Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.stsci.edu/hst/cos/documents/isrs/ISR2010_15.pdf
Дата изменения: Unknown
Дата индексирования: Tue Feb 5 14:39:50 2013
Кодировка:
Поисковые слова: столовая гора

Operated for NASA by AURA

SP CE A TELESCOPE SCIENCE INSTITUTE

Instrument Science Report COS 2010-15 (v1)

Early Results from the COS Spectroscopic Sensitivity Monitoring Programs
Rachel A. Osten1 , Parviz Ghavamian1, Sami-Matias Niemi1 , Derck Massa1 , Steve Penton2 , Alessandra Aloisi1, Tony Keyes1, Steve Osterman2 , Charles Proffitt3
Space Telescope Science Institute, Baltimore, MD 2 CASA, University of Colorado, Boulder, CO Space Telescope Science Institute/Computer Sciences Corporation, Baltimore, MD August 2, 2010
1

3

ABSTRACT We report initial results from the Cosmic Origins Spectrograph spectroscopic sensitivity monitoring calibration programs. For the NUV channel we use both internal and external measurements to determine the time dependence of any sensitivity changes, while for the FUV channel we use only external targets. Time-dependent sensitivity (TDS) trends for the NUV channel appear to be wavelength-independent but gratingdependent. Internal Grating Efficiency Tests (GETs) performed in the NUV show that the bare-Al gratings G225M and G285M are exhibiting declines in efficiency, while the Al+MgF2 -coated G185M and G230L gratings appear to be quite stable. The changes in efficiencies for the bare-Al gratings were first noted on the ground and continue in orbit. GET data obtained on the ground indicate efficiency declines relative to the G230L grating of -1.8 ± 0.9%/year for G225M and -5.4 ± 0.6%/year for G285M. The onorbit NUV TDS data obtained using external targets confirms the trends seen with the internal wavecal exposures. In the first 9 months of Cycle 17, the weighted means of the TDS slopes using external targets for G230L and G185M are -1.1 ± 0.4%/year and -0.8 ± 0.4%/year, respectively, while the G225M and G285M gratings on the NUV channel have exhibited on-orbit declines of -3.3 ± 0.3%/year and -10.8 ± 0.2%/year, respectively. We have considered the effect which potentially inaccurate target acquisition had on early FUV monitoring observations. The absolute values of the corrections are small (generally <1%) and the uncertainty in their values is larger than the value of
Operated by the Association of Universities for Research in Astronomy, Inc., for the National Aeronautics and Space Administration.

the correction itself, and so we do not apply them. All gratings on the FUV channel are experiencing wavelength-dependent sensitivity degradation, which is worse at longer wavelengths. The behavior is slightly different for each detector segment, possibly reflecting differences in the photocathodes. The G130M grating is exhibiting an average decline in sensitivity of -4.6 ± 0.3%/year for segment A, and -6.3 ± 0.2%/year for segment B, while the G160M grating is on average declining in sensitivity at a rate of -11.2 ± 0.8%/year for segment A, and -12.0 ± 0.4%/year for segment B. Sensitivity degradation of the G140L grating ranges from -4.9 ± 2.3%/year in the interval 955-970 ° ° A, to -10.9 ± 3.5%/year in the interval 1700-1800 A. The implications of the observed sensitivity changes for Cycle 17 and 18 COS users are presented. A discussion of the cause of the sensitivity degradations is deferred to a later document, as analysis and investigation is still ongoing.

Contents
· Introduction (page 2) · Sensitivity Monitoring of the NUV Channel (page 3) · NUV Grating Efficiency Tests (page 4) · NUV External Target TDS (page 5) · Sensitivity Monitoring of the FUV Channel (page 11) · Summary and Implications for Cycle 18 (page 16) · Acknowledgments (page 24) · Change History (page 26) · References (page 26) · Appendix (page 27)

1. Introduction
The on-orbit absolute flux calibration of the Cosmic Origins Spectrograph (COS) was performed in the summer of 2009 during the Servicing Mission Observatory Verification (SMOV). Results for the NUV and FUV channels are reported in Massa et al. (2010a) and Massa et al. (2010b), respectively. Following these determinations, two calibration programs in Cycle 17 (11896 and 11897) are monitoring the spectroscopic sensitivity of each detector/grating combination to characterize subsequent time-dependent changes in sensitivity due to e.g. contaminants or detector/grating performance issues. On the Instrument Science Report COS 2010-15 (v1) Page 2

NUV channel, measurements of the grating efficiency with the internal wavelength calibration lamp have been computed since 2003 and continue on-orbit. External spectrophotometric white dwarf standard stars are used for the external on-orbit monitoring of both NUV and FUV channels. This Instrument Science Report describes initial results from the monitoring of external targets and supporting internal observations. Spectral datasets of external targets from HST calibration programs 11896 and 11897 were processed with the latest version of calcos (here 2.12). The one-dimensional extracted spectra of net count rate versus wavelength for each epoch are used, and ratioed to the initial spectrum of the same grating/central wavelength. The ratioed spectra are then averaged in wavelength to provide a TDS data point; wavelength ranges for averaging were chosen to avoid the edges of the spectrum. These computations are done separately for each spectral stripe (program 11896) or segment (program 11897). For the FUV G140L spectra TDS values were also computed for segment A in smaller ° 100 A bins. Any changes of these mean values with time constitute the time-dependent sensitivity (TDS) changes of a given mode, here assumed to be piece-wise continuous linear functions of time. The slopes of these fits give the TDS trend in units of % per year. Since the TDS values are averages of ratioed spectra, we determined the error in that value due to random variations by computing the standard error of the mean of all the bins which are used in the ratio. This is performed separately for each spectral stripe (program 11896) or segment (program 11897) of the data. Due to the large number of pixels used in the average, this error due to statistical fluctuations is generally a very small number, and the errors are dominated by systematics. We estimated the magnitude of the systematic error by computing the error in the data needed to achieve a reduced chi-squared near 1 for the linear fit. Note that this method assumes that a linear fit is the correct functional form. This method of estimating the systematic errors generally gives consistent results for different central wavelengths of the same grating. Both values of error are plotted for TDS data. Any trends seen in analysis of the pipeline processed extracted spectra were confirmed by manually extracting spectra from two-dimensional spectral images and recalculating the TDS trends, in order to rule out any systematics in the pipeline. Note that the white dwarfs used for spectroscopic sensitivity monitoring are spectrophotometric standards, chosen primarily for their relative brightness at these wavelengths. To tie the monitoring targets to the standards used for absolute flux calibrations in the SMOV programs (described in Massa et al. 2010a and 2010b), we performed additional observations of these spectrophotometric standards within 8 hours of one of our regular monitoring observations in overlapping configurations. The scope of this ISR is limited to the regular monitoring, as not all modes have had these tied observations.

2. Sensitivity Monitoring of the NUV Channel
The sensitivity monitoring of the NUV channel is performed using both internal and external measures. Section 2.1 describes internal monitoring of the sensitivity performed on the ground and in orbit, while section 2.2 describes on-orbit sensitivity monitoring Instrument Science Report COS 2010-15 (v1) Page 3

Table 1. NUV On-orbit Grating Efficiency Test Exposures Grating G1 8 5 M G1 8 5 M G2 2 5 M G2 2 5 M G2 2 5 M G2 2 5 M G2 8 5 M G2 8 5 M G2 3 0 L G2 3 0 L Central Wavelength Exposure Time ° (A) (s) 1 2 2 2 2 2 2 2 2 3 9 0 1 2 3 4 6 6 6 3 8 1 8 1 9 1 1 3 3 6 6 0 6 7 0 0 7 7 5 0 210 195 300 225 75 300 90 300 30 105

using external targets. 2.1 NUV Grating Efficiency Tests The Grating Efficiency Test (GET) obtains internal lamp exposures of NUV gratings to gauge the relative decline in efficiency of the gratings by comparing the count rates of lines in common between the gratings being compared. The ground-based grating efficiency test was designed in 2003 after thermal vacuum testing suggested a possible decline in efficiency in the two NUV bare Aluminum gratings G225M and G285M (Wilkinson 2003). The G225M and G285M gratings are bare Al, and can react with oxygen and develop a thin oxide coating. The GET starts with 20 minutes of waiting time to ensure that the lamp has cooled from any prior use, and is performed in two sections, with 2025 minutes of lamp cooling time between the two sections. Each section contains 10 exposures of NUV gratings designed to obtain the same S/N of various strong emission lines throughout the NUV band. GETs were performed on the ground every six months from 20032009 to characterize the stability of the NUV gratings, and were performed in a nitrogen purge environment. To date two additional GETs have been performed on orbit, one as part of program 11496 and another as part of program 12052. All exposures use Pt-Ne lamp 2. In late 2006, the exposure times of the GETs were reduced to limit the impact of the tests on the lifetime of the Pt-Ne lamp 2. The details of the reduction of GET exposure times, and an analysis of the early GET results are reported in Penton (2006). The sequence and duration of exposures during on-orbit and ground-based GETs taken after Jan. 1, 2007 are the same, and are listed in Table 1.

Instrument Science Report COS 2010-15 (v1) Page 4

A number of efficiency ratios were defined using the range of grating/central wavelengths observed, for which an appropriate emission line was in common. Figure 1 shows the GET results for these NUV grating pairs, including both ground-based and on-orbit data. The data for a given ratio are normalized to the first data point. The G185M and G230L gratings are coated with Al+MgF2 , while the G225M and G285M gratings are bare Al. The fits exclude the first data point in 2003, as it was derived from a collection of exposures which did not have the same setup as subsequent GETs with regards to lamp history. In the panels exhibiting the G285M/G230L ratios, there are two obviously discrepant data points which are outliers to the linear trend: these data were obtained in late 2008/early 2009 while COS was at the Kennedy Space Center. These data may have been affected by the increase in humidity at Cape Canaveral. This reveals that conditions during these two GETs may be different than those under which the other data points were obtained, and we subsequently excluded GET data taken on these dates from any linear fits to the normalized efficiency ratios. Table 2 lists the slopes derived for the GET data. The discrepant points have been excluded from the fits, as described above. In addition, the slopes have been derived for ground and on-orbit data separately. Because there are only two on-orbit data points, we only quote this slope to one significant digit. An additional visit of program 12052 in fall 2010 will provide a third on-orbit measurement. The ratio of the two Al+MgF2 -coated gratings appear to be stable in time, while the bare Al gratings relative to the Al+MgF2 gratings exhibit an efficiency decline. Figure 2 summarizes the grating efficiency ratio trends as a function of wavelength for ground data excluding the points described above. The loss of relative efficiency for a given grating appears to be constant across wavelength for that grating, and this loss is a different value for the G225M and G285M grating. The change in grating efficiency observed on the ground was thought to be due to continued growth of an oxide layer, and initial expectations were that this would cease once in an on-orbit environment. There are too few data points obtained on-orbit to permit a meaningful comparison of any differences between ground data and on-orbit data. The weighted mean of the trends for G225M relative to G230L indicate a decline of -1.8±0.6%/year for pre-launch data, while the weighted mean of the trends for G285M relative to G230L indicate a decline of -5.4±0.6%/year. The weighted mean of the G185M grating relative to G225M is 2.2±0.8%/year, while for G225M relative to G285M it is 3.2±1.4%/year.

2.2 NUV External Target TDS The spectrophotometric standard star used for monitoring of NUV low-resolution spectroscopic modes is the white dwarf WD1057+719, while for the medium-resolution spectroscopic modes the standard star is G191B2B. The spectral energy distributions of WD1057+719 and G191B2B are shown in Figures 3 and 4, respectively. It was discovered during SMOV that there was a vignetting of counts from external targets in the Instrument Science Report COS 2010-15 (v1) Page 5

Figure 1. Results from the NUV GET, displaying grating efficiency versus time for grating pairs. The title to each sub-plot lists the gratings and central wavelengths used in each ratio, at the stated wavelength. Data points are plotted with blue circles and connected by a blue line. The magenta line is the linear fit to all data, excluding outliers as described in section 2.1. The red line is the fit to only ground-based data, again excluding the outliers, while the green line is the slope computed for on-orbit data. The data have been normalized to the first data point. The values of the slopes are listed in Table 2.

Instrument Science Report COS 2010-15 (v1) Page 6

Figure 2. Summary of NUV GET results from line ratios in different gratings as a function of wavelength as measured on the ground. The three discrepant ground-based data points described in the text have been excluded from the fits which are summarized here. The legend indicates the gratings used in the ratios. For ratios at different wavelengths made with the same two gratings, the results are consistent, indicating that any observed change in the relative grating efficiency is wavelength independent but grating dependent.

Instrument Science Report COS 2010-15 (v1) Page 7

Table 2. Wavelength ° (A) 2 2 2 2 2 2 2 2 2 2 2 2
a

NUV Grating Efficiency Test Results All Dataa Ground Dataa On-orbit Data (%/year) (%/year) (%/year) 1.9±0.6 0.1±0.5 -2.5±0.8 2.6±0.5 -2.5±1.2 -7.0±0.8 5.0±1.5 -2.5±1.2 -7.0±1.0 4.9±1.2 -8.2±1.9 -6.1±0.7 2.2±1.1 0.6±0.6 -1.8±1.4 2.3±1.2 -2.2±1.9 -5.7±1.1 3.0±2.0 -1.5±1.6 -5.4±1.2 3.4±2.0 -5.3±1.2 -5.4±1.0 1 -2 -6 5 -2 -9 10 -2 -5 5 -40 -4

Grating Pair grating1 (cenwave1 )/ grating2 (cenwave2 ) G1 8 5 M G1 8 5 M G2 2 5 M G1 8 5 M G2 2 5 M G2 8 5 M G2 2 5 M G2 2 5 M G2 8 5 M G2 2 5 M G2 8 5 M G2 8 5 M (1986) (2010) (2217) (2010) (2390) (2617) (2390) (2410) (2637) (2410) (2617) (2637) / / / / / / / / / / / / G2 2 5 M G2 3 0 L G2 3 0 L G2 2 5 M G2 3 0 L G2 3 0 L G2 8 5 M G2 3 0 L G2 3 0 L G2 8 5 M G2 3 0 L G2 3 0 L (2186) (3360) (3360) (2217) (2635) (2635) (2617) (2635) (2635) (2637) (2635) (2635)

1 1 1 1 4 4 4 5 5 5 6 7

0 3 3 3 9 9 9 1 1 1 2 5

0 0 0 0 0 0 0 0 0 0 0 0

Outliers have been excluded as described in Section 2.1.

left hand portion of NUV spectra, for x pixels less than about 200 (Ake et al. 2010). The effect appears to be caused by the optical beam partially missing the NCM3 camera mirror, and is grating dependent. Possible offsets of the target with respect to aperture center may cause slight changes in the amount of this vignetting, so we excluded the wavelengths covered by these pixels from TDS calculations. The observations obtained through the end of April in each grating/central wavelength configuration are described in the Appendix. The observing strategy for the NUV TDS calibration program (11896) is to monitor two central wavelength settings of the low-resolution grating (G230L) approximately monthly, in order to discover any wavelength-dependent throughput changes. The longest G230L central wavelength setting is observed roughly quarterly to reach the longest wavelengths covered by the COS NUV channel, although with poorer S/N due to the target's declining flux at these wavelengths. One central wavelength of the G185M grating is also observed quarterly. These two gratings are Al coated with MgF2 , and are expected to have stable behavior in orbit, based on ground-based GET results (§2.1). Because of the relative efficiency declines seen in the bare Al gratings on the ground (§2.1), and concern about capturing any on-orbit changes, the NUV sensitivity monitoring includes monthly observations of two central wavelengths of each of the G225M and G285M. Quarterly observations of these gratings span the entire wavelength range available for that grating. The NUV channel obtains data in three spectral stripes, and TDS trends are deInstrument Science Report COS 2010-15 (v1) Page 8

Figure 3. Net spectra of WD1057+719, the spectrophotometric standard star used for sensitivity monitoring of the G230L and G160M modes. Initial spectra from each grating and central wavelength used in monitoring are shown to illustrate the relative count rates and wavelength coverage. For the NUV spectra, each setting has stripes A, B, and C plotted, while for FUV spectra, segments A and B are shown. The G160M data are discussed in section 3.

Instrument Science Report COS 2010-15 (v1) Page 9

Figure 4. Net spectra of G191B2B, the spectrophotometric standard star used for sensitivity monitoring of the medium-resolution NUV modes. Initial spectra from each grating and central wavelength used in monitoring are shown to illustrate the relative count rates and wavelength coverage. Each setting has stripes A, B, and C plotted.

Instrument Science Report COS 2010-15 (v1) Page 10

termined for each stripe separately. Figures 5 and 6 display the TDS trends for a representative central wavelength setting of each grating. The figures for the remaining combinations of grating and central wavelength are displayed in the Appendix. The TDS trends for the G230L and G185M gratings indicate relatively little degradation of sensitivity. Note that stripe C of the G230L grating records second-order light, which appears with 2030 times fewer counts compared with the same wavelength region in first order. Stripe B of G230L/3360 also includes contributions from second order light, as discussed in Massa et al. (2010a). The behavior of second order light also shows little degradation, although the error bars are larger than for the first order spectra. The TDS results for the NUV channel are listed in Table 3. The TDS results for the bare Al gratings exhibit an obvious downward trend. The weighted mean of the TDS data for the first order spectra measured with the G230L grating is -1.1 ± 0.4%/year, while for the G185M grating it is -0.8 ± 0.4%/year. The weighted mean of the TDS data for the G225M grating is -3.3 ± 0.3%/year, while for the G285M grating it is -10.8 ± 0.2%/year. The implications of these sensitivity degradations are discussed in Section 4.

3. Sensitivity Monitoring of the FUV Channel
The spectrophotometric standard star for FUV G140L and G130M modes is WD0947+857, while for the G160M mode it is WD1057+719. Figure 7 displays the spectral energy distribution of WD0947+857, while Figure 3 shows that of WD1057+719 (it is also used in NUV sensitivity monitoring). The individual observations obtained to date in each grating/central wavelength configuration are described in the Appendix. Note that observations from programs other than the FUV Spectroscopic Sensitivity Monitoring Program (11897) have been used where the target and instrumental configuration is identical to that used in Program 11897. The observing strategy for Program 11897 initially was to monitor the low-resolution spectroscopic modes monthly, in order to discover issues related to sensitivity degradation over the widest wavelength range. Observations with the medium-resolution spectroscopic modes were performed roughly quarterly, to confirm the behavior seen in overlapping wavelength ranges of the G140L modes. However, by January 2010 it was apparent that the G140L modes were showing an apparent decline in sensitivity which was steeper at longer wavelengths (see below). The limited observations of the mediumresolution modes appeared consistent with the trends seen in the low-resolution data. In order to confirm these results for the medium-resolution modes, as part of the Cycle 17 supplemental calibration program additional visits of the G130M and G160M sensitivity monitoring were added so that the monitoring frequency of the medium-resolution modes is now also monthly. A subset of central wavelengths of the G160M grating, spanning the shortest and longest central wavelengths used in the quarterly monitoring visits, are observed in these monthly visits. The switch to monthly visits of the G130M and G160M gratings began with monitoring observations in late March 2010. Several months into Cycle 17, as results from the target acquisition evaluation proInstrument Science Report COS 2010-15 (v1) Page 11

° Figure 5. (left) TDS results for the 2635 A central wavelength setting of G230L. Spectral stripes A through C are plotted from top to bottom, respectively. Note that for G230L spectra, stripe C records second order light. The wavelength ranges used in averaging each stripe are listed in the title to each sub-plot, while the slopes and errors from a linear fit to the data for each stripe are listed in the sub-plot. The error bars resulting from statistical fluctuations in the average of the ratioed spectra are illustrated with thick lines, "top-hat" style, while the thin vertical line is the error in the data needed ° to achieve 2 1 for the linear fit. (right) TDS results for G185M 1921 A central wavelength. Format is the same as in the plots to the left.

Instrument Science Report COS 2010-15 (v1) Page 12

Table 3. Grating G G G G G G G G G G G G G G G G G G G G G G G G G G G G G G 2 2 2 2 2 2 2 2 2 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 8 8 8 2 2 2 2 2 2 2 2 2 8 8 8 8 8 8 8 8 8 0 0 0 0 0 0 0 0 0 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 L L L L L L L L L M M M M M M M M M M M M M M M M M M M M M Channel N N N N N N N N N N N N N N N N N N N N N N N N N N N N N N U U U U U U U U U U U U U U U U U U U U U U U U U U U U U U V V V V V V V V V V V V V V V V V V V V V V V V V V V V V V

NUV TDS Summary Wavelength Range ° A 1 2 1 1 2 1 2 3 2 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 6 5 8 7 8 9 1 2 2 8 9 0 0 1 2 3 4 4 1 2 3 4 6 7 6 7 8 9 0 1 0 3 3 5 4 9 8 7 0 0 1 1 7 7 7 0 0 9 9 9 9 8 0 2 1 2 4 7 8 9 0 7 6 3 7 0 4 0 3 7 0 5 7 6 8 7 2 8 7 5 3 6 5 6 0 8 6 2 2 2 1 2 1 2 3 2 2 3 2 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 7 8 9 0 1 1 4 5 3 8 9 0 1 2 3 3 4 5 2 3 4 5 6 7 6 7 8 0 1 2 5 4 8 6 5 4 9 8 5 3 3 4 0 0 0 3 2 2 2 2 1 2 3 5 4 5 7 0 1 2 0 0 0 0 0 0 0 0 0 5 5 0 5 0 0 0 6 0 5 0 5 0 5 5 0 5 5 0 0 0 Slope %/year 1.9±4.8 -1.2±0.8 -1.1±2.7 0.3±1.3 -1.9±1.3 -4.7±6.4 -1.1±0.5 -4.2±8.4 -7.2±8.4 -1.1±0.7 -0.8±0.7 -0.6±0.7 -3.4±2.9 -3.4±1.4 -3.7±1.4 -2.6±1.4 -2.9±1.4 -2.4±0.7 -3.3±0.5 -4.1±0.5 -2.8±0.7 -12.2±1.4 -11.0±1.4 -11.8±0.9 -11.8±1.4 -11.6±1.4 -12.0±0.9 -10.6±0.3 -10.8±0.3 -9.1±1.1

Central Wavelength ° A 2 2 2 2 2 2 3 3 3 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 6 6 6 9 9 9 3 3 3 9 9 9 1 1 1 4 4 4 3 3 3 6 6 6 7 7 7 0 0 0 3 3 3 5 5 5 6 6 6 2 2 2 8 8 8 1 1 1 0 0 0 1 1 1 3 3 3 9 9 9 5 5 5 0 0 0 0 0 0 1 1 1 6 6 6 0 0 0 6 6 6 7 7 7 9 9 9 4 4 4

Instrument Science Report COS 2010-15 (v1) Page 13

° Figure 6. (left) TDS results for G225M 2186 A central wavelength. (right) TDS results ° for G285M 2617 A central wavelength. Format is the same as in Figure 5. gram came in (Penton et al. 2010), there was a concern that the target acquisition strategy being used for the G140L and G130M observations, a dispersed light G140L/1105 peakup, was not placing the target in the most optimal location, i.e., at aperture center. It was determined later that the target acquisition strategy for the G160M monitoring observations, a dispersed light G160M/1577 peakup, was also placing the spectra in an off-nominal location. The most robust target acquisition strategy is to use an NUV imaging target acquisition. Monitoring visits with the G140L and G130M gratings taken after January 2010 use an NUV imaging target acquisition strategy, while monitoring visits with the G160M grating taken after February 2010 use this updated target acquisition strategy. The offset between the wavecal aperture (WCA) and primary science aperture (PSA) is a constant which is related to the difference in cross-dispersion location of the wavecal spectrum and external target spectrum, so any systematic shift of the external target spectrum in cross-dispersion coordinates relative to the WCA implies a shift of the target within the PSA. By looking at the relative location of the spectra in cross dispersion coordinates we confirmed that there is a systematic offset of spectra taken with the two different target acquisition strategies; this amounts to roughly 4 pixels in the cross-dispersion direction. The pipeline program calcos v2.12 uses spectral extraction regions which are based on the default value of the WCA-PSA offset in cross-dispersion location, and does not center the spectrum within the extraction region. We confirmed that the shift of the spectrum by these amounts ( 4 pixels) does not significantly afInstrument Science Report COS 2010-15 (v1) Page 14

Figure 7. Net spectra of WD0947+857, the spectrophotometric standard star used for TDS monitoring of the G140L and G130M modes. Initial spectra from each grating and central wavelength used in TDS monitoring are shown to illustrate the relative count rates and wavelength coverage. Data for segments A and B are shown, except for the case of G140L/1105 where only data from segment A are collected.

Instrument Science Report COS 2010-15 (v1) Page 15

fect the extracted net count rates using the default extraction regions, as the extraction regions are fairly wide; ignoring the offset of the spectrum in cross-dispersion location results in a difference of <0.1% in the net count rates extracted compared to using a specialized extraction region centered on the spectrum. These displacements relative to the WCA can lead to possible throughput losses as discussed in Ghavamian et al. (2010). We examined the effect that miscentering in the aperture has on the time-dependent sensitivity through a possible change in throughput. Ghavamian et al. (2010) determined the change in throughput as a function of spectrum cross-dispersion location relative to aperture center for a limited set of gratings/central wavelengths. The results are based on sparse emission lines, which may lead to flat field and repeatability differences of as much as 3%. The shape of the throughput curve may not apply to a bright continuum source. For each spectrum in the grating/central wavelength combination for which these throughput measurements were made (G140L/1230, G130M/1309, G160M/1589), we determined the location of each spectrum in crossdispersion direction, and compared the throughputs made with dispersed light and NUV imaging acquisitions. Figures 8 through 10 display the throughput values at the crossdispersion locations of spectra for these gratings and central wavelengths. Data taken with both the old and new target acquisitions fall in the flat part of the curves, and the differences in throughput at these values are generally less than 1%, which is smaller than the uncertainties in the throughput measurements themselves. As the uncertainties are larger than the corrections, and they could lead to spurious results in attempting to correct for any throughput losses due to spectrum misalignment in the aperture arising from the target acquisition strategy, we do not attempt to correct for this effect. Figures 11 through 13 show the TDS data for a representative central wavelength of each FUV grating; data for additional central wavelengths in the monitoring program can be found in the Appendix. In addition to the averages done across each detector segment, for the G140L spec° tra we also computed averages in 100 A bins, to explore any wavelength dependence to the TDS results. Figure 11 displays the TDS results for G140L/1230 in segment A ° in 100 A bins. The FUV TDS results are listed in Table 4. There does appear to be a wavelength dependence to the sensitivity degradation it has the sense that longer wavelengths exhibit larger declines in sensitivity. For G140L data we computed the ° weighted mean of the TDS results in each 100 A bin on segment A. These values range ° from -7.2 ± 1.5%/year for 1300 (A) 1400 to -10.9 ± 3.5%/year for 1700 ° ) 1800. The weighted mean of the TDS points for the G130M grating (both seg (A ° ments) is -5.6 ± 0.2%/year, while for segment A (12901460 A) it is -4.6 ± 0.3%/year ° ) it is -6.3 ± 0.2%/year. The weighted mean of the and for segment B (11401310 A TDS points for the G160M grating (both segments) is -11.8 ± 0.3%/year, while for ° ° segment A (15801790 A) it is -11.2 ± 0.8%/year, and for segment B (13901600 A) it is -12.0 ± 0.4%/year.

Instrument Science Report COS 2010-15 (v1) Page 16

Figure 8. Throughput as a function of cross-dispersion location, for G140L/1230 spectra in segment A, as measured by Ghavamian et al. (2010). The sensitivity of G140L/1230 in segment B is already low, so corresponding curves were not computed. Asterisks show throughput as a function of cross-dispersion determined in SMOV program 11490. Error bars computed from adding in quadrature uncertainties due to Poisson statistics and a systematic error of 3% due to repeatability for these measurements are shown. Thick solid line is an interpolation of these data using quadratic spline fitting. Colored vertical lines indicate the location in fully corrected cross-dispersion coordinates of different TDS spectra. Spectra taken early in Cycle 17 using a dispersed light target acquisition systematically fall near pixel 498 in cross-dispersion location. After switching to an NUV imaging target acquisition algorithm, the spectra have moved to a location near y pixel 494. Any difference in throughput is very small; see discussion in section 3.

Instrument Science Report COS 2010-15 (v1) Page 17

Figure 9. Same as Figure 8, but for the G130M/1309 setting. Calculations for both segment A (top) and segment B (bottom) are shown. The observations made with the dispersed light target acquisition place the spectra near pixel 488 in cross-dispersion in segment A (546 in segment B), while spectra obtained with an NUV imaging target acquisition place the spectra near 484 in segment A (542 in segment B). The odd shape of the G130M curves may result from the dispersed light target acquisition used to acquire the data, as well as the spectral energy distribution of the sparse emission line source used to make the throughput curves.

Instrument Science Report COS 2010-15 (v1) Page 18

Figure 10. segment A dispersed l in segment acquisition

Same as Figure 8, but for the G160M/1589 setting. Calculations for both (top) and segment B (bottom) are shown. The observations made with the ight target acquisition place the spectra near pixel 484 in cross-dispersion A (541 in segment B), while spectra obtained with an NUV imaging target place the spectra near 481 in segment A (536 in segment B).

Instrument Science Report COS 2010-15 (v1) Page 19

Figure 11. TDS data for G140L/1230. The title to each panel lists the wavelength ranges over which the average ratio is computed. The first five panels are from data obtained ° on segment A, shown averaged in 100 A bins, while the last panel is from data falling on segment B. The error bars are as in Figure 5.

Instrument Science Report COS 2010-15 (v1) Page 20

Figure 12. TDS data for G130M/1309. Data from segment A are shown in the top panel, while data from segment B are shown in the bottom panel. Symbols are as in Figure 5.

Instrument Science Report COS 2010-15 (v1) Page 21

Figure 13. TDS data for G160M/1589. Symbols and lines are as described in the caption to Figure 12.

Instrument Science Report COS 2010-15 (v1) Page 22

Table 4. FUV TDS Summary Grating G G G G G G G G G G G G G G G G G G G G G G G G G G G G G 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 4 4 4 4 4 4 4 4 4 4 4 4 4 3 3 3 3 3 3 6 6 6 6 6 6 6 6 6 6 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 L L L L L L L L L L L L L M M M M M M M M M M M M M M M M Channel FU FU FU FU FU FU FU FU FU FU FU FU FU FU FU FU FU FU FU FU FU FU FU FU FU FU FU FU FU VB VA VA VA VA VA VA VA VA VA VA VA VA VA VB VA VB VA VB VA VB VA VB VA VB VA VB VA VB Central Wavelength ° A 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1 1 1 1 1 1 2 2 2 2 2 2 3 3 2 2 3 3 5 5 5 5 6 6 6 6 6 6 3 0 0 0 0 0 0 3 3 3 3 3 3 0 0 9 9 2 2 7 7 8 8 0 0 1 1 2 2 0 5 5 5 5 5 5 0 0 0 0 0 0 9 9 1 1 7 7 7 7 9 9 0 0 1 1 3 3 Wavelength Range ° A 955970 1250180 1300140 1400150 1500160 1600170 1700180 1260180 1300140 1400150 1500160 1600170 1700180 1310145 1220129 1290142 1140120 1330146 1230131 1580174 1390155 1590175 1400155 1610177 1410158 1620178 1421159 1630179 1440160 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 Slope %/year -4.9±2.3 -9.7±2.2 -7.3±2.2 -8.9±1.3 -11.3±2.9 -9.2±2.4 -11.7±5.1 -8.3±2.3 -7.1±2.0 -7.3±1.8 -8.9±2.7 -9.3±3.7 -10.1±4.9 -4.8±0.6 -6.5±0.3 -3.5±0.6 -5.6±1.0 -5.4±0.6 -5.6±0.6 -10.5±1.8 -11.4±0.7 -11.4±2.2 -12.1±1.1 -11.3±1.5 -12.2±0.5 -11.7±2.2 -12.4±1.1 -11.3±1.6 -11.8±0.7

Instrument Science Report COS 2010-15 (v1) Page 23

4. Summary and Implications for Cycle 18
Tables 3 and 4 list the TDS slopes derived from all the gratings and central wavelengths monitored as part of programs 11896 and 11897, respectively, and Figure 14 displays the results as a function of wavelength. Several conclusions can be drawn from this. On the NUV channel, any observed sensitivity decline appears to be wavelength-independent but grating-specific, primarily affecting the bare Aluminum gratings, and with different amounts of degradation. This is consistent with the results obtained from the NUV GET described in section 2.1. On the FUV channel, the observed sensitivity degradation as shown in Figure 14 is larger at longer wavelengths. While the behaviors seen for each grating are broadly consistent with this observed wavelength trend, the magnitude is slightly different for each grating and detector segment. This is most apparent in Figure 14 in two regions: from ° 13001500 A the weighted mean of G130M segment A TDS data is -4.6 ± 0.4%/year ° compared to a value of -8.0 ± 0.8%year for G140L segment A, and from 14001600 A the weighted mean of G160M segment B TDS data is -12.0 ± 0.3%/year, compared to a value of -8.7 ± 0.9%/year for G140L segment A. There is a difference in the weighted mean TDS values for segments A and B, respectively, for the G130M and G160M gratings, as described in section 3. The FUV spectra are not currently flat-fielded, and observations made on the same detector segment but with different gratings fall on different parts of the detector. Thus some differences in behavior of different gratings at the same wavelengths and detector segments may arise from detector effects which are not currently being corrected. The different TDS values for gratings falling on the same detector segments at a common wavelength may be caused by slight differences between the two photocathodes. The results described in the above sections have been incorporated into a new TDS reference file which can be used within calcos to correct the initial on-orbit absolute sensitivity determination for the actual sensitivity at the time of a particular observation. These reference files were delivered on July 14, 2010. For the NUV gratings, the TDS corrections are grating-specific but wavelength-independent, and use the weighted mean TDS values given in Section 2.2. For the FUV gratings, wavelength-, detector-, and grating-specific corrections are calculated. These files should correct fluxes to an accuracy of ±2%. Pipeline processing of COS datasets now routinely apply these corrections. For COS observers who have already obtained observations, these corrections can be applied by manually setting the TDSCORR option to PERFORM and changing TDSTAB to specify the new reference file, or re-extracting pipeline-processed spectra from the archive. The Exposure Time Calculator on the web (ETC v18.2) uses these updated sensitivities, projected midway into Cycle 18, for observers to estimate their exposure time requirements.

Instrument Science Report COS 2010-15 (v1) Page 24

Figure 14. Summary of TDS slopes as a function of wavelength. All gratings and central wavelengths used in the NUV and FUV spectroscopic sensitivity monitoring programs are shown. For the FUV data, open symbols indicate measurements made on segment A, while filled symbols indicate measurements made on segment B. Low resolution NUV and FUV data are illustrated in green, while FUV M-modes are shown in blue and NUV M-modes are shown in red. For the G230L gratings, stripes in which contributions from second order light can be found are additionally indicated with an X. These values have been used to construct a TDS reference file which is implemented in calcos as of June 14, 2010; see discussion in Section 4.

Instrument Science Report COS 2010-15 (v1) Page 25

Acknowledgments
We gratefully acknowledge other COS+STIS team members; discussions with Tom Ake and Dave Sahnow in particular have helped us understand the various aspects of the time-dependent sensitivity changes reported in this document.

Change History for COS ISR 2010-15
Version 1: August 2, 2010- Original Document

References
Ake, T. et al. 2010 COS Instrument Science Report 2010-03 Ghavamian, P. et al., 2010 COS Instrument Science Report 2010-09 in prep. Keyes, C. D. & Penton, S. V. 2010, COS Instrument Science Report 2010-14 Massa, D., et al. 2010a, COS Instrument Science Report 2010-01 Massa, D. et al., 2010b, COS Instrument Science Report 2010-02 Penton, S. 2006, http://cos.colorado.edu/ControlDocs/COS-11-0045.pdf Wilkinson, E. 2003 http://cos.colorado.edu/ControlDocs/COS-05-0004.pdf

Instrument Science Report COS 2010-15 (v1) Page 26

Appendix
Table 5 lists the individual observations in the NUV Spectroscopic Sensitivity Monitoring Program through the beginning of May, 2010. Figures 15 through 17 display the TDS trends for additional central wavelengths of the NUV gratings. Table 6 lists the individual observations in the FUV Spectroscopic Sensitivity Monitoring Program through the beginning of May, 2010. Figures 18 through 21 display the TDS trends for additional central wavelengths of the FUV gratings. Table 5.: NUV TDS observations, processed with calcos v2.12 Grating Central Wavelength Target °) (A G2 3 0 L 2635 WD1057+7 WD1057+7 WD1057+7 WD1057+7 WD1057+7 WD1057+7 WD1057+7 WD1057+7 WD1057+7 GD1 5 3 GD1 5 3 G2 3 0 L 2950 WD1057+7 WD1057+7 WD1057+7 WD1057+7 WD1057+7 WD1057+7 WD1057+7 WD1057+7 WD1057+7 GD1 5 3 GD1 5 3 G2 3 0 L 3360 WD1057+7 WD1057+7 WD1057+7 WD1057+7 WD1057+7 WD1057+7 GD 1 5 3 Rootname 1 1 1 1 1 1 1 1 1 9 9 9 9 9 9 9 9 9 l l l l l l l l l l l l l l l l l l l l l l l l l l l l l b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b d d d d d d d d d d d d d d d d d d d d d d d d d d d d d 0 0 0 0 0 0 0 0 0 9 1 0 0 0 0 0 0 0 0 0 9 1 0 0 0 0 0 0 9 1gyq 2ot q 3km q 4i nq 5j bq 6mrq 7x1q 8p4q 9dbq t dt q 0dj q 1hvq 2p7q 3koq 4j nq 5j dq 6 m wq 7x8q 8p6q 9dgq tdrq 0dl q 1hpq 1hxq 4j l q 4j pq 7x3q 7xbq t dt q Observation Date Au g . 1 4 , 2 0 0 9 Sept. 4, 2009 Sept. 29, 2009 Nov. 3, 2009 Dec. 1, 2009 Jan. 6, 2010 Feb. 4, 2010 March 3, 2010 March 29,2010 March 29, 2010 April 30, 2010 Au g . 1 4 , 2 0 0 9 Sept. 4, 2009 Sept. 29, 2009 Nov. 3, 2009 Dec. 2, 2009 Jan. 6, 2010 Feb. 4, 2010 March 3, 2010 March 29, 2010 March 29, 2010 April 30, 2010 Au g . 1 4 , 2 0 0 9 Au g . 1 4 , 2 0 0 9 Nov. 3, 2009 Nov. 3, 2009 Feb. 4, 2010 Feb. 4, 2010 March 29, 2010 Continued on n texp (s) 540 480 480 540 480 480 540 480 480 1318 54 768 762 765 760 753 737 746 738 735 149 149 999 998 968 967 947 945 1318 ext page

1 1 1 1 1 1 1 1 1

9 9 9 9 9 9 9 9 9

1 1 1 1 1 1

9 9 9 9 9 9

Instrument Science Report COS 2010-15 (v1) Page 27

Table 5 continued fro Grating Central Wavelength Target ° (A) GD 1 5 3 G1 8 5 M 1921 G191B2B G191B2B G191B2B GD7 1 G2 2 5 M 2186 G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B G2 2 5 M 2410 G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B G2 2 5 M 2306 G191B2B G191B2B G191B2B GD7 1 G2 8 5 M 2617 G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B

m previous page Rootname Observation Date l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d d 10dnq a1thq a2khq a3baq 2tegq abruq a1tjq ncxxq a2kjq adstq aep4q a3bcq afceq agb8q ahcyq abs1q a1tnq ncxzq a2knq adsvq aep6q a3bgq afcgq agbaq ahd0q a1tlq a2klq a3beq 2telq abs3q a1tsq ncy1q a2kpq adt2q aep8q a3biq afclq agbcq April 30, 2010 Sept. 16, 2009 Nov. 3, 2009 Feb. 8, 2010 Feb. 8, 2010 Sept. 1, 2009 Sept. 16, 2009 Oct. 15, 2009 Nov. 3, 2009 Dec. 3, 2009 Jan. 6, 2010 Feb. 8, 2010 Mar. 8, 2010 April 5, 2010 April 26, 2010 Sept. 1, 2009 Sept. 16, 2009 Oct. 15, 2009 Nov. 3, 2009 Dec. 3, 2009 Jan. 6, 2010 Feb. 8, 2010 Mar. 8, 2010 April 5, 2010 April 26, 2010 Sept. 16, 2009 Nov. 3, 2009 Feb. 8, 2010 Feb. 8, 2010 Sept. 1, 2009 Sept. 16, 2009 Oct. 15, 2009 Nov. 3, 2009 Dec. 3, 2009 Jan. 6, 2010 Feb. 8, 2010 Mar. 8, 2010 April 5, 2010 Continued o

texp (s) 1318 130 130 130 130 241 241 241 241 241 241 241 241 241 241 225 225 225 225 225 225 225 225 225 225 213 213 213 213 330 330 330 330 330 330 330 330 330 n next page

Instrument Science Report COS 2010-15 (v1) Page 28

Table 5 continued fro Grating Central Wavelength Target ° (A) G191B2B G2 8 5 M 2739 G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B G191B2B GD7 1 G191B2B G191B2B G191B2B G2 8 5 M 3094 G191B2B G191B2B G191B2B

m previous page Rootname Observation Date l l l l l l l l l l l l l l l b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b d d d d d d d d d d d d d d d ahd2q abs5q a1tuq ncy3q a2krq adt4q aepaq a3bkq 2tesq afcnq agbeq ahd4q a1txq a2ktq a3buq April 26, 2010 Sept. 1, 2009 Sept. 16, 2009 Oct. 15, 2009 Nov. 3, 2009 Dec. 3, 2009 Jan. 6, 2010 Feb. 8, 2010 Feb. 8, 2010 Mar. 8, 2010 April 5, 2010 April 26, 2010 Sept. 16, 2009 Nov. 3, 2009 Feb. 8, 2010

texp (s) 330 440 400 440 400 440 440 400 560 440 440 440 1230 1230 1230

Table 6.: FUV TDS observations, processed with calcos v2.12 Grating Central Wavelength ° (A) G1 4 0 L 1105 Target WD WD WD WD WD WD WD WD WD WD WD WD WD WD WD 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 +8 +8 +8 +8 +8 +8 +8 +8 +8 +8 +8 +8 +8 +8 +8 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 Rootname l l l l l l l l l l l l l l l b b b b b b b b b b b b b b b b b b b b b b b b b b b b b b 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 3 4 5 6 7 8 9 0 1 3 4 5 6 7 j wq lm q d3q jjq axq bqq n2q pfq fzq j yq m dq d5q jlq brq c2q Observation Date Sept. 15, 2009 Sept. 29, 2009 Nov. 2, 2009 Dec. 1, 2009 Jan. 4, 2010 Feb. 1, 2010 March 3, 2010 April 1, 2010 May 1, 2010 Sept. 15, 2009 Sept. 29, 2009 Nov. 2, 2009 Dec. 1, 2009 Jan. 4, 2010 Feb. 1, 2010 Contin texp (s) 200 200 200 200 200 200 200 200 200 160 160 160 160 160 160 ued on next page

G1 4 0 L

1230

Instrument Science Report COS 2010-15 (v1) Page 29

Table 6 continued Grating Central Wavelength Target ° (A) WD0947+857 WD0947+857 WD0947+857 WD0947+857 G1 3 0 M 1309 WD0947+857 WD0947+857 WD0947+857 WD0947+857 WD0947+857 G1 3 0 M 1291 WD0947+857 WD0947+857 WD0947+857 WD0947+857 WD0947+857 G1 3 0 M 1327 WD0947+857 WD0947+857 WD0947+857 WD0947+857 WD0947+857 G1 6 0 M 1577 WD1057+719 WD1057+719 WD1057+719 WD1057+719 WD1057+719 WD1057+719 G1 6 0 M 1600 LDS749B L DS 7 4 9 B WD1057+719 WD1057+719 WD1057+719 WD1057+719 LDS749B WD1057+719 WD1057+719 WD1057+719 WD1057+719 G1 6 0 M 1623 LDS749B WD1057+719 WD1057+719

from previous page Rootname Observation Date l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l bb908n5q be701hhqa bb909pl q bb910g2q bb901k0q bb905j nq bb908neq bb909pnq bb910g7q bb901k2q bb905j t q bb908ngq bb909ppq bb910g9q bb901k6q bb905j vq bb908nj q bb909ps q bb910gbq bb916l dq bb917k9q bb918seq bb9x1m qq bb919dxq bb9x2dhq a9t71hfqb a9t12e5qb bb916lfq a9r02dlqc a9r02dtqc bb917kbq bb92t l s q bb918s gq bb9x1m s q bb919dzq bb9x2dkq a9t12emqb bb916l hq bb917kdq March 3, 2010 March 21, 2010 April 1, 2010 May 1, 2010 Sept. 15, 2009 Dec. 1, 2009 March 3, 2010 April 1, 2010 May 1, 2010 Sept. 15, 2009 Dec. 1, 2009 March 3, 2010 April 1, 2010 May 1, 2010 Sept. 15, 2009 Dec. 1, 2009 March 3, 2010 April 1, 2010 May 1, 2010 Sept. 8, 2009 Nov. 13, 2009 Jan. 28, 2010 March 31, 2010 April 19, 2010 April 19, 2010 Au g . 6 , 2 0 0 9 Au g . 1 7 , 2 0 0 9 Sept. 8, 2009 Sept. 14, 2009 Sept. 14, 2009 Nov. 13, 2009 Nov. 14, 2009 Jan. 28, 2010 March 31, 2010 April 19, 2010 April 19, 2010 Au g . 1 7 , 2 0 0 9 Sept. 8, 2009 Nov. 13, 2009 Contin

texp (s) 160 81 160 160 365 365 365 365 365 310 310 310 310 310 420 420 420 420 420 330 330 330 330 330 330 780 1200 390 330 490 390 2010 390 390 390 390 1200 460 460 ued on next page

Instrument Science Report COS 2010-15 (v1) Page 30

Table 6 continued from previo Grating Central Wavelength Target Rootname ° (A) LDS749B l bb92t l uq WD1057+719 lbb918siq WD1057+719 lbb9x1muq WD1057+719 lbb919e1q WD1057+719 lbb9x2drq G1 6 0 M 1589 WD1057+719 lbb916ljq WD1057+719 lbb917kfq WD1057+719 lbb918skq WD 1057+719 lbb919e3q G1 6 0 M 1611 WD1057+719 lbb916llq WD1057+719 lbb917khq WD1057+719 lbb918smq WD 1057+719 lbb919e5q Observation was made using dispersed light target acquisition. a Additional monitoring observation from program 12096. b Observations from SMOV program 11492. c Observation was made as part of SMOV program 11494.

us page Observation Date Nov. 14, 2009 Jan. 28, 2010 March 31, 2010 April 19, 2010 April 19, 2010 Sept. 8, 2009 Nov. 13, 2009 Jan. 28, 2010 April 19, 2010 Sept. 8, 2009 Nov. 13, 2009 Jan. 28, 2010 April 19, 2010 This is discussed in m

texp (s) 2300 460 460 460 460 360 360 360 360 420 420 420 420 ore detail in §3.

Instrument Science Report COS 2010-15 (v1) Page 31

Figure 15. TDS data for additional central wavelengths of G230L, 2950 (left) and 3360 ° A(right) central wavelengths. Stripes A through C are shown top to bottom.

Instrument Science Report COS 2010-15 (v1) Page 32

Figure 16. TDS data for additional central wavelengths of G225M, 2410 (left) and 2306 ° A(right) central wavelengths. Stripes A through C are shown top to bottom.

Instrument Science Report COS 2010-15 (v1) Page 33

Figure 17. TDS data for additional central wavelengths of G285M, 2739 (left) and 3094 ° A(right) central wavelengths. Stripes A through C are shown top to bottom.

Instrument Science Report COS 2010-15 (v1) Page 34

° Figure 18. TDS data for G140L/1105. Data are shown averaged in 100 A bins. Symbols are the same as in Figure 11.

Instrument Science Report COS 2010-15 (v1) Page 35

Figure 19. TDS data for additional central wavelengths of G130M, 1291 (left) and 1327 (right) central wavelengths. Symbols are the same as in Figure 11.

Figure 20. TDS data for additional central wavelengths of G160M, 1577 (left) and 1600 (right) central wavelengths. Symbols are the same as in Figure 11.

Instrument Science Report COS 2010-15 (v1) Page 36

Figure 21. TDS data for additional central wavelengths of G160M, 1611 (left) and 1623 (right) central wavelengths. Symbols are the same as in Figure 11.

Instrument Science Report COS 2010-15 (v1) Page 37