Cycle 20 COS Calibration Plan. Jerry Kriss & COS/STIS Team 9/13/2012

Transcription

1 Cycle 20 COS Calibration Plan Jerry Kriss & COS/STIS Team 9/13/2012

2 COS Cycle 20 Usage Statistics Based on Phase II Submissions COS orbits comprise of 24.1% all prime exposure time (2832 GO orbits) in Cycle 20 when considering all instruments. (Compared to 17.6% in Cycle 19.) 98% of the total COS prime observing time is Spectroscopy science exposures. 92% of the COS prime spectroscopy observing time is FUV science exposures. There are no COS SNAP programs in Cycle 20. Instruments* Prime Exposure Time % SNAP Exp Time % ACS COS STIS WFC FGS NICMOS N/A N/A * Based on Merle Reinhart s Cycle 20 statistics in /home/reinhart. Exposure times have some inaccuracies due to the way sub-exposures are treated in the database query.

3 COS Cycle 20 Usage Statistics Based on Phase II Submissions Configuration Total science exposures breakdown by Mode and GraMng Grating % COS Prime Exposures Cycle 20 % COS Prime Exposures Cycle 19 COS/FUV G140L 29.8% * 6.6% G130M 38.5% ** 52.5% G160M 18.1% 29.7% COS/NUV G230L 1.4% 1.4% G185M 6.0% 9.8% G225M 0.0% 0.0% G285M 0.7% 0.0% * 77% of total G140L observamons go to exposures using CENWAVE ** 26.5 % of total G130M observamons go to exposures in the blue Modes : CENWAVEs 1222 (20.5%), 1055 (5%), 1096 (1%).

4 COS Cycle 20 Usage Statistics Based on Phase II Submissions Total Science Exposure Time Breakdown by FP- POS TIME- TAG/ FP- POS Exposure Time (s) Percentage of Total Exposure Time (%) ALL Total % Our FP- POS RecommendaMons are followed by our PIs (nice ).

5 COS Cycle 20 Calibration and Monitor Orbits Request Prop. ID Title External Internal Frequency FUV Monitors FUV Spectroscopic Sensitivity Monitor 33 9x1 using GD71, 12x2 using WD Cycle 19 Allocation 34 COS Cycle 17: 20 programs 149 external orbits 446 internal orbits 615 total orbits Cycle 20 Requests FUV Detector Dark Monitor 260 5x52 (5/wk) 130 FUV Internal/External Wavelength Scale Monitor 2 1x1/M + 1x1/L 6 FUV Detector Recovery After Anomalous Shutdown (13-17) 32 COS Observations of Geocoronal Ly α Emission 4 5 NUV Monitors Monitoring of COS ACQ/IMAGE performance 2 2x1 NUV Spectroscopic Sensitivity Monitor 6 3x1/L, 3x1/M 8 NUV Detector Dark Monitor 52 52x1 (2/alt. wk) 52 NUV Internal/External Wavelength Scale Monitor 1 1x1 3 NUV MAMA Fold Distribution 1 * 1x1 1 NUV Detector Recovery After Anomalous Shutdown (4) COS Cycle 18: 11 programs 65 external orbits 5 parallel orbits 183 internal orbits 253 total orbits COS Cycle 19: 10 programs 51 external orbits 193 internal orbits 244 total orbits COS Cycle 20: 11 programs 44 external orbits 313 internal orbits 357 total orbits * Indicates orbits > 1800s. ** Indicates orbits > 1800s and a mixture of external and internal orbits. Indicates external parallel orbits allocated with a STIS program. () Indicates contingency orbits not included Cycle 20 request. Number of COS Cycle 20 GO External Orbits = 481

6 FUV Spectroscopic Sensitivity Monitor P.I. Azalee Bostroem Purpose Description Fraction GO/GTO Programs Supported Observations Monitor sensitivity of each FUV grating mode to detect any change due to contamination or other causes. The FUV gratings are the most heavily used on COS, and have also experienced several changes in the time-dependent spectroscopic sensitivity since launch. Since these trends appear to be grating-, segment-, and wavelengthdependent, it is necessary to monitor them frequently to capture the changes. The main goal is to track the time dependence of sensitivity as a function of wavelength. Obtain exposures in all FUV gratings every month. Every month there will be 2 visits totaling 3 orbits (except May-July when GD71 is unavailable). The 1 orbit visit will cover the G130M/1096/FUVB, G160M/1577/FUVA, and G160M/1623/FUVA central wavelengths. The 2 orbit visit will cover G130M/1222, G130M/ 1291, G130M/1327, G160M/1577/FUVB, G160M/1623/FUVB, G140L/1105/FUVA, and G140L/1230 central wavelengths. These comprise the reddest and bluest central wavelengths of each grating with additional coverage of the new G130M blue modes. 92% of COS total exposure time. 33 external orbits Analysis 10 FTE weeks Products ISR, Time-Dependent Sensitivity Reference File and a summary in the end of cycle ISR. Accuracy Goals SNR of 15 per resel at wavelength of least sensitivity Scheduling & Special Requirements Changes from Cycle 19 Monitoring should be monthly. Hide FUVA turn-off of GD71 visits in GS-ACQ. (GD71 is unavailable May July.) New targets reduce the lifetime impact. Removed M-mode middle central wavelengths: G130M/ 1309 and G160M/1600. Switched from G140L/1230 to G140L/1280. The Increased efficiency permits us to add monitoring for G130M/1096 and G130M/1222.

7 TDS Decline Over Time vs. Solar Activity

8 Improvements for Cycle 20 Reduce the lifetime impact: Eliminate monitoring of the middle central wavelengths for medium resolution modes. Simulations verify that this does not reduce the final accuracy. Use targets with flatter SEDs to avoid overexposing the detector to meet minimum S/N requirements. Monitor G160M segments separately to avoid overexposing the short-wavelength segment. Add G130M/1096/FUVB and G130M/1222 to monitor the short wavelengths of G130M. GOs are using these modes in Cycle 20. Monitor G140L/1280 (rather than 1230) as it is now preferred.

9 Continuing with 3 Orbits per Month Analyzed moving from 3 orbits per month to 2 orbits per month by eliminating monitoring of the middle central wavelengths of all gratings. We would need to cut 60% of the exposure time on all central wavelengths of the WD0308 target. Peak of the solar cycle à most extreme slopes and slope changes. These are our first monitoring observations at the new lifetime position using new targets and a return to a monthly cadence. TDS observations have had a variety of unforeseen uses (e.g. gainsag, flat fields, optimal extraction). High S/N observations required to characterize the instrument to 2% for all calibration. Despite extensive analysis we remain unconvinced that we can achieve 2-3% relative flux accuracy in 2 orbits. To verify, we re-analyzed our current data using the new scenarios. We also used Monte Carlo simulations to investigate alternative cadence and S/N scenarios.

10 Changes to Modes Monitored in Cycle 20 Grating Cenwave Segment Cycle 19 Cycle 20 G130M 1096 FUVB 1222 FUVA/FUVB 1291 FUVA/FUVB 1309 FUVA/FUVB 1327 FUVA/FUVB G160M 1577 FUVA 1577 FUVB 1600 FUVA/FUVB 1623 FUVA 1623 FUVB G140L 1105 FUVA 1230 FUVA/FUVB 1280 FUVA/FUVB

11 Lifetime Impact From FCAL3 (12806) observations: Target/Segment Percent lifetime used in 1 visit for brightest resel Percent lifetime used in 1 visit for the median resel GD71/FUVA 0.28% 0.005% GD71/FUVB 0.01% % WD /FUVA 0.37% % WD /FUVB 0.37% % Total/ FUVA 0.41% 0.01% Total/FUVB 0.37% 0.04%

12 FUV Detector Dark Monitor P.I. Justin Ely Purpose Description Fraction GO/GTO Programs Supported Observations Perform routine monitoring of FUV XDL detector dark rate. The main purpose is to look for evidence of a change in the dark rate, both to track on-orbit time dependence and to check for a developing detector problem. Monitor the FUV detector dark rate by taking TIME-TAG science exposures with no light on the detector. Five times every week a 22-min exposure is taken with the FUV detector with the shutter closed. The length of the exposures is chosen to make them fit in Earth occultations. All orbits < 1800s. 92% of COS total exposure time. 260 internal orbits Analysis Products 4 FTE weeks Improve dark correction method, ETC updates, bad-pixel table updates as needed, ISR, and a summary in the end of cycle ISR. Accuracy Goals Obtain enough counts to track changes on timescales of ~1-2 months. Scheduling & Special Requirements 5 times every week during Earth occultations. Changes from Cycle 19 Double frequency of observations to a set of 5 every week.

13 Cycle 20 Programs Sensitive to Dark Current Some of the larger combined exposure times according to current Cycle 20 statistics. These are some of the programs that would most benefit from higher frequency for the dark monitor and an improved dark correction scheme. PID Target Grating Cenwave Total Exptime (s) HS G130M SDSS G140L MASS-J G140L SDSS-J G160M DFGRSS394Z150 G130M DFGRSS394Z150 G130M SDSS-J G230L

14 FUV Dark Current The COS FUV dark current is increasing with solar activity.

15 FUV Dark Current The COS FUV dark current in the extraction region decreases as the gain sags. This decrease is around 10% per year, and is different for each detector segment. The dark rate in the background regions does not decrease, so dark is oversubtracted. Increases due to solar activity and decreases due to gain sag evolve on 2-month time scales. Increasing the frequency of the monitors will: Build up enough signal on 2-month timescales to more accurately track and correct the sag. This can be done by implementing either a scaling factor to the dark spectrum or by delivering master-darks from which the dark spectrum can be extracted. Provide much more information on the typical pulse-height values of dark counts across the detector, which will help facilitate better PHA screening of dark counts. Allow us to provide users with master-dark frames that they can use for custom dark corrections.

16 FUV Internal to External Wavelength Scale Monitor P.I. Paule Sonnentrucker Purpose Description Fraction GO/GTO Programs Supported Observations This program monitors the offsets between the wavelength scale set by the internal wavecal versus that defined by absorption lines in external targets. This program monitors the offset between the internal and external wavelength scales with observations every 6 months. This offset is referred to as "DELTA" in the wavelength dispersion reference file and corrects for the shift between the WCA and PSA in TV03 versus the shift between the WCA and PSA in orbit : (WCA-PSA)_TV03 - (WCA - PSA)_orbit. Analysis of TV data indicates that this DELTA (offset) is cenwave and FPPOS independent for a particular grating, but it is grating dependent. To verify and monitor this, this program observes some cenwaves at different FPPOS. All orbits > 1800s. 92% of COS total exposure time. 2 external orbits and all orbits > 1800s. Analysis 3 FTE weeks Products Possible update to wavelength dispersion reference file and a summary in the end of cycle ISR. Accuracy Goals Scheduling & Special Requirements Changes from Cycle 19 G140L 150km/s, pixels G130M 15km/s, pixels G160M 15km/s, pixels ORIENT for some exposures to avoid bright field targets. These observations are taken every 6 months. Only one observation (in April) is required in Cycle 20. AV 75 is used with the 3 FUV gratings in 2 orbits as opposed to using 2 different targets for the M and L gratings for a total of 3 orbits.

17 COS Observations of Geocoronal Lyman α Emission P.I. Sean Lockwood Purpose Description COS observations are in parallel with the STIS MAMA Spectroscopic Sensitivity Monitor observations. COS obtains G130M and G140L spectra of the geocoronal Lyman α emission feature with S/N ratios sufficient to trace the line wings. We have received requests from GOs for high-s/n observations of the geocoronal Lyman α line profile observed with G130M in order to model and subtract the line wings from their spectra. Three programs in Cycle 20 require high S/N Ly α profiles. Observations to date provide insufficient airglow data to construct such profiles. We propose airglow observations totaling 10 ks for the most-used CENWAVE, The data will be archived, but must be reduced by the GOs themselves. Fraction GO/GTO Programs Supported Observations Analysis 9% of COS total exposure time. COS: 4 parallel orbits COS: 0 FTE weeks Products For COS, there are no products. Observers must reduce these data themselves. Accuracy Goals G130M, SNR of 1.5 per pixel at 1213 A. Scheduling & Special Requirements Changes from Cycle 19 COS observations are obtained as parallels with STIS external calibration observations. 1 less parallel orbit requested this cycle.

18 FUV Detector Recovery After Anomalous Shutdown P.I. Tom Wheeler Purpose Description Fraction GO/GTO Programs Supported Observations Analysis This proposal is designed to permit the conservative, safe, and orderly recovery of the COS/FUV detector after an anomalous HV shutdown. Anomalous shutdowns can occur because of bright object violations, which trigger the Count Rate Protection Monitor or the Global Software Monitor. Anomalous shutdowns can also occur because of hardware anomalies or failures. The cause of the shutdown should be thoroughly investigated and understood prior to recovery. Wait intervals are required after each test for data analysis. COS event Flag 3 is used to prevent inadvertent FUV usage. 92% of COS total exposure time. This is a contingency procedure only. This is a contingency proposal only. A maximum of orbits would be needed (based on final HV setting.) 0.4 FTE weeks Products Accuracy Goals Scheduling & Special Requirements Special commanding is required. Changes from Cycle 19 After the anomaly event in April/May 2012, the recovery procedure has been modified.

19 Monitoring of COS ACQ/IMAGE performance P.I.: S. Penton Purpose Description Fraction GO/GTO Programs Supported Observations Analysis Re-measure and monitor the WCA-to-SA offsets for the NUV ACQ/IMAGE modes. [SA = Science Aperture = PSA or BOA] There are four NUV ACQ/IMAGE mechanism combinations: 2 science apertures (SAs: PSA & BOA) x 2 mirror modes (MIRRORA & MIRRORB). The choice of combination is driven by the target brightness. During SMOV, the PSA+MIRRORA offset was determined by an aperture scan; the other offsets were bootstrapped from this offset. A target of appropriate brightness was centered using an NUV ACQ/IMAGE with PSA+MIRRORA, then COS obtained an image using PSA+MIRRORB. This process was repeated on another target for PSA/MIRRORB + BOA/ MIRRORA, and finally a third target was used for BOA/MIRRORA+BOA/MIRRORB. This program repeats this process, and obtains PSA and BOA spectra of the targets to track any changes in the spectroscopic WCA-to-SA offsets. 100% of COS total exposure time. 2 one-orbit visits, with 1 one-orbit visit contingency. The PSA+MIRRORA and PSA+MIRRORB centerings are periodically tested in the SIAF file verifications of HST program If for some reason this program has not been run with the current SIAF file, the contingency visit would be needed to provide the bootstrap back to PSA+MIRRORA. 3 FTE weeks for analysis, creating new offsets, and writing an ISR. Products Accuracy Goals Scheduling & Special Requirements Updated NUV imaging WCA-to-SA offsets, NUV & FUV Spectroscopic WCA-to-SA offsets. Imaging WCA-to-SA offsets are known to better than 0.5 pixels. Spectroscopic WCA-to-SA offsets are known to 0.5 XD pixel. Should be executed annually and after each COS SIAF adjustment. Changes from Cycle 19 New program.

20 NUV Spectroscopic Sensitivity Monitor P.I. Azalee Bostroem Purpose Monitor sensitivity of each NUV grating mode to detect any change due to contamination or other causes. The bare-aluminum gratings on COS are known to degrade with a rate which has been fairly steady since the start of on-orbit operations. Another goal is to track time dependence of sensitivity as a function of wavelength. Description Obtain exposures in all NUV gratings G230L, G185M, G225M, and G285M 3 times a year. We will monitor the following modes: G230L/2635, G230L/2950, G185M/1786, G185M/1921, G225M/2186, G285M/2617, and G285M/ These central wavelengths constitute the reddest and bluest central wavelengths containing only first order light with the exception of G225M remains unused in Cycle 20. NUV usage continues to be minimal and the trends are stable, therefore we will continue to monitor the central wavelength used in the previous cycle and cut down the cadence to 3 times a year. Fraction GO/GTO Programs Supported Observations 8% of COS total exposure time. 6 external orbits Analysis 5 FTE weeks Products ISR, Time-Dependent Sensitivity Reference File and a summary in the end of cycle ISR. dependence to TDS reference file. Add wavelength Accuracy Goals SNR of 30 per resel at the central wavelength except for G285M with SNR 26 per resel Scheduling & Special Requirements Space observations at 4-month intervals Changes from Cycle 19 Reduce monitoring to 3 times per year

21 Stability of the NUV Modes G185M/1921 G225M/2186 G285M/3094 G230L/2635

22 Wavelength Dependence of COS NUV TDS Trends

23 NUV Detector Dark Monitor P.I. Justin Ely Purpose Description Fraction GO/GTO Programs Supported Observations Perform routine monitoring of MAMA detector dark current. The main purpose is to look for evidence of a change in the dark, both to track on-orbit time dependence and to check for a developing detector problem. Monitor the NUV detector dark rate by taking TIME-TAG science exposures without illuminating the detector. Twice every other week a 22-min exposure is taken with the NUV (MAMA) detector with the shutter closed. The length of the exposures is chosen to make them fit in Earth occultations. All orbits < 1800s. 8% of COS total exposure time. 52 internal orbits Analysis 4 FTE weeks Products ETC updates, bad-pixel table updates as needed, ISR, and a summary in the end of cycle ISR. Accuracy Goals 5% accuracy in the dark rate. Scheduling & Special Requirements Two separate exposures during Earth occultations every other week. Changes from Cycle 19 No changes.

24 NUV Dark Current

25 NUV Internal to External Wavelength Scale Monitor P.I. Paule Sonnentrucker Purpose Description Fraction GO/GTO Programs Supported Observations This program monitors the offsets between the wavelength scale set by the internal wavecal versus that defined by absorption lines in external targets. This program monitors the offset between the internal and external wavelength scales with observations every 6 months. This offset is referred to as DELTA in the wavelength dispersion reference file and corrects for the shift between the WCA and PSA in TV03 versus the shift between the WCA and PSA in orbit: (WCA-PSA)_TV03 - (WCA-PSA)_orbit. Analysis of TV data indicates that this DELTA is cenwave and FP-POS independent for a particular grating, but it is grating and stripe dependent. To verify and monitor this, this program observes some cenwaves at different FP-POS. All orbits > 1800s. 8% of COS total exposure time. 1 external orbits and all orbits > 1800s. Analysis 3 FTE weeks Products Possible update wavelength dispersion reference file and a summary in the end of cycle ISR. Accuracy Goals Scheduling & Special Requirements G230L 175km/s, pixels G185M 15km/s, pixels G225M 15km/s, pixels G285M 15km/s, pixels These observations are taken every 6 months. Only 1 visit is required in Cycle 20. Changes from Cycle 19 Decrease frequency from 3x per year to 2x per year.

26 Purpose Description Fraction GO/GTO Programs Supported Observations NUV MAMA Fold Distribution P.I. Tom Wheeler The performance of the MAMA microchannel plates can be monitored using a MAMA fold analysis procedure that provides a measurement of the distribution of charge cloud sizes incident upon the anode giving some measure of change in the pulse-height distribution of the MCP, and therefore, MCP gain. While illuminating the detector with a flat field, the valid event (VE) rate counter is monitored while various combinations of row and column folds are selected. The process is implemented using a time-tag exposure and special commanding. This proposal executes the same steps as Cycle 18 proposal and is described in COS TIR All orbits > 1800s. 8% of COS total exposure time. 1 internal orbit and all orbits > 1800s. Analysis 0.1 FTE weeks Products The fold analysis findings are reported to the COS Science Team and V. Argabright of Ball Aerospace after completion of the analysis, typically one-two weeks after execution of the test. A summary in the end of cycle ISR. Accuracy Goals Position of the peak in the fold distribution can be measured to about 5% accuracy from this procedure. Scheduling & Special Requirements Special commanding is required. Changes from Cycle 19 No changes.

27 NUV Detector Recovery After Anomalous Shutdown P.I. Tom Wheeler Purpose Description Fraction GO/GTO Programs Supported Observations This proposal is designed to permit the safe and orderly recovery of the COS/NUV MAMA detector after an anomalous shutdown. This is accomplished by using slower-than-normal MCP and PC high-voltage ramp-ups and diagnostics. Anomalous shutdowns can occur because of bright object violations, which trigger the Global Hardware Monitor or the Global Software Monitor. Anomalous shutdowns can also occur because of MAMA hardware anomalies or failures. The cause of the shutdown should be thoroughly investigated and understood prior to recovery. Twentyfour hour wait intervals are required after each test for MCP gas desorption and data analysis. COS event flag 2 is used to prevent inadvertent MAMA usage. 8% of COS total exposure time. This is a contingency procedure only. This is a contingency proposal only. 4 orbits would be needed. Analysis 0.4 FTE weeks Products Accuracy Goals Scheduling & Special Requirements Special commanding is required. Changes from Cycle 19 No changes. The procedure executed as planned in 4 orbits for the recovery in May 2012.

28 Backup Slides

29 Detector Illumination for New vs. Old TD Targets

30 Shallow TDS Trend:

31 Steep TDS Trend:

32 Monte Carlo Tests of Cadence and Required S/N Cadence (days) Slope Times= Slope Mag= Ftol= (200,60) (200,45) (200,30) (200,60) (200,45) (200,30) (200,60) (200,45) (200,30) Local Max 6.0 Final Max Deviation (400,60) (400,45) (400,30) (400,60) (400,45) (400,30) S/N (400,60) (400,45) (400,30) , 3 Allowed slopes varied from 5 60% per year with break points varying from days (using uniform distributions). Our proposed monitoring program gives the following S/N ratios used for our analysis: G130M: 137 in 5 Å bins G160M: 124 in 5 Å bins G140L: 96 in 20 Å bins Achieving a maximum deviation of 2% in our final calibration analysis requires a 30-day cadence at S/N = 100. (The G140L requirement is <3%.)

33 Flux Error Histograms We used the current set of TDS data to compare different observation strategies. Shorter observation times were simulated by truncating the corrtag files appropriately. 2 orbits per month Two orbits per month requires shortening observation times by 60%. The residual relative flux errors in this scenario triple the 3-orbit scenario. 3 orbits per month Three orbits per month gives residual relative flux errors that meet our requirements. (<2% for 95% of all data points.)