1

Comparative Performance of a 30m Groundbased GSMT and a 6.5m

(and 4m) NGST

NAS Committee of Astronomy & Astrophysics9th April 2001

Matt MountainGemini Observatory/AURA NIO

2

Overview

• Science Drivers for a GSMT• Performance Assumptions

– Backgrounds, Adaptive Optics and Detectors

• Results– Imaging and Spectroscopy

• compared to a 6.5m & 4m NGST

– A special case, • high S/N, R=100,000 spectroscopy

• Conclusions

3

GSMT Science Case“The Origin of Structure in the Universe”

From the Big Bang… to clusters, galaxies, stars and planets

Najita et al (2000,2001)

4

Mass Tomography of the Universe

z~0.5

Existing Surveys + Sloan

z~3

Hints of Structure at z=3(small area)

100Mpc (5Ox5O), 27AB mag (L* z=9), dense sampling

GSMT 1.5 yr

Gemini 50 yr

NGST 140 yr

5

Tomography of Individual Galaxies out to z ~3

• Determine the gas and mass dynamics within individual Galaxies• Local variations in starformation rate Multiple IFU spectroscopy R ~ 5,000 – 10,000

GSMT 3 hour, 3 limit at R=5,000

0.1”x0.1” IFU pixel(sub-kpc scale structures)

J H K 26.5 25.5 24.0

6

Probing Planet Formation with High Resolution Infrared Spectroscopy

Planet formation studies in the infrared (5-30µm):

Planets forming at small distances (< few AU) in warm region of the disk

Spectroscopic studies:

Residual gas in cleared region emissions Rotation separates disk radii in velocity High spectral resolution high spatial resolution

8-10m telescopes with high resolution (R~100,000) spectrographs can detect the formation of Jupiter-mass planets in disks around nearby stars (d~100pc).

S/N=100, R=100,000, >4m

Gemini out to 0.2pc sample ~ 10s

GSMT 1.5kpc ~100s

NGST X

7

30m Giant Segmented Mirror Telescope concept

Typical 'raft', 7 mirrors per raft

Special raft - 6 places, 4 mirrors per raft

1.152 m mirror across flats

Circle, 30m dia.30m F/1 primary, 2m adaptive secondary

GEMINI

8

GSMT Control ConceptLGSs provide full sky coverage

Deformable M2 : First stage MCAO, wide field seeing improvement and M1 shape control

10-20’ field at 0.2-0.3” seeing

1-2’ field fed to the MCAO module

M2: rather slow, large stroke DM to compensate ground layer and telescope figure,

or to use as single DM at >3 m. (~8000 actuators)

Dedicated, small field (1-2’) MCAO system (~4-6DMs).

Focal plane

Active M1 (0.1 ~ 1Hz)619 segments on 91 rafts

9

GSMT Implementation concept- wide field (1 of 2)

Barden et al (2001)

12

GSMT Implementation concept- MCAO/AO foci and instruments

MCAO opticsmoves with telescope

Narrow field AO ornarrow field seeing limited port

MCAO Imagerat vertical Nasmyth

elevation axis

4m

Oschmann et al (2001)

14

MCAO Optimized Spectrometer

• Baseline design stems from current GIRMOS d-IFU tech study occurring at ATC and AAO– ~2 arcmin deployment field

– 1 - 2.5 µm coverage using 6 detectors

• IFUs– 12 IFUs total ~0.3”x0.3” field

– ~0.01” spatial sampling R ~ 6000 (spectroscopic OH suppression)

15

Quantifying the gains of NGST compared to a groundbased telescope

• Assumptions (Gillett & Mountain 1998)• SNR = Is . t /N(t): t is restricted to 1,000s for NGST

• Assume moderate AO to calculate Is , Ibg

• N(t) = (Is . t + Ibg. t + n . Idc .t + n . Nr

2)1/2

• For spectroscopy in J, H & K assume “spectroscopic OH suppression”

• When R < 5,000 SNR(R) = SNR(5000).(5000/R)1/2

and 10% of the pixels are lost

Source noise background dark-current read-noise

16

Space verses the Ground

Takamiya (2001)

17

Adaptive Optics enables groundbased telescopes to be competitive

For background or sky noise limited observations:

S Telescope Diameter .

N Delivered Image Diameter

Where: is the product of the system throughput and detector QE

is the instantaneous background flux

18

Adaptive Opticsworks well

19

Modeling verses Data

20 arcsec

M15: PSF variations and stability measured as predicted

GEMINI AO Data

Mod

el R

e su l

ts

2.5 arc min.

20

Quantitative AO Corrected Data

• AO performance can be well modeled• Quantitative predictions confirmed by observations

• AO is now a valuable scientific tool:

• predicted S/N gains now being realized

• measured photometric errors in crowded fields ~ 2%

Rigaut et al 2001

21

•Tomographic calculations correctly

estimated the measured atmospheric phase

errors to an accuracy of 92%

–better than classical AO

–MCAO can be made to work

Multi-Conjugate Adaptive Optics

MCAO

2.5 arc min.

Mod

el r

esu

lts

22

AO Technology constraints (50m telescope)

r0(550 nm) = 10cm No. of Computer CCD pixel Actuator pitch S(550nm) S(1.65m) actuators power rate/sensor

(Gflops) (M pixel/s) 10cm 74% 97% 200,000 9 x 105 800

25cm 25% 86% 30,000 2 x 104 125

50cm 2% 61% 8,000 1,500 31 SOR (achieved) 789 ~ 2 4 x 4.5

Early 21st Century technology will keep AO confined to > 1.0mfor telescopes with D ~ 30m – 50m

23

MCAO on a 30m: summary

• MCAO on 30m telescopes should be used m• Field of View should be < 3.0 arcminutes,

• Assumes the telescope residual errors ~ 100 nm rms• Assumes instrument residual errors ~ 70 nm rms

– Equivalent Strehl from focal plane to detector/slit/IFU > 0.8 @ 1 micron– Instruments must have:

• very high optical quality• very low internal flexure

(m) Delivered Strehl

1.25 0.2 ~ 0.4 1.65 0.4 ~ 0.6 2.20 0.6 ~ 0.8

9 Sodium laser constellation4 tip/tilt stars (1 x 17, 3 x 20 Rmag)

PSF variations < 1% across FOV

Rigaut & Ellerbroek (2000)

24

Modeled characteristics of a 30m GSMT with MCAO (AO only, m) and a 6.5m NGST

Assumed detector characteristics

m <m 5.5m <m

Id Nr qe Id Nr qe

0.01 e/s 4e 80% 10 e/s 30e 40%

Assumed encircled-energy diameter (mas) containing energy fraction

30M 1.2m 1.6m 2.2m 3.8m 5.0m 10m 17m 20m(mas) 23 29 41 34 45 90 154 181NGST 1.2m 1.6m 2.2m 3.8m 5.0m 10m 17m 20m (mas) 100 100 82 138 182 363 617 726

25

Comparative performance of a 30m GSMT with a 6.5m NGST

1 101E-3

0.01

0.1

1

10

Comparative performance of a 30m GSTM with a 6.5m NGST

S/N

Ga

in (

GS

MT

/ N

GS

T)

Wavelength (microns)

R=5 R=1,000 R=10,000

Assuming a detected S/N of 10 for NGST on a point source, with 4x1000s integration

GS

MT

a

dv

an

tag

eN

GS

T a

dva

nta

ge

R = 10,000 R = 1,000 R = 5

26

Comparative performance of a 30m GSMT with a 4m NGST

1 10

0.01

0.1

1

10

Comparative performance of a 30m GSTM with a 4.0m NGST

S/N

Ga

in (

GS

MT

/ N

GS

T)

Wavelength (microns)

R=5 R=1,000 R=10,000 R = 10,000 R = 1,000 R = 5

Assuming a detected S/N of 10 for NGST on a point source, with 4x1000s integration

GS

MT

a

dv

an

tag

eN

GS

T a

dva

nta

ge

27

Observations with high Signal/Noise, R>30,000 is a new regime

- source flux shot noise becomes significant

10 1000.1

1

10

100

1000

10 1000.1

1

10

100

1000

GSMT 30m

Com

par

ativ

e n

ois

e c

ont

ribu

tion

s a

fter

firs

t 1

,000

s

(ele

ctro

ns)

1/2

Target S/N after 4,000s

Detector Background Source

4.6m Spectroscopy at R=100,000

NGST 6.5m

Target S/N after 4,000s

Detector Background Source

28

High resolution, high Signal/Noise observations

1 10

0.01

0.1

1

10

17.0

12.3

4.6

Molecular line spectroscopy S/N = 100

S/N

Gai

n (

GS

MT

/ N

GS

T)

Wavelength (microns)

R=10,000 R=30,000 R=100,000

Detecting the molecular gas from gaps sweptout by a Jupiter mass protoplanet, 1 AU from a 1 MO young star in Orion (500pc) (Carr & Najita 1998)

GSMT observation ~ 40 mins (30 mas beam)

29

Conclusions

6.5m 4.0m Comments

1. Camera 0.6 – 5 mDeep imaging from space; consistent image quality, IR background, even for < 2.5m if D>4.0m

2.MOS

R=1,000

1.2 – 2.5m

2.5 – 5.0 m

NGST MOS still competitive for < 2.5m only if D~6.0m (consistent image quality, coverage)

3.Camera

Spec. R=1500

5 – 28 m

5 – 28 m

Clear IR background advantage observing from space, even for D~4m

and R< 30,000

4. IFU

R=5,000

1.2 – 2.5m

2.5 – 5.0 m

Detector noise limited for < 2.5m D2 advantage for groundbased GSMT

For >2.5m, NGST wins even D~4m

D2 advantage for groundbased GSMT

For <12m

A advantage of GSMT,technology challenges from space (fibers)

NGST advantage GSMT advantage X

X

X

X

NGST

NG

ST

In

stru

men

t

High S/N, R~100,000 spectroscopy

WF MOS Spectroscopy m

XX

X X