APOD: Galaxy Einstein Ring (2016 Apr 20)

[geckzilla]

Curious that the quoted size is a few arc seconds yet the Hubble image is highly pixellated. With Hubble's resolution of 0.05 arc seconds, it surely shouldn't appear like this. Are the merged images on vastly different scales?

The pixel size for the WFC3/IR channel is 0.13 arcsec. Does it make more sense now?

Actually, it still isn't obvious what's going on. Here's the synthesized ring image combined with the actual IR/F160W Hubble image:

gal.jpg

Here we can see the actual pixels that correspond to the visible light image of the lensing galaxy. The image seen in today's APOD apparently consists of a mathematical fit of the two datasets, resulting in synthetic pixels (of irregular dimension) that don't directly correlate to either of the original datasets.

When I looked at that same image in the paper the pixels didn't seem to be at the right scale, either. They look about twice as large as they should be.

[Chris Peterson]

When I looked at that same image in the paper the pixels didn't seem to be at the right scale, either. They look about twice as large as they should be.

In the image, they're 0.26 arcsec/pixel, which is exactly half the native resolution of the camera. That strongly suggests that the image is binned 2x2, either in the original data or in subsequent processing. You're the master of tracking down HST images. Maybe you can find the original science images.

[DorenaScientist]

Using modern computer techniques such as ray tracing, is it possible to remove the distortion and obtain a direct picture of the lensed object?

Probably not. A gravitational lens doesn't produce an image like a conventional lens; even if the lens were perfectly characterized, there are normally multiple sources that could produce the same sort of ring. But we have no way of perfectly characterizing the lens, so that leaves even more possibilities.

Oh, but check this out from NRAO Outreach: https://vimeo.com/15897134 This is just animation, but it is using sophisticated data analysis and algorithms to render it, and NRAO is producing it. Cool.

[geckzilla]

Using modern computer techniques such as ray tracing, is it possible to remove the distortion and obtain a direct picture of the lensed object?

Probably not. A gravitational lens doesn't produce an image like a conventional lens; even if the lens were perfectly characterized, there are normally multiple sources that could produce the same sort of ring. But we have no way of perfectly characterizing the lens, so that leaves even more possibilities.

Oh, but check this out from NRAO Outreach: https://vimeo.com/15897134 This is just animation, but it is using sophisticated data analysis and algorithms to render it, and NRAO is producing it. Cool.

I'm guessing you meant to post this instead of skateboarding: https://vimeo.com/158971342

[geckzilla]

When I looked at that same image in the paper the pixels didn't seem to be at the right scale, either. They look about twice as large as they should be.

In the image, they're 0.26 arcsec/pixel, which is exactly half the native resolution of the camera. That strongly suggests that the image is binned 2x2, either in the original data or in subsequent processing. You're the master of tracking down HST images. Maybe you can find the original science images.

I was already looking at them. It's a tiny galaxy occupying a tiny fraction of the frame in a seemingly unremarkable image. Red box for reference: http://geckzilla.com/astro/SDP.81.png

And don't be fooled into thinking any of those arcs are lensing images. Whatever they are, they don't line up with the radio data. I have no idea how they even knew to look there.

[alter-ego]

...
And don't be fooled into thinking any of those arcs are lensing images. Whatever they are, they don't line up with the radio data. I have no idea how they even knew to look there.

SDP.81 was discovered in 2010 by the Herschel Space Observatory (FIR to SM wavelengths).

Using data from the Herschel Space Telescope, Negrello et al. (p. 800) showed that by searching for the brightest sources in a wide enough area in the sky it was possible to detect gravitationally lensed submillimeter galaxies with nearly full efficiency.

A relatively rapid search revealed SDP.81 as a highly probable, gravitationally lensed source, and was subsequently included in ALMA's Long Baseline Campaign (2014). The abundance of substructure in the ring probably made it the candidate-of-choice for Hezaveh et al.

When I looked at that same image in the paper the pixels didn't seem to be at the right scale, either. They look about twice as large as they should be.

In the image, they're 0.26 arcsec/pixel, which is exactly half the native resolution of the camera. That strongly suggests that the image is binned 2x2, either in the original data or in subsequent processing.

The paper mentions that binning was necessary to make the analyses tractable. Once a good fit was obtained for binned (pixelated) data, the same visibility binning and noise estimations were applied to mock (simulated) data. Then a best-fit smooth model could be found. This probably explains the smoothness of the APOD's Einstein Ring, but I don't understand why the paper used a binned (pixelated) HST image. Maybe it's simply meant to show the binned FoV used in the analyses?

[Chris Peterson]

In the image, they're 0.26 arcsec/pixel, which is exactly half the native resolution of the camera. That strongly suggests that the image is binned 2x2, either in the original data or in subsequent processing.

The paper mentions that binning was necessary to make the analyses tractable. Once a good fit was obtained for binned (pixelated) data, the same visibility binning and noise estimations were applied to mock (simulated) data. Then a best-fit smooth model could be found. This probably explains the smoothness of the APOD's Einstein Ring, but I don't understand why the paper used a binned (pixelated) HST image. Maybe it's simply meant to show the binned FoV used in the analyses?

I saw those references. The paper is pretty dense, and I'm not a radio astronomer, so there's some terminology I don't understand. I have the impression that the radio data binning under discussion isn't spatial, however, but in the frequency or phase domains, and that the radio data in spatial form (as we see it in the image) is quite a bit higher resolution than the optical data, even unbinned.

[geckzilla]

I went ahead and did some simple processing to make a better version of the Hubble image of the galaxy. I included the radio contours from the paper in the image. You can see them as the little white ring around the center galaxy.

SDP.81 by Judy Schmidt, on Flickr

[alter-ego]

In the image, they're 0.26 arcsec/pixel, which is exactly half the native resolution of the camera. That strongly suggests that the image is binned 2x2, either in the original data or in subsequent processing.

The paper mentions that binning was necessary to make the analyses tractable. Once a good fit was obtained for binned (pixelated) data, the same visibility binning and noise estimations were applied to mock (simulated) data. Then a best-fit smooth model could be found. This probably explains the smoothness of the APOD's Einstein Ring, but I don't understand why the paper used a binned (pixelated) HST image. Maybe it's simply meant to show the binned FoV used in the analyses?

I saw those references. The paper is pretty dense, and I'm not a radio astronomer, so there's some terminology I don't understand. I have the impression that the radio data binning under discussion isn't spatial, however, but in the frequency or phase domains, and that the radio data in spatial form (as we see it in the image) is quite a bit higher resolution than the optical data, even unbinned.

Yeah, I'm not a radio astronomer either, and dense is an understatement!
From what I've read, a "visibility" is the product of frequency data sets between various antennas. 2D spatial images are obtained by Fourier transforming the visibility (frequency) files. (For the SDP.81 analysis, two bands 6 & 7, treated as spectral channels, were used; after binning each channel containing ~10⁶ visibilities). Binned frequency data will result in binned transformed data (pixelated images).

I don't follow your comment about comparing radio data in spatial form to optical data, but I believe I understand ALMA spatial resolution limits:
The maximum reduced spatial resolution for SDP.81 image is 23mas (listed elsewhere). For a maximum 16km baseline, the theoretical spatial resolution limits for Band 6 (λ_center=243GHz) and Band 7 (λ_center=342GHz) are ≈18mas and ≈12mas respectively (ALMA Basics and ALMA Front End). ALMA's Long Baseline Campaign achieved 15km, so these resolutions are compatible with the claimed 23mas SDP.81 resolution. However, the paper mentioned maximum 2km baseline data, and also mentioned that the best-fit parameters were found with ≤2km baselines:

"We first use only the
subset of baselines shorter than 2 km in our initial MCMC
analysis to localize the neighborhood of the best fit. This
Figure 4. The SDP.81 system. Grayscale shows HST/WFC3 F160W data,
while red contours show ALMA continuum emission in band 6.
approach greatly expedites our MCMC optimization, since
many evaluations of the likelihood are required to fully search
the highly multidimensional parameter space of our smooth
model. Once this initial localization has been achieved, we
can use the full dataset.

Does the full data set mean a 15km baseline? It must because a 2km baseline yields a maximum resolution ~8x larger, or 144mas and 96mas respectively for Bands 6 and 7 - poorer resolution by 4x to 6x.
Interestingly, in the HST overlap image, I'm thinking the red Band 6 continuum contours may correspond to the 15km baseline data sets. Contour features are 10% or less of the 0.26" HST pixel size.

[alter-ego]

I went ahead and did some simple processing to make a better version of the Hubble image of the galaxy. I included the radio contours from the paper in the image. You can see them as the little white ring around the center galaxy.

Hey, really nice job with the image make-over! I read your comments in the Flickr post.
I wanted to check the lensed arc locations at the NIR and SM wavelengths; I used a zoomed-in view of the HST image (eso1522d.​tiff) and the red ALMA SM contour map from the paper. The comparison below looks pretty good. Possibly a better looking overlap might be had by translating the contour map a few pixels toward the upper right, but I was satisfied with the initial overlap quality. Basically, the brightest NIR arc positions reasonably agree with the SM arc positions. I didn't analyze your composite well enough to see the difference, so I'm not sure what looks wrong to you. There is some uncertainty in creating this composite. I believe my overlap error could be a couple pixels.

Click to view full size image 1 or image 2

I was a bit surprised that the pixel resolution = 90mas and not 130mas. I doubled checked the FoV and image orientation the STScI Digitized Sky Survey - all are right on.

[geckzilla]

The NIR lensed images really don't look like they line up with the radio contours at all to me. They look a bit better aligned with your image but not really with mine. I dunno. It was so off I was questioning if the north orientation was even right.

[Ann]

I went ahead and did some simple processing to make a better version of the Hubble image of the galaxy. I included the radio contours from the paper in the image. You can see them as the little white ring around the center galaxy.

SDP.81 by Judy Schmidt, on Flickr

That's a really really nice image, Geck. It's great to see the whole galaxy cluster, not just the lensing and the lensed galaxies.

The galaxy cluster image looks moderately "true color" to me. Do you know what the filters were?

Ann

[geckzilla]

That's a really really nice image, Geck. It's great to see the whole galaxy cluster, not just the lensing and the lensed galaxies.

The galaxy cluster image looks moderately "true color" to me. Do you know what the filters were?

F160W and F110W. The colors are barely even visible without modifying the saturation. It's one of those things where even though it's not super colorful, I still had to boost it immensely.

[alter-ego]

The galaxy cluster image looks moderately "true color" to me. Do you know what the filters were?

Just goes to show you how "true color" appearances may have nothing to do with the individual filters. Both filters transmit light that's not visible to the unaided eye. The image colors are the result of processing choices.

[geckzilla]

The galaxy cluster image looks moderately "true color" to me. Do you know what the filters were?

Just goes to show you how "true color" appearances may have nothing to do with the individual filters. Both filters transmit light that's not visible to the unaided eye. The image colors are the result of processing choices.

I would argue the image colors are the result of the light frequencies sent from the galaxies. I didn't try to force the elliptical galaxies to look redder or for the arms of spiral galaxies to look bluer. That's how they really were in the data. I just increased the separation to make the colors much more easily perceptible. That shows you that a computer has no problem "seeing" the colors. It's there in the numbers.