Host galaxy association for LSST transients: request for comments

Tags: #<Tag:0x00007f27af9bbc10> #<Tag:0x00007f27af9bbb48> #<Tag:0x00007f27af9bba80>

The Data Management System Science Team proposes to add to the DIAObject table a total of four new elements that will provide more scientifically useful associations for transients and their potential host galaxies. This proposal is documented in RFC-695 and in DMTN-151, and summarized below.

Comments and questions about this proposal for new parameters related to DIAObjectObject host galaxy association are welcome.

Currently, the two DIAObject table elements (Table 3, DPDD) that are most relevant to the association of transients with their potential host galaxies are:

  • nearbyObj (uint64[6]) “Closest Objects (3 stars and 3 galaxies) in DR database.”
  • nearbyObjDist (float[6]) “Distances to nearbyObj.” [arcsec]

Scientifically, this is inadequate for two main reasons:

  1. The radial distance in arcsec matters less than a separation distance which accounts for the galaxies’ spatial luminosity profile (e.g., Sullivan+06).
  2. Transients in the local volume might not be associated with their large extended host galaxies, especially those in the outskirts of their hosts.

To relieve the first issue, it is proposed that the following two elements should be computed during Alert Production and added to the DIAObject table:

  • potentialHost (uint64[3]) – the objectId for the three Object catalog extended sources with the lowest separation distances
  • potentialHostSeparation (float[3]) – the separation distances for each potentialHost

The separation distance would be calculated with respect to the transient location using the second moments of each Object's luminosity profile, as described in Section 3. of DMTN-151.

To relieve the second issue, Data Management will adopt an established catalog of low-redshift galaxies (e.g., from the Revised New General Catalogue and Index Catalogue) and do a cross match during Alert Production in order to report the name and the radial sky separation of the nearest catalog galaxy:

  • potentialLowzHost (str[1]) – the catalog name of the nearest “low-z” potential host
  • potentialLowzHostSeparation (float[1]) – the distance to potentialLowzHost [arcsec]

If the community provides a catalog of names and coordinates that they would prefer to be used, it will be adopted. If that catalog includes more information about the galaxy then the value of potentialLowzHostSeparation could be more sophisticated than the radial distance in arcseconds. For example, if distances are included then the offset in kpc could be provided; if sizes are included the number of effective radii could be reported; or if the second moments are included then the separation could be the same as used for potentialHostSeparation.

Thanks for raising this issue.

Indeed, closest objects may be inadequate in particular for extra-galactic transients. The proposal of adding potential host information using the luminosity profiles will be very valuable. Even more if it is possible to enrich this information with host redshifts if available from a catalogue and separation distance in kpc.