?

It's like a magnifying glass as far as I know. Hubble is being used to look at them because they let us see farther than we normally would. I have absolutely no idea how but it also seems to reduce the redshift. I could be completely misunderstanding it, though. Lensed galaxies always look pretty strange because of the distortion they go through. I think maybe if when processing a galaxy cluster you could balance the colors for the lensed z-depth only and they would look more normal. I probably need to try this myself at some point. In fact, I will make it a point to look at the CLASH data and pick one to do soon.

[Ann]

Judging from how I read the caption, the host galaxy can actually be seen in this picture, and it appears to be much less red (or magenta) than the supernova. The best explanation I can think of is that the supernova created an incredible amount of either infrared or redshifted visible light, but not very much UV light. The galaxy, on the other hand, emits a lot of UV light, which is seen as blue, yellow and orange light in today's APOD.

Here's a careful enlargement from the color managed original:

tiberius.jpg

The supernova itself shows 46% of its total energy emitted from 491-748 nm, 25% of its total energy emitted from 362-397 nm, and 29% of its total energy emitted from 203-292 nm. That makes the UV component a little high, perhaps, but this seems largely consistent with most supernovas.

The galaxy itself is a little harder to figure. Most of it is highly red biased, which is absolutely normal, meaning that the emitted light was substantially white (as is the case with virtually all galaxies). The blue edge, however, has most of its energy emitted in UV. Perhaps this galaxy is actually a pair in collision, or something else is going on producing a front of intense star formation. Even so, I'd expect more intensity in the visible than the IR, so something odd is happening there.

Thanks, Chris. The supernova is not at all UV deficient, then, but a large part of the galaxy is strangely UV bright.

Ann

[Chris Peterson]

I think maybe if when processing a galaxy cluster you could balance the colors for the lensed z-depth only and they would look more normal. I probably need to try this myself at some point. In fact, I will make it a point to look at the CLASH data and pick one to do soon.

Try it with this one. Map the 1050 filter to blue, the 1100 filter to green, and the 1250, 1400, and 1600 filters to red. That will be quite close to the emitted ranges for z=1.14. Then you'll just have to play around with color balance.

[Chris Peterson]

I'm thinking that when we make our own lenses, we make them in a shape that is convenient, to produce good optical results. However, these natural gravitational lenses are not shaped for our purposes. So, I'm wondering if that produces the kind of "optics" you are reporting. Are most gravitational lenses of a similar "shape", perhaps like looking at something through a spherical crystal ?

No, gravitational lenses are nothing like our optical lenses. An optical lens has no power through its optical center, and increases in power as you move towards its edge. A gravitational lens has its maximum power in the center, and the power drops off with radial distance. If you were to make an optical equivalent of a gravitational lens, it would look something like the base of a wine glass, with the maximum curvature in the center and the minimum at the edges (and a singularity at the very center).

Gravitational lenses can't bring a distant object to focus; optically, they have no focal plane.

[JuanAustin]

I've noticed thru the years of being an avid fan of APOD that photos that appear in their original format and sources are often opposite when they appear in APOD. I guess it doesn't matter in the grander scheme of life, but it's kinda annoying and i have to wonder why go thru the trouble of flipping and reorientating the originals? Not a big deal, just saying. thanks

[geckzilla]

I've noticed thru the years of being an avid fan of APOD that photos that appear in their original format and sources are often opposite when they appear in APOD. I guess it doesn't matter in the grander scheme of life, but it's kinda annoying and i have to wonder why go thru the trouble of flipping and reorientating the originals? Not a big deal, just saying. thanks

To me it seems it's just an aesthetic choice by one of the editors and other times it's to bring the subject into focus. I've never seen any flipping done but sometimes things are rotated, stretched, or cropped and they are almost invariably resized.

[NGC3314]

On the discovery of gravitational lensing - the popular-level book The Sky at Einstein's Feet devoted 2 chapters to the discovery and exploitation of gravitational lensing. I think it's now print-on-demand, but most of it can be found from Google Books here. There were, as so often happens, a lot of twists and turns along the way.

[alter-ego]

No, gravitational lenses are nothing like our optical lenses. An optical lens has no power through its optical center, and increases in power as you move towards its edge. A gravitational lens has its maximum power in the center, and the power drops off with radial distance. If you were to make an optical equivalent of a gravitational lens, it would look something like the base of a wine glass, with the maximum curvature in the center and the minimum at the edges (and a singularity at the very center).

Gravitational lenses can't bring a distant object to focus; optically, they have no focal plane.

A good visual helps with this analogy:

Click to play embedded YouTube video.

[geckzilla]

I think maybe if when processing a galaxy cluster you could balance the colors for the lensed z-depth only and they would look more normal. I probably need to try this myself at some point. In fact, I will make it a point to look at the CLASH data and pick one to do soon.

Try it with this one. Map the 1050 filter to blue, the 1100 filter to green, and the 1250, 1400, and 1600 filters to red. That will be quite close to the emitted ranges for z=1.14. Then you'll just have to play around with color balance.

Ok, not all of the filters covered that particular area. There's a problem! The 1100 filter did not reach so I was forced to use the next one up. They are not separated far enough for a pleasing range of color, but you can already see that actually, this supernova is not leaning toward red at all. It's slightly blue-green assuming I didn't do this incorrectly. Also it is very apparent this is a merger. The two cores are plain as day. The western one is slightly yellower. I left the two smaller nearby galaxies for context. They're probably all interacting.

Supernova behind Abell 383. North is up, east is left.

Here's a link to the overall image at 50% size, ~2MB file. This is minimally processed. Colors were balanced and seams were made less apparent. Some egregious anomalies were removed.
http://www.geckzilla.com/astro/Abell_383.jpg

[southern cross]

Thanks Chris, for the insight, especially about the impact of ccds.