Na Laser Guide Stars for CELT

CfAO Workshop on Laser Guide Stars

99/12/07

Rich Dekany

California Extremely Large Telescope (CELT)

Early concept– 30 m diameter, filled, segmented hyperbolic primary mirror

– Alt/az mount (fully steerable)

– To be built at excellent site

– Extensive instrumentation to exploit various AO modes• Individually or in combination

CELT AO performance requirements

– Undefined, but strong function of science requirements• For example, what are sufficient statistics for studies of galaxy

formation in early Universe?

– Possible AO modes:• High order AO (HOAO) corrected field (1’ - 2’ diameter)

– Strehl > 50%+- 5% at 1m (/D = 6.9 mas)

– Requires MCAO

• Low order AO (LOAO) corrected field (2’ - 6’ diameter)– MCAO tradeoff between uniformity of correction and rolloff with field)

• Full sky coverage with 2 m AO subapertures – Classic NGS (no MCAO)

– Implemented w/ deformable secondary to minimize # of surfaces

• Active control (ACS) only

CELT AO preliminaries

Assumptions– Mauna Kea seeing profile (Keck AO Set B, 0.55m)

r0 = 0.2 m

0 = 4 arcsec

k= 160 arcsec

0 = 3 ms

k = 140 ms

d0 = 7 m (Assume Na LGS’s only)

L0 = 50 m

– No use of NGS’s for high order wavefront information• NGS stars for tip/tilt determination utilize the entire 30m aperture

CELT LGS issues

CELT goal is total wavefront error of 133 nm rms– Since LGS’s remain expensive, allow a generous 75 nm for laser

specific errors• Assume FA dominates LGS errors

– Focal anisoplanatism error = fa = (D/d0)5/6 [rad] --> D < 6 m

– This calls for ~ 24 Na LGSs

• Other multiple LGS errors to worry about– Coma anisoplanatism, propagation of conic tilt errors, beacon position

uncertainty

– Tip/tilt NGS’s• Down to mV = 20 implies N = 2-3 stars per square arcmin

CELT LGS issues (continued)

Other error budget terms– Keeping focal anisoplanatism as the performance driver requires

• Small bandwidth and measurement errors:– 1.5 kHz high order sampling (to get 75 Hz -3db bandwidth)

– ~ 10 W average power per beacon, each delivered into ~ 1 arcsec2 focus» Yields ~ 100 photons per measurement (0.41 cm subaps, 0.3 total transmission)» Result is ~50 nm rms measurement error (Sandler, et. al.)

• Fitting error– N ~ 10,000 efficient d.o.f. correction for highest order mirror

» Other MCAO mirrors may be ~ 1,000 actuator range

• Compute speed > 20x performance of highest-order DoD AO system demonstrated to date

CELT LGS open issues

Entire LGS implementation dependent on capability to accurately describe, model, and understand MCAO– For example, what is the role of NGS in high order MCAO?

• Can lesser, but sufficient, sky coverage may be available with fewer (or even no) LGS beacons

– Probability of finding N=5 guides stars of mV < 14 within 6 arcmin diameter is ~ 10%

» Can MCAO be extended to exploit such large area?» What is degree and uniformity of correction?

– What information beyond tip/tilt can be gained from NGS’s» What can be gained using variable integration times for each NGS?» For example, what is performance of an AO system which slows as much as

necessary to get good SNR on brightest 5 stars in field?

– How can all this be modeled (requires supercomputing resources)?

CELT LGS issues (continued)

AO design strategy– To assure technology readiness, encourage Na LGS development

today• Discussion this afternoon: “How to do this?”

– Resolve major sky coverage issues ASAP• LGS issues permeate nearly all design issues

– For example, site selection» Latitudinal/seasonal/diurnal variations in column density and height, frequency of

‘enhanced Na layers’, infrastructure, safety, neighborly relations, etc.

• Emphasize modeling effort in near term• Validate models

– Multiple NGS wavefront sensing

– Gemini MCAO system