CEA DSM Dapnia SAp

Flux calibrationof the Photometer

Koryo Okumura, Marc Sauvage,

Nicolas Billot, Bertrand Morin

DSM/DAPNIA/Sap

November 8th 2007 - Koryo Okumura - PACS Sience Verification Review (Phase 3 @ MPE Garching) : Flux calibration of the PhotometerCEA DSM Dapnia SAp 2

Stable signal for a better photometry

• How stable is the signal?– Response?– Offset?

• Response variation is quite small– less than 0.1%

• Offset is the main component of the drift seen in the noise spectral density– about some hundreds of micro volts in 3

hours

November 8th 2007 - Koryo Okumura - PACS Sience Verification Review (Phase 3 @ MPE Garching) : Flux calibration of the PhotometerCEA DSM Dapnia SAp 3

Response and offset monitoring

November 8th 2007 - Koryo Okumura - PACS Sience Verification Review (Phase 3 @ MPE Garching) : Flux calibration of the PhotometerCEA DSM Dapnia SAp 4

Drift seen in low frequency measurements

November 8th 2007 - Koryo Okumura - PACS Sience Verification Review (Phase 3 @ MPE Garching) : Flux calibration of the PhotometerCEA DSM Dapnia SAp 5

Linearity and non-linearity of the response

1 % of responsivity loss for ≈ 0.1 pW/pixel ≈ 3 Jy/pixel

November 8th 2007 - Koryo Okumura - PACS Sience Verification Review (Phase 3 @ MPE Garching) : Flux calibration of the PhotometerCEA DSM Dapnia SAp 6

Flat field derivation using data from ILT3 (1)

OGSE BBCS2 CS1

Chopper scan

Median value for each pixel on this interval

2 OGSEBB fluxes

2 medianimages

1 responsivity image

Responsivity = <responsivity image>, Flat_field =Responsivity image

Responsivity

November 8th 2007 - Koryo Okumura - PACS Sience Verification Review (Phase 3 @ MPE Garching) : Flux calibration of the PhotometerCEA DSM Dapnia SAp 7

Flat field derivation using data from ILT3 (2)

• Chopper scan on OGSE of the field of view measurements are used

• Median 2D image is considered to have a flat flux level at the central chopper interval

• Difference of 2 median images of different fluxes (pixel to pixel response to a uniform brightness distribution through the central field-of-view distortion)

• This divided by the flux difference computed from OGSE temperatures provides a pixel to pixel responsivity 2D map

• The average over the valid pixels gives the mean responsivity (scalar) and the flat field

November 8th 2007 - Koryo Okumura - PACS Sience Verification Review (Phase 3 @ MPE Garching) : Flux calibration of the PhotometerCEA DSM Dapnia SAp 8

Field of view measurements during ILT3Blue with 2 different filters

T_ogse

[kelvin]

70 um

[pW/pixel]

100 um

[pW/pixel]

dFlux

[pW/pixel]

10 0.0000 0.0004 0.0003

15 0.002 0.029 0.026

20 0.055 0.281 0.227

22 0.132 0.530 0.399

25 0.382 1.142 0.761

30 1.424 2.953 1.529

35 3.694 5.884 2.191

40 7.612 9.951 2.340

November 8th 2007 - Koryo Okumura - PACS Sience Verification Review (Phase 3 @ MPE Garching) : Flux calibration of the PhotometerCEA DSM Dapnia SAp 9

Flux difference on the blue detectorwith 2 different filters and different OGSE temperatures

Flux 1 [pW/pixel] Flux 2 [pW/pixel] dFlux [pW/pixel]

0.0000 0.0004 0.0003

0.0004 0.0023 0.0020

0.0023 0.0287 0.0264

0.0287 0.0548 0.0260

0.0548 0.1317 0.0769

0.1317 0.2815 0.1498

0.2815 3815 0.1000

0.3815 0.5303 0.1488

0.5303 1.1421 0.6118

1.1421 1.4238 0.2817

1.4238 2.9532 1.5293

2.9532 3.6937 0.7405

3.6937 5.4843 2.1906

5.8843 7.6117 1.7275

7.6117 9.9514 2.3396

November 8th 2007 - Koryo Okumura - PACS Sience Verification Review (Phase 3 @ MPE Garching) : Flux calibration of the PhotometerCEA DSM Dapnia SAp 10

Flux difference on the red detectorwith 2 different OGSE temperatures

Flux 1

[pW/pixel]

Flux 2

[pW/pixel]

dFlux

[pW/pixel]

0.043 0.712 0.669

0.712 3.107 2.395

3.107 4.693 1.586

4.693 7.760 3.067

7.760 14.574 6.814

14.574 23.207 8.633

23.207 33.294 10.088

November 8th 2007 - Koryo Okumura - PACS Sience Verification Review (Phase 3 @ MPE Garching) : Flux calibration of the PhotometerCEA DSM Dapnia SAp 11

Noise propagation in the photometry

• Signal = flat_field x Flux + offset• dSignal = flat_field x dFlux

• Each time we do a multiplication or division, a relative variance is added

• dFlux should be large to reduce the noise, but should be small enough to stay within the valid linear range

November 8th 2007 - Koryo Okumura - PACS Sience Verification Review (Phase 3 @ MPE Garching) : Flux calibration of the PhotometerCEA DSM Dapnia SAp 12

Which responsivity and flat field to use?

• Standard flat field and CS flux:– OGSE flat field : dSignal / dFlux– CSs flux : dSignal / flatField

• Calibration block flat field:– CSs flat field : dSignal / dFlux

Calibration block Observation Calibration block

November 8th 2007 - Koryo Okumura - PACS Sience Verification Review (Phase 3 @ MPE Garching) : Flux calibration of the PhotometerCEA DSM Dapnia SAp 13

How do we use the CS calibration blocks?

• Standard flat field

• Flux = Signal / OGSE_flat_field

• Sum of relative variances of :

– OGSE flux

– OGSE signal

– Data signal

• CSs flat field

• Flux = Signal / CSs_flat_field

• Sum of relative variances of :

– OGSE flux

– OGSE signal

– Cal CSs signal

– Data CSs signal

– Data signal

November 8th 2007 - Koryo Okumura - PACS Sience Verification Review (Phase 3 @ MPE Garching) : Flux calibration of the PhotometerCEA DSM Dapnia SAp 14

Flat field and responsivity as calibration file

• The standard flat field should be used for the flux calibration of the data

• The flat field depends on the flux level• We need flat fields with low noise• The flat field has to be interpolated from

available flux level to a real background flux level

• Flat field of the red detector contains the electrical cross-talk

• Flat field of the red detector is a poor quality because of the offset drift in the measurements

• How can we measure in orbit a good flat field, if necessary?