LSST Filter Development LSST Conceptual Design Review Tucson, AZ Sept 17-20, 2007 Kirk Gilmore SLAC

LSST Filter DevelopmentLSST Filter Development

LSST Conceptual Design ReviewTucson, AZ

Sept 17-20, 2007

Kirk GilmoreSLAC

LSST Conceptual Design Review September 17-20, 2007 Tucson, AZ

LSST Filter DesignLSST Filter DesignParametersParameters

1. 1. Beam that is incident on the filter has a focal ratio Beam that is incident on the filter has a focal ratio of f/1.25 with a 61.5% obscuration.of f/1.25 with a 61.5% obscuration.

2. 2. The filter is concentric about the chief ray so that allThe filter is concentric about the chief ray so that all portions of the filter see the same angle of portions of the filter see the same angle of incidence range, 14.2º to 23.6ºincidence range, 14.2º to 23.6º

______________________________________________


Lenses FiltersClear Aperture Dims L1 L2 L3 u g r i z y

Surface 1 vertex to FPA 1031.950 537.080 88.500 149.500 149.500 149.500 149.500 149.500 149.500Surface 2 vertex to FPA 949.720 507.080 28.500 123.300 128.360 131.700 133.800 135.300 136.000Center thick. 82.230 30.000 60.000 26.200 21.140 17.800 15.700 14.200 13.500Clear aperture rad. 775.000 557.500 346.000 375.000 375.000 375.000 375.000 375.000 375.000Surface 1 spherical rad. 2824.000 1.000E+15 3169.000 5624.000 5624.000 5624.000 5624.000 5624.000 5624.000Surface 2 spherical rad. -5021.000 -2529.000 -13360.000 -5513.000 -5564.000 -5594.000 -5612.000 -5624.000 -5624.000Sagitta of Surface 1 108.424 0.000 18.945 12.516 12.516 12.516 12.516 12.516 12.516Sagitta of Surface 2 -60.172 -62.214 -4.481 -12.769 -12.651 -12.583 -12.543 -12.516 -12.516Thick. at Clr Aperture 33.977 92.214 45.536 26.453 21.275 17.867 15.727 14.200 13.500

Approx Physical Dims L1 L2 L3 u g r i z yFreeboard 20.000 19.000 19.000 20.000 20.000 20.000 20.000 20.000 20.000Outer diameter 1590.000 1153.000 730.000 790.000 790.000 790.000 790.000 790.000 790.000Sagitta of Surface 1 114.212 0.000 21.090 13.889 13.889 13.889 13.889 13.889 13.889Sagitta of Surface 2 -63.338 -66.585 -4.987 -14.169 -14.039 -13.963 -13.918 -13.889 -13.889Sagitta of centroid 88.775 33.292 13.039 14.029 13.964 13.926 13.903 13.889 13.889Virtual edge thick. 31.356 96.585 43.897 26.480 21.290 17.875 15.730 14.200 13.500Actual edge thick. 28.974 89.422 43.897 26.480 21.290 17.875 15.730 14.200 13.500Volume (m^3) 0.113 0.066 0.022 0.013 0.010 0.009 0.008 0.007 0.007Mass (kg) 249.5 145.046 47.844 28.404 22.878 19.235 16.946 15.313 14.558*All dimensions in mm except as noted "Approx Physical Dims" are for reference only

Optical Design Parameters


Optical Element TolTolerance L1 L2 L3 FPA Filter Unit Type+

X 100.000 100.000 100.000 100.000 microns 2 Y 100.000 100.000 100.000 100.000 microns 2 Z 100.000 100.000 100.000 100.000 100.000 microns 2 Theta_X 0.0070 0.0100 0.0150 0.0167 0.015 degrees 2 Theta_Y 0.0070 0.0100 0.0150 0.0167 0.015 degrees 2 Surface 1 Spherical Radius 2.000 4.000 6.000 mm 1 Surface 1 Curvature 2.0E-08 1/mm 1 Surface 1 Conic 0.0020 -- 1 Thickness 0.200 0.250 0.500 0.500 mm 1 Surface 2 Spherical Radius 3.000 2.000 100.000 6.000 mm 1 Surface 2 Conic 0.0002 -- 1 Surface 2 A3*r^6 TBD mm^-5 1 Wedge 20.000 20.000 30.000 30.000 arcsec 2*All values are the half-amplitude value of a +/- tolerance+ Type 1: prescription errors Type 2: rigid-body placement errors Type 3: residuals from compensator palcement

Optic Tolerences


SDSSSDSSBand-pass TransitionBand-pass Transition

Half Maximum Transmission Half Maximum Transmission Wavelength (nm)Wavelength (nm)

g r i z g r i z Blue side Blue side 402 552 693 840 402 552 693 840

Red side Red side 548 693 851 -548 693 851 -

LSSTLSSTBand-pass TransitionBand-pass Transition

Half Maximum Transmission Half Maximum Transmission Wavelength (nm)Wavelength (nm)

u g r i z u g r i z y1 y2 y3 y1 y2 y3 Blue side Blue side 330330 400 552 691 818 400 552 691 818 960 970 970960 970 970 Red side Red side 400 552 691 818 922 400 552 691 818 922 1030 1020 10501030 1020 1050

G-Balmer break @400. OI line@ 557R-matches SDSSI-red side short of sky emission @826Z-red side stop before H2O bandsY–options detailed


Filter y1 and y3: Wider, redder choicesThe blue side is just long-ward of H2O bands.. CCD response steep slope - absorption coefficient of silicon falling rapidly as wavelength increases. The absorption coefficient is strongly affected by temperature. Temperature control and uniformity will be important for photometric accuracy with this filter. Filter y2: This is the narrower choice. Y2 does not extend as far to the red as the y1 filter. The red side excludes nearly all of the 930 to 960 nm H2O bands. The cutoff makes the response less sensitive to temperature of the CCD..

Y1 – 960 to 1030Y2 – 970 to 1020 (orig)Y3 – 970 to 1050 (open)


Filter band-pass is based on a combination ofscientific considerations

Filter Band Pass Transitions

Half-Maximum Transmission Wavelength

Blue Side Red Side Comments

U 350 400 Blue side cut-off depends on AR coating

G 400 552 Balmer break at 400 nm

R 552 691 Matches SDSS

I 691 818 Red side short of sky emission at 826 nm

Z 818 922 Red side stop before H2O bands

Y 948 1060 Red cut-off before detector cut-off

Uniform depositionrequired at 1% levelover entire filter

• 75 cm dia.• Curved surface• Filter is concentric about the chief Filter is concentric about the chief ray so that all portions of the filter see ray so that all portions of the filter see the same angle of incidence range, the same angle of incidence range, 14.2º to 23.6º14.2º to 23.6º

Specs

LSST Ideal Filter Passbands

0

10

20

30

40

50

60

70

80

90

100

300 400 500 600 700 800 900 1000 1100

Wavelength (nm)

Efficiency (%)

u g r i z y


LSST system throughput parameters

LSST System Throughput

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

100.0

300 400 500 600 700 800 900 1000 1100

Wavelength (nm)

System Throughput (%)

u

g r i z y

atmo

detector

optics


LSST system spectral throughput in the six filter bands

Wavelength (nm)

Sy

ste

m t

hro

ug

hp

ut

(%)

Includes sensor QE, atmospheric attenuation, optical transmission functions

LSST G- Band Composite Graph

0.0000

10.0000

20.0000

30.0000

40.0000

50.0000

60.0000

70.0000

80.0000

90.0000

100.0000

300 400 500 600 700 800 900 1000 1100 1200

Wavelength

% Transmittance

Lens/Mirror ThrouIghput*100 G(ave)-Ideal G(ave)-LSST G(ave)-Atmosphere Atmosphere LSST Detector G-Band Final

G-Band

R-Band Angular Data

0

10

20

30

40

50

60

70

80

90

100

500 550 600 650 700 750

WL

%Transmittance

ang=14.2 ang=16.08 ang=17.96 ang=19.84 ang=21.72 ang=23.6

Lens and Mirror Throughput/100 Lens and Mirror Throughput*100 R Band-IdealLens#3-AVE/100Primary/100Secondary/100Tertiary/100 Lens/Mirror Throughput WL Lens/Mirror Throughput*100 WL R-14.2(Ideal)R-23.6(Ideal)

0.379654 0.833583 0.833583 0.833583 0.153825955 300 15.38 300 36.6643 8.8580.39081 0.833922 0.833922 0.833922 0.159392441 301 15.94 301 37.187 1.64330.402537 0.834269 0.834269 0.834269 0.165324958 302 16.53 302 20.101 1.26930.414827 0.834624 0.834624 0.834624 0.171633648 303 17.16 303 21.0061 1.27620.427681 0.834986 0.834986 0.834986 0.178334226 304 17.83 304 34.2281 0.78650.441095 0.835356 0.835356 0.835356 0.185441396 305 18.54 305 16.8827 0.39770.455059 0.835734 0.835734 0.835734 0.192967841 306 19.30 306 2.0069 0.11110.469559 0.836118 0.836118 0.836118 0.200924906 307 20.09 307 1.0141 0.06610.484579 0.836508 0.836508 0.836508 0.2093242 308 20.93 308 1.2441 0.090.500098 0.836905 0.836905 0.836905 0.218175766 309 21.82 309 0.851 0.15820.516088 0.837307 0.837307 0.837307 0.227485873 310 22.75 310 0.5857 0.19240.53252 0.837716 0.837716 0.837716 0.237261594 311 23.73 311 0.1701 0.19890.549357 0.83813 0.83813 0.83813 0.247504733 312 24.75 312 0.0659 0.07360.566558 0.838548 0.838548 0.838548 0.258214727 313 25.82 313 0.061 0.04590.584076 0.838972 0.838972 0.838972 0.269387658 314 26.94 314 0.11 0.06610.601859 0.839401 0.839401 0.839401 0.28101465 315 28.10 315 0.166 0.20660.619851 0.839833 0.839833 0.839833 0.293082243 316 29.31 316 0.2122 0.85120.637991 0.84027 0.84027 0.84027 0.305572885 317 30.56 317 0.1233 1.0120.656213 0.840711 0.840711 0.840711 0.318461754 318 31.85 318 0.0482 0.3030.67445 0.841156 0.841156 0.841156 0.331720161 319 33.17 319 0.0415 0.19260.69263 0.841604 0.841604 0.841604 0.345311176 320 34.53 320 0.0802 0.30050.710681 0.842056 0.842056 0.842056 0.3591939 321 35.92 321 0.3157 1.19120.728528 0.84251 0.84251 0.84251 0.373318046 322 37.33 322 1.1314 8.6690.7461 0.842968 0.842968 0.842968 0.387632555 323 38.76 323 0.5986 14.5353

0.763322 0.843428 0.843428 0.843428 0.402074328 324 40.21 324 0.2014 14.94330.780127 0.843891 0.843891 0.843891 0.416581754 325 41.66 325 0.1625 23.40670.796449 0.844356 0.844356 0.844356 0.431085909 326 43.11 326 0.3075 32.24710.812226 0.844823 0.844823 0.844823 0.445515075 327 44.55 327 1.4525 11.46620.827404 0.845292 0.845292 0.845292 0.45979808 328 45.98 328 9.5569 13.87960.841932 0.845764 0.845764 0.845764 0.473861353 329 47.39 329 13.5281 44.1260.85577 0.846236 0.846236 0.846236 0.487632381 330 48.76 330 13.8242 37.740.868879 0.846711 0.846711 0.846711 0.501040591 331 50.10 331 25.3203 45.45720.881236 0.847187 0.847187 0.847187 0.514022475 332 51.40 332 27.0566 45.92810.892825 0.847664 0.847664 0.847664 0.526521101 333 52.65 333 10.2254 37.86230.903635 0.848142 0.848142 0.848142 0.538483223 334 53.85 334 14.6052 32.26790.913663 0.848621 0.848621 0.848621 0.549863436 335 54.99 335 46.2692 19.35280.922915 0.849101 0.849101 0.849101 0.560626487 336 56.06 336 37.1887 22.10550.931402 0.849581 0.849581 0.849581 0.570742638 337 57.07 337 46.5319 41.26270.939142 0.850063 0.850063 0.850063 0.580196064 338 58.02 338 45.8883 34.77550.946158 0.850544 0.850544 0.850544 0.588972489 339 58.90 339 37.8818 38.18210.952477 0.851026 0.851026 0.851026 0.597072299 340 59.71 340 31.7262 50.672

=((1/R3)+(1/S3)+(1/T3)+(1/U3)+(1/V3)+(1/W3)-5)^-1

Typical pass-band spreadsheet

U filter specification issues:1) Where is the optimal blue edge of the u filter? How should this blue edge be defined - by filter or by atmosphere?2) Where is the optimal red edge of the u filter?3) What is the permissible out-of-band leakage?4) What is the permissible in-band variability in the filter bandpass? 5) Is this a spatial variation as well as wavelength variation?

G band leak vs. reduced throughput:Reduced throughput in g will have huge effect on solar system object detection ➢ Kirk - calculate throughput and limiting magnitude to reduce red leak to SRD limits ➢ Lynne - with reduced limiting mag calc expected number of NEOs / other SS objects.G band leak may affect photos calculations for galaxies ➢ Lynne - ask Andy Connolly, see if additional test to run for this g band leak.➢ Zeljko - tell Lynne if there are other issues for g band leak.➢ Kirk - estimate cost to reduce g band leaks to SRD

Tasks - Some ongoing-some answered

Y band filter specification issues:

1) What is the optimal red edge for the Y filter?2) What is the optimal blue edge for the Y filter?

Independent vs. non-independent filter tradeoffs:Red and blue edge of Y filter independent.Interactions with atmosphere, detector efficiency.

Location of Red edge:Absolute limit at red edge of Y filter is set by silicon at 1.1micron. If push all the way to 1.1 micron, then must control temperature of chip closely to avoid variations in detector efficiency.

ACTIONS:➢ Kirk - specify what kind of detector variations we might expect (X% variation with a Y change in Temperature?) Atmospheric emission in OH lines occurs near the red edge of the Y filter - variation in these lines through the night will change the sky background

➢ Lynne - see if MODTRAN4 predicts these kinds of variations and what level of variations we could expect. (mag variation of sky background with typical variations in sky). This does not affect sources, but will essentially affect limiting magnitude. How will this affect calibration in Y?We would like to go as far to the red as possible, to measure quasars at the highest possible redshifts.

➢ Lynne - calculate Y2 & Y-1&3 magnitude for quasars at increasing redshift.Interactions with atmosphere:We don't actually know how the sky background varies, or even what the sky background really is in Y band.

ACTIONS:ﾘﾘ Lynne - check with calibration & opsim teams on progress of Y band measurement. Add their results to the input on Y filter. Check that measurements are consistent with MODTRAN4 predictions used in analysis for Yband edges.

I have put a new copy of the ETC on the LSST website, with the baseline sky magnitudes from the studies I did this Spring. The lunar phase information has also been upgraded based on the CTIO .9m data from Smith and Tucker.

This version will now be called ETC 4.3.0. This version replaces any beta versions of 4.3 that some of your have been running.

I put a warning screen on the sign-on page so that people would realize things have changed.. For a week or so, you will need to click in on the hyperlink "Version 4.3".

Here are the adopted solar min, max, and avg values for dark sky, airmass = 1

Values are ugrizy AB magnitudes, at the telescope (not extinction corrected). y = Y2, the 970-1020 y filter.

private static double[] solarminMag4_3 = { 22.95, 22.32, 21.43, 20.40,19.22, 18.12 };

private static double[] solaravgMag4_3 = { 22.80, 22.20, 21.30, 20.30,19.10, 18.01 };

private static double[] solarmaxMag4_3 = { 22.65, 22.10, 21.13, 20.12,18.92, 17.90 };

New ETC

Seeing = 0.500 n source type z Y1 Y2 Y3400 elliptical-galaxy 0 16.51 14.26 17.11400 elliptical-galaxy 1 16.55 14.30 17.36400 elliptical-galaxy 2 15.88 14.15 17.54Seeing = 0.750 n source type z Y1 Y2 Y3400 elliptical-galaxy 0 11.08 9.59 11.49400 elliptical-galaxy 1 11.11 9.62 11.65400 elliptical-galaxy 2 10.65 9.52 11.78Seeing = 1.000 n source type z Y1Y2 Y3400 elliptical-galaxy 0 8.32 7.21 8.63400 elliptical-galaxy 1 8.34 7.23 8.75400 elliptical-galaxy 2 8.00 7.15 8.85Seeing = 1.250 n source type z Y1Y2 Y3400 elliptical-galaxy 0 6.66 5.77 6.91400 elliptical-galaxy 1 6.68 5.79 7.01400 elliptical-galaxy 2 6.41 5.73 7.08

S/N Calculations in Y-band

By Seeing

n source type z Y1 Y2 Y3400 elliptical-galaxy 0 8.32 7.21 8.63400 elliptical-galaxy 1 8.34 7.23 8.75400 elliptical-galaxy 2 8.00 7.15 8.85400 spiral-galaxy 0 8.34 7.21 8.61400 spiral-galaxy 1 7.74 7.30 7.75400 spiral-galaxy 2 8.25 7.20 8.66400 G5V 0 8.39 7.25 8.48400 G5V 1 8.33 7.22 8.65400 G5V 2 7.86 7.12 9.00

By Source

n source type z Y1 Y2 Y3 25 elliptical-galaxy 1 2.09 1.81 2.19 50 elliptical-galaxy 1 2.95 2.56 3.10 75 elliptical-galaxy 1 3.61 3.13 3.79100 elliptical-galaxy 1 4.17 3.62 4.38125 elliptical-galaxy 1 4.66 4.04 4.89150 elliptical-galaxy 1 5.11 4.43 5.36175 elliptical-galaxy 1 5.52 4.78 5.79200 elliptical-galaxy 1 5.90 5.11 6.19225 elliptical-galaxy 1 6.26 5.42 6.57250 elliptical-galaxy 1 6.60 5.72 6.92275 elliptical-galaxy 1 6.92 6.00 7.26300 elliptical-galaxy 1 7.22 6.26 7.58325 elliptical-galaxy 1 7.52 6.52 7.89350 elliptical-galaxy 1 7.80 6.77 8.19375 elliptical-galaxy 1 8.08 7.00 8.48400 elliptical-galaxy 1 8.34 7.23 8.75

By Num of Exposures

From: S.Holland, SPIE 2006

QE for a High-rho vs a thin partially depleted CCD

High-rho study contract device ## (vendor X)

measured at BNLLSST requirements: Design Min Measured

Broadband Optimized for minimum fringing at 1 m

Wavelength Quantum Efficiency Allowable Target 400 nm 55% 60% 600 80 85 800 80 85 900 60 85 1000 25 45

LSST Sensor Sensor Thickness Study

1 0 0 1 4 0 1 8 0 2 2 0 2 6 0 3 0 0

2 5 0

2 0 0

1 5 0

1 0 0

5 0

3 0 0

8 0 %

6 0 %

4 0 %

2 0 %

T e m p e r a t u r e , K

Thickness,

μ

m

E x p e c t e d r a n g e

1 0 0 1 4 0 1 8 0 2 2 0 2 6 0 3 0 0

2 5 0

2 0 0

1 5 0

1 0 0

5 0

3 0 0

8 0 %

6 0 %

4 0 %

2 0 %

T e m p e r a t u r e , K

, Thickness

μ

m

E x p e c t e d r a n g e

LSST Sensor Sensor Thickness Study

OH Emission

• Source - Bright airglow produced by a chemical reaction of hydrogen and ozone in the Earth’s upper atmosphere

• Band system is due in part to emission from vibrationally excited OH radicals produced by surface interactions with ground-state oxygen atoms.

• Emission can vary 10-20% over a 10 minute period• Ramsey and Mountain (1992) have reported

measurements of the nonthermal emission of the hydroxyl radical and examined the temporal and spatial variability of the emission.

YY1

OH Emission at OH Emission at NIRNIR

Y2

0.00E+00

5.00E+01

1.00E+02

1.50E+02

2.00E+02

2.50E+02

3.00E+02

3.50E+02

4.00E+02

4.50E+02

5.00E+02

6000 7000 8000 9000 10000 11000

Series1

Wavelength (Å)

OH Emission

-10

0

10

20

30

40

50

800 850 900 950 1000 1050 1100 1150 1200

Wavelength

% Transmittance

Y1 930.1060 Y2 970.1020 Y3 970.open redshifted elliptical combined sky sed Atmosphere

Comparison of Y1, Y2, and Y3


z Y

Atmospheric HAtmospheric H22O BandO Band

Atmospheric Transmission

0.000

0.100

0.200

0.300

0.400

0.500

0.600

0.700

0.800

0.900

1.000

3000 4000 5000 6000 7000 8000 9000 10000 11000 12000

Wavelength

%TPalomar

Cerro Pachon

Flux ComparisonPalomar/Cerro Pachon

0

0.2

0.4

0.6

0.8

1

1.2

6000 6500 7000 7500 8000 8500 9000 9500 10000 10500

Series1

LSST G- Band Composite Graph

0.0000

10.0000

20.0000

30.0000

40.0000

50.0000

60.0000

70.0000

80.0000

90.0000

100.0000

300 400 500 600 700 800 900 1000 1100 1200

Wavelength

% Transmittance

Lens/Mirror ThrouIghput*100 G(ave)-Ideal G(ave)-LSST G(ave)-Atmosphere Atmosphere LSST Detector G-Band Final

G-Band


QuickTime™ and aTIFF (LZW) decompressor

are needed to see this picture.

Atmospheric Extinction

QuickTime™ and aTIFF (LZW) decompressor

are needed to see this picture.


M13 in Y-band

-no flat-noisy bias/dark-no cosmic ray rejection-no slope-fitting

COATINGS TEST PROGRAMReflectivity curves after 3 reflections

20

30

40

50

60

70

80

90

100

300 400 500 600 700 800 900 1000 1100 1200

Wavelength (nm)

% Reflectivity

gr

i

z

Y

Protected Ag“Aged” bare Al

“Fresh” bare Al

LLNL Protected Ag

Apache Point Mirror Data

80

82

84

86

88

90

92

94

96

98

100

300 800 1300 1800 2300

Wavelength

% Reflectance

APCH#6 APCH#7 APCH#8 APCH#9 APCH#10 APCH#11 R(ave)

SDSS Primary Mirror Witness Samples

Ghost analysis shows worst case is double-reflection from thinnest spectral filter

Relative intensity of ghost image to primary image

I = [ S / G]2 R1 R2 , S = image diameter = 0.020 mm

G = ghost image diameter = 14 mm

R = surface reflectivities = 0.01

I = 2.0 x 10 –10 = ~ 24 visual magnitude difference

Ghost halo: 14 mm

Detector plane Double-reflection in filter

________________________________________________

Filter Set SimulationsSummary

Filter Depth (in 20s) u 22.5 b 24.5 (1/2 g) g 24.3 r 24.2 I 23.4 z 22.4 y 21.5

Input SED’s valid for z<1

photoz relation for I-band: single pass < 23.4 200 passes < 25.0 400 passes < 26.65 all S/N=10

Some basic conclusions:1. u-band reduces the scatter for z < 0.5 sources2. y-band keeps the scatter tight to z~1.6

Documents

LSST Filter Development LSST Conceptual Design Review Tucson, AZ Sept 17-20, 2007 Kirk Gilmore SLAC