In Torrey+15, it is mentioned that "image field of view is set to be 10 times the stellar half-mass radius for the galaxy". Is that rule valid for any redshift? In that case, I cannot understand the scale of e.g. subhalo # 9989 at z=2, as shown in Figure 1 of Wellons+15. I have downloaded the corresponding stellar_mocks/broadband.fits, the FoV should be 15.5 kpc, right?. I can roughly measure the FoV of the Wellons+15's cutout as ~ 6 kpc, far from the stated 30 kpc. Am I missing a scale factor?
Thank you
Sarah Wellons
14 May '17
The images in Wellons+15 were generated separately from the Torrey+15 mocks, so the 10x half-mass radius FoV rule does not apply. The FoV of that image should be 30 kpc as stated. The FoV in the mock catalog should be 15.5 kpc, as you mention. How are you getting your ~6 kpc measurement of the Wellons+15 image? You can tell just by visually estimating the half-light radius that the FoV must be at least 10x that.
Ignacio G. de la Rosa
15 May '17
I download the broadband_9989_cam3.fits. After some mirror+rotation, I can recognize the features (arms+arcs) of the image shown in Figure 1 and work out the scale and calculate the ~ 6 kpc. It looks odd that, with a canonical 15.5 kpc FoV, this downloaded image seems to span a larger FoV than the published 30 kpc one. Similar comment can be made for # 51056. These galaxies are pretty structureless, so I can be absolutely wrong on the features identification. It would help to e.g. know the scale and subhalo number of the lower row galaxies, full of traceable features.
Hmm.... ok, I think I have downloaded the same file that you did, but I have never actually interacted with the public database before so take what I say with a grain of salt. I have not figured out how to display the image properly yet (...theorist here, sorry...) BUT I managed to look into the header and there's a line in there that says "HIERARCH linear_fov = 119.919133 / [kpc] Linear field of view in y-dir at origin", which would seem to agree with your assessment that the FoV is larger than the one in my image. I am not sure why it is this size - let me consult with Paul Torrey and see if I can get a better answer for you.
Ignacio G. de la Rosa
19 May '17
Thank you! Your answer has helped to solve the problem: the idea is to check the "linear_fov [kpc]" in the image header. By doing that, I have discovered that:
(i) There is an almost fixed (size independent) linear_fov ~ 119.919133 kpc for all the images at z>0
(ii) At z=0, the "linear_fov" = 10 x "halfmassrad_stars" [kpc]. However, "halfmassrad_stars" is given in units of [ckpc/h]. I ignore why "h" has not been taken into account.
Sorry for this further simple question: is it possible, just from the public database information, to create subhalo pictures with spatial distribution of the in-situ and ex-situ star particles, like in Wellons+16 (Figure 3)?
Sarah Wellons
20 May '17
Ah great, I'm glad that has been cleared up! It would be possible, but unfortunately fairly complicated, to make the in-situ/ex-situ images you describe using the public data. You would need to traverse the merger tree back to the time that each star particle was formed to determine whether they belong to the main progenitor of the galaxy in question, which could get somewhat computationally intense. There are a number of ways you could do this, but for an individual galaxy I think the fastest procedure might be to use the merger tree to find the main progenitor at each snapshot, then step along the main branch and make a running tally of the IDs of the in-situ star particles by identifying at each step all the star particles which were formed since the last snapshot. After doing this for each progenitor along the main branch, any star particle in the galaxy at z=0 whose ID is not on that list would be ex-situ.
In Torrey+15, it is mentioned that "image field of view is set to be 10 times the stellar half-mass radius for the galaxy". Is that rule valid for any redshift? In that case, I cannot understand the scale of e.g. subhalo # 9989 at z=2, as shown in Figure 1 of Wellons+15. I have downloaded the corresponding stellar_mocks/broadband.fits, the FoV should be 15.5 kpc, right?. I can roughly measure the FoV of the Wellons+15's cutout as ~ 6 kpc, far from the stated 30 kpc. Am I missing a scale factor?
Thank you
The images in Wellons+15 were generated separately from the Torrey+15 mocks, so the 10x half-mass radius FoV rule does not apply. The FoV of that image should be 30 kpc as stated. The FoV in the mock catalog should be 15.5 kpc, as you mention. How are you getting your ~6 kpc measurement of the Wellons+15 image? You can tell just by visually estimating the half-light radius that the FoV must be at least 10x that.
I download the broadband_9989_cam3.fits. After some mirror+rotation, I can recognize the features (arms+arcs) of the image shown in Figure 1 and work out the scale and calculate the ~ 6 kpc. It looks odd that, with a canonical 15.5 kpc FoV, this downloaded image seems to span a larger FoV than the published 30 kpc one. Similar comment can be made for # 51056. These galaxies are pretty structureless, so I can be absolutely wrong on the features identification. It would help to e.g. know the scale and subhalo number of the lower row galaxies, full of traceable features.
Many thanks for your answer
Hmm.... ok, I think I have downloaded the same file that you did, but I have never actually interacted with the public database before so take what I say with a grain of salt. I have not figured out how to display the image properly yet (...theorist here, sorry...) BUT I managed to look into the header and there's a line in there that says "HIERARCH linear_fov = 119.919133 / [kpc] Linear field of view in y-dir at origin", which would seem to agree with your assessment that the FoV is larger than the one in my image. I am not sure why it is this size - let me consult with Paul Torrey and see if I can get a better answer for you.
Thank you! Your answer has helped to solve the problem: the idea is to check the "linear_fov [kpc]" in the image header. By doing that, I have discovered that: (i) There is an almost fixed (size independent) linear_fov ~ 119.919133 kpc for all the images at z>0 (ii) At z=0, the "linear_fov" = 10 x "halfmassrad_stars" [kpc]. However, "halfmassrad_stars" is given in units of [ckpc/h]. I ignore why "h" has not been taken into account.
Sorry for this further simple question: is it possible, just from the public database information, to create subhalo pictures with spatial distribution of the in-situ and ex-situ star particles, like in Wellons+16 (Figure 3)?
Ah great, I'm glad that has been cleared up! It would be possible, but unfortunately fairly complicated, to make the in-situ/ex-situ images you describe using the public data. You would need to traverse the merger tree back to the time that each star particle was formed to determine whether they belong to the main progenitor of the galaxy in question, which could get somewhat computationally intense. There are a number of ways you could do this, but for an individual galaxy I think the fastest procedure might be to use the merger tree to find the main progenitor at each snapshot, then step along the main branch and make a running tally of the IDs of the in-situ star particles by identifying at each step all the star particles which were formed since the last snapshot. After doing this for each progenitor along the main branch, any star particle in the galaxy at z=0 whose ID is not on that list would be ex-situ.