STScI Logo

2014 Calibration Workshop
Poster Presentations

Listing of Poster Abstracts

WFC3/UVIS Charge-Transfer-Efficiency Losses: Mitigation and Correction
Dr.  Jay Anderson (STScI)
Installed on HST in 2009, WFC3/UVIS has been subjected to the harsh radiation environment of space for more than five years. One consequence of that radiation damage is charge-transfer-efficiency (CTE) losses and these losses became apparent much more quickly than expected for WFC3/UVIS because of the instrument's low intrinsic dark current and the low backgrounds in the ultraviolet. Several years ago, we developed a pixel-based model to describe the WFC3/UVIS charge-transfer process, similar to the model that was developed for ACS/WFC. This model showed us that a low-level of post-flash (12 electrons) significantly improves CTE, and many observations now take advantage of this preventative strategy. Furthermore, the model has been used to develop a standalone script available to observers to correct for CTE losses in their images. We have recently taken some calibration data to allow us to improve the pixel-based CTE-correction model in anticipation for the model's inclusion in the WFC3 calibration pipeline and report on those results here.
Bridging STIS's Neutral Density Desert
Thomas Ayres (University of Colorado)
For UV-bright continuum sources, like early-type stars, the Global Count Rate (GCR) limit on STIS's MAMA cameras becomes an important consideration and limitation. A "safe" source must be counting below the 200 kcps (kilo-counts per second) limit, say 100 kcps. When divided among the typically 40 echelle orders of about 500 spectral resolution elements each, such a continuum source will deposit roughly 5 cps per resol. This is about two orders of magnitude less (efficient) than if just the other bright-limit, Local Count Rate, was the limiting factor. The maximum S/N for a GCR-safe observation is about 100 in one orbit (2.5 ks exposure), which sounds like a lot, but we are talking about UV-bright stars and a 2.4-m telescope; and state-of-the-art ground-based echelle spectrographs routinely achieve S/N>200. To match that with STIS implies four, or more, orbits per wavelength setting. With about seven mixed-resolution settings needed to cover the full 115-310 nm interval, that adds up to Medium-size project per target. Thus, the global limit already negatively impacts the possibility to carry out large spectroscopic surveys with STIS, if one wants to achieve full UV coverage at the highest practical S/N on a diverse sample of objects. But, it gets worse. If a target is predicted to be close to the CGR limit, or exceeds it, a neutral density (ND) filter must be used. Two such slits are supported for STIS echelle work: 0.2x0.05ND2 (0.01 transmission) and 0.3x0.05ND3 (0.001). If the target just exceeds the GCR limit, say at 300 kcps, you would have to use the ND2 slit. The 100x attenuated GCR will be safe, but very low: about 50x smaller than it would have been had the source been counting safely below the limit, say 150 kcps. To reach the same S/N as for the safe case now would require 50 orbits, instead of one; cruel reality of Poisson statistics. If just one wavelength setting, of the perhaps seven, is GCR-limited, the candidate target would be rendered unfeasible; a potentially crippling limitation on a broad survey that aspired to include representative interesting objects in an unbiased way. The limitations imposed by having only ND2 and ND3 in the STIS toolkit inspired a Cycle 19 pilot study (GO-12567) to validate three "available but unsupported" ND-filtered long slits (31x0.05): NDA (ND=0.6), NDB (1.0), and NDC (1.3). The program not only evaluated slit throughputs vs. wavelength utilizing observations of the hot white dwarf G191B2B; but also demonstrated -- by high-res (E140H) observations of UV-bright, sharp-lined Vega -- that there was no significant bleeding of light between orders, or negative impact on the spectral LSF, by the tall slits compared with the normal short echelle apertures (0.2 arcsec high). The attenuation factors of the newly validated long slits nicely bridge the gap between the default clear echelle apertures and ND2, and in fact have enabled surveys like current Cycle 21 Large Treasury Program "Advanced Spectral Library (ASTRAL): Hot Stars" (GO-13346).
Space Telescope Imaging Spectrograph Update
Dr.  John Biretta (STScI)
We review changes in the performance, calibration, and support of the Space Telescope Imaging Spectrograph since Servicing Mission 4 in 2009. Both the dark current and charge transfer efficiency (CTE) of the CCD detector continue to slowly degrade due to on-orbit radiation damage. There is now a large population of permanent warm and hot pixels. Nonetheless, the dark calibration remains very good; we discuss comparisons of the dark current subtraction residuals with the expected Poisson noise. While CTE is degraded, recent work on pixel-based corrections appears extremely promising. The MAMA detectors continue to work well. The NUV detector initially showed very high dark current after SM4, but it has since decreased and is currently at ~130% of 2004 levels. Observers are encouraged to use Time-Tag mode with the MAMA detectors, as it allows removal of data segments with high background (variable geocoronal emission, etc.), and provides additional scientific value to the HST archive. Long-term sensitivity decreases noted for most modes prior to SM4 continue along the same trajectories after SM4. The largest decreases are in the blue (e.g. G230LB), reaching about 10% loss since the start of the mission. These changes are automatically corrected in the calibration pipeline with good accuracy using Time Dependent Sensitivity tables. Two recent developments are useful for observing bright targets: Work is underway to better characterize effects which limit photometric accuracy of the CCD detector for targets near and above full-well saturation. We have also begun supporting observations with three additional neutral density slits; the exact throughput reduction factors are wavelength dependent, but are approximately 6, 14, and 33 for the 31x0.05NDA, 31x0.05NDB, and 31x0.05NDC slits, respectively. In addition, new tests have been performed on the coronagraphic finger BAR5 and the corner of BAR10 at inner working angles of ~0.2 arcseconds.
ACS/WFC Geometric Distortion and Its Linear Time Dependency
David Borncamp (Space Telescope Science Institute)
We present a new geometric distortion model and its time dependency for the Wide Field Camera (WFC) of the HST Advanced Camera for Surveys (ACS). The primary goal of this study is to derive a new Instrument Distortion Coefficient (IDC) table for superior alignment and combination of WFC images. This will allow the STSDAS and DrizzlePac software to accurately align and combine ACS images with any HST images. The calibration observations of globular cluster 47Tuc taken through the ACS/WFC F435W, F606W and F814W filters before and after Servicing Mission 4 over a wide range of the HST roll angles were used for astrometric calibration. DrizzlePac had to be modified to properly handle the implementation of the new time-dependency and its results are presented here.
The WFC3/UVIS Dark Calibration and CCD Monitor
Matthew Bourque (Space Telescope Science Institute)
Wide Field Camera 3 (WFC3) has been one of the primary HST instruments since its installation during Servicing Mission 4 in May 2009 and continues to perform well. As one of two channels available on WFC3, the UVIS channel contains two e2v CCDs and is designed for its ultraviolet and visible light detecting capabilities. UVIS dark frames are routinely acquired in order to monitor the dark current behavior and hot pixel growth. They are also used to generate dark reference files, which are used in the calibration pipeline to provide the most current calibration for science data, and are typically available to the user community 2-5 days after initial processing. We discuss the long-term evolution of permanent hot pixels (increasing at ~100/chip/day and currently occupying ~3.5% of each chip) and background dark current (increasing at ~1.4 e-/hr/year/pix, and currently ~6 e-/hr/pix). We outline the procedure for reference file generation, which involves aggressive cosmic ray and hot pixel masking, and discuss future changes to the algorithm that will improve the calibration product. Lastly, we discuss the effects that Charge Transfer Efficiency and post-flashing have on the hot pixel population. Dark monitoring will continue throughout and beyond the current HST proposal cycle as a means of evaluating the overall health and stability of WFC3/UVIS.
A new HST FGS Astrometry capability
Art Bradley (Spacecraft System Engineering Services)
A new HST FGS calibration mode was developed and executed on-orbit. This new mode employs a series of grids, each composed of 351 x 351 horizontal and vertical steps, offset in such a way as to densely sample the 2-dimensional intensity profile of the collimated beam of a point source over the 5x5 arc second instantaneous field of view of the FGS.
Gaia: Initial CCD Radiation Damage State and Comparison of the CTI Mitigation Scheme to that Applied to HST Data
Dr.  Cian Crowley (ESA/ESAC)
The European Space Agency's Gaia mission was launched in December 2013. It will continue to operate at L2 for a nominal 5 years, rotating slowly to scan the sky such that its two optical telescopes will repeatedly observe more than one billion stars. The focal plane contains 106 large area CCDs continuously operating in a mode where the line transfer rate and the satellite rotation speed are in synchronisation (TDI mode). One of the greatest challenges in the data-analysis is dealing with the effects radiation damage to the CCDs which will cause charge deferral and image shape distortion. This is particularly important because of the extreme accuracy requirements of the mission. We present preliminary results on the diagnosis of the early-stage radiation damage to the detectors (both in the parallel and serial direction) obtained during the commissioning phase, and also compare the results to those obtained from on-ground tests. The overall plan for mitigation of these effects in the data-processing is outlined and compared with the CTI (Charge Transfer Inefficiency) mitigation schemes that are applied to HST data.
The devil is in the details: improving COS calibration for high precision reverberation mapping
Gisella De Rosa (Ohio State University)
Reverberation mapping is a powerful technique that can help us understand the geometry and kinematics of the emitting gas in the proximity of supermassive black holes (AGN broad line regions). The limiting factor in such a technique is the ability of obtaining a precise and stable flux calibration for the AGN spectra. Our group is currently analyzing HST reverberation mapping data for a nearby AGN: NGC 5548. The program was executed during Cycle 21 (GO Program: 13330). Data were collected daily for 179 orbits using the COS FUV detectors. While the standard reduction pipeline guarantees flux uncertainties at approximately the 5% level, this is not sufficient for our scientific goals. We present here the results of a thorough investigation aimed at improving the COS wavelength and flux calibrations. In particular, we have developed an improved strategy for handling the systematics in the computation of the COS sensitivity functions and their time dependency, that enables us to reach a relative flux calibration that is better than 2%. We were also able to characterize the dependency of residuals due to fixed-pattern noise correction (p-flat) on signal to noise ratio (SNR): by stacking spectra at different SNR we determined the impact of the latter calibration step in various regimes.
The WFC3 Two Chip Solution: New Zeropoints and Flatfields for WFC3/UVIS
Dr.  Susana Deustua (STScI)
The two WFC3 UVIS chips are independent detectors, with different properties. To better track the individual behavior of the two chips, including inherent differences in color terms, the WFC3 team developed a chip-dependent approach to photometric calibration, such that each chip will now have a unique set of zeropoint values and each chip/filter combination will have its own flat. New header keywords, PHTFLAM1 and PHTFLAM2, the inverse sensitivity for chip 1 and chip 2 (in units of ergs cm-2 A-1 DN-1), will be written to the image headers via the IMPHTTAB image photometry reference file. CALWF3 will be updated with a new calibration switch FLUXCORR, which will scale chip 2 by the ratio of the inverse sensitivity of the two chips (PHTRATIO). The resulting FLT and DRZ products (in elecrons/sec) will therefore be continuous across the full detector field of view. Users will apply a single zeropoint (corresponding to chip 1) to the full frame image. For targets which may have significant color terms, users performing photometry on chip 2 should back out the PHTRATIO correction before applying chip-specific zeropoints and/or color terms.
A New Generation of WFC3/IR Dark Calibration Files
Mr.  Michael Dulude (STScI)
In mid to late 2013, we released a new generation of WFC3/IR dark calibration files for all allowed full-frame and subarray observing modes. These new dark calibration files are based on data collected during observing cycles 17, 18, 19 and 20 (2009-2013). Compared to the previous generation of IR dark calibration files, the signal to noise ratio improved by a factor of 3-11 and error values were reduced by 44-98%. This poster summarizes the properties of the new dark calibration files, and how they compare the their predecessors. In addition, we briefly describe the processes used to create and validate these calibration files. The files are currently in use in the calibration pipeline, applied to any data retrieved via the Mikulski Archive for Space Telescopes (MAST).
Advanced Time-domain Calibrations and Data-reductionTechniques with HST/COS
Justin  Ely (STScI)
The Hubble Space Telescope/Cosmic Origins Spectrograph (HST/COS) detectors have a time-tag mode of observation in which the arrival time of each photon is recorded individually. Although the COS calibration pipeline (CalCOS) makes use of this capability in many aspects of routine processing, there remains a number other ways that this information can be used to improve the calibration for specific science cases. This has led to the development of tools and techniques to perform additional calibrations that are not part of the standard data products output by CalCOS, but have been made available to the user. Here we demonstrate a few of these techniques including day/night filtering, extracting spectra on sub-exposure timescales, producing photometric light-curves, and performing additional dark-count screening.
Longterm photometric trends in WFC3/UVIS
Catherine Gosmeyer (STScI)
The WFC3 photometric throughput is routinely monitored in order to assess its stability as a function of time and wavelength in both channels. The data consist of spectrophotometric standard star subarray images taken in a subset of filters. The current observation cadence, once every five weeks, is deliberately out of phase with the monthly anneal procedures in order to sample the phase space between anneals. Aperture photometry of the UVIS channel data show no evidence for throughput degradation due to contamination and there has been effectively no change in throughput in F275W and F336W. Other filters, however, have shown small trends: a ~0.1-0.2%/year increase in F225W, F218W and ~0.1-0.2%/year decrease in F814W and F850LP. At visible wavelengths (e.g. F606W, F547M) the throughput change is somewhat larger, 0.3-0.4% /year decline in throughput, similar to what has been seen previously with ACS/WFC.
STIS Coronagraphy: Teaching an Old Coronagraph New Tricks
Dr.  Carol  Grady (Eureka Scientific and GSFC)
Using HST/STIS broadband optical coronagraphy and a multi-aperture, multi-spacecraft roll observing strategy, we have completed the observational phase of a program to study the spatial distribution of dust in a well-selected sample of 10 debris and one “mature” protoplanetrary disks, all with HST pedigree, using STIS. With this observational strategy, we achieve a median S/N=48 per resolution element for the bright debris ring associated with HD 181327. We are able to probe the interior regions of the debris systems with inner working distances as close as 5 - 6 AU from their host star for the nearest systems, and simultaneously resolve disk substructures well beyond those distances corresponding to the giant planet and Kuiper belt regions within our own solar systemSolar System. These observations also reveal diffuse low-surface brightness dust at larger stellocentric distances, observations of which remain a technical challenge to the most aggressive and advanced ground based techniques and facilities.
The WFC3/UVIS Gain Monitor and Low Sensitivity Pixel Population
Ms.  Heather  Gunning (STScI)
The Wide Field Camera 3 (WFC3) was installed on the Hubble Space Telescope during Servicing Mission 4 (SM4) in May 2009 and has been performing well on-orbit. WFC3 has two observational channels, IR and UV/visible (UVIS). Since installation, the WFC3 team has been diligent in monitoring the health of both channels. The UVIS channel consists of two 2051x4096 silicon-based detectors assembled in a 2x1 mosaic. The absolute gain of the UVIS channel is used as a proxy for the on-orbit evolution of the UVIS channel. We present the results of the gain stability monitor for Cycle 20 (PID: 13168) and Cycle 21 (PID: 13561) and compare to previous measurements. Each month, the UVIS channel undergoes an annealing process, where the detector is warmed to 20+ C, repairing 20-30% of the current total hot pixel population and is essential for the health and stability of the UVIS detector. In addition to the hot pixel population there exists another variant sensitivity population, which has lowered sensitivity and develops over the time between anneals. The lowered sensitivity pixels are a unique set each anneal cycle and show no evidence of being pixels that oscillate between low and expected sensitivity. We discuss the ongoing analysis of this population including general characteristics, wavelength dependence, and mitigation options.
Progress Toward a STIS Pixel-Based CTI-Correction
Mr.  Sean Lockwood (STScI)
The CCD on board Hubble's Space Telescope Imaging Spectrograph (STIS) has been accumulating radiation damage on-orbit for the past 17 years. This damage manifests itself as increased charge transfer inefficiency (CTI), with charge traps temporarily capturing photo-electrons during the read-out process and releasing them downstream. These CTI-induced trails not only reduce the number of photons counted from a given source, but contaminate other pixels as well. CTI also has detrimental effects on faint source detection and astrometry. While CTI effects for spectroscopic observations may be mitigated by placing the spectrum closer to the CCD readout register, this is not always possible when fully exploiting the spatial coverage of STIS (e.g. extended or multiple sources observed with long slits) or mining archival data. To help reduce the impact of CTI on science, the STIS team has been developing a stand-alone pixel-based CTI-correction tool based on the CALACS algorithm (Anderson and Bedin 2010). Here we present an overview of the steps required to apply the CTI-correction to scientific data, including the creation of CTI-corrected reference files, and describe measurements of absolute CTI levels derived from dark data taken with alternate readout amplifier A that is used to verify the efficacy of the correction. We also detail an auto-correlation technique used to fine tune our correction parameters and demonstrate the correction on a multi-source observation contaminated by cosmic ray CTI-trails.
WFC3/UVIS Flat Field Accuracy & Improved Solutions for UV Filters
Jennifer Mack (STScI)
To quantify the accuracy of the UVIS flats, a white dwarf standard was stepped across the detector field of view in a subset of filters. At visible wavelengths (F336W, F438W, F606W, F814W), the stepped photometry is consistent to +/-0.7% rms and +/-1.3% peak to peak. For the UV filters (F218W, F225W, F275W, F280N), the ground test flats were obtained in ambient conditions require a correction for sensitivity residuals with temperature. The stepped photometry varies by +/-3.3% peak-to-peak for these filters, and the residuals correlate with the UV 'crosshatch' pattern. We have developed a model to correct this residual pattern, and the revised flats correct the photometry to +/-1.5% peak to peak. Further testing using dithered star cluster observations is in progress, and we expect to deliver a new set of UV flats by Fall 2014.
Flux Calibration of the COS Blue Modes
Dr.  Derck Massa (Space Science Institute)
We present GO COS FUV observations of the Helium wd WD0308-565 using the COS FUV CENWAVE = 1055 and 1096 (Blue) modes. Together with a model atmosphere derived from a fit to STIS observations, these data are used to derive a new flux calibration for the blue modes. We describe the calibration, its internal and external consistency and specific issues unique to these modes.
Continuum subtracted Halpha images from HST F656N data
Dr.  Divakara Mayya (INAOE, Mexico)
No ground-based Halpha images can compete with the quality of structures of ionized gas seen in F656N and F606W filter images of the HST. However, obtaining a continuum-free Halpha morphology from these images is not straight-forward due to the absence of off-band Halpha images. We have successfully created continuum-free Halpha images of M81 and M82 using the broad-band continuum images as off-band Halpha images. I would like to present these results in this contribution.
Next Generation SED Standards: Establishing A Faint DA White Dwarf Network with HST
Dr.  Gautham Narayan (NOAO)
Photometric calibration uncertainties are the dominant source of error in current type Ia supernova dark energy studies, and other forefront cosmology efforts, e.g., photo-redshifts for weak lensing mass tomography. Modern 'all-sky' surveys require a network of calibration stars with 1) known SEDs (to properly and unambiguously account for filter differences), and 2) that are on a common photometric zero-point scale. HST enables us to establish this essential network of faint spectrophotometric standards, by eliminating the time-variable Earth's atmosphere, and by exploiting the well-understood energy distributions of DA white dwarfs. We have been awarded 60 orbits to observe 26 DA WD targets in Cycle 22 (GO 13711) that will have SNR ~200 in PanSTARRS and the Dark Energy Survey, and LSST images in the future. We will obtain high S/N ground-based spectra of Balmer lines to constrain the two parameters (temperature and log(g)) that determine the SED. We use UVOIR HST WFC3 photometry to set the overall flux scale for each source, and determine any applicable reddening, allowing this sample to be used as flux standards at wavelengths well beyond the range of HST, and in any arbitrary passband. This sample will constitute the gold-standard with which to directly calibrate, and cross-validate the survey data products. In our Cycle 20 program (GO 12967) we obtained 18 orbits to observe 9 DA white dwarfs. Here we present preliminary results from that program using 6 DA WDs with HST photometry and Gemini GMOS spectroscopy. We demonstrate sub-percent precision between the different objects. Our program extends the precision photometric heritage from HST to the benefit of essentially all existing and upcoming surveys, standardizes (spectro)photometry across bservatories and facilities, and directly addresses one of the current barriers to understanding the nature of dark energy.
Charge Transfer Efficiency in WFC3/UVIS: Monitoring and Corrections from Star Clusters
Dr.  Kai Noeske (ESA/STScI)
Using calibration fields in the star clusters NGC6791 and NGC104/47 Tuc, we have been monitoring the evolution of the charge transfer efficiency (CTE) of the WFC3/UVIS detectors, starting 5 months after the WFC3 installation, ranging from October 2009 to the present; this program will continue through future cycles. Earlier epochs had shown a strong evolution of the CTE, amounting to more than 0.1 mag CTE loss per year for stars with a total flux of 1000 electrons, at large distances from the readout amplifiers, in short-exposure zero sky background images. Brighter stars, and sources in longer exposures and/or higher backgrounds, are less severely affected. We find the previous trends to continue with time, although there is evidence for a flattening in the CTE evolution. We present an update to the CTE monitoring results, as well as an update to the empirical correction model that corrects CTE losses for point source aperture photometry as a function of observation date, source flux, local background level, and source distance on the detector from the readout amplifiers. The data we collect are also currently analyzed to independently test the pixel-based CTE correction algorithm for UVIS (Anderson et al.), and the efficiency of CTE mitigation through post-flash introduced backgrounds in UVIS images.
The IR background as seen by WFC3
Dr.  Nor Pirzkal (STScI)
We present a new and improved characterization of the sky background light as seen by WFC3 in the the near-infrared. We determine an empirical model of the zodiacal background from thousands of images obtained using WFC3 since its installation on board of HST. While the structure of the background, as parametrized as a function of sun angle and ecliptic latitude, is similar to the model contained in the WFC3 ETC, this new model extends the model to smaller values of sun angle with an accuracy of ~0.1 e-/s/pix for F098M, F105W, F125W, F140W, and F160W. We also present a characterization of the Earth-glow background as a function of the HST pointing orientation with respect to the Earth limb. Interestingly, we identify a strong emission line component of the background from He I at 1083 nm in the upper atmosphere that can significantly increase the background in both IR grisms and in broad-band filters sensitive to this wavelength, even well above (> 40 deg) the bright Earth limb. Accounting for these effects will be important for optimizing the efficiency of potential future deep integrations with the WFC3/IR grisms and bluer broad-band filters.
Changes to the Spectral Extraction Algorithm at the Third COS FUV Lifetime Position
Dr.  Charles Proffitt (STScI/CSC)
The current COS extraction algorithm simply adds up the flux in a box significantly larger than the height of the point spread function in the cross-dispersion direction. If any pixel within the extraction height is flagged as bad in a column, the entire column is discarded from the combined extracted spectrum in the X1DSUM files. The Third COS FUV Lifetime Position (LP3) will be located only 2.5" away from the original lifetime position (LP1), a separation significantly smaller than the 3.5" spacing between LP1 and LP2. This closer spacing is desirable to keep the spectral resolution as high as possible, but it will also cause the far wings of spectral images observed at LP3 to potentially overlap with low gain regions created by the heavy usage of the detector at LP1. Any of these pixels which have lost more than 5% of their sensitivity will be flagged as bad. In practice, however, the total flux loss for spectra centered at LP3 that overlap these regions will be quite small, both because many gain sagged locations retain most of their initial sensitivity, and because they will be located in the far wing of the LP3 profile, which contain a small fraction of the total flux. To avoid unnecessarily discarding columns affected by such pixels, an algorithm is needed that can judge whether the effects of the gain sagged pixels on the extracted flux are significant. However, small scale detector distortions and drifts in the focus and event locations (Y-walk) would limit the robustness of any straightforward encircled energy correction or optimal extraction algorithm. The solution adopted for pipeline use was tailored specifically for the COS FUV data characteristics and involves a number of steps. (1) Residual geometric distortion in the cross dispersion direction is first removed using a library of 1D "trace" centroid vectors. (2) A library of smoothed 2D reference profiles is aligned with the individual spectral observation. (3) Based on the aligned reference profile, a wavelength dependent inner zone which contains about 80% of the flux is defined, as well as an outer zone that contains 99.5% of the flux, (the actual fractions used to define the zones are adjustable via a reference file). (4) In each column, all counts in both zones are added, but the column is only marked as bad if one of the pixels in the inner zone is flagged as bad. We present draft version of the trace and profile libraries that will be used, and provide also provide comparisons of extractions made using the new and old algorithms.
Moving COS FUV Spectra to the Third COS FUV Lifetime Position
Dr.  Charles Proffitt (STScI/CSC)
In February 2015, the location of COS FUV spectra on the detector is expected to be moved from the second COS FUV Lifetime Position (LP2) to the third COS FUV Lifetime Position (LP3). This move is needed to mitigate the continuing effects of gain sag in the microchannel plate detectors. With the LP2 move (July 2012) spectra were shifted perpendicularly to the dispersion direction by the equivalent of +3.5", or about 41 pixels, from the original LP1 location. With the LP3 move, spectra will be shifted below the original LP1 position by about –2.5”, or about –31 pixels, which preserves the spectral resolution to the greatest extent possible and also maximizes the detector area available for future lifetime positions. The modes with the widest profiles perpendicular to the dispersion, G130M 1055 and 1096, will remain at LP2, while all other modes will be moved to LP3. Improved extraction techniques are being developed to ensure that the gain sagged regions near LP1 do not affect the spectral quality of point source observations at LP3. Typical spectral resolving power (λ/Δλ) for the modes that are moved to LP3 may decline by about 15% relative to the performance at LP1. Observations of spatially extended sources may also be slightly impacted by the proximity to the gain sagged regions near the original LP1 location and care may be needed not to confuse gain sag artifacts with real spectral features for such sources. Instrument performance is otherwise expected to be similar to that seen during Cycle 21.
New Gapless COS G140L Mode for Low Astigmatism Far-UV Observation
Mr.  Keith Redwine (Johns Hopkins University)
The demonstrated sensitivity of COS G140L mode to wavelengths below 1150 angstroms has opened a new window to the universe previously inaccessible to Hubble, and enabled a number of compelling science investigations. To take full advantage of this new mode, we successfully proposed to calibrate a new observing mode, wherein CENWAVE was set to 800 A, thereby providing contiguous wavelength coverage from 900 to 1850 A on the Segment A detector of COS. This new mode has three significant advantages: 1) a ~2 times lower background from reduced astigmatism at the short wavelength end where the effective area is lowest, which will boost the S/N of background limited observations; 2) contiguous spectral coverage on a single detector segment, which will simplify tracking of flux, wavelength and flat-field calibrations; and 3) a more efficient use of observing time for programs requiring the full far-UV wavelength coverage by eliminating the need for a grating change. We describe the flux and wavelength calibration of the new mode, done using the calibration target AV 243, a stable 06V star in the Small Magellanic Cloud. We observed AV243 in Cycle 19, and also utilized archival spectra of AV234 previously acquired by the Far-Ultraviolet Spectroscopic Explorer (FUSE) and the Faint Object Spectrograhp (FOS), which guided our analysis of the new CENWAVE 800 mode of COS G140L. We determine the best focus position and characterize the astigmatism as a function of wavelength, as well as provide a wavelength calibration, effective area, and background equivalent flux.
Characterizing and correcting distortion due to walk in the COS FUV detector
Dr.  David Sahnow (STScI)
Since the installation of the Cosmic Origins Spectrograph (COS) in 2009, more than 1×10^10 photon events have been collected by the cross-delay line (XDL) detector in the instrument’s Far Ultraviolet (FUV) channel. Each incident photon has the effect of slightly decreasing the gain of the microchannel plates (MCPs) in the region where it hits. As the gain of the MCPs change, the effects of walk, or the dependence of detected position on gain, lead to a distortion of the two-dimensional spectral image. The instrument design and sensitivity, along with the spectral energy distributions of the targets observed, combine to give an uneven illumination pattern on the detector, which leads to distortions over a range of spatial scales. These distortions can affect the data in several ways, from target acquisition accuracy to the flux accuracy of spectra extracted by the calibration pipeline. We discuss the effects of the walk on the data and our efforts to mitigate it by managing the high voltage across the MCPs. We also discuss our recently derived walk correction.
Temporal Evolution of the HST/COS Sensitivity
Dr.  Hugues Sana (STScI)
We review the history of on-orbit changes in the sensitivity of the Cosmic Origins Spectrograph (COS) on board the Hubble Space Telescope. Despite a significant decline since Servicing Mission 4, the COS Far Ultraviolet (FUV) sensitivity is still outstanding, and COS remains the instrument of choice for most spectroscopic observations of faint FUV targets. Since the beginning of on orbit operations in May 2009, COS has experienced long periods (~1 year) of relatively shallow sensitivity loss ( < 10% per year) and short periods of steeper decline, most notably at the end of 2011 where, for a few months, the sensitivity of some modes declined at rates of up to 35% per year. The rate of decline varies both with time, wavelength and detector segments. It is suspected that at least one component of these changes is related to atomic oxygen from the residual atmosphere at HST's orbital altitude reacting with the CsI photocathode of the open-faced COS FUV XDL detector. We will discuss how the observed throughput changes correlate with solar activity and with the inferred atmospheric density at HST's orbit and will compare the on-orbit measurements with laboratory studies of CsI sensitivity to atomic oxygen exposure. For the COS NUV channel, the G185M, G230L, and imaging modes, which use MgF2 coated aluminum optics, have shown only small changes over time, with sensitivities at most wavelengths remaining within a few percent of the values seen at launch. This contrasts with the bare aluminum G225M and G285M gratings, which have shown an ongoing linear degradation over time that began prior to launch. The G225M has been declining at about 3% per year, while the G285M throughput has been dropping by 10% to 12% per year. By early 2015, the throughput as a function of wavelength for the G285M grating will be down from its initial values by more than a factor of two.
Dr.  Laurence Taff (Taff & No Associates)
The discrepancy between the heretofore-successful accretion disc theory for the prototypical dwarf nova, i.e., SS Cygni, and a Hubble Space Telescope Fine Guidance Sensor-determined annual parallax of 159 pc for it, revealed a significant deficiency with the instrument’s astrometric calibration methodology. A VLBL-determined parallax of 114 pc “… reconciles the source behavior with our understanding of accretion disc theory in accreting compact objects” (Miller-Jones et al. 2013, pg. 350). Because, the lower distance is consistent with the disc instability model’s prediction of its absolute magnitude during an outburst. The Miller-Jones et al. (2013) result also agrees with Schreiber & Gänsicke’s (2002) prediction of 117 pc. Given this, one is forced to question whether other flaws exist in the processing of FGS data. Multiple astronomical and mathematical failings specific to the Astrometry Team’s processing of FGS measures are brought to the fore, there are grounds for believing so. Whence, a new, astronomically and mathematically sensible, astrometric calibration of the FGSs is required and delivered herein. It can both maximize the utility of the observations already acquired as well as those obtained in the future. I recapitulate two applications of the procedure that have clearly demonstrated its efficacy for visual and for radio data (the Appendix has documentation for dozens of other instances including one utilizing gamma-ray observations). The adjustment philosophy, and implementation, rests on the fact---as opposed to unfounded statistical hypotheses---that a star can only be in one place at one time plus the method of infinitely overlapping circles.
Motivations and challenges while using an unsupported mode : observations with the E140H grating and the 52
Dr.  Frederic Vincent (LATMOS)
Using the E140H echelle grating with the 52" x 0.5" slit on STIS allow to combine high spectral resolution with strong signal-to-noise ratio. This unsupported mode is useful to observe extended sources as the interplanetary hydrogen or planets (e.g. Jupiter). However this mode results into overlapping orders and the lack of standard line spread function, which creates difficulties to calibrate wavelengths and fluxes. Calibration observations have been made with the HITM2 lamp, using the 0.2" x 0.5" slit to have the same slit width but a smaller aperture to avoid exceeding the bright object protection limits on the FUV-MAMA detector. We present the results and then discuss their use for astrophysical observations and the possible need for other calibration observations.
A Path to NIST Calibrated Stars over the Dome of the Sky
Prof.  Peter Zimmer (Univ. of New Mexico)
The UNM Measurement Astrophysics group is currently testing two mobile instruments, a 0.5m multi-wavelength backscatter lidar and a 100mm stellar spectroradiometer, that are designed to help create a new set of standard stars calibrated to NIST laboratory standards. Initially this is being done for bright stars across the wavelength range 350nm to 1050nm at 1nm spectral resolution with measurement accuracy goal of better than 1% per spectral resolution element by calibration to NIST silicon detectors. The design, operation and calibration of the lidar (the Facility Lidar for Astronomical Measurement of Extinction - FLAME) and spectroradiometer (the Astronomical Extinction SpectroPhotometer - AESoP) are detailed. We have chosen this pairing of measurements, lidar and spectroradiometry, because they are most sensitive to different components of Earth's atmosphere, the variable transmission of which is the single largest impediment to ground-based radiometric calibration. FLAME has been designed to generate calibrated absolute transmission at three well-chosen laser wavelengths, allowing the spectrum of atmospheric transmission to be pinned at those wavelengths. AESoP, by virtue of having an absolutely-calibrated throughput, makes accurate measurements of the bottom-of-the-atmosphere spectra of bright stars. MODTRAN models allow us to bridge the gap between the two and obtain the top-of-the-atmosphere spectral energy distribution of the star as well as the spectrum of atmospheric transmission along the observation path.