PROSPECTS FOR THEEXTRAGALACTIC DISTANCE SCALE

WITH TMT

GEORGE P. & CYNTHIA WOODS MITCHELL INSTITUTE

FOR FUNDAMENTAL PHYSICS & ASTRONOMY

DEPARTMENT OF PHYSICS & ASTRONOMY

TEXAS A&M UNIVERSITY

LUCAS MACRI

OUTLINE

Ø Motivation

� Current status of the Extragalactic Distance Scale

� The landscape in the mid-2020s

� Possible TMT programs

SUZUKI+ (2012); RIESS, MACRI+ (2016)

MOTIVATION: WHAT IS THE NATUREOF DARK ENERGY?Ω Λ

ΩM

0.4 0.6 0.8

0.2

0.4

0.6

0.8

1.0

1.2

1.0

H0

0.20

BAO

w = P/rc2

SUZUKI+ (2012); RIESS, MACRI+ (2016)

MOTIVATION: WHAT IS THE NATUREOF DARK ENERGY?Ω Λ

ΩM

0.4 0.6 0.8

0.2

0.4

0.6

0.8

1.0

1.2

1.0

H0

0.20

wa

w0-1.5 -1 -0.5

-1

0

1

BAO

AFT

ER

CU

EST

A+

(201

6)

� C16: w0 ±0.18; wa ±0.6� DES: w0 ±0.08; wa ±0.3� LSST: w0 ±0.05; wa ±0.1

Coupled with additionalpriors (such as H0)

w(a)=w0+wa(1-a)w = P/rc2

MOTIVATION FOR FURTHERIMPROVEMENT IN H0

BASED ON WEINBERG+ (2012)

FIG

URE

OF

ME

RIT

100

1000

None 0.512

ACCURACY OF H0 PRIOR (%)

DES

LSST

22% increase in FoM

44% increase in FoM

Figure of Merit:

[σ(w0)×σ(wa)]-1

H(Z) VIA BAO� BAO (calibrated via CMB) can map H(z)� Can be used to predict H0 for an assumed cosmology

AFTER BUSCA+ (2012) & CUESTA+ (2016) BLA

KE

+ (2

011,

201

2); C

HUA

NG

& W

AN

G(2

012)

;RE

ID+

(201

2); X

U+

(201

2)

PlanckLCDM

50

60

70

80

90

H(z

)/(1

+z)

[km

s-1M

pc-1

]

1 2 z0

H(Z) VS. H0

� Compare predicted CMB/BAO H0 for an assumedcosmology against measurement (Cepheids+SNe)

� 3.7s difference: systematic error(s) or “New Physics”?

AFTER BUSCA+ (2012) & CUESTA+ (2016) BLA

KE

+ (2

011,

201

2); C

HUA

NG

& W

AN

G(2

012)

;RE

ID+

(201

2); X

U+

(201

2)

50

60

70

80

90

PlanckLCDM

H(z

) / (1

+z)

[km

s-1M

pc-1

]

1 2 z

RIESS, MACRI+ (2016)

0

H(Z) VS. H0

� Compare predicted CMB/BAO H0 for an assumedcosmology against measurement (Cepheids+SNe)

� 3.7s difference: systematic error(s) or “New Physics”?

CUESTA+ (2016); RIESS, MACRI, HOFFMANN+ (2016)

0 0.5 1z

60

70

65

80

H(z

) / (1

+z)

[km

s-1M

pc-1

]

75SH0ES

DVORKIN+ (2014); RIESS+ (2016, 2018)

“NEW PHYSICS” BEYOND LCDM?

� 3.7s tension between predictions from Planck (LCDM) and local measurements of H0 may be a hint of additional relativisticspecies in earlyUniverse

� “Dark radiation” can be parameterized asan increase in Neff

OUTLINE

✓Motivation

Ø Current status of the Extragalactic Distance Scale

� The landscape in the mid-2020s

� Possible TMT programs

CURRENT STATUS

� Cepheids◦ 0 ≲ log P(d) ≲ 2 ; -2 ≲ MI ≲ -8 ; -3 ≲ MK ≲ -9◦ Best method to estimate distances for moderately-inclined

galaxies with recent star formation◦ Observable to D~10 Mpc with 8-m telescopes

(Fassnaugh+’15; Hoffmann & Macri ‘15)

◦ Observable to D~40 Mpc with HST (Hoffmann, Macri, Riess+ 2016)

◦ Used to calibrate SNe Ia & determine H0 to 2.3%(Riess, Macri, Hoffmann+ 2016; Riess, Casertano, Yuan+ 2018)

H-B

AN

DM

AG

NIT

UD

E

PERIOD (DAYS) RIESS, MACRI, HOFFMANN+ (2016)

CEPHEIDS TO 40 MPC WITH HST

� Cepheids in 19 hosts of modern SNe Ia + 4 calibrators� >1500 with homogeneous HST photometry◦ +800 with ground-based observations

CEPHEIDS TO 40MPC WITH HSTR

IESS

, MA

CRI

, HO

FFM

AN

N+

(201

6)

CURRENT STATUS

� RR Lyraes◦ -0.7 ≲ log P(d) ≲ 0 ; 0 ≲ MI,MK ≲ -1◦ Old, low-mass stars: present in any galaxy type◦ Currently limited to M31 + Sculptor (D ≲2 Mpc) w/HST

(Contreras-Ramos+ ‘13; Rejkuba+ ‘11)

� Miras◦ 2 ≲ log P(d) ≲ 3 ; -6 ≲ MK ≲ -11◦ Low/Intermediate-mass stars: present in any galaxy type◦ Recent absolute calibrations using LMC, N4258

(Yuan, Macri, He+ 2017; Huang, Riess, Hoffmann+ 2018)

LMC-BASED CALIBRATIONS

� All variables can be absolutely calibrated in LMC◦ Distance to 2% via eclipsing binaries (Pietrzynski+ ’13)

◦ Discovered by OGLE surveys (Soszynski+ ’08ab, ‘09ab)

◦ NIR light curves from LMCNISS, VMC, othersMACRI+ (2015)

KS

MA

GN

ITU

DE

JHK

SM

AG

NIT

UD

ES

PHASE

MURAVEVA+ (2018)

PHASEMJD

IJH

K M

AG

NIT

UD

ES

YUAN+ (2017)

LMC-BASED CALIBRATIONSM

AG

NIT

UD

E

PERIOD [DAYS]

First overtone Cepheids Fundamental mode Cepheids

Pop II Cepheids

O-rich Miras

MA

CR

I+(2

015)

; BH

AR

DW

AJ+

(201

6); Y

UA

N+

(201

7A)

LMC-BASED CALIBRATIONS:RR LYRAES

MURAVEVA+ (2015)

LOG PERIOD (DAYS)

KS

MA

GN

ITU

DE

[FE/H] (DEX)

OUTLINE

✓Motivation

ü Current status of the Extragalactic Distance Scale

Ø The landscape in the mid-2020s

� Possible TMT programs

THE LANDSCAPE IN MID-2020S

� Gaia◦ Final data release of extended mission (parallaxes, phot, RVs)

� JWST◦ Cycle 4(?) under way

� LSST◦ Firmly in survey operations

� TMT◦ First light!

THE LANDSCAPE IN MID-2020S

Ø Gaia◦ ~9,000 Galactic Cepheids; P-L zeropoint to 0.3-0.6%

� JWST◦ Could detect Cepheids to D~50 Mpc, but time-consuming…

� LSST◦ Miras in ~200 galaxies within D~15 Mpc

� TMT◦ First light!

GAIA DISCOVERY OF & PARALLAXES TOMILKY WAY CEPHEIDS

WINDMARK, LINDEGREN & HOBBS (2011)

� There are ~20,000 Cepheids in the Milky Way;Gaia will discover and accurately measure ~9,000

� Uncertainty in Period-Luminosity relation parameters:◦ Slope: 0.1-0.2%◦ Zeropoint: 0.3-0.6%

� Range reflects uncertaintiesdue to dust corrections

THE LANDSCAPE IN MID-2020S

ü Gaia◦ ~9,000 Galactic Cepheids; P-L zeropoint to 0.3-0.6%

Ø JWST◦ Could detect Cepheids to D~50 Mpc, but time-consuming…

� LSST◦ Miras in ~200 galaxies within D~15 Mpc

� TMT◦ First light!

JWST OBSERVATIONS

� JWST+NIRCam improves over HST+WFC3◦ 4× finer sampling & 3× resolution◦ Similar FoVs (123” vs 130”)

JWST OBSERVATIONS

� Imagine this HST image with 4× better samplingand 3× higher resolution…

N3972, D~25 MpcHST/WFC3-IR F160WPI: R. Foley

PRO

CE

SSIN

GBY

L. M

AC

RI

JWST OBSERVATIONS

� Everything is prettier in color J

PRO

CE

SSIN

GBY

L. M

AC

RI

N3972, D~25 MpcHST/WFC3-UVIS+IRPIs: R. Foley+A. Riess

PRO

CE

SSIN

GBY

STSC

I

JWST OBSERVATIONS

� JWST+NIRCam improves over HST+WFC3◦ 4× finer sampling & 3× resolution◦ Similar FoVs (123” vs 130”)

� But it’s still a modest aperture telescope…◦ 3hr to SNR~10 for P=20d Cepheid @ 50 Mpc in J+K◦ At least 10 epochs needed to obtain periods◦ Large overheads…

THE LANDSCAPE IN MID-2020S

ü Gaia◦ ~9,000 Galactic Cepheids; P-L zeropoint to 0.3-0.6%

ü JWST◦ Could detect Cepheids to D~50 Mpc, but time-consuming…

Ø LSST◦ Miras in ~200 galaxies within D~15 Mpc

� TMT◦ First light!

� Plentiful in all galaxies ➝ go beyond face-on spirals� Large amplitudes in I-band ➝ relatively easy to detect

� Very luminous in NIR → powerful distance indicator

WHY MIRAS?

First overtone Cepheids Fundamental mode CepheidsO-rich Miras

MACRI+(2015); BHARDWAJ+(2016); YUAN+(2017A)

SOSZYNSKI+ (2009)

DAYS

Other pulsating variables

KS

MA

GI M

AG

PERIOD [D]

� LSST will be sensitive enough to detect over105 Miras in ~200 galaxies with D < 15 Mpc

� How to detect periodic but irregular variablesusing sparsely-sampled light curves?

� Develop & test novel periodogram techniquewith existing high-cadence observations(He, Yuan, Huang+ 2016)

� Apply to sparser observations of M33(Yuan, Macri, He+ 2017; Yuan, Macri, Lin+ in prep)

WHY MIRAS?

GAUSSIAN PROCESS PERIODOGRAM

HE

, YUA

N+

(201

6)

DECOMPOSITION OF MIRA LIGHT CURVE (OGLE LMC)

GAUSSIAN PROCESS PERIODOGRAM

HE

, YUA

N+

(201

6)

APPLIED TO NOISIER & SPARSER SIMULATED LIGHT CURVE

Successfully recoveredperiod for 74% of sample

RESULTS FROM M33

YUAN+ (2017B)

� Searched for Miras among 2.4×105 stars in M33◦ Based on I-band data only, spanning ~7 years◦ Used Random Forest classifier trained on 18 features

� Discovered >1800 Mira candidates (Yuan, Macri, He+2017)

EXPAND METHODOLOGY TO NIR

YUAN, MACRI, LIN+ (IN PREP)

� JHKS observations of Miras in M33 (Gemini, KPNO) supplemented with literature (Javadi+2015, UKIRT)

◦ Fit multi-wavelength model, using P from I-band

EXPAND METHODOLOGY TO NIR

� JHKS observations of Miras in M33 (Gemini, KPNO) supplemented with literature (Javadi+2015, UKIRT)

◦ Fit multi-wavelength model, using P from I-band

YUA

N, M

AC

RI, L

IN+

(INPR

EP)

OUTLINE

✓Motivation

ü Current status of the Extragalactic Distance Scale

ü The landscape in mid-2020s

Ø Possible TMT programs

TMT OBSERVATIONS

� TMT+IRIS improves over JWST+NIRCam◦ ~9× finer sampling, ~5× better resolutionà greatly reduce impact of crowding & blending

� FoV considerably smaller…◦ 34” vs 123”

� But much larger aperture!◦ 5, 15 min to SNR~10 for P=20d Cepheid @ 50 Mpc in J&K◦ 1, 2 hr for same object @ 100 Mpc

TMT OBSERVATIONS

� TMT’s great advantage over JWST:Designed from the start to be very agile

� Slew from one target to another and be ready to observe in a matter of minutes◦ Critical to enable time-series programs that may only require

a few hours of observing time spread over a few semesters◦ Will work best if part of a multi-partner adaptive queue

POSSIBLE TMT PROGRAMS

� Cepheid or Mira distances to objects of interest◦ Hosts of rare/interesting objects from LIGO/ZTF/LSST that

require precise luminosity calibration� Absolute photometric calibration with 10-m and/or WFIRST

◦ Next-gen bulge luminosity vs. black hole mass relations

POSSIBLE TMT PROGRAMS

� Cepheid or Mira distances to objects of interest◦ Hosts of rare/interesting objects from LIGO/ZTF/LSST that

require precise luminosity calibration� Absolute photometric calibration with 10-m and/or WFIRST

◦ Next-gen bulge luminosity vs. black hole mass relations

� Mira populations in different environments◦ Follow-up LSST discoveries

POSSIBLE TMT PROGRAMS

� Cepheid or Mira distances to objects of interest◦ Hosts of rare/interesting objects from LIGO/ZTF/LSST that

require precise luminosity calibration� Absolute photometric calibration with 10-m and/or WFIRST

◦ Next-gen bulge luminosity vs. black hole mass relations

� Mira populations in different environments◦ Follow-up LSST discoveries

� RR Lyrae in Virgo?◦ Massive investment of time…