Astrophysical false alarms in high contrast imaging surveys · PDF fileAstrophysical false alarms in high contrast imaging surveys ... Polarisation Simultaneous on same ... Solar System

DRH Workshop – Heidelberg 13-15 Apr 2010

Astrophysical false alarms in high contrast

imaging surveys

– The SPHERE/NIRSUR case –

G. Chauvin

IPAG – Institut de Planétologie et Astrophysique de Grenoble

P. Delorme (IPAG), A. Vigan, A. Zurlo, C. Moutou (LAM), M. Bonnefoy

(MPIA), M. Bonavita,, S. Desidera, D. Mesa (INAF/Padova), and

the SPHERE consortium

Planet Validation Workshop – LAM – 13-15th May 2013


I- SPHERE/NIRSUR survey

II- Lessons from the past

III- Classes of contaminants

IV- False alarm diagnostics

V- Adopted strategy


Outline

DRH Workshop – Heidelberg 13-15 Apr 2010 Planet Validation Workshop – LAM – 13-15th May 2013

I- SPHERE/NIRSUR

VLT/SPHERE ((Beuzit et al. 08)

• Scientific Goal: High-Contrast Imager for Giant Planets & Disks

• Description:

- SAXO, XAO system (ITTM-DM and DTTS, PTTS)

- NIR (YJHK): IRDIS (Dual imaging Spectrograph)

and IFU 3D-spectroscopy

- VIS: ZIMPOL (Imaging Polarimeter)

- Coronagraphs: Classical Lyot, A4P and ALC

- Differential imaging capabilities


I- SPHERE/NIRSUR


I- SPHERE/NIRSUR

ZIMPOL IRDIS IFS

FoV Sq 3.5‟‟ (instantaneous) Up to 4‟‟ radius (mosaic)

Sq 11‟‟ Sq 1.77‟‟

Spectral Range 0.5 – 0.9 μm 0.95 – 2.32 μm 0.95 – 1.35/1.65 μm

Spectral information

BB, NB BB, NB Slit spectro: 50/400

50 / 30

Linear Polarisation

Simultaneous on same detector, x 2 arms, exchangeable

Simultaneous dual beam, exchangeable

x

Coronography: no /4Q / Lyot

Rotation at Nasmyth: Pupil-stab. (instrument fixed wrt tel.) Field-stab (slit spectro, long DIT…) No rotation: minimize crosstalk…)

AO sensitivity for high contrast: R <9.5 for NIR; R<9 for R; R<7.8 for whole VIS

Separation with improved contrast: 2 - 20 λ/D, ie 30-300 mas in R, or 80 – 800 mas in H

Mode switching: not VIS and NIR in same night


I- SPHERE/NIRSUR

SPHERE GTO • 260 nights over 4 – 6yrs

• Scientific programs:

. Disks (20n) . Reflected Light (10n)

. NIRSUR (200n) . SS & Evolved stars (10n)

. Add/Tests (10n)

• NIRSUR:

Survey of >400 nearby stars (AFGKM and <1 Gyr)

Occurrence of giant planets at wide orbits (>10 AU)

Giant planet atmospheres, Evolution, Planetary systems Dynamics…


I- SPHERE/NIRSUR

High-Contrast Imaging of Exoplanets 1/ XAO: Atmospheric turbulence (90% Strehl in H-band)

2/ Diffraction pattern: Halo suppression

3/ Quasi-static instrumental aberrations

& post-processing

XAO, Sr~90% Diffraction + static

aberration correction

Speckle Calibration,

Differential Methods

Contrast ~ 10-3-10-4 Contrast ~ 10-5 Contrast ~ 10-7


I- SPHERE/NIRSUR

NIRSUR Observing Strategy • IRDIFS Mode:

• IRDIS-DBI H2H3 (12.5” FoV) & IFS-Y-J (1.77” FoV)

simultaneously

• Atmosphere: Limiting seeing and coherence time for AO

• Star-brightness: AO performances (R < 9.5)

• Meridian-constraint:

• maximize field rotation for ADI

• Minimize the quasi-static aberrations

• Companion faintness: Observing sequence of 1-2hrs

• Dedicated calibrations: photometric & Astrometric field






V- Adopted strategy


Outline

DRH Workshop – Heidelberg 13-15 Apr 2010 NIRSUR Workshop – Padova 3 – 6 May 2010

Past Deep Imaging Surveys . Magnitude limit: Hlim = 21– 23 (HA = 8-9; ΔH=14-15) (Chauvin et al. 2010)

. 65 stars observed; 240 False Alarms (FAs) . FA rate ~ 55%

28”


NaCo (VLT)


False alarms

Programme Instrument FoV Filter Nb b (o) Fc(%) Reference

Young stars VLT/NaCo 28” H/K 63 27.3 55.0 (Chauvin et al. 2010)

Young stars HST/NICMOS 19” H 45 9.9 57.7 (Lowrance et al. 2005)

Planet hosts Keck 5” K 40 -11.5 16.0 (Luhman et al. 2002)

Young & Int-old Gemini/NIRI 22” H 85 16.9 64.7 (Lafreniere et al. 2007)

Field contamination from past DI surveys:




Proper motion study of CC-1 .

Primary star proper & parallactic

motion. Follow-up at several epochs

>> co-moving object

Orbital motion

>> physically bound

But do we want to conduct

systematic follow-up observations?

Identification: Astrometric Follow-up



Identification: Impact on the survey size . 200 GTO nights

. Obs.Time/target

. FAs Contamination (Fc)

“Systematic astrometric follow-up”: Fc = 50% 66n follow-up

Fc = 33% 50n follow-up

. Nb Visits

. Target List (Nb stars)

. Nb Follow-up Visits

Nb Targets = Nb Visits / ( 1 + Fc)

Nb “Discovery“ Nights = 200. / ( 1 + Fc)




Identification: Impact on the survey size . 200 GTO nights

. Obs.Time/target

. FAs Contamination (Fc)

>> Identification Strategy: impact the Nb Targets observed

our chance of planet detection

the statistical significance of our complete sample

. Nb Visits

. Target List (Nb stars)

. Nb Follow-up Visits

Nb Targets = Nb Visits / ( 1 + Fc)

Nb “Discovery“ Nights = 200. / ( 1 + Fc)








V- Adopted strategy


Outline



1. Solar System Bodies: (B. Carry; F. Marchis)

. Asteroids: km- to 100-m size, G2 solar-type spectra

. TNOs: km-size bodies; H20 (CH4...) ice-composition

(Barucci et al. 2005)


1. Solar System Bodies: (B. Carry; F. Marchis)

. Asteroids: km- to 100-m size, G2 solar-type spectra

. TNOs: km-size bodies; H20 (CH4...) ice-composition

. Identification:

High-proper motion: μ(„‟/min) = 140/d(AU) >> smearing!

930mas/s main-Belt (2.5 AU); 50 mas/s TNOs (50 AU)

Ephemerids: - SkyBoT (http://vo.imcce.fr/webservices/skybot/)

- JPL/Horizons (http://ssd.jpl.nasa.gov/?horizons)

>> implemented with success for NaCo, Keck !

. How many asteroids/TNOs FAs? (>> could simulate the expected nb)

NIRSUR Workshop – Padova 3 – 6 May 2010




2. Galactic: Field Stars • Galactic Population Synthesis:

The Besançon Model (Robin et al. 2003)

The TRILEGAL Model (Girardi et al. 2005)

• Predictions calibrates using Hipparcos observations and recent

large scale surveys in optical and NIR.

• Friendly-tool to simulate Field Stars Contamination:

Inputs: Catalogue and magnitude limit (JH...)

Outputs: > Star counts / sq.deg

> properties: SpT, Teff, Age, FeH, ppm....





2. Galactic: Field Stars • The NIRSUR survey simulation

597 stars from the TDB (and their gal. coordinates)

Hlim = 23 (ΔH = 16–17 for typ. H = 6.5)

Cumulated star counts

density Map / sq.deg

(for Hlim<23)



2. Galactic: Field Stars • NIRSUR: Cumulated star counts

• In average, 7 field stars (H < 23) per IRDIS FoV



2. Galactic: Field Stars • NIRSUR: Complete survey field stars contamination

• 52% of IRDIS FoVs, contaminated by 1 field star with H < 23



2. Galactic: Field Stars • NIRSUR: Contaminant predicted properties



2. Galactic: Field Stars • NIRSUR: Contaminant predicted properties

Field MS M dwarfs!



3. X-Galactic: (M. Turatto; P. Delorme)

• Low- and high-redshift galaxies

HST/NICMOS deep field study

Galaxies counts & morphology (H < 24)

(Teplitz et al. 1998)




• Low- and high-redshift galaxies

HST/NICMOS deep field study

Galaxies counts & morphology (H < 24)

(Teplitz et al. 1998)

• Counts = 2.1 X-Gal/FoV

IRDIS (H < 23):

. Morphology:

z = 0.3, 3kpc Ø = 0.6”

z = 1-4, 4-8kpc Ø = 0.4-1.0”





• QSOs (Active nucleus galaxies)

Chandra Survey: 60 QSOs over 300 sq. arcmin; (Barger et al. 2002)

> QSO counts: (J<23) 0.01 QSO / IRDIS FoV

• Low & High-z Star forming regions

Giant HII regions (200pc size); At z = 0.3 - 4, Ø = 25 – 45 mas

• Intergalactic Globular Clusters

ACS Deep field (Williams et al. 2007)

> IGC counts (H < 22): 0.04 IGC / IRDIS FoV

• High-z Super Novae

SNe computation (Dahlen & Fransson et al. 1999, SNe rates)

> SNe counts: 0.004 SNe / IRDIS FoV


False alarms summary, Classe Type Contamination

(Nb/IRDIS FoV)

Comments

Solar System Asteroids/TNOs ? smearing, ephemerids

Galactic Field Stars 7 M-stars for H = [17-25]

point-like. same H2H3 color as L-type EGPs

X-Gal Low & High-z Gal. 2. resolved

QSOs 0.01 negligible

X-SFRs ? negligible

IGCs 0.04 negligible

SNe 0.004 negligible

>> Major source of astro-FAs: background field M-dwarfs








V- Adopted strategy


Outline


Identification, available diagnostics:

1. Database/Preparatory observations:

cross-correlation with on-line catalogues

building a Deep Imaging Database

2. NIRSUR spectro-photometric diagnostics

ability to identify false alarms

3. Astrometric follow-up

Astrometric precision & target proper motion

IV- False alarms diagnostics



False Alarm Operation Diagramme

1 Visit

Detection

Non-detection

New

Identified

Data Analysis NIRSUR Diagnostics

Identified

Un-identified

Follow-up TDB

TDB

Identified

1. 2. 3.


Priority?




1. Database/Preparatory observations

1.1 Literature: compilation of published surveys

(catalogue, instrument, observing mode, detection performances,

source identification...)

1.2 Collecting Deep Imaging Data (DIVA)

Status: currently 400 stars (300 potential targets for NIRSUR)

Interface with the NIRSUR Target Database

1.3 NaCo Large Programme (Preparatory survey)

ADI H-band observation: 110 new stars

NIRSUR Workshop – Padova 3 – 6 May 2010 Planet Validation Workshop – LAM – 13-15th May 2013



2. NIRSUR diagnostics

Observing modes

. IRDIS DBI:

H2H3, H2, H3 photometry

. IFS (0.95-1.35 μm):

Y-J Rec. photometry

LRS

IRDIS DBI (H2H3)

IFS (0.95-1.35 μm)

12.5’’ 1.77’’

Outer FoV

Inner FoV





2.1 Outer “IRDIS alone” FoV

Single IRDIS detection:

H2 or H3 or H2H3

Cannot discriminate!

Double IRDIS detection:

H2 and H3 (H2 and H2H3...)

Can easily discriminate cool

T2 to Y EGPs from field stars.

>> Problem for young L-type EGPs!

>> Expect a lot of M field stars!




2.2 Inner “IRDIS/IFS ” FoV

Single IFS detection:

Y-J low-res spectra

(Sp. indexes, cross-libs)

High SNR: can well discrim-

inate L-,T- and Y-types

from FAs

.

IFS Extraction simulations (Dino):

L0-type EGP, 10-6 (ΔJ = 15) at 0.5”

>> low SNR: more difficult

for L-types than T, Y-types

>> Smaller FoV (less FAs)



2. NIRSUR diagnostics Summary:

. Outer “IRDIS alone” FoV

. Important Nb of FAs/FoV (Fc ~ 50%)

. Single detection no diagnostics

. Double detection no discrimination btw warm L-types

(early-T) EGPs and field M stars

. Inner “IRDIS/IFS” FoV

. Less FAs/FoV (Fc of a few %)

. “IFS” detection good discrimination

low S/N, more difficult for L-types

. Combined “IRDIS/IFS” improved diagnostics




3. Astrometric Follow-up

TDB, NIRSUR target proper motion

Easy confirmation over a

few months with a ~5mas

accuracy



Available diagnostics – summary (Contaminants: mostly field M dwarfs, 52% Outer FoV)

1. Archive/TDB: - relevant for Outer FoV

- Follow-up obs. reduction to be quantified

2. NIRSUR Diagnostics: - handicap for the Outer FoV: L (early-T) type EGPs

- Inner FoV: less problematic

(lower contamination and better diagnostics)

>> enable to derive the CC (Mass, Δproj) if bound!

3. Astrometry: - Firm identification (a few months)

- But, impact Nb of targets observed







V- Adopted strategy


Outline


V- Strategy & Planet Validation

Adopted strategy 1. Scientific Goals Detection & statistical completeness

2. Sample selection: Age, distance, SpT, FeH, (Galactic latitude)

Selection: Detection or Statistics driven?

Expected Fc ~50%

BUT, we want the Follow-up to be < 30%

3. Physical window Focus on close and low-mass companions

Physical cutoff: < 200 AU (r < 4”) and < 60 Mjup



V- Strategy & Planet Validation

Adopted strategy 4. Candidate Ranking: 1/ Top-priority: CH4 or spectral signatures

for Follow-up 2/ Others, focus on the closest and lightest EGPs

Predicted mass & distance; Contamination Probability;

Ranking: merit function with (M, a) cutoff

depending on predicted M, a and Fc

5. Follow-up: Total should not follow-up more than 30%

Priority according to the candidate ranking

Complete on a physical distance parameter space



Outline Adopted strategy

6. Characterization: a/ Primary refined parameters (Age, d)

> Alice Zurlo’s Talk

b/ Atmosphere

IFS and IRDIS extended diagnostics

(spectroscopy and additional photometric filter)

c/ Predicted Physical properties (Mass)

d/ Orbital properties (sma, ecc.)

e/ Circumstellar environment (disk, planets…)

Documents

Astrophysical false alarms in high contrast imaging surveys · PDF fileAstrophysical false alarms in high contrast imaging surveys ... Polarisation Simultaneous on same ... Solar System