photometric errors overestimated by a factor of ~1.7-1.9

[djones1040]

Hi Ken, I think this might be something Mark has to work on, but recording it as a bug at least. The chi^2 of pre-explosion epochs for SNe in YSE is approximately 0.3-0.35 (looking at test cases of 2020juq, 2020ioz, 2020ivg, 2020inp). This translates to overestimated errors by nearly a factor of two, and from looking at photpipe reductions of these SNe this appears to apply to both the regular photometry and the forced photometry. Here's one example file:

This is most likely hurting our discovery rates at high redshift by up to a factor of three, so potentially a pretty critical bug to fix.

[mhuber0]

Yeah, this is a problem for Mark and have added an IPP ticket for it. If you send me the photopipe (and might as well IPP) photometry I can do an internal check and some IPP tests.

[djones1040]

sent via email! Thanks for taking a look.