SOFIE Overview Mark Hervig Larry Gordley GATS SOFIE

SOFIE Overview Mark Hervig Larry Gordley GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 1 -1 AIM Science Team Meeting 23 Jan 2007

SOFIE Solar Occultation 8 channel differential radiometer 16 bands, 290 nm - 5 m wavelength Coverage from 65 - 85 latitude Retrieval Precision Required / CBE Altitude Range (km) PMC Interference (signal fraction) T 5 / 3. 5 K 1 - 95 0% O 3 100 / 7 ppbv 15 - 105 50% H 2 O 0. 6 / 0. 04 ppmv 15 - 100 0% CO 2 10 / 7 ppmv 15 - 100 0% CH 4 50 / 38 ppbv 15 - 90 90% NO 1. 0 107 / 1. 1 107 cm-3 80 - 110 20% PMCs V: 5 10 -5 / 1. 8 10 -5 km-1 V: 5 10 -5 / 1. 1 10 -7 km-1 cloud - GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 2 -2 AIM Science Team Meeting 23 Jan 2007

SOFIE Changes The SOFIE steering mirror assembly (SMA) failed when vibrated above design limits during S/C integration & test in June 2006. The steering mirror was replaced with a rigid mirror. The impact: 1) SOFIE pointing now relies on the spacecraft AC system. AIM requirements are met. 2) Complete calibration and TVAC testing were performed in November 2006. Confirmed that nothing else was damaged. Small changes in mirror reflectance spectra resulted in little or no change. GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 3 -3 AIM Science Team Meeting 23 Jan 2007

SOFIE Pointing • SOFIE Steering Mirror Assembly provided: - Ability to point and hold FOV at sun center during entire occultation - Agility to accomplish calibration maneuvers during an event • Using the spacecraft to point SOFIE provides: - Sufficient stability to achieve required science objectives. - Capability to accomplish calibration maneuvers, though some will require multiple events and increased demand from flight operations. • The SOFIE sun sensor remains functional and precise: sub-arcsec knowledge GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 4 -4 AIM Science Team Meeting 23 Jan 2007

SOFIE Pointing Then and Now The steering mirror provided Lockdown: sun center, < 1 arcsec uncertainty FOV drift: <1 arcsec during an occultation. FOV Pointing stability is important because solar intensity changes across the disk. exaggerated drift The spacecraft provides Lockdown: sun center, <2. 5 arcmin uncertainty FOV drift: <12 arcsec through PMC altitudes GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 5 -5 AIM Science Team Meeting 23 Jan 2007

SOFIE Errors Related to Pointing Effects of FOV lockdown uncertainty and drift were characterized by combining Spacecraft pointing errors & SOFIE signal simulations • Signals simulated by convolving 2 D FOV over 2 D solar source. • Drift-induced V signals were determined as worst case for: All lockdowns with 2. 5 arcmin of center 12 arcsec drift in all directions from lockdown. • Drift-induced V signals were then corrected using: Sun sensor pointing knowledge (0. 88 arcsec 1 error) Measured limb darkening curve (w/ appropriate noise) • The remaining signal error is now carried in SOFIE performance estimates. • Note: expected steady drifts can be removed with lower uncertainties by fitting and extrapolating in time. GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 6 -6 AIM Science Team Meeting 23 Jan 2007

Impact of Spacecraft Pointing Drift on SOFIE SNR Worst case induced V uncertainty Margin after Center V SNR RSS of (quasi-systematic in noise units) Chan. Wavelength Required / CBE 12 arc-sec drift in any direction random + ( m) Margin (random) from lockdown at 2. 5 arcmin* systematic 1 / O 3 s 0. 291 1 1. 0 e 4 / 210. 9 0. 08 12. 8 2 / O 3 w 0. 330 3 / PMC s 0. 867 2 1. 0 e 6 / 5. 9 1. 79 0. 6 4 / PMC w 1. 04 5 / H 2 O w 2. 46 3 2. 5 e 4 / 16. 6 0. 02 15. 7 6 / H 2 O s 2. 62 7 / CO 2 s 2. 79 4 3. 0 e 5 / 1. 4 0. 19 1. 3 8 / CO 2 w 2. 94 9 / PMC w 3. 06 5 1. 0 e 5 / 3. 0 0. 16 2. 7 10 / PMC s 3. 12 11 / CH 4 s 3. 38 6 4. 0 e 5 / 1. 3 0. 23 1. 3 12 / CH 4 w 3. 48 13 / CO 2 s 4. 32 7 4. 0 e 5 / 1. 5 0. 17 1. 5 14 / CO 2 w 4. 65 15 / NO w 5. 01 8 3. 0 e 5 / 0. 9 0. 26 0. 9 16 / NO s 5. 32 *Lockdown was moved within a radius of 2. 5 arcmin and the signal change for 12 arcsec drift (over 10 seconds / 25 km) was calculated. Numbers shown are the worst case for all possible lockdowns. Note: the drift induced d. V signal uncertainty is the error in the corrected signal considering a regression to the signal change vs FOV drift, the measurement noise (from calibration), and the pointing error (from calibration). Band / Target This analysis captures all of the errors and is our CBE. GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 7 -7 AIM Science Team Meeting 23 Jan 2007

Predicted SOFIE Retrieval Performance Retrieval performance meets or exceeds AIM science requirements Geophysical Parameter Precision Required / SMA Margin / Rigid Mirror Margin Altitude Range, km Required / Current Vertical Resolution, km Required / Current NIR cloud extinction* 5 x 10 -6 km-1 / 5. 9 / 0. 6 78 – 85 / 75 – 90 3 / 1. 5 IR cloud extinction 5 x 10 -5 km-1 / 3. 0 / 2. 7 78 – 85 / 75 – 90 3 / 1. 5 temperature 5 K / 1. 5 / 1. 4 70 – 90 / 15 - 95 3 / 1. 5 O 3 mixing ratio 0. 1 ppmv / 50 / 12. 8 78 – 90 / 15 - 105 3 / 1. 5 H 2 O mixing ratio 0. 6 ppmv / 17 / 14. 0 78 – 90 / 15 - 100 3 / 1. 5 CO 2 mixing ratio 10 ppmv / 1. 5 / 1. 4 80 – 100 / 15 - 100 3 / 1. 5 CH 4 mixing ratio 0. 05 ppmv / 1. 3 / 1. 4 30 – 90 / 15 - 90 3 / 1. 5 NO mixing ratio 107 cm-3 / 0. 9 80 – 95 / 80 - 110 5/3 *This measurement is not an AIM Level 1 science requirement GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 8 -8 AIM Science Team Meeting 23 Jan 2007

On-Orbit Calibration with S/C Pointing In-flight calibrations are accomplished by directing the FOV through various maneuvers: Routine Calibration (every event): Solar scan to characterize the solar limb darkening curve (SLDC) Periodic Calibration (~monthly): Nonlinearity, Sun sensor FPA flat field, Boresight, FOV mismatch Difference signal gain, Low-pass filter response The spacecraft can perform the required maneuvers. Some calibrations will now require statistical analysis. Elevated but realistic requirements on SOFIE operations and Mission operations. (Greg has more on this) GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 9 -9 AIM Science Team Meeting 23 Jan 2007

Recent SOFIE Calibration SOFIE underwent final TVAC testing and Calibration in Nov 2006. All items characterized at nominal operating temperature: • Noise • Gain • Nonlinearity • FOV, knife edge & point source • RSR, in-band & select out-of-band • Sun sensor GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 10 - 10 AIM Science Team Meeting 23 Jan 2007

SOFIE Calibration: Signal-to-Noise SOFIE noise levels were characterized using measurements taken with the chamber aperture closed. Comparisons with past data indicate little or no change. Chan. 1 / O 3 s 1 2 3 SNR margin = 4 (215 V gain / V) / (SNR required) 5 6 No Change, all channels have margin > 1, except 8. Band / Target 7 8 2 / O 3 w 3 / PMC s 4 / PMC w 5 / H 2 O w 6 / H 2 O s 7 / CO 2 s 8 / CO 2 w 9 / PMC w 10 / PMC s 11 / CH 4 s 12 / CH 4 w 13 / CO 2 s 14 / CO 2 w 15 / NO w 16 / NO s GATS V SNR Margin Oct 2005 Nominal V SNR Margin Feb 2006 Nominal V SNR Margin Nov 2006 Nominal 146. 2 148. 1 210. 9 5. 2 5. 8 5. 9 15. 7 17. 4 16. 6 1. 3 1. 4 1. 3 3. 0 1. 2 1. 3 1. 5 1. 7 1. 5 0. 9 SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 11 - 11 AIM Science Team Meeting 23 Jan 2007

SOFIE Calibration: RSR In-Band RSR 50% Response Limits Specification Jan 2006 (PER) Nov 2006 Cuton - Cutoff ( m) Cuton-Cutoff ( m) Band Measurement 1 O 3 Strong 0. 2857 – 0. 2941 0. 2858 – 0. 2972 0. 2867 – 0. 3008 2 O 3 Weak 0. 3226 – 0. 3333 0. 3227 – 0. 3375 0. 3237 – 0. 3381 3 Particle Strong 0. 8475 – 0. 8772 0. 8506 – 0. 8831 0. 8510 – 0. 8830 4 Particle Weak 1. 0101 – 1. 0526 1. 015 – 1. 059 5 H 2 O Weak 2. 427 – 2. 475 2. 436 – 2. 488 6 H 2 O Strong 2. 577 – 2. 632 2. 592 – 2. 644 2. 592 – 2. 643 7 CO 2 Strong 2. 740 – 2. 794 2. 758 – 2. 813 8 CO 2 Weak 2. 907 – 2. 967 2. 910 – 2. 968 9 Particle Strong 3. 030 – 2. 091 3. 035 – 3. 094 3. 038 – 3. 097 10 Particle Weak 3. 160 – 3. 226 3. 152 – 3. 219 3. 155 – 3. 221 11 CH 4 Strong 3. 333 – 3. 401 3. 345 – 3. 422 3. 347 – 3. 424 12 CH 4 Weak 3. 472 – 3. 546 3. 445 – 3. 514 3. 447 – 3. 516 13 CO 2 Strong 4. 255 – 4. 444 4. 219 – 4. 428 4. 221 – 4. 429 14 CO 2 Weak 4. 630 – 4. 740 4. 581 – 4. 712 4. 583 – 4. 714 15 NO Weak 4. 951 – 5. 051 4. 959 – 5. 053 4. 962 – 5. 055 16 NO Strong 5. 263 – 5. 376 5. 236 – 5. 396 5. 238 – 5. 399 GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 12 - 12 AIM Science Team Meeting 23 Jan 2007

SOFIE Calibration: Nonlinearity Oct 2005 Comparison of nonlinearity calibration data from Oct 05, Feb 06, and Nov 06 show consistent results. Example plots for band 8 are shown here. Feb 2006 GATS Nov 2006 SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 13 - 13 AIM Science Team Meeting 23 Jan 2007

SOFIE Calibration: FOV MIC 1 rectangular aperture was scanned across the SOFIE aperture in discrete steps. The derivative of measured response yields the FOV response curve. Requirements (FWHM): EL: 1. 8 arcmin Chan. 1 2 3 Band 3 -16 Averages, Feb 2006: EL: 1. 90 arcmin 4 Band 3 -16 Averages, Nov 2006: EL: 1. 83 arcmin 5 6 Note: the Nov 2006 cal was done at higher spatial resolution than previously. Changes are small and within the calibration uncertainty. 7 8 GATS Band / Target EL Dimension (arcmin, FWHM) Feb 2006, day 51 EL Dimension (arcmin, FWHM) Nov 2006 1 / O 3 s 1. 86 2 / O 3 w 1. 96 3 / PMC s 1. 83 1. 86 4 / PMC w 1. 77 1. 85 5 / H 2 O w 1. 73 1. 65 6 / H 2 O s 1. 87 1. 68 7 / CO 2 s 2. 00 1. 80 8 / CO 2 w 1. 99 1. 76 9 / PMC w 1. 85 1. 70 10 / PMC s 1. 81 1. 59 11 / CH 4 s 2. 03 1. 93 12 / CH 4 w 1. 86 1. 66 13 / CO 2 s 2. 00 2. 02 14 / CO 2 w 1. 96 2. 07 15 / NO w 1. 97 1. 96 16 / NO s 1. 98 2. 07 SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 14 - 14 AIM Science Team Meeting 23 Jan 2007

SOFIE Calibration: Sun Sensor The sun was viewed for several minutes on 6 Nov 2006. The SOFIE sun sensor measured the image location, and stayed in fine track for most of the test. Without the SMA, the solar image drifted across the FPA. Drift was removed using a linear trend, and the tracking precision was estimated as the standard deviation of the mean at 2 Hz. Example analysis of elevation data are shown at right. Results from Nov 2006: Elevation precision = 0. 70 arssec. Azimuth precision = 0. 54 arssec. Results from Nov 2005: Elevation precision = 0. 32 arssec. Azimuth precision = 1. 68 arssec. GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 15 - 15 AIM Science Team Meeting 23 Jan 2007

November 2006 Calibration Summary Data collected to address all critical performance aspects Results indicate excellent system performance & little or no changes System Functionality Requirement November 2006 Calibration Results Signal to Noise Margin > 1 Response Nonlinearity < 0. 5% V Gain As specified No change Radiometer Gain * As specified Small ( 5%) changes at short wavelengths (as expected due to new mirror reflectance spectra) Background Signals N/A No change RSR - Bandpass As specified No change RSR – Out of Band Energy Desired < 1% No change FOV As specified No change Pointing Knowledge < 1 arcsec in elevation Slight Improvement (< 1 arcsec in azimuth & elevation) No Change Measured margin > 1 (except NO channel, 0. 9) No change < 0. 5% calibration uncertainty on all bands * Does not affect science. Balance adjustment can accommodate changes of 100% or more before science is impacted. GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 16 - 16 AIM Science Team Meeting 23 Jan 2007

Science Products / Post processing PMC identification: • Find PMCs in SOFIE profiles using threshold approach • Determine cloud heights: top, base, & peak • Status: baseline version complete • Issues: may require sophistication such as spectral identification PMC ice mass: • Determine ice mass content in PMCs (direct proportionality to IR extinction) • Status: complete • Issues: small dependence on particle shape & refractive index PMC size distribution retrievals: • Retrieve a 3 -parameter size distribution (lognormal or gaussian) • Status: baseline version complete • Issues: choice of refractive indices, particle shape / aspect ratio… GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 17 - 17 AIM Science Team Meeting 23 Jan 2007

Summary Testing indicates that damage to SOFIE was limited to the steering mirror. Calibration shows little or no change in SOFIE performance. Spacecraft pointing allows SOFIE measurements to meet AIM requirements. (Backup Slides Follow) GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 18 - 18 AIM Science Team Meeting 23 Jan 2007

SOFIE Pointing Solar intensity varies spatially across the disc Ø S 1 Ro R 1 R 2 The strength and shape of this variation is a function of wavelength Ø Radiance 2 Radiometer signals vary simply with spatial variations in the solar limb darkening (SLD) curve Ø Ø Difference signals are affected by Differences in the SLD at each wavelength. Distance Along Solar Disc FOV missmatch between band pairs 2 < 1 For S: Ro - R 2 >> Ro - R 1 GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 19 - 19 AIM Science Team Meeting 23 Jan 2007

SOFIE Status Instrument: SOFIE is integrated on the spacecraft. S/C TVAC is complete. S/C vibe is complete. No issues or non-conformances Data Processing Software (Retrievals): Level 0: Level 1: Level 2: GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 20 - 20 AIM Science Team Meeting 23 Jan 2007

Outline 1) Overview of SOFIE changes 2) SOFIE status 3) Calibration results 4) Predicted retrieval performance 5) Summary GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 21 - 21 AIM Science Team Meeting 23 Jan 2007

SOFIE Nonlinearity Calibration Detector response nonlinearity has been calibrated numerous times. The CSM flight ND filter was absent in the Oct 2005 results, allowing the entire dynamic range to be sampled. The Oct 2005 results are therefore considered our current baseline. The ND filter was in place for later results, allowing about 1/3 of the dynamic range to be covered. Changes from the baseline Oct 2005 results are small, and generally within the calibration uncertainty. Chan. 3 4 5 6 7 8 Band / Target Nonlinearity (%) error* (%) at 215 counts Oct 2005 Nonlinearity (%) error* (%) at 215 counts Feb 2006 Nonlinearity (%) error* (%) at 215 counts Nov 2006 5 / H 2 O w 5. 9 0. 3 4. 7 1. 3 4. 2 1. 2 6 / H 2 O s 5. 1 0. 3 6. 1 1. 0 6. 9 0. 7 7 / CO 2 s 31. 4 0. 2 33. 4 0. 8 35. 7 0. 5 8 / CO 2 w 28. 0 0. 4 31. 2 0. 7 31. 2 0. 3 9 / PMC w 2. 6 0. 2 6. 5 0. 9 7. 7 0. 7 10 / PMC s 5. 3 0. 2 6. 5 0. 7 4. 4 0. 4 11 / CH 4 s 5. 7 0. 2 8. 0 0. 6 7. 1 0. 3 12 / CH 4 w 8. 4 0. 2 10. 5 0. 8 8. 3 0. 4 13 / CO 2 s 16. 4 0. 4 19. 1 0. 8 17. 4 0. 2 14 / CO 2 w 10. 3 12. 1 0. 5 12. 3 0. 2 15 / NO w 6. 3 0. 2 10. 4 0. 5 8. 5 0. 2 16 / NO s 7. 2 0. 2 11. 1 0. 5 9. 7 0. 2 *The nonlinearity error is the associated signal error when viewing the unattenuated sun. GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 22 - 22 AIM Science Team Meeting 23 Jan 2007

Impact of Spacecraft Pointing Drift on SOFIE SNR HALOE demonstrated steady time dependent drifts, if this is the case for SOFIE… Chan. 1 2 3 4 5 6 7 8 Band / Target 1 / O 3 s 2 / O 3 w 3 / PMC s 4 / PMC w 5 / H 2 O w 6 / H 2 O s 7 / CO 2 s 8 / CO 2 w 9 / PMC w 10 / PMC s 11 / CH 4 s 12 / CH 4 w 13 / CO 2 s 14 / CO 2 w 15 / NO w 16 / NO s Center V SNR Wavelength Required / CBE ( m) Margin (random) 0. 291 0. 330 0. 867 1. 04 2. 46 2. 62 2. 79 2. 94 3. 06 3. 12 3. 38 3. 48 4. 32 4. 65 5. 01 5. 32 Baseline Margin IF drift-induced noise is a Including drift-induced steady function of time noise 1. 0 e 4 / 210. 9 27. 6 12. 8 1. 0 e 6 / 5. 9 1. 4 0. 6 2. 5 e 4 / 16. 6 16. 2 15. 7 3. 0 e 5 / 1. 4 1. 3 1. 0 e 5 / 3. 0 2. 7 4. 0 e 5 / 1. 3 4. 0 e 5 / 1. 5 3. 0 e 5 / 0. 9 GATS SOFIE – October 22, 2003 SOFIE Overview, Mark. PDR Hervig 23 - 23 AIM Science Team Meeting 23 Jan 2007