stardice

First interesting results from the StarDICE experiment

center

fom

Instrumental calibration for Rubin

  • Exploiting the full statistical power of a Rubin SNe Ia survey sets very stringent constraints on calibration uncertainties
  • Computation made for 10years of well cadence survey (Hazenberg 2019)
  • σ(ZP) < 1 mmag
  • σ(\lambda) < 0.1 nm
stubbs

From astrophysical to physical standards

  • Astrophysical flux scale is defined by the numerical model of the SED of 3 DA white dwarfs.
  • Evolution of the integral of theoretical model in the 6 Rubin bands (C. Stubbs 2026, private com.)
  • Excluding band, the change in color is in the range 5-10 mmag
  • Motivate calibration based on laboratory standards
stardice

A quick history of StarDICE

  • 2016: Concept proposal to LSST-France
  • 2016-2019: Test with an "SnDICE" source (Hazenberg 2019)
  • 2020: MoU for a permanent installation at OHP
  • 2022: Measurement with a CBP (Souverin et al. 2025)
  • 2023: Telescope first light at OHP
  • 2024: Installation LED T152
  • 2025: in-situ tests with MonoDICE
  • 2026: Measures with calibrated MonoDICE
monodice

Last MonoDICE version

Monochromatic point source

  • Full pupill illumination
  • Point-like (1mm @ 113m)
  • Quasi-monochromatic (1nm)

Improvements

  • Precise motorized mount
  • Alignment laser
  • Temperature monitoring
  • NIST Calibrated
kappa

NIST calibration

Simpler than LEDs

  • Monochromatic
  • Simple photodiode ratios
  • Swappable amplifiers

coverage 300-1100nm

  • Cover all 6 rubin filter
  • Hole @950nm (no flux)
  • 0.1%/nm (stat) 350-1100nm
  • limited by NIST @ all
kappa

Understanding systematics

Succesfull tests

  • Electrometer swap (gray)
  • Current vs charge
  • Linearity
  • Electrometer Temp

Ongoing work

  • understanding important T dependance
  • Geometrical beam effects (alignment, photodiode ed
monodice_data

2026 absolute measurement

Secured 2 successful nights

  • 2 + 4 full scans in all 6 bands + EMPTY
  • Very visible varying atmospheric transmissions
  • We did not manage to obtain the exact source-telescope distance
tcp_fit

Transmission scale fit

Scale uncertain in CBP measurements

  • Due to "pupill stitching"
  • Can be refit with these data
  • Along with the atmosphere
  • Linear evolution of aerosols

tcp_fit

CBP/MonoDICE Agreement on filter fronts is impressive

  • < 0.1 nm in general
  • Worst offender is the blue front of r 0.3nm ()
  • Demonstration that pupil stitching works for filter shape (good news for Rubin CBP)
tcp_fit

First NIST-calibrated filter scale measure

  • Full-pupil and in-situ
  • Single night
  • Drops to when fixing the atmosphere
  • CBP measurement: ~120000 images (1 week cont.)

bottom:90%
Two nights

different atmosphere

different evolution

consistent answers

Limitations

  • Still large correlations with atmospheric parameters
  • Assume linear time evolution of aerosols
  • Trust the shape of the EMPTY CBP measurement

Signs of disagreement with the CBP shape

  • Residuals in y band
  • Non physical value for the parameter

Ultimate in-situ measurement needs simultaneous measurement with two distances

Conclusion

StarDICE has accumulated a survey of 60000 images of two primary standards over 60 nights.

  • Still observing in the current configuration until august 2026
  • In conjunction with the FIR instrument

We finally cracked in-situ calibration

  • Proof that the instrument has been reasonably stable since its
    shipping from LPNHE
  • Last campain with MonoDICE in Sept 2026.

Preparing the publication for this phase in parallel

Conclusion

Upgrades

  • MITI grants for the design and construction of a new FIR instrument
  • "2-distance" or airborne MonoDICE in reflexion.

Notes

fit