LOR Example 1: A Science Use-Case for PZ

Title: LOR Example 1: A Science Use-Case for PZ
Contributors: Melissa Graham
Co-signers: N/A

This is not a real LOR. This is an example of the minimum contents that the Rubin Data Management team is looking for in these letters. Many LOR written to describe a science-use case for LSST PZ will probably be longer than this minimum, but LOR at this level are still helpful.

0. Summary Statement
This example LOR uses PZ for transient host galaxies for the science case and recommends that an estimator which can deliver “PZ with <X% uncertainty at redshift z“ be shortlisted. X and z are being used instead of some estimated values in order to emphasize that this is just an example. Someone should please write a real LOR for real transient host galaxy science with real values!

1. Scientific Utility
I would use PZ both as an approximate distance estimator and as a galaxy type classifier for volumetric supernova rate analyses. Having PZ for LSST transient host galaxies will allow me to compare the rates of different transients in different types of environments, like rich galaxy clusters, like I did with the Supernova Legacy Survey (SNLS) data in Graham+2008. My primary science interests are extragalactic transients, and I would not make my own custom PZ or otherwise have access to an LSST PZ catalog.

2. Outputs
The minimum set of PZ outputs required for my science goals are point estimates and uncertainties, along with some kind of galaxy type (a best fit spectrum or SED, or a classification). The full posteriors can help a lot when parameterizing host galaxy environments in terms of clustering on a variety of size scales (as in, e.g., Graham 2010). Flags that allow me to easily identify potentially erroneous PZ estimates would be especially helpful.

3. Performance
I estimate that the minimum PZ quality to enable volumetric transient rate analyses correlated with large scale structure is PZ with <X% uncertainty (i.e., the errors on the point estimates) at redshift z. A value of <X2% uncertainty would be acceptable at higher redshift z2. It is also important that the rate of catastrophic outliers be <X3%, to enable robust volumetric studies in faint red host galaxies at low and high redshift.

4. Technical Aspects