Space Warps Talk

SN Refsdal

  • Capella05 by Capella05 moderator

    (Atel 21 Nov 2014; 22:02 UT)

    We report the discovery of a multiply imaged, gravitationally lensed
    supernova, forming an Einstein cross around a z=0.544 early-type
    galaxy in the galaxy cluster MACS J1149.6+2223. The redshift of the
    supernova host galaxy is 1.491.

    http://archive.stsci.edu/pub/ffsn/macs1149/finder_lens_MACS1149.png

    New paper: Multiple Images of a Highly Magnified Supernova Formed by an Early-Type Cluster Galaxy Lens arXiv:1411.6009 from Patrick L. Kelly (UCB), Steven A. Rodney (JHU), Tommaso Treu (UCLA), et al.

    We report the discovery of the first multiply-imaged
    gravitationally-lensed supernova. The four images form an Einstein
    cross with over 2" diameter around a z=0.544 elliptical galaxy that is
    a member of the cluster MACSJ1149.6+2223. The supernova appeared in
    Hubble Space Telescope exposures taken on 3-20 November 2014 UT, as
    part of the Grism Lens-Amplified Survey from Space. The images of the
    supernova coincide with the strongly lensed arm of a spiral galaxy at
    z=1.491, which is itself multiply imaged by the cluster potential. A
    measurement of the time delays between the multiple images and their
    magnification will provide new unprecedented constraints on the
    distribution of luminous and dark matter in the lensing galaxy and in
    the cluster, as well as on the cosmic expansion rate.

    The MACS J1149.6+2223 galaxy cluster (Image: NASA, ESA, and M. Postman (STScI), and the CLASH team)

    New paper: Predicted properties of multiple images of the strongly lensed supernova SN Refsdal from Masamune Oguri (University of Tokyo)

    We construct a mass model of the cluster MACS J1149.6+2223 to study
    the expected properties of SN Refsdal, the first example of a
    gravitationally lensed supernova with resolved multiple images
    recently reported by Kelly et al. We find that the best-fit model
    predicts six supernova images in total, i.e., two extra images in
    addition to the observed four Einstein cross supernova images S1--S4.
    One extra image is predicted to have appeared about 17 years ago,
    whereas the other extra image is predicted to appear in about one year
    from the appearance of S1--S4, which is a testable prediction with
    near future observations. The predicted magnification factors of
    individual supernova images range from $\sim 18$ for the brightest
    image to $\sim 4$ for the faint extra images. Confronting these
    predictions with future observations should provide an unprecedented
    opportunity to improve our understanding of cluster mass
    distributions.

    Posted

  • Capella05 by Capella05 moderator

    We thought it would be a nice test of SpaghettiLens to try modelling this candidate to see if we were able to determine similar time delays (as referenced here )

    As with all our collabrative modelling attempts can you:

    • Only have a single model in a post.
    • List the parent of the model.
    • The ID of the model you have just generated
    • The changes you have made to the model
    • The results of your changes.
    • Thoughts for improvement.

    For Example:

    • Parent - 4516
    • Revised Model - 4568
    • Action - Moved the minima to the NE of the galaxy and added a mass point.
    • Result - The arc is more clearly defined from 7 to 11:30, not too happy about the predicted counter image at 2:45
    • Thoughts - Perhaps a cleaner mass distribution?

    Happy Modelling 😃

    Posted

  • Capella05 by Capella05 moderator

    To get us started:

    • Parent - Original
    • Revised Model - 12683
    • Action - Just a basic model to get us started.

    ..

    Time delay plot:

    enter image description here

    Posted

  • c_cld by c_cld

    other start http://mite.physik.uzh.ch/data/012692 Pixel Radius:11
    revised http://mite.physik.uzh.ch/data/012693 Pixel Radius:12

    Posted

  • Capella05 by Capella05 moderator

    • Parent - 12683
    • Revised Model - 12742
    • Action - Added the redshift values, increased pixel radius to 10. Changed shear to 0.21
    • Result - contour map and mass distribution has improved.
    • Thoughts - the synthetic image is less point like, will wait for the time delay graph.

    ..

    To post images you can use the following - this will keep all the images in the same scale 😃

     <IMG SRC="http://mite.physik.uzh.ch/result/012742/img2.png" WIDTH="220">

    Posted

  • ElisabethB by ElisabethB moderator

    • Parent - 12742
    • Revised Model - 12775
    • Action - Pushed the maxima down a little bit
    • Result - cleaner mass distribution.

    Posted

  • c_cld by c_cld

    Looks excellent!
    could you give levels (z, px, shear ) used?

    And please remind me how to downsize images by editing your post, Thanks 😃

    Posted

  • ElisabethB by ElisabethB moderator

    I only changed the shear to 0.2. (z 0.5/1 and px:8 )

    as for posting smaller images see Capella's previous post 😄

    Posted

  • c_cld by c_cld in response to ElisabethB's comment.

    Thanks Els 😃

    • Parent - 12775

    • Revised Model - 12781

    • Action - move contours close to plot, more important input redshifts for time delays zl=0.54 zs=1.49, px to 12

    • Result - images brightness increase, rounder mass core but external noiser mass with px 12

    Posted

  • Capella05 by Capella05 moderator

    I was chatting with Phil yesterday, and he reminded me we should take the cluster into account. I think we should consider placing some mass points in the model.

    Posted

  • ElisabethB by ElisabethB moderator

    • parent : 12775
    • revised model : 12805 - shear 0.2, z 0.5/1, px 10
    • action : introduced two mass points
    • result : image looks weird, mass distribution clean

    Posted

  • ElisabethB by ElisabethB moderator

    • parent : 12781
    • revised model : 12832
    • action : introduced two other mass points
    • result : not too sure, I'm just posting it because the mass distribution looks okay-ish

    remark: putting in mass points will have a disastrous effects on the synthetic image.

    Posted

  • Capella05 by Capella05 moderator

    I will get Rafi to have a look, all my models with mass points are not worthy to be posted.

    Posted

  • ElisabethB by ElisabethB moderator

    • parent : 12832
    • revised model : 12834
    • action : introduced another mass point
    • action : another crappy synthetic image with a reasonably clean contour map and mass distribution !

    Posted

  • c_cld by c_cld

    • parent : 12848 from Budgie
    • revised model : 12866
    • action : mass point toward brightest central galaxy of cluster ( outside image), input redshifts, px to 10
    • result : plausible model even it's doubtful to get the convergence and shear right with the superposed circles and marks on image input. ( some MACS1149.6+2223 HLA cutout is certainly preferable )

    Posted

  • c_cld by c_cld

    MACS1149+2223 RA = 177.396220 Dec = 22.403040 [11:49:35.093 +22:24:10.94] with location of SN Redsfal around small elliptical just above central star.

    Such Images from HLA could be downloaded in FITS format.

    http://hla.stsci.edu/cgi-bin/fitscut.cgi?red=hst_12068_a4_wfc3_ir_f140w&blue=hst_12068_a4_wfc3_ir_f105w&size=ALL&output_size=768&format=jpg&asinh=1&autoscale=99.50

    MACS1149+2223

    http://hla.stsci.edu/cgi-bin/fitscut.cgi?red=hst_12068_a2_wfc3_ir_f160w&blue=hst_12068_a2_wfc3_ir_f110w&size=ALL&output_size=768&format=jpg&asinh=1&autoscale=99.50

    enter image description here

    http://hla.stsci.edu/cgi-bin/fitscut.cgi?red=hst_12068_b8_wfc3_ir_f160w&blue=hst_12068_b8_wfc3_ir_f125w&size=ALL&output_size=768&format=jpg&asinh=1&autoscale=99.50

    enter image description here

    http://hla.stsci.edu/cgi-bin/fitscut.cgi?red=HST_9722_08_ACS_WFC_F814W&blue=HST_9722_08_ACS_WFC_F555W&size=ALL&output_size=768&format=jpg&asinh=1&autoscale=99.50

    enter image description here

    Happily, spikes of star don't blurred the gravitational lensing elliptical 😃

    Posted

  • c_cld by c_cld

    • parent 12866
    • revised model 12897
    • action : shear to 0.09 , px to 12
    • result : clean mass map

    Posted

  • ElisabethB by ElisabethB moderator

    • parent : 12866
    • revised model 12900
    • action: pushed the maximum down and downsized the mass point
    • result: cleaner mass distribution

    Posted

  • ElisabethB by ElisabethB moderator

    • parent 12900
    • revised model 12902
    • action: deleted the mass point
    • result: an even cleaner massdistribution. But on second look it looks a lot like the parent.

    Posted

  • Budgieye by Budgieye moderator

    http://mite.physik.uzh.ch/data/012686 just to make sure I could find web
    site, saddle points sloppy, wrong redshift

    http://mite.physik.uzh.ch/data/012837 moved points closer to green
    circles, forgot to use correct redshift, poor mass image, unbalanced contour plot

    Posted

  • c_cld by c_cld

    As I questionned the suitability of an annotated image (Kelly's paper) for modelling, I propose to pick the input from a HST dataset.

    For example I attached cropped image of MACS1149.6+2223 from dataset ICA524050 (instrument WFC3 F140W ) taken 2014-11-11 02:48:21 - 03:13:22

    SN Redsfal
    SN Redsfal

    3 dots of SN on top of elliptical at center!

    Posted

  • Budgieye by Budgieye moderator


    http://mite.physik.uzh.ch/data/012842 I moved the centre of the galaxy one pixel to the left, and disconnected points in mass distribution have disappeared, but contour plot looks lighter on the right side

    Posted

  • Budgieye by Budgieye moderator


    . http://mite.physik.uzh.ch/data/012843 I moved the centre another pixel to the left, mass looks nearly symmetrical except for a slight up and down stretch,

    Posted

  • Budgieye by Budgieye moderator


    http://mite.physik.uzh.ch/data/012844 I moved the centre one pixel up, it didn't make much difference to mass plot

    Posted

  • Budgieye by Budgieye moderator


    http://mite.physik.uzh.ch/data/012845 I moved the center point down, mass plot looking jagged, so maybe that was wrong.

    --

    Posted

  • Budgieye by Budgieye moderator


    I went back to 012843

    http://mite.physik.uzh.ch/data/012846 I pulled S4 down a pixel, not much
    change

    http://mite.physik.uzh.ch/data/012847 I moved S1 one pixel to left, not
    much change

    http://mite.physik.uzh.ch/data/012848 I moved S3 to the right, not much
    change

    Posted

  • Budgieye by Budgieye moderator


    For fun.

    http://mite.physik.uzh.ch/data/012851 I added a point mass and got a
    weird mass plot that looked like modern art. The synthetic image was
    strange, but it slightly resembled the image if you include the red
    circles which are supposed to be some other lensed object. A coincidence
    I suppose.

    Maybe I had gone as far as I could go.


    Posted

  • Budgieye by Budgieye moderator

    Then I looked at Claude's http://mite.physik.uzh.ch/data/12692. which is my

    http://mite.physik.uzh.ch/data/12850

    I liked his synthetic image, but his mass distribution had disconnected points
    were not as good as mine. All the points in his image seemed slightly to
    the right of mine. So I moved his centre of the galaxy to the left. I
    was surprised how long his simulation took, much longer than mine. I got
    http://mite.physik.uzh.ch/data/012849 and the mass distribution very
    very poor and synthetic image very bad. I tried moving all his points to
    the left, but results were very bad. http://mite.physik.uzh.ch/data/012850

    Posted

  • c_cld by c_cld

    • parent: 12805
    • revised model: 12912
    • action: input redshifts zl 0.54 zs 1.49, px to 12, same shear 0.2; no change in points/contours but erase mass points
    • result: similar round mass map (higher px resolution), brighter lensed SN images, brightest slight shift from S1 to S2

    Posted

  • Capella05 by Capella05 moderator in response to C_cld's comment.

    Wow -love that last one! The synthetic image is almost spot on.

    Nice one 😃

    Posted

  • ElisabethB by ElisabethB moderator in response to C_cld's comment.

    • parent : 12912
    • revised model : 12991
    • action : dropped the maximum a bit

    Posted

  • ElisabethB by ElisabethB moderator

    • parent 12991
    • revised model 12995
    • actions : introduced a mass point and shifted the saddle point to the bottom in between of of the green and red circle.

    Posted

  • c_cld by c_cld in response to ElisabethB's comment.

    In aiming to get likely time delays, I think we have to model this SN Refsdal with the best information available.

    Your revised models don't keep the known redshifts ( and px) which could change significantly the time delays in my opinion.

    About the mass point, the main galaxy of the cluster is outside the image input (see HST cutout). Your point is located in a lensed spiral arm of the galaxy close to the elliptical lens.

    I won't follow you on that model and I would have preferred to model the better gray cutout of investigator Kelly:

    http://www.newscientist.com/article/dn26652-spotted-first-quadruple-star-image-produced-by-gravity.html#.VIBAkTGsWSp

    SpaghettiLens is a splendid toy 😃

    Posted

  • Capella05 by Capella05 moderator

    In aiming to get likely time delays, I think we have to model this SN Refsdal with the best information available.
    Your revised models don't keep the known redshifts ( and px) which could change significantly the time delays in my opinion.

    Perhaps we should start posting the parameters used to generate our models?

    As for the mass points - I agree with Claude.

    Posted

  • ElisabethB by ElisabethB moderator

    I wasn't too happy with the last ones either, just trying out things . But a good idea to include the parameters.

    Posted

  • ElisabethB by ElisabethB moderator

    • parent : 12775
    • revised model : 13114
    • action : dropped maximum, added poin mass
    • z 0.5/1, px 8 shear 0.2

    Posted