June 2016 STAN
Updates to the COS/NUV Wavelength Calibration
The long-term drift of the OSM2 (Optics Select Mechanism 2) has been slowly shifting the wavelength ranges covered by the COS/NUV channel (see March 2015 edition of the COS/STIS STAN). As a result, the lamp flash data taken with each science observation have been slowly drifting with respect to the NUV lamp template obtained during the Science Mission Operation Verification (SMOV) phase in 2009. The search range, defined in the WCPTAB, used in the pipeline as part of the cross-correlation between the lamp flash exposures and lamp templates (WAVECORR) has consequently needed regular updates. The last WCPTAB update was on July 31st, 2013. Monitoring of the shifts applied by the WAVECORR has shown that the search range once again needs to be updated. Several NUV spectroscopic datasets taken with the G185M and G225M gratings since February 2014 have suffered from erroneous zero-point shifts applied during the WAVECORR calibration step, because the true shift was outside the search range. This calibration issue has been mitigated by increasing the search range of the cross correlation from 57 pixels to 90 pixels in the dispersion direction for data taken with the G185M and G225M gratings. This change has been applied to the new WCPTAB reference file and only affects data taken since February 2014. The list of those programs affected by this update can be found in this document.
Additionally, due to lack of suitable calibration and/or science data, the zero-point of the dispersion solutions had not been updated with on-orbit values after the Science Mission Observatory Verification phase in 2009 for all G185M settings, and for select cenwave/stripe combinations of the G225M, G285M and G230L gratings listed below. This lack of proper calibration can result in a significant (up to 1 Angstrom) mis-registering of the wavelength for spectral features detected in the affected configurations. The COS team delivered a new DISPTAB reference file, which contains preliminary updates to the zero-points for the affected G225M, G285M, and G230L combinations in the list below (except for the G185M modes for which no updates have been made at this time). This preliminary update improves the zero-point accuracy from ~1 Angstrom (100 km/s) to about 0.3 Angstrom (30 km/s). Since for all affected G225M, G285M, and G230L settings, the on-orbit zero-point value (d in the DISPTAB) is known for at least one stripe, the unknown zero-points are assigned the value from the closest stripe. This update does not provide the wavelength zero-point accuracy (15 km/s for the M-gratings, 175 km/s for G230L) promised in the instrument handbook (COS Instrument Handbook Section 5.1.11).
The G185M zero-points have not been updated on orbit since no data is currently available for these settings. The calibration data to 1) derive zero-point offsets for the G185M modes, and 2) to improve the zero-point accuracy to within 15 km/s for the affected G225M and G285M settings, and to 175 km/s for the affected G230L settings are not yet available, but will be acquired over the summer 2016. The COS team will deliver a new DISPTAB containing updated zero-points for all settings in the list below once the calibration data are analyzed.
PIs of COS/NUV programs that use any of the configurations affected by either with the new WCPTAB reference file (see here for a list of PIDs) or by the new DISPTAB reference file (see list of combinations below), should re-retrieve their data from the HST archive hosted by MAST at archive.stsci.edu. This will ensure that the most up-to-date COS/NUV calibrated products are applied. Users will be notified once the updated calibrations are available for the NUV settings affected by the zero-point issue.
List of NUV cenwave/stripe combinations affected by the improper calibration of wavelength zero points:
NUV Modes with uncalibrated wavelength zero-points NOT updated in the new DISPTAB (for lack of suitable calibration data):
All cenwaves and stripes
NUV Modes with previously uncalibrated wavelength zero-points with updated preliminary zero-points in the new DISPTAB:
- 2186 – NUVA, NUVB
- 2217 – NUVA, NUVB
- 2233 – NUVA, NUVB
- 2250 – NUVA, NUVB
- 2268 – NUVA, NUVB
- 2283 – NUVA
- 2306 – NUVA
- 2325 – NUVA
- 2339 – NUVA
- 2996 – NUVC
- 3035 – NUVC
- 3057 – NUVC
- 3074 – NUVC
- 3094 – NUVC
- 2635 – NUVA, NUVC
- 2950 – NUVA, NUVC
- 3000 – NUVA, NUVC
- 3360 – NUVB, NUVC
New STScI Science Software Released
AstroConda, a new software repository, is now available through STScI’s Science Software Branch. AstroConda provides tools and utilities commonly used in astronomy as well as HST-specific software. AstroConda is a replacement for the previously released Ureka environment. We strongly encourage anyone who runs the HST pipelines to use AstroConda, especially COS users. The most recent CalCOS build, version 3.1.7, is only available in AstroConda and cannot be used with Ureka.
Additionally, we encourage users to use CalCOS with Python 3.5. CalCOS version 3.1.7 was optimized for Python 3.5 usage. Due to a difference in the native python rounding function, CalCOS products using Python 2.7 will yield slightly different results than those obtained using Python 3.5.
Full details on the new AstroConda software environment and how to install it can be found at: http://astroconda.readthedocs.io
Questions about AstroConda and the included STScI-based and Gemini-based software can be directed to: https://stsci.service-now.com/hst