Skip to main content
August 1, 2023

About This Article

In this STAN, we announce the delivery and implementation of updates to the reference files for the STIS NUV time dependent sensitivity and for selected STIS spectroscopic and imaging modes contained in the third installment of the STIS flux recalibration effort.

NUV Time Dependent Sensitivity (TDS) Reference File Updates

The time dependent sensitivity reference file (TDSTAB) is used by CALSTIS during the conversion from counts to cgs (centimeter–gram–second) flux as part of the spectral extraction (X1DCORR) and spectral image rectification (X2DCORR). These reference files provide the information to correct for time-varying wavelength-dependent changes in the sensitivity for the different spectral elements, as well as temperature-dependent sensitivity variations for each detector. After analyzing the latest TDS monitoring data for the NUV-MAMA G230L/2376 setting (program ID = 16960), the STIS team identified a change in the sensitivity trends for wavelength bins ≥2400 Å (see Figure 1). An updated NUV TDSTAB (77o1532oo_tds.fits) was delivered on July 24, 2023 to replace the existing post-SM4 (USEAFTER date of May 01, 2009) reference file, primarily impacting wavelengths ≥2400 Å. The improvements introduced by the new NUV reference file are of the order of ~1-1.5% at the reddest wavelengths for observations taken after January 01, 2019. Any observations taken before this date remain unchanged. Users working with data at wavelengths longer than 2400 Å taken in the last few years should ensure that their observations have been calibrated with the latest NUV TDSTAB.

We remind users that the STIS higher resolution modes (echelle modes and first-order M-modes) adopt the TDS trends as characterized by the first-order L-mode observations. The accuracy of this approach is monitored for a limited selection of higher resolution spectral settings. Overall, investigations have shown that adopting the L-mode sensitivity variation trends for the rest of the STIS settings provides accurate flux calibrations (STIS ISR 2017-06), with a few exceptions. Following the same approach, the new NUV TDSTAB reference file includes similar updates for all NUV modes.

We advise users to confirm that their observations are using the latest reference file by examining the TDSTAB keyword in the primary header of their science observations. For any NUV observation downloaded before the release of this new TDSTAB we recommend that users retrieve their calibrated products from the MAST archive one more time, as these archive products will then be using the latest reference files.

We also note that the STIS team is currently monitoring new emerging NUV trends for wavelength bins 1800-2000 Å. At these short wavelengths, the last two monitoring observations, taken on March 18 and June 1 of 2023, show a sharp decrease in sensitivity (up to 2%) after more than five years of a slowly but steadily increasing trend. The team is investigating this new pattern to confirm and better characterize its presence.

For any issues related to the updated NUV TDS reference file please contact the STIS Team through the HST Help Desk.

TDS trends for the NUV G230L/2376 setting for wavelength bin 2400 Ă…
Figure 1:  TDS trends for the NUV G230L/2376 setting for wavelength bin 2400 Å. The red and yellow symbols show the old and updated TDSTAB corrections, respectively. We note that the new TDSTAB reduces the residuals by ~1% for observations taken after 2019.


Updated STIS Flux Calibration (Release 3)

The STIS team recently delivered a new set of reference files, including new photometric throughputs and echelle blaze shift coefficients (PHOTTAB), ripple tables (RIPTAB), spectral traces (SPCTRTAB), and image photometry tables (IMPHTTAB). The new delivery is part of the larger STIS Flux Recalibration effort to incorporate recent improvements to the atmospheric models for the standard stars (Bohlin et al. 2020, CALSPECv11) and the accompanying re-examination of the Vega flux. This third release includes updates to various high- and medium-resolution echelle settings (post-Servicing Mission 4: E230H/2263, E230H/2713, pre-Servicing Mission 4: E140M/1425,  E140H/1271, E140H/1416, E230M/1978, E230M/2707, E230H/2263, E230M/2415, E230H/2713), and NUV imaging modes (F25QTZ, F25SRF2, F25CN182, F25CN270). This latest release also includes updates to the spectral traces for some of the secondary echelle modes (E140H/1271, E140H/1307, E140H/1343, E140H/1489, E230M/2415, E230H/2463, E230H/2713, E230H/2812, E230H/2912).

Echelle modes

This third release includes three different updates to the echelle modes: (1) new photometric throughputs and blaze shift coefficients for post-Servicing Mission 4 (SM4; in 2009) E230H/2263, E230H/2713 data, (2) updated photometric throughputs for pre-SM4 E140M/1425, E140H/1271, E140H/1416, E230M/1978, E230M/2415, E230M/2707, E230H/2263, E230H/2713 data, and (3) a new characterization of the spectral traces for the secondary echelle modes E140H/1271, E140H/1307, E140H/1343, E140H/1489, E230M/2415, E230H/2463, E230H/2713, E230H/2812, E230H/2912.

Post-SM4 updates

The various STIS E230H settings provide high-resolution echelle spectra with coverage between ~1600 and 3100 Å.  The post-SM4 throughputs for the E230H modes were last updated using the CALSPECv7 spectral models (STIS ISR 2012-01). As for our recent updates to the E230M throughputs (January 2023 STAN), new sensitivity curves and blaze shift coefficients for the primary E230H/2263 and secondary E230H/2713 modes were derived through the analysis of observations of the standard DA white dwarf G191-B2B.  The newly derived throughputs for these two modes were obtained following a similar approach as that described in STIS ISR 2022-04 . The updated throughputs, based on the CALSPECv11 models, increase the NUV fluxes by ~3% around 1700 Å and by ~2% at 3000 Å. Additionally, the previous version of the NUV PHOTTAB did not flux calibrate orders 272 and 300 (covering wavelength ranges of 2830-2845 Å and 2560-2580 Å, respectively, when using the secondary E230H/2713 mode). The recently delivered PHOTTABs now flux calibrate these additional orders. We note, however, that because these additional orders are located on the bottom and top edges of the detector, one of the two pre-defined background regions for each order falls off the detector.  The background contributions can be underestimated, and the data quality (DQ) flag 2048 (>30% of background pixels rejected by sigma-clipping) can be assigned to all points in the order. While the calibrated fluxes of high signal-to-noise observations are not strongly impacted by this issue (see orders 272 and 300 in right panel of Figure 1), the STIS Team is currently working on updating the 1-D extraction parameter reference file (XTRACTAB) to define more appropriate background regions for these two orders present in E230H/2713 observations.

We highlight that the newly updated reference files (PHOTTABs and RIPTABs), which improve the flux calibration of the E230H/2263 and E230H/2713 settings, replace the previous versions and impact only post-SM4 observations (see Table 1).

Calibrated E230H/2713 observations for the standard G 191-B2B
Figure 1.  Calibrated E230H/2713 observations for the standard G 191-B2B. The corresponding order numbers are shown towards the bottom of the spectra. The green solid line displays the CALSPECv11 model for this standard star, and the dashed lines show the +/- 5% intervals. The left-hand panels show the calibrated products using the old PHOTTAB and RIPTTAB files. The right-hand panels show the calibration using the recently delivered reference files, better aligned to the CALSPECv11 fluxes. We also include the additional newly calibrated orders 272 and 300.


Detector Updated setting Old PHOTTAB New PHOTTAB Old RIPTAB New RIPTAB USEAFTER date

E140M/1425, E140H/1271, E140H/1416

64719319o_pht.fits 77o1827ho_pht.fits     1996 Oct
FUV-MAMA E140M/1425, E140H/1271, E140H/1416 64719310o_pht.fits 77o1827ko_pht.fits     1999 Mar
FUV-MAMA E140M/1425, E140H/1271, E140H/1416 6471930ro_pht.fits 77o18270o_pht.fits     2001 Jul

E230M/1978, E230M/2707, E230H/2263, E230H/2415, E230H/2713

64719316o_pht.fits 77o18276o_pht.fits     1996 Oct
NUV-MAMA E230M/1978, E230M/2707, E230H/2263, E230H/2415, E230H/2713 64719313o_pht.fits 77o1826to_pht.fits     2001 Jul
NUV-MAMA E230H/2263, E230H/2713 6cg2020mo_pht.fits 77o1827fo_pht.fits 6cg2020po_rip.fits 77o1827bo_rip.fits 2009 May
NUV-MAMA E230H/2263, E230H/2713 6cg2020oo_pht.fits 77o18278o_pht.fits     2018 Mar


Table 1.  Replaced and recently released photometric throughput files and ripple table for the spectroscopic modes included in the third flux recalibration release.

Pre-SM4 updates

The throughputs used to calibrate pre-SM4 echelle observations were last updated using CALSPECv4 models (STIS ISR 2007-01 ). The process for deriving updated throughputs for echelle post-SM4 data has been described in detail in STIS ISR 2012-01 and STIS ISR 2022-04 . Briefly, multi-nodal splines are fit to the inverse sensitivity functions (ISF; net count rates of a given standard star divided by the corresponding CALSPEC model), after removing the time-dependent sensitivity component, the ISFs are then used to define new throughputs taking into account the aperture and spectral extraction box size, as well as the collecting area of the telescope. Typically, when updating the post-SM4 throughputs, the STIS team also updates the blaze shift coefficients as needed. Because the blaze shift coefficients for pre-SM4 data have been well characterized and are currently static, the team adopted a slightly different approach for updating the throughputs for the echelle modes to account for the CALSPECv11 improvements to the standard star G 191-B2B.

For wavelengths between ~1200-3000 Å, the increase in flux of the CALSPECv11 G 191-B2B model, compared to that from CALSPECv4, is of the order of ~1-2.5% (Figure 2). To improve the flux calibration of pre-SM4 echelle observations, the STIS team has begun updating the pre-SM4 throughputs by applying a scaling factor derived from the difference between the new and old models. In Figure 3 we show the percent difference between the CALSPECv11 model and pre-SM4 E230M/1978 observations calibrated using the old PHOTTAB (based on CALSPECv4 models; grey squares), along with the difference observed in an exposure calibrated with the new scaled PHOTTAB (dark green squares). The new scaled PHOTTABs provide fluxes with an absolute accuracy of  ~2-2.5% — well below the promised absolute accuracy of 8% (see STIS Instrument Handbook Section 16.1).

For the present release, the team has updated the throughputs based on this scaling method for the pre-SM4 prioritized echelle modes: primary modes E140M/1425, E140H/1271, E140H/1416, E230M/1978, E230M/2707, E230H/2263, and secondary modes E230M/2415, E230H/2713 (see Table 1).

Percent difference between the CALSPECv11 and CALSPECv4 models for the standard star G 191-B2B
Figure 2.  Percent difference between the CALSPECv11 and CALSPECv4 models for the standard star G 191-B2B. We highlight the wavelength coverage for the E230M/1978 setting for reference with the observed changes in the new calibrations shown Figure 3.
Percent difference between the CALSPECv11 model and a pre-SM4 E230M/1978 exposure
Figure 3.  Percent difference between the CALSPECv11 model and a pre-SM4 E230M/1978 exposure calibrated with the old PHOTTAB (grey squares), and calibrated with the new scaled PHOTTAB (dark green squares).

Updated spectral traces

The shapes of the spectral traces for the different echelle orders vary with grating and wavelength. For the primary echelle modes, the STIS team has accurately characterized the shape and its variations on a configuration basis. The STIS echelle secondary mode traces, on the other hand, have been defined as straight lines — which in some cases might not be ideal for accurately extracting the flux of a given order (particularly near the edges, where many orders tend to exhibit more complex shapes). Information on the spectral traces — the 1-D spectrum trace table — is stored in the SPTRCTAB reference file and is used during the X1DCORR calibration step to locate and extract 1-D spectra for each echelle order. In Figure 4 we show an example of the current extraction box (centered on the current straight trace) in blue, for order 285 of the secondary mode E140H/1562.

The STIS team has improved and updated the spectral traces for a prioritized list of secondary modes: E140H/1271, E140H/1307, E140H/1343, E140H/1489, E230M/2415, E230H/2463, E230H/2713, E230H/2812, E230H/2912. The new traces were defined by adopting gaussian process regression methods. This approach automatically determines length scales for individual orders. In Figure 4, the dashed orange lines show the updated extraction box (centered on the new trace) for order 285 in the E140H/1562 setting. Tests have shown that these updated traces can improve the flux estimates by 1-4%, particularly around the edges of the orders. Users aiming to achieve the highest possible flux accuracies are advised to confirm that their observations were calibrated with the latest SPTRCTAB available (see Table 2).

2-D image showing spectral order 285 for secondary mode E140H/1562
Figure 4.  2-D image showing spectral order 285 for secondary mode E140H/1562. The old spectral extraction box (centered on the trace in the reference file SPTRCTAB) is shown with solid blue lines. The optimized and updated extraction box (centered on the new trace for this particular order) is shown with dashed orange lines.


Detector Updated Setting Old SPCTRTAB New SPTRCTAB USEAFTER date
FUV-MAMA E140H/1271, E140H/1307, E140H/1343, E140H/1489 q8l14503o_1dt.fits 77o18272o_1dt.fits 1996 Oct
FUV-MAMA E140H/1271, E140H/1307, E140H/1343, E140H/1489 q8l14504o_1dt.fits 77o1827do_1dt.fits 1999 Mar
NUV-MAMA E230H/2463, E230H/2713, E230H/2812, E230H/2912, E230M/2415 q8l14501o_1dt.fits 77o18279o_1dt.fits 1996 Oct
NUV-MAMA E230H/2463, E230H/2713, E230H/2812, E230H/2912, E230M/2415 q8l14502o_1dt.fits 77o1827jo_1dt.fits 2000 Jun


Table 2.  Replaced and recently released 1-D spectrum trace tables (SPCTRTAB) for the secondary modes included in the third flux recalibration release.

As a final note on the calibration of echelle observations, we remind users that the STIS Team has released a Python module, stisblazefix, which can help improve the alignment of the blaze shift function — giving better flux agreement in the order overlap regions — on an individual-exposure basis.

NUV imaging modes

STIS imaging observations are processed through the IRAF-independent CALSTIS pipeline using the Image Photometry Table, IMPHTTAB, to determine the photometric parameters (PHOTFLAM, PHOTBW, and PHOTPLAM) across all three detectors. The throughputs for the different imaging modes were last updated in the early 2000s (STIS ISR 2003-01 and STIS ISR 2004-05). The STIS team is now working on updating the sensitivities used to derive the new photometric parameters in the IMPHTTAB.  In this flux recalibration release, we include IMPHTTAB files created using updated throughputs accounting for the latest stellar atmospheric model improvements in CALSPECv11. Similar to the methods described in STIS ISR 2003-01, the throughputs were updated through the analysis of a set of calibration targets (e.g., HS 2027+0651, NGC 6681). The latest version of the IMPHTTAB files increase the NUV fluxes by ~1-2.5%. This initial imaging flux recalibration update includes the prioritized NUV-MAMA filters: F25CN182, F25CN270, F25QTZ and F25SRF2. See Table 3 for more details on the new reference file names and corresponding USEAFTER dates.

Detector Updated Setting Old IMPHTTAB New IMPHTTAB USEAFTER date
NUV-MAMA F25QTZ, F25SRF2, F25CN182, F25CN270 y2i1649no_imp.fits 77o1826ro_imp.fits 1996 Oct
NUV-MAMA F25QTZ, F25SRF2, F25CN182, F25CN270 59l1632po_imp.fits 77o18274o_imp.fits 2007 Jan


Table 3. Replaced and recently released Image Photometry Tables (IMPHTTAB) for the NUV modes included in the third flux recalibration release.

All of the archival data impacted by this new release have now been reprocessed. Users who require the most accurate flux calibrations are advised to re-retrieve their science observations and confirm that the calibrated products were made using of the latest reference files (see Tables 1, 2, and 3). For any questions regarding the updated reference files delivered as part of this recalibration release, please contact the HST help desk. Please consult the STIS Flux Recalibration webpage for additional information on the status of the on-going and future flux calibration updates.



HST Help Desk 

Please contact the HST Help Desk with any questions.