• Slides can be found here: 20220119_ZTF_SN1a_ubercal_update.pdf

ZTF Ia Phone-con: 2022-Jan19

Meeting Agenda:

1. General news [Mickael R]

2. Sub-WG updates [Melissa, Benjamin, Georgios, Suhail]

7. AOB / Close

-

Participants (24):

• Mat S, Mickael R, Joel J, Adam M, Ariel G, Bastien C, Benjamin R, Dominique F, Estelle R, Fabrice F, Georgios D, Jacco T, Jeremy L, Julian B, Kate M, Leander L, Maxime D, Melissa A, Nora N, Philippe R, Simeon R, Suhail D, Valery B, Jakob N

General News:

• ZTF is back on the telescope. Observing to start imminently. The plan is also to take star-flats on the first night.
• Ariel is also thinking about a joint DES/ZTF meeting

wg-simulations update (Melissa):

• The simulation WG (6 people currently) met yesterday to discuss progress, overlaps and starting-points.
• Starting points:
  • There are two sets of observing logs (SkyVision v Caltech). They aren't the same....
  • Jacco and Bastien have looked into this. There are differences at the 10% level, but no-obvious coherent pattern.
  • The differences are particularly different for 2018. 2019 seems much better though.
  • The plan is to get to one unified set of logs, so we all know that we're simulating SN at the same time.
• What values to use?:
  • Half the team are using limiting magnitudes from the science images; the other from the difference images
  • Can we not all use the same numbers?
  • The difference image limiting magnitudes should be 'best', but it's rumoured that these have issues. No one knows why though. Let's figure this out.
  • Ariel: shouldn't ubercal give us this?
    • Yes, but it'll only be valid for SMP :).
• The uncertainties:
  • Again half the team are using the default simsurvey values of ZP=30, gain=1: meaning that every observation in effect has the same ZP. This isn't true (the actual ZP varies and is ~26), and futher the gain is in fact 6.2.
  • Converting through, this means a difference in poisson error between the defaults and the truth of ~15 (see slack for this calculation ;)).
  • Again we need to figure this out. Bastien and Melissa have gathered the ZP for every image, and are investigating the effect of gain in comparison to the data. (This is where things get murky, so the answer is left for next time).
• Next meeting is in two weeks: hoping to have one set of logs

wg-external update: TRGB (Suhail):

• A first draft of the proposed MNRAS letter is available: see slack!
• This is a proof-of-concept work using one SN (ZTF21abiuvdk) with a pre-existing TRGB measurement.
• The constraints from this one object measure H0 with ~9% uncertainty. The aim is to push this out quickly to apply for HST time to get another 3 distances.
• This is a fast-declining SNIa: is this a weirdo? It has a clear secondary-max. Maybe it's transitional, but should still be useful for cosmology.
  • Mickael: this would likely make the Riess sample.
  • Kate's student (Luke H) is writing a paper on this.
    • He's fitting the spectra with TARDIS, and also finds that the object is definitely transitional, so maybe it's not standardly standardisable? That's a larger scale question.... if this object was at z=0.1 then it would make a HD.
    • He also finds that the TRGB distance is over-estimated: it makes the object too bright/hot.
    • Luke could, maybe, give an update next week.
  • In summary, it would make a standard cosmological analysis, so let's use it, but let's revisit 'standardisation', and in particular sub-type standardisation / selection, when the full ZTF sample is ready :).
• To date there are 120 events with z<0.02, and thus could be used to anchor any measurement.
• Comments please by this time next week (26th Jan).
• The HST deadline is the end of March. It'll be in collaboration with the Chicago group (W. Freedman)

wg-calibration update (Benjamin R):

• He's working on Ubercal.
• This is the calibration system that pulls the entire sky into one consistent ZTF system. i.e. it's not the calibration that puts us on an external system, but the system that makes all ZTF values self consistent.
  • It's thus looking at relative magnitudes.
• This is a big data problem: all stars across the sky have to be fit simultaneously.
  • The (x,y) position of every star is used. i.e. the fit is done in pixel space.
• Currently looking at 2000 sq. deg. over a 2 week period.
• The ZP is fitted on a per-quadrant basis, which improves when using aperture photometry.
  • There are still unexplained residuals (dust?) though?
  • Including star-flat photometry from Estelle removes a lot of these problems.
  • Residuals are at the 0.01 level. The remaining scatter is maybe noise dominated?
• The next step is to do it over the whole sky (30,000 sq. deg.) and over a 6 month period.
  • The volume of data is clearly an issue, but not unsolvable.
• Combining ubercal with Scene-Model Photometry (Leander L and Nicolas R) will give us calibrated light-curves! That's the goal for the next couple of months.
• Ariel/Jakob: there are some bad regions of the sky, whilst the high-cadence field seems amazing?
  • Is there a limit to the ZTF photometry?
  • Fabrice: lots of issues to be considered (e.g. variable dithering across the sky): lets see ;).
• Jakob: have you looked at time-dependence?
  • Nope! It's firmly on the list :).
  • Come join the fun ;).

Closing remarks:

• Back next week :).