Instrument Science Reports (ISRs)

Observations of HST spectrophotometric standard stars show that the COS NUV detector has a time-dependent sensitivity (TDS) that must be monitored and accounted for in flux calibration. Regular observations monitor the changes in sensitivity for three NUV gratings: G230L, G185M, and G225M. The fourth NUV grating, G285M, was removed from routine monitoring because its sensitivity has become too low. G285M continues to be available but unsupported for General Observer programs. In this ISR, I present the results of the Cycle 30, Cycle 31, and Cycle 32 NUV TDS programs.

We present updated time-dependent sensitivity (TDS) corrections for the far-ultraviolet (FUV) channel on board the Cosmic Origins Spectrograph (COS). Updated TDS corrections were required due to the recently derived geometric distortion, delta-geometric, X-walk, and Y-walk corrections introduced in the COS calibration pipeline. The new high-voltage-dependent sensitivity (HVDS) correction was also applied to all COS FUV data via the HVDSCORR module in CalCOS and incorporated into the TDS corrections. Monitoring data were binned in wavelength and fit using a piecewise linear function of time. The analysis included TDS monitoring data of calibrator white dwarfs from the start of monitoring in September 2009 until February 2024. Comparison of the recalibrated data with CALSPEC models WD0308-565 and GD71 showed that the time-dependent flux calibration is within the 5% absolute and 2% relative requirements for all FUV gratings. The residuals between data and model over time show 2-10% improvements across all wavelength bins for all Lifetime Positions (LPs).

The new Hubble Spectroscopic Legacy Archive (HSLA) provides coadded spectra of individual targets that have been observed with the Cosmic Origins Spectrograph (COS) and the Space Telescope Imaging Spectrograph (STIS) over their operating lifetime. HSLA uses data available in the Mikulski Archive for Space Telescopes (MAST). It automatically produces coadds whenever new data become publicly available or when there is newly recalibrated data. HSLA defines individual targets by their associated coordinates, accounting for proper motions, and uses SIMBAD, NED and the Phase II observing proposals to obtain astronomical classifications for each object. Coadded spectra are produced for each observing mode. In the case of COS far-ultraviolet observations there is one coadded spectrum for each lifetime position (LP). Additionally, a spectrum spanning the entire wavelength range covered by the observations is produced by abutting the spectra from a selection of individual modes. For each individual target, HSLA also provides a human-readable metadata file with key information that can be used in searches or for further exploration of the data. The HSLA project also makes the code used for coadding spectra publicly available along with several other tools (using Jupyter notebooks) for custom coaddition required in special cases. In this report we will describe the main components of HSLA and provide a brief description of how the data and metadata can be accessed.

In preparation for the activation of Lifetime Positions 7 (LP7) and 10 (LP10) on the Cosmic Origins Spectrograph (COS) Far-Ultraviolet (FUV) channel aboard the Hubble Space Telescope (HST) in Cycle 33, we conducted a series of focus sweep observations in December 2024 and January 2025 to optimize the focus settings for the COS G130M/1222, 1291, 1055, and 1096 modes at LP7 and the G160M modes at LP10. The focus varies with lifetime position, requiring a re-determination of the Optics Select Mechanism 1 (OSM1) absolute focus position for each grating and central wavelength configuration. Using spectra from the subdwarf stars Feige 48 and AG+81-266, we performed a spectral line-width analysis via autocorrelation functions of focus offsets relative to an estimated absolute focus value to identify the OSM1 positions that produced the sharpest absorption features. Our results yielded final absolute focus values of -993, +37, and -2556 focus steps for G130M/1222, 1291, and 1096 at LP7, and +124 focus steps for G160M/1600 at LP10. We then extrapolated from these results to find the focus values for G130M/1055 and the remaining G160M central wavelengths. The LP7 values were incorporated into the COS flight software table pcmech_OSMTbl and the LP10 values were included within the commanding instructions to ensure continued optimal spectral resolution and instrument performance.

We present an update to the time dependent sensitivity (TDS) corrections for the near-ultraviolet (NUV) channel aboard the Cosmic Origins Spectrograph (COS). These new TDS corrections account for wavelength dependence, whereas the previous the TDS corrections were assumed to be grating and stripe dependent only. The TDS corrections are derived by monitoring flux calibration standard white dwarfs twice a year and fitting a linear function to the monitoring data binned in wavelength as a function of time. The analysis included TDS data from the start of monitoring in August 2009 until August 2023. Additionally, to increase the wavelength coverage of the G225M settings, we included in our analysis calibration observations from Cycle 17 in conjunction with Cycle 30 supplemental monitoring data. Comparison of the data with the CALSPEC models G191-B2B and WD1057+797 and comparison between the old and new calibrations applied to HIP 66578 showed that the new TDS corrections significantly improved the flux calibration. The new NUV TDS corrections maintain the flux accuracies to within the 5% absolute calibration requirements for most central wavelengths (cenwaves). The wavelength-dependent TDSTAB (8a818589l tds.fits) was delivered to operations on 8 October 2024.

The wavelength calibration procedure employed by CalCOS relies on measuring the separation in the dispersion direction between Pt-Ne lamp spectra that are observed concurrently with science exposures (or immediately before and after a science exposure in some cases), and a template Pt-Ne spectrum for which the dispersion solution of the Primary Science Aperture is defined. Template spectra for each Cenwave/Segment/FPPOS combination are stored in the LAMPTAB reference files, which are specific for each lifetime position (LP). Improvements to the geometric distortion, delta-geometric, X-walk, and Y-walk corrections applied to the Cosmic Origins Spectrograph far ultraviolet detector necessitated updates to all “downstream” reference files used by CalCOS, including the LAMPTAB files at all LPs. This document describes the re-creation of the LAMPTAB files.

While generally stable with time, the count rate over the duration of a COS exposure can occasionally vary, either due to the intrinsic nature of the object being observed or due to issues with the telescope assembly or COS detector. Depending on the nature of this variability, reduction of these observations may result in poor flux calibration. We use a method called Local Outlier Factor to identify COS exposures that exhibit significantly variable count rates, and classify them based on their variability. Out of all COS spectroscopic TIME-TAG data (FUV and NUV) taken through Spring 2025, we identify a total of 260 variable exposures, 11 of which exhibit variability due to telescope/detector issues, and the rest exhibiting variability due to the intrinsic variable nature of the object. We demonstrate that the problematic exposures can be corrected using the BADTCORR step in CalCOS. We provide lists of all variable exposures in addition to the explicit “bad time” intervals that can be entered into the BADTTAB reference file to correct the 11 problematic exposures.

Due to a light leak through the flat-field calibration aperture of the Cosmic Origins Spectrograph (COS), some observations with the FUV detector require the wavelength calibration aperture to move to a position where the lamp can safely flash on the detector and track wavelength solution drifts. This is applicable for any science observations taken above 5.4” in relation to Lifetime Position (LP) 1 (such as at LP6, and LP7). This procedure, also known as SPLIT-wavecal, incur significant additional overheads due to this movement. To minimize these overheads, the standard pipeline adopts a simulated wavelength calibration lamp flash at the 600-s exposure time based on historical COS data. In this report, we analyze historical COS data between 2009 and 2024 to re-derive SPLIT-wavecal parameters that define the simulated lamp flash. We find no time dependence in the values of these parameters within the range 2009-2024 and refine the wavelength solution drift of G130M and G160M Observations.

We report on the analysis and implementation of updates to the Bad Pixel Table (BPIXTAB) reference file for the Hubble Space Telescope (HST) Cosmic Origin Spectrograph (COS) far-ultraviolet (FUV) detector before the moves to Lifetime Positions 5 and 6 (LP5 and LP6). The BPIXTAB is a critical component of the COS calibration pipeline that identifies problematic or less-than-optimal detector regions through data quality (DQ) flagging. We employed a semi-automated detection procedure to identify new very low response regions (DQ = 16) and low response regions (DQ = 1024) at and near the areas of the new lifetime positions. Additionally, we adjusted the boundaries of several existing DQ regions to account for distortions in pixel coordinates. Technical and scientific testing confirmed that the updated reference files correctly flagged problematic areas and function properly within the CalCOS pipeline. The updated BPIXTAB reference files for the LP5 and LP6 eras were delivered to the HST Calibration Reference Data System in November 2021 and September 2022.

The COS FUV detector comprises two microchannel plate segments (FUVA, FUVB) with cross delay line anodes to report the dispersion (x) and cross-dispersion (y) location of the charge cloud generated from a photon event. The electronic settings used to measure the locations were optimized during ground testing, but as the detector ages with usage, these positions shift due to the lowered pulse height amplitude (PHA) of the charge cloud, a phenomenon known as “walk”. In this work, we use a wide range of archival COS data across Lifetime Positions (LPs) 1–5 to determine the corrections needed to remove walk in the y direction. Prior to this work, a Y-walk correction that was independent of detector location was applied to all COS FUV data. However, we find that the Y-walk shifts, dy, become increasingly steeper with PHA at locations above LP1 and shallower below. We use these dy measurements to create a YWLKFILE reference file that applies a correction based on YCORR location. This new correction, when combined with the updated GEOFILE, results in straighter spectral traces at LPs 1–6, and ensures that low gain events registered at the LP furthest below LP1 (i.e. LP4) remain in the spectral extraction region as expected.

The COS FUV detector comprises two microchannel plate segments (FUVA, FUVB) with cross delay line anodes to report the dispersion (x) and cross-dispersion (y) location of the charge cloud generated from a photon event. The electronics used to measure the locations were optimized during ground testing, but as the detector ages with usage, these positions shift due to the lowered pulse height amplitude (PHA) of the charge cloud, a phenomenon known as “walk”. In this work, we determine the corrections needed to remove walk in the x direction. We find that, while the shifts are ∼1-2 pixels for photon events at and near the gain nominal values of PHA = 10-12, events with PHA < 7 can have X-walk of up to ∼7 pixels, particularly low PHA events occurring on the left side of FUVA. We have used these measurements to create a new XWLKFILE reference file that contains the corrections to be applied by the CalCOS XWLKCORR module to correct these shifts.

We describe the derivation of a new geometric distortion correction reference file for the far-ultraviolet (FUV) detector of the Cosmic Origins Spectrograph (COS) aboard the Hubble Space Telescope (HST). The geometric distortion correction aims to achieve a uniform plate scale across both segments of the FUV detector by applying a transformation from thermally-corrected coordinates to a coordinate system where pixels measure 6 μm °ø 24 μm. The previous version of the geometric distortion correction reference file (GEOFILE) was known to have problems at some locations that negatively impacted the accuracy of fluxes and wavelengths in calibrated data products, hence the motivation to derive a new geometric distortion correction. To derive the distortion corrections in both the dispersion and cross-dispersion directions (dx and dy, respectively), we utilized Pt-Ne lamp spectra obtained during the Thermal Vacuum testing campaign in 2003. This process involved comparing the observed locations of individual emission lines from the Pt-Ne spectra with ray-trace models. Through interpolation and extrapolation we generated the reference file image arrays corresponding to the full active area of the detector. The new GEOFILE addresses previous artifacts and sub-optimal corrections, significantly enhancing performance across multiple detector regions in conjunction with newly implemented walk and delta-geometric corrections.

The geometric distortion correction—one of the earliest steps in the calibration of COS FUV data—enforces a uniform plate scale across the detector such that all digital pixels correspond to the same physical size. The walk correction, which follows the geometric distortion correction, ensures that photons observed at different gain levels are properly aligned on the detector. During the effort to measure X-walk (walk in the dispersion direction) and re-measure geometric distortion, it was determined that coherent, small-scale (≤ 5 pixels) errors in the dispersion direction remained, even after application of new geometric distortion and X-walk corrections. The decision was made to remove these residuals using a delta-geometric correction, to be applied after the initial geometric distortion correction, and before the X-walk correction within the flow of the CalCOS pipeline. Here we describe the motivations for using a delta correction, the data used to measure the correction, and the analysis used to create a new DGEOFILE reference file.

The process of assigning incident photons to digital pixels on the far ultraviolet (FUV) detector of the Cosmic Origins Spectrograph (COS) suffers from distortions. Nominally these are removed via the geometric distortion correction and walk correction steps within the CalCOS pipeline, but over time it has become apparent that at some locations on the detector the existing corrections are inadequate. Here we describe the overall effort to revise the COS FUV geometric distortion and walk corrections, first explaining the need for new corrections, and then presenting results that demonstrate improved calibration.

Beginning in HST cycle 25, all COS FUV target acquisition algorithms work by stepping the target across the aperture, measuring the flux at each dwell point, calculating a flux-weighted centroid based on those measurements, and slewing the telescope to that centroid position. This entire procedure is performed by the flight software, and no images are recorded. To confirm that the target acquisition procedures are working as expected, each year we run a series of tests that simulate this process by acquiring exposures at specified offsets in the cross-dispersion direction. Here, we present the results of these tests from cycle 32.