I'm interested in rendering the emission line surface brightness from several line species for TNG galaxies and gas in their local environment. I'm planning to use these rendered images as inputs to simulated telescope data.
The web api for the visualization provides a very nice tool to do this for a handful of sources -- I can run a query to identify sources of interest, and then manually construct a url for downloading arrays of (for example) OIII or Lyman-alpha emission. My typical request ends up looking like this:
My particular application would benefit from being able to do this at a larger scale (a few 10^4 sources would be ideal). The provided web service is quite slow, however, and I've noticed that the remote connection frequently hangs.
It seems that the modeling and rendering for this is done with CLOUDY and on the fly. Would it be possible to expose the software you're using to do this on the Jupyterhub server? If this is already documented somewhere, I'd be grateful for a link.
Dylan Nelson
11 May '21
Hello Eric,
Glad that you find this functionality useful - because it is coming from code which hasn't yet been released/documented (although this is a future goal), this is intended more as an exploratory tool. If you would like to do serious science based on emission line mocks, I would certainly suggest to develop a more sophisticated pipeline.
As you say, this is using CLOUDY, with a number of assumptions/chosen parameters. It is slow since it has to read many of the fields of the gas cells each time, to derive the emission. And yes, unfortunately it isn't yet available on the Lab, but you are of course welcome to use the API - in this case, please in serial (i.e. no more than 1 request at a time). Thus a few 10^4 sources is realistically too much.
I'm interested in rendering the emission line surface brightness from several line species for TNG galaxies and gas in their local environment. I'm planning to use these rendered images as inputs to simulated telescope data.
The web api for the visualization provides a very nice tool to do this for a handful of sources -- I can run a query to identify sources of interest, and then manually construct a url for downloading arrays of (for example) OIII or Lyman-alpha emission. My typical request ends up looking like this:
https://www.tng-project.org/api/TNG100-1/snapshots/67/subhalos/50/vis.hdf5?partType=gas&partField=sb_H--1-6562.81A&size=2&sizeType=arcmin
My particular application would benefit from being able to do this at a larger scale (a few 10^4 sources would be ideal). The provided web service is quite slow, however, and I've noticed that the remote connection frequently hangs.
It seems that the modeling and rendering for this is done with CLOUDY and on the fly. Would it be possible to expose the software you're using to do this on the Jupyterhub server? If this is already documented somewhere, I'd be grateful for a link.
Hello Eric,
Glad that you find this functionality useful - because it is coming from code which hasn't yet been released/documented (although this is a future goal), this is intended more as an exploratory tool. If you would like to do serious science based on emission line mocks, I would certainly suggest to develop a more sophisticated pipeline.
As you say, this is using CLOUDY, with a number of assumptions/chosen parameters. It is slow since it has to read many of the fields of the gas cells each time, to derive the emission. And yes, unfortunately it isn't yet available on the Lab, but you are of course welcome to use the API - in this case, please in serial (i.e. no more than 1 request at a time). Thus a few 10^4 sources is realistically too much.