1 SNIa Calibration from May 2012 Chicago Calibration meeting.

1

SNIa Calibration

from May 2012 Chicago Calibration meeting

2

Current State of SN Systematics

PhotometricCalibration Uncertainties Dominate!

From Sullivan et al. 2011

3

Our ability to determine cosmological parameters with DL(z) is completely degenerate

with our ability to perform precise photometric calibration.

This is currently the main systematic limitation to SN cosmology.

4

Current State of SN Cosmology

SNLS

5

Current State of SN Cosmology

SNLS

6

Current State of SN Cosmology

SNLS

7

Discovery data1998

20 distant SNe 10% precision

ESSENCE,SNLS

…2009

200 distant SNe 3 % precision

PanStarrs, DES2011-2015

2000 SNe 1% precision (?)

LSST2018

20,000 SNe < 1% goal

Next Steps on Dark Energy: Bigger and Better Imaging Surveys

Broadband photometry: “Metrology and Meteorology”

Four aspects to the photometry calibration challenge:

1. Relative instrumental throughput calibration

2. Absolute instrumental calibration (Best controlled)

3. Determination of atmospheric transmission

4. Determination of line of sight extinction

Historical approach has been to use spectrophotometric sources (known S()) to deduce the instrumental and atmospheric transmission,

but this (on its own) has become problematic if we need % precision:

- integral constraints are inadequate,

- we don’t know the source spectra to the required precision.

Source Atmosphere Instrumental transmission

Extinction above the atmosphere

9

Atmospheric Transmission

Burke et al, ApJ 720, 811B (2010)

10

Potential Color calibration approaches1. Terrestrial black-body sources, using triple point of

metals, and Vega as the transfer standard

2 . Theoretical models of stellar spectra DA white dwarf stars, with 20,000K < T <80,000K.

Theoretical models depend only on log (g) and TModel of Vega plays a role as wellBut beware of extinction effects

3. Statistical assemblage of stars, en masse. Color-color diagrams, ubercalibration tie to another photometric system.

4. Shift the calibration approach entirely, and base it on well-characterized detectors.

Not mutually exclusive

How to address this ?

1. Explicit measurement of atmospheric transmission.

2. Explicit measurement of instrumental response function.• use a tunable laser in conjunction with NIST photodiode standard

3. Explicit determination of atmospheric transmission• multi-narrowband imager• dispersed imager• balloon-borne sources (2012)

4. Re-assess Galactic extinction. • Schlafly & Finkbeiner (2011) used SDSS to revise SFD dust map,

correction factor of 0.78 at 1 micron!

5. Try to shift away from celestial calibrators entirely.

13

NIST Cal Photodiode Spectral Responsivity

NIST photodiode responsivity measurements - InGaAs spectral responsivity uncertainty of 0.1% (1s) for 1.0< <l 1.7 mm

Photodiode detectors extremely stable over time - Si stability exceeds 15 years, thus far - InGaAs stability exceeds 10 years, thus far

Standard Detectors − not standard sources − are the calibrator of choice - increased precision in the photodetector calibration, - ease of use, - repeatability of standard detectors relative to standard laboratory sources

Eppeldauer, Metrologia 2009updated InGaAs figure: courtesy Keith Lykke (NIST)

14

15

Precise Filter Determination

16

Collimated Source Measurements

17

Atmospheric Transmission

Burke et al, ApJ 720, 811B (2010)

The Variable Aspects of Atmosphere

1. Ozone satellite data

2. Water Vapor EW of water lines

dual-band GPS

differential narrowband

3. Aerosols stellar monitor

balloon-borne lasers

4. Clouds local zeropoint adj.

19

Objective grating atmospheric monitor

20

PanSTARRS-1 throughput

Tonry et alarXiv:1203.0297

Closing the loopInstri

InstrumentalSensitivity

Instri

Atmospheric Transmission

Instri

Spectrophotometric Standards

Instri

Precisephotometry

With this initial effort, come within 5% rms of matching ab initio calculations with observations

(eg, Tonry et al, arXiv:1203.0297, 2012)

Summary (Chris Stubbs- May 2012)

1. SN cosmology is stalled until we improve calibration• Determination of luminosity distance vs. z is completely

degenerate with our ability to calibrate photons(l). 2. We need to determine 3 things:

• instrumental sensitivity function• atmospheric transmission • extinction along the line of sight

3. A relative determination suffices. Don’t need absolute flux scale (zeropoint), since this is degenerate with M SN.

4. I am dubious about any celestial spectrophotometric standard below the 1% level.

5. We are making (slow) progress towards implementing a relative calibration based on laboratory standards.

23

Go to Space?

HST, Gaia, Euclid

-Open issues still for absolute colour calibration

- K corrections (precise inter and filter calibration)

- SN model - Standard stars and detectors

The future (>2020): multiprobe DE projects(LSST, KDUST,…)

Absolute Color Calibration:The 5 Step Plan

1. Establish a standard candle - transfer NIST calibration standard to the source input to telescope

4. Monitor ACCESS sensitivity - NIST calibrated on-board lamp tracks sensitivity throughout the program

3. Transfer NIST calibrated standard to the Stars - Observe Standard Stars with the calibrated ACCESS payload

2. Transfer NIST calibrated standard to the ACCESS payload - calibrate ACCESS payload with NIST certified laboratory irradiance standards

5. Fit Stellar Atmosphere Models to the flux calibrated observations - confirm performance; refine and extend Standard Star models

25

Euclid SN survey ?

Basic goal: a significant gain over existing SN surveys

In particular SNLS and DESEuclid has the potential to provide the first NIR survey for SNe from

space

Provides an independent Euclid probe of cosmology

With 6 months of observing time, the most interesting option is the

“AAA survey”

Reaches high redshift : up to z ~ 1.5

Cannot be done from the ground

“AAA” survey[Simulations by P. Astier, K. Maguire, S.Spiro]

A dedicated Euclid SN survey 6 months total Euclid time

split into two 6-month seasons

(observing ~half time) to provide

reference images

10 sq deg

4 day cadence

Increased imaging exposure times:

y,J,H=1200, 2100, 2100s (no spectra)

Simultaneous ground-based i and z-

band

Provides 1700 well measured SNeIa with

0.75 < z < 1.5

Complemented with low- and mid-z

ground based surveys (not

simultaneous)

Above: example Euclid lightcurve at z=1.5 and predicted DETF FoM

Union2(current)

DES

LSST in1 year

OPTICAL SN samples

JWSTE-ELT

Euclid*(schedule permitting)

CSP(current)

OPTICAL and NIR SN samples

ACCESS

30

ACCESSAbsolute Color Calibration Experiment for Standard Stars

M.E. Kaiser & the ACCESS Team

Calibration and Standardization of Large Surveys and Missions in Astronomy and Astrophysics

16 April 2012

Status, Calibration Strategy, and Design Performance

This work supported by NASA grant NNX08AI65G

16 April 2012

Current Standard Star Uncertainties

Uncertainty floor (circa 2007) in the fundamental stellar standards is 2%. across the 0.35 - 1.7 m bandpass (Bohlin 2007, Cohen 2007)

Judicious selection of standard stars - Observe existing (known) standard stars - Vega (A0V) - absolute VIS NIR std, bright (V=0.026), pole-on-rotator => variety of thermal zones, complex - Sirius (A1V) - IR std, bright (V=-1.47) - BD +17o4708 (sdF8) simpler spectra, SDSS std, fainter +,- - HD 37725 (A3V) - absolute calibrator for IR satellites, possible alternate target: HD84937 (F5V) - Minimize spectral features & enable robust modeling - Flux level chosen to minimize calibration transfers

Major uncertainty contributors: - Earth’s atmosphere Sol’n:dedicated monitoring or observe above the atmosphere - Stellar models: describe & extend the data Sol’n: Improved stellar models - need data constraints & test wrt NIST at the 1% level

A single stellar calibrator spanning the full bandpass introduces less error than when two stellar calibrators are required to span the bandpass. (Lampton 2002, Kim et al., 2004 )

32

Observe above the Earth’s Atmosphere

Sounding Rocket observes completely above the Earth’s atmosphere - eliminates problem of measuring residual atmospheric abs’pn seen by balloons - OH arises at 70 km; typical balloon altitude: 39 km, rocket altitude: 300 km - OH airglow emission lines are 10-100X stronger than 13th mag star - continuous spectral calibration across the 0.35 - 1.7 m bandpass

Balloon: OH introduces additional complexity - increased statistical noise & systematics from bkrnd subtraction - increased instrument costs to avoid scattered OH airglow

Rocket disadvantage: Flight times are short (~400 sec) Limits faintest standard to ~ 9th magnitude (BD+17o4708) with <1% uncertainty

Establish repeatability: Two flights per target - Vega & Sirius 12h apart - four flights of 2 targets each

33

ACCESS: Optical Design

Spectrograph:Slit: 1mm (33 arcsec/mm)

Grating: Concave,

Blaze angle:1.65o

Utilize multiple orders

1st : 0.9 – 1.9 mm

2nd : 0.45 – 0.95 mm

3rd : 0.30 – 0.63 mm

Cross disperser:

Prism spherical figure

Telescope: F/15.72 Dall-Kirkham

Primary figure: ellipse

393 mm (15.47in) diameter

Secondary figure: sphere

Coatings: MgF2 over Al

34

ACCESS Payload - Spectrograph

The End

35