16
Low order wavefront sensor Low order wavefront sensor trade study trade study Richard Clare Richard Clare NGAO meeting #4 NGAO meeting #4 January 22 2007 January 22 2007

Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

  • View
    219

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

Low order wavefront sensor trade studyLow order wavefront sensor trade study

Richard ClareRichard Clare

NGAO meeting #4NGAO meeting #4

January 22 2007January 22 2007

Page 2: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

2

Outline

• Background on sky coverage simulator• Assumptions and parameter set chosen for NGAO• What spectral band should we use for the LOWFS?• How many LOWFS do we need? • What modes should the LOWFS measure?• What is the sky coverage for different science cases?• What is the effect of the LGS asterism radius on partial

correction and sky coverage?

Page 3: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

3

Sodium LGS at 90km

Discrete layers of turbulence, describedby Zernikes, a.

NGS at infinitygenerated with guide star statistics(Bahcall-Soneira, Spagna models)

a(1)

a(3)

a(2)

•Calculate transformation matrices from

LGS, NGS, science points to aperture

naTb nn

aTb LL

aTb ss

Modeling overview

Page 4: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

4

Simulator methodology

• Calculate atmospheric tip/tilt error with minimum variance estimator from transformation matrices and covariance matrices of atmosphere & noise

• Optimize sampling frequency to balance servo lag and noise

• Choose combination of NGS that gives lowest total error• Monte Carlo over many NGS constellations• Generate cumulative density functions of performance

Page 5: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

5

Checking against an AO Simulation

• Compare to LAOS for 4 asterisms for an 8m telescope with no windshake, no sodium tracking error, integrator control, and 10 phase screens for each asterism

– Generally good agreement, but LAOS results somewhat poorer with noise

Median tip/tilt error (nm)

Asterism LAOS without noise

Sky cov. without noise

LAOS with noise

Sky cov. with noise

Good (equilateral)

71±9 58 115±11 84±2

25th percentile

76±13 86 88±9 119±0

Median 106±16 114 200±15 178±5

75th percentile

192±33 221 480±52 334±4

Page 6: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

6

Simulation Parameters/Assumptions• Finite outer scale (75m)

• Mauna Kea (7 layer) turbulence and velocity profile

• First 6 Zernike orders considered, only tip/tilt errors are evaluated

• Detector pixels are seeing-limited in V band (0.5 arc sec)

and diffraction-limited in J/H/K bands (λ /D rads)

• NGS are partially corrected in J/H/K bands. Not in V band.

• Integral control with g=0.5

• 7 LGS asterism (1 on-axis, 6 in a ring) = Ralf’s asterism 7a

• LGS measurements are noise-free

• Limiting magnitude is chosen to be 19 for all spectral bands

• At zenith

• Read noise = 10 e

• Run over 500 NGS constellations

Page 7: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

7

Median Field of View

J=16.4

Field Galaxies case: Latitude=30 deg

J=17.1

J=19.0

J=17.4

J=18.7

J=16.6

Page 8: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

8

Example Cumulative Density Function

Errors are in nm. 1 mas =12.1 nm for a 10m telescope

Field Galaxies science caseJ band

30th percentile=107nm

Page 9: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

9

Choice of Spectral Band

• Trade-off between:

1. Partial correction

2. Sky background

3. Zeropoint (number of photons)

4. Spot size

As λ increases, tip/tilt estimate

Page 10: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

10

Spectral BandErrors are in nm. 1 mas =12.1 nm for a 10m telescope

Tip/tilt error (nm)

Spectral band

10th percentile Median 90th percentile

V 198 326 1140*

J 80 131 221

H 78 127 212

K 99 161 261

For field galaxies science case and 1 TTFA + 2TT sensors

Page 11: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

11

NGS Patrol Field Diameter

For field galaxies science case, J band, and 1 TTFA + 2TT sensors

Page 12: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

12

LOWFS number & order

Tip/tilt error (nm)

LOWFS 10th percentile Median 90th percentile

1 TT 152 209 359

1 TTFA 125 215 312

3 TT 90 146 263

1 TTFA + 2TT

80 131 221

Errors are in nm. 1 mas =12.1 nm for a 10m telescope

For field galaxies science case, and J band

TT=tip/tilt (ie 1x1), TTFA=tip/tilt/focus/astigmatism (2x2)

Page 13: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

13

Different science cases

• Three science cases chosen from NGAO proposal• Science cases have different higher order error, galactic

latitude and science field size

1. Goods N (218 nm, 45 deg, 1.09 arc min)

2. Narrow Field (86 nm, 10 deg, 0.178 arc min)

3. Field Galaxies (173nm, 30 deg, 0.7 arc min)

Page 14: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

14

Degree of partial correction

• Partial correction depends on LGS asterism radius and higher order error from science case

Goods N (218nm)Field galaxies (173nm)Narrow Field (86nm)

Page 15: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

15

Median results for science cases/asterisms

Median tip/tilt error (nm)

Science Case

LGS radius

=7”.2

LGS radius

=21”.6

LGS radius

=35”.9

Goods N 317 284 277

Narrow Field 116 96 94

Field Galaxies

156 131 127

Errors are in nm. 1 mas =12.1 nm for a 10m telescope

Page 16: Low order wavefront sensor trade study Richard Clare NGAO meeting #4 January 22 2007

16

Conclusions

• IR WFS (either J or H) is preferable to visible• Multiple NGS WFS significantly improve tip/tilt

estimate• Measuring focus with 1 of the tilt sensors also

helps• A wider LGS asterism improves partial

correction over the field and hence sky coverage