SPACE TELESCOPE SCIENCE INSTITUTE

SPACETELESCOPESCIENCEINSTITUTE

Operated for NASA by AURA

The On-Orbit Performance of the Cosmic Origins Spectrograph

Alessandra Aloisi(for the COS STScI team and the COS IDT)

21 July 2010

Aloisi – 21 July 2010 Slide 1 of 15

Highlights

Quick Overview of the Instrument Sensitivities and their Time Dependence Gain Sag of FUV Detector On-orbit Line Spread Function Darks and their Evolution Flat Fields Wavelength Calibration COS-to-FGS Alignment COS Documentation

Instrument Overview

Aloisi – 21 July 2010 Slide 2 of 15

FUV spectroscopy: R = 3,000 - 20,000 [1150 - 2050 Å] NUV spectroscopy: R = 2,000 - 20,000 [1700 - 3200 Å] Some NUV imaging capabilities (mostly used for TA)

Aloisi – 21 July 2010 Slide 3 of 15

COS Sensitivity

Superb in FUV Comparable to STIS in NUV NUV sensitivity in G225M & G285M is as expected at launch from grating degradation Within 20% of ground values

New COS & STIS on-orbit sensitivities

Aloisi – 21 July 2010 Slide 4 of 15

COS has Sensitivity down to the Lyman Limit

McCandliss et al. 2009 ApJL, submitted

G140L ModeR ~ 2,500

McCandliss et al. 2010, ApJ, 709, L183 Mode currently

supported in Cycle 18 (G140L @ 1280 Å Segment B)

Additional F & characterization in Cycle 17 Calibration program 12081

Two new G130M settings at R ~ 5,000 (1055 Å & 1096 Å)

S. Osterman’s talk

COS sensitivity is regularly monitored as part of Cycle 17 & 18 Calibration Programs

NUV Sensitivities - gratings keep following pre-launch trends Bare-Al gratings G225M and G285M continue to show grating-dependent

but wavelength-independent decline on orbit at a rate similar or higher than pre-launch rates (up to ~ 10% per year )

MgF2-coated gratings G185M and G230L continue to be stable FUV Sensitivities – non localized wavelength-dependent decline up to

~ 12% per year at longer λ similar for all gratings suggestive of CsI photocathode aging (with each segment of the XDL detector having different rates)

COS TDS trends already included in ETC (v18.2) and in CALCOS

See R. Osten’s talk fore more details Aloisi – 21 July 2010

Slide 5 of 15

Time Dependence of COS Sensitivity

Gain Sag of COS FUV Detector

FUV cumulative images and maps of pulse height distributions as a function of time now available (fully implemented into pipeline for OPUS build 2010.4)

Evidence of detector gain sag (localized effect) Gain after launch higher than on the ground

HV lowered during SMOV Evolution of gain sag faster than expected

Consistent with on-orbit gas contamination around launch time

We are probably re-scrubbing the detector on-orbit Gain sag is already slowing down in most heavily

exposed areas Very localized loss up to a few % in most heavily

affected areas (geocoronal Lyα) In the process to identify mitigative actions

See D. Sahnow’s poster C2Aloisi – 21 July 2010

Slide 6 of 15

Aloisi – 21 July 2010 Slide 7 of 15

COS On-Orbit FUV LSF

FUV spectral resolution not quite as expected due to MFWFEs

See P. Ghavamian’s poster C5

LSF power is distributed from core to wings

Effect is wavelength dependent,

being most extreme in the FUV

R ~ 16,000 vs 20,000 @ 1300 Å

— 1150 Å (G130M)— 1450 Å (G130M)— 1750 Å (G160M)— Gaussian 6.5 pixel FWHM (no MFWFEs)

Aloisi – 21 July 2010 Slide 8 of 15

COS On-Orbit NUV LSF

Contribution of MFWFE

present also in NUV

Almost negligible longward of 2500 Å

NUV spectral resolution at > 2500 Å as expected

Aloisi – 21 July 2010 Slide 9 of 15

COS FUV Darks

FUV dark rate away from SAA measured during SMOV was as expected from pre-launch estimates, i.e. ~ 1.5-2 x 10-6 cts/sec/pixel, when pulse-height filtering (4-30) is applied

FUV dark rate up to ~ 30 times higher close to the SAA boundary (before SAA model adjustment in May 2010)

Several weak features found for Segment B (Segment A is featureless)

TIME-TAG: most of the features removed by pulse-height filtering (4-30) during ground processing (already implemented into CALCOS)

ACCUM: suspect regions flagged; features too weak to affect very bright objects observed in this mode

FUV dark rate monitored weekly No evolution with time

See D. Sahnow’s poster C2

G130M @ 1309 Å (Segment B)

Aloisi – 21 July 2010 Slide 10 of 15

COS NUV Darks

NUV dark rate away from SAA measured during SMOV was 6x10-5 cts/sec/pixel, significantly below prelaunch predictions (~ 20x10-5 cts/sec/pixel)

NUV dark rate up to ~ 15 times higher close to the SAA boundary until SAA model adjustment in May 2010 (no long-term phosphorescence in MAMA window after SAA passage)

NUV dark exposures relatively featureless NUV dark rate monitored weekly

Approximately linear increase with time, once temperature fluctuations are taken into account

Currently slightly higher than prelaunch estimates, i.e. ~ 35x10-5 cts/sec/pixel

Still several times lower than current STIS NUV MAMA values

See W. Zheng’s poster S1

Aloisi – 21 July 2010 Slide 11 of 15

COS NUV Flat

NUV on-orbit flat (internal lamp) confirmed consistency with ground flat at higher S/N (alignment better than 1 pixel)

NUV flat field currently used Combination of ground and on-orbit flats Includes on-orbit vignetting (depression up to 15-

20%) in first ~ 200 pixels Vignetting currently implemented separately for

M (G185M, G225M, G285M) and L (G230L) gratings

NUV flat field correction currently applied in CALCOS

NUV flat allows to achieve S/N > 100 per resel when combined with 4 different FP-POS positions with high S/N data

NUV ground flat

NUV flat for on-orbit data

Aloisi – 21 July 2010 Slide 12 of 15

COS FUV Flat

Vallerga et al. 2001, SPIE 4498, 141

Sum of the 8 rows indicated above

“Berkeley” Flat of Segment A (whole detector, no grid wires)

2D ground flats (Berkeley + TV03) still to be validated with on-orbit data

No 2D on-orbit flat with adequate S/N currently available

FUV flat field correction currently not applied in CALCOS

CALCOS currently omits grid wires and other defects when creating x1dsum files

1D iterative technique combined with 4 different FP-POS positions looks promising and allows to achieve S/N > 100 per resel with high S/N data

Implementation of 1D flats on extracted spectra underway

New 1D flats for all FUV gratings to be obtained in Cycle 17 Calibration program 12086

See T. Ake’s talk for more details on 1D flats

Aloisi – 21 July 2010 Slide 13 of 15

COS On-Orbit Wavelength Calibration

TV03 demonstrated that dispersion solutions of PSA and WCA differ from each other

PSA/WCA wavelength dispersion solutions obtained in TV03 TV06 demonstrated that dispersion solution difference between

WCA and PSA depends on optical alignment SMOV data used to:

infer new on-orbit wavecal templates measure on-orbit zeropoints compared to ground calibration

New CALCOS reference files that update wavelength calibration for most settings delivered in January 2010 (G185M still to be completed)

On-orbit wavelength scales for most settings meet requirements 15 km/s for the M gratings 150 km/s (175 km/s) for G140L (G230L) gratings In a few cases, extra offsets of up to 10 pixels in FUV appear in

part of the spectrum (nature & frequency of such localized distortions still under investigation)

Cycle 17 monitoring indicates that PSA to WCA on-orbit offsets are stable with time

See C. Oliveira’s poster C4

Aloisi – 21 July 2010 Slide 14 of 15

COS-to-FGS Alignment and TA COS-to-FGS alignment optimized during

SMOV and Cycle 17; SIAF tables updated Figure includes centering slews in all TA

sequences of GTO/GO programs since update of SIAF table in March 2010

Alignment now appears to be accurate to within ~ 0.1”– Initial pointing now limited by accuracy of

GSC2 catalogue (~ 0.2”) Reconsidered the need for an ACQ/Search at

start of each visit for Cycle 18 Phase II– Change will save time, especially for

acquisitions of very faint targets– For faint targets, ACQ/Search can have

problems with variable background– Coordinates need to be accurate within 0.4”

COS Documentation

On-orbit COS performance summarized in 10 published ISRs and 6 soon-to-be-published ISRs:

http://www.stsci.edu/hst/cos/documents/isrs

Preliminary SMOV updates included into the Instrument Handbook for Cycle 18 (published in January 2010): http://www.stsci.edu/hst/cos/documents/handbooks/current/cos_cover.html

SMOV and Cycle 17 Calibration updates currently implemented into the Instrument Handbook for Cycle 19 (for publication in December 2010)

Visit our web pages at: http://www.stsci.edu/hst/cos

Aloisi – 21 July 2010 Slide 15 of 15