Unreasonably large negative/positive epochs for some events

[jparrent]

Hey @guillochon, below is the list of objects with either highly negative or highly positive epochs that we looked over. I have loosely organized them by whether or not I think the problem is due to an inaccurate date of max (e.g., b/c of "C" photometry points), a problem of dates on WISeREP ("W"), and so on. There are some of these where the epoch will need to be estimated with something like SNID, in which case we might want to add a field/flag that highlights that the date of maximum has been determined by estimating the post/pre-max epoch of any available spectra and working backwards/forwards. I can jump to this if & when you feel it appropriate to do so.

Uncertain Max ("C")

• [x] CSS120418:125049+222228
• [x] CSS120419:143825+175814
• [x] CSS120419:143825+175814
• [x] CSS140128:100158+255444
• [x] CSS150413:141540+464138
• [x] CSS140424:123226+154840
• [x] SN2012dv
• [x] SN2012fl
• [x] SN2012gd
• [x] SN2016aqt
• [x] SN2016flv
• [x] LSQ14bcj
• [x] SN2013hg

W -- Questionable date for the spectra

• [ ] iPTF14bdn
• [ ] SN2001cp -- the earliest spectrum.
• [ ] SN2013hn -- the two earliest spectra.

Uncertain Max -- Other

• [ ] SN2006fw -- the latest spectrum
• [ ] SN2005A -- date of max is being taken from the Y-band when it should be something like the B-band.
• [ ] SN2006dd -- this one appears to be off b/c the max date is taken from the photometry with a max value, whereas their is a gap in the photometry, which I would think is closer to max; i.e. might need to fit a curve to this one.
• [ ] SN2007ca -- similarly above, date of max is taken from Ks band whereas it should be taken from the B-band.
• [ ] SN2013hk -- [...] I-band date of max. Not that B-band is available near max.
• [ ] SN2016aak -- [...] R-band [...]
• [ ] SN2016coj -- [...] "G" [...]
• [ ] iPTF15lp
• [ ] SN1999gj
• [ ] SN2000dd
• [ ] SN2003ag
• [ ] SN2015bq

The following list are simply events with questionable epochs given the limited photometry available; unclear how to address outside of SNID estimates.

• [ ] iPTF13dkx
• [ ] iPTF14ikw
• [ ] iPTF15asr
• [ ] iPTF15cnq
• [ ] LSQ14zz
• [ ] LSQ15bxe
• [ ] PS15dpq
• [ ] PTF11gdh - max date is being estimated from R band photometry, all that's available.
• [ ] PTF11htb
• [ ] PTF11jaw
• [ ] PTF11jgq
• [ ] PTF11khk
• [ ] SN1999ch
• [ ] SN1999fz
• [ ] SN2003jz
• [ ] SN2006te
• [ ] SN2008ee
• [ ] SN2008fg
• [ ] SN2009ag
• [ ] SN2011dz
• [ ] SN2011ff
• [ ] SN2012er
• [ ] SN2013V
• [ ] SN2014Q
• [ ] SN2015B
• [ ] SN2016ajq
• [ ] SN2016dho
• [ ] SN2016eiy
• [ ] SN2016fnr
• [ ] SN2016gsb

[guillochon]

@jparrent I believe the C-band-related issues should be fixed now, I only checked a few but that should be corrected.

[jparrent]

@guillochon Awesome, I will check these out later this evening.

[jparrent]

@guillochon Alright, I've updated the list. We can check out the rest in the coming weeks and I think I'll start playing around with SNID on something these at home to see how accurate the estimates for epochs are going to be.

[guillochon]

Hey @jparrent, so the date of max is interesting. We have a separate "date of visual maximum" date now being computed that is based on either B or V (or similar), but currently we just use "date of maximum" for plotting on the page. It's not clear to me what the best selection is.

Really what we need to do is some basic LC fitting to determine a better date of max, or SNID, or both. Using individual photometric observations clearly isn't cutting it for some of the events.