HST STIS Advanced Spectral Library Project

Thomas Ayres (for the ASTRAL Co-Investigators), University of Colorado (CASA)

ASTRAL is Hubble Space Telescope Treasury Project to create high-resolution, high-S/N, full-coverage UV spectral "atlases" of representative bright stars using high-performance Space Telescope Imaging Spectrograph. In 2010–2011, eight iconic late-type stars were targeted (146 orbits); then in 2013–2015, twenty-one early-type objects (230 orbits). Subsequently, additional stars with suitable archival material were added.

!! NEWS !!

Contents of Catalog

Top-level object tables, below, summarize stellar characteristics, mostly from SIMBAD, and link to several layers of processed data.


ASTRAL Cool Stars

HD NumberNameα2000δ2000VB−VTypeϖCALSTIS Date
61421ALP-CMI114.825+05.2250.340.40F5IV-V284.62017-JUN-23 + NOV-02
128620ALP-CEN-A219.902−60.8340.010.71G2V754.82017-JUN-24 + SEP-28
128621ALP-CEN-B219.896−60.8381.330.88K1V796.92017-JUN-24 + SEP-28

ASTRAL Hot Stars

HD NumberNameα2000δ2000VB−VTypeϖCALSTIS DateNotes
2017-JUN-23double: 0.05" sep
66811ZET-PUP120.896−40.0032.25−0.28O4I(n)fp3.012017-JUN-23classic P-Cygni wind
46223HD-4622398.038+04.8237.320.13O4V((f))0.582017-JUN-23MK standard
2017-JUN-23MK standard
36959HR-188683.754−06.0095.67−0.23B1Vv3.122017-JUN-23sharp-lined, `normal' B
37479SIG-ORI-E84.696−02.5946.66−0.18B2Vp2.282017-JUN-23broad-lined, magnetic star
52089EPS-CMA104.656−28.9721.51−0.13B1.5II8.052017-JUN-23low ISM
160762IOT-HER264.866+46.0063.80-0.17B3IV7.172017-JUN-23sharp-lined, abundance standard
215573XI-OCT342.595−80.1235.31−0.12B6IV6.332017-JUN-23sharp-lined, `normal' B
1757341-ARI42.495+27.2603.61−0.09B8Vn19.72017-JUN-23broad-lined, LISM
87901ALP-LEO152.092+11.9671.40−0.16B8IVn41.12017-JUN-23broad-lined, LISM
176437GAM-LYR284.735+32.6893.25−0.05B9III5.262017-JUN-23broad-lined, MK standard
172167ALP-LYR279.234+38.7830.030.00A0V130.22017-JUN-23,24photometric standard, LISM
48915ALP-CMA101.287−16.716−1.470.01A0V379.22017-JUN-24Am, brightest star, LISM

Metal-Deficient Turn-Off Stars, and Related Objects

HD Numberα2000δ2000VB−VTypeϖ[Fe/H]CALSTIS Date

Calibration White Dwarfs

Nameα2000δ2000VB−VTypeϖCALSTIS Date
G191-B2B76.378+52.83111.69−0.25DA.8 C18.892017-MAY-20
BD+28°4211327.796+28.86410.58−0.33sdO2VIIIHe58.802017-MAY-20,21 + OCT-19

First layer down (linked through Column 1) holds final spectrum (or spectra) for each target, covering all (or significant part) of 1150–3100 Å range.

"Final spectrum" links down to subordinate spectra. Lowest layers of tree are sub-exposures; next layer up, co-added "o-type" basic exposures; then setting-specific "E-type" co-added spectra; top layer holds broad-coverage "U-type" spliced spectra.

Each page displays preview of spectrum. At top level, graphical timeline of observations is color-coded by mode. FITS files of fundamental data are linked for downloading. Also, for top-level datasets, "ETC-ready" file is highly streamlined version of final spectrum, intended for HST Exposure Time Calculator (or any adhering to HST standard), but NOT SUITABLE FOR ANALYSIS (see: ETC Summary). Finally, ASCII versions of co-added data files are available at ASCII.

Fits Formats. "o-type" files (individual observations) have basic header information in 0-th extension describing target, exposure properties, and splice points from order-merging; and one or more trailing data extensions. If observation had only single sub-exposure, there is one extension (EXTEN=1), containing WAVE – wavelength (Å); FLUX – flux density (erg/cm²/s/Å); ERROR – photomeric error (same units as flux density); and DQ – data quality (0 for no issues; higher values to flag various conditions: bad pixels, camera blemishes, gaps, and so forth). If two or more sub-exposures, EXTEN=1 holds Stage ZERO co-added spectrum; subsequent extensions have parameters of individual sub-exposures: sub#n in EXTEN=n+1. EXTEN=0 header now contains additional information, concerning cross-correlation template and derived velocity shifts for Stage ZERO co-addition.

"E-type" co-added and "U-type" spliced spectra have similar FITS structure to single-exposure o-type, consisting of just two extensions. 0-th extension again lists basic information concerning target and exposure properties; and summaries of cross-correlation templates, velocity shifts, splice points (for U-types), and flux scale factors. Extension 1 contains spectral parameters for co-added and/or spliced spectrum. In all cases, including o-types, most refined dataset always is EXTEN=1.


ASTRAL is a Hubble Space Telescope (HST) Large Treasury Project, whose aim was to collect high-quality ultraviolet spectra of representative bright stars utilizing high-performance Space Telescope Imaging Spectrograph. In Cycle 18 (2010–2011), ASTRAL focused on eight iconic late-type stars, devoting 146 HST orbits to the effort. In Cycle 21 (2013–2015), program shifted to warm side of H-R diagram, capturing 21 diverse early-type objects with allocation of 230 orbits. Main objective was to record well-known bright stars, like Procyon, Betelgeuse, Sirius, and Vega, with broad uninterrupted UV coverage (1150–3100 Å), at highest signal-to-noise and highest echelle spectral resolution achievable within allotted orbits and observing constraints. These UV "atlases" have enormous interpretive value in their own right, and complement efforts from ground-based observatories, which now routinely achieve comparably high resolution and S/N in optical and near-infrared spectra of bright stars.

Broad ultraviolet coverage was achieved by splicing together STIS echellegrams taken in multiple FUV (1150–1700 Å) and NUV (1600–3100 Å) grating settings. S/N was maximized by spreading total desired exposure depth in each setting over 2–5 separate "visits." Because of slight randomness in STIS grating positioning mechanism, and because of changing telluric and spacecraft motions, each independent exposure shifts slightly on detector and thus experiences different "fixed pattern noise." Combining independent spectra post-facto mitigates these systematics, improving S/N. Further, observing sequences had at least one exposure of each setting immediately after target-centering "peak-up" so that velocity zero point, which can be affected by thermal drifts, would be as accurate as possible. Other exposures of that type can be registered to reference observation by cross-correlation. If practical, a few exposures of each sequence were taken through the photometric aperture (0.2″×0.2″), and again near a peak-up, to ensure that radiometric scale of final spliced spectrum would be close to true absolute level.

Post-processing of ASTRAL spectra followed protocols developed for earlier StarCAT project, extensive catalog of STIS echelle spectra of objects classified as "stars." Full description in Ayres (2010: ApJS 187, 149), which author strongly encourages consulting as general introduction before attempting to use ASTRAL spectra for analysis purposes. At same time, ASTRAL experience has uncovered number of observation-related issues, which have required modifications and additions to original protocols.

Reduction Steps

Processing begins with CALSTIS pipeline. Several key reference files were modified (by author) based on post-SM4 measurements of wavelength calibration images, to update coefficients of polynomial dispersion relations, and detector locations of echelle orders, for the 44 supported echelle settings. New lamp emission-line atlas was introduced. Some photometry files modified to include orders present on detectors, but not included in contemporary reference files. Time-evolution coefficients for blaze shifts, based on post-SM4 WD calibration spectra, were updated.

Post-processing begins with CALSTIS "x1d" output file, tabulation of extracted wavelengths, flux densities, photometric errors, and data quality flags for up to several dozen orders of particular grating setting (e.g., E140M-1425, where first part is echelle mode and second is central wavelength [CENWAVE] in Å). x1d file contains at least one — sometimes several — sub-exposures, which were treated as separate observations. Initial processing included up-dated post-facto correction for subtle wavelength distortions identified in previous study of STIS dispersion relations (see, e.g., "Deep Lamp Project"); Bayesian reformulation of photometric error; and trimming of E230 settings to avoid un-flagged "drop-outs" at edges of low orders. Also, spurious tilts of spectra, due to unfortunate interaction between narrow slits and telescope "breathing," were corrected, if necessary. x1d orders then were merged, averaging overlap regions weighted by individual sensitivity functions s λ, but accounting for bad pixels and wavelength gaps. In parallel, "active blaze correction" determined optimum blaze shift to balance fluxes in overlaps between adjacent orders. If blaze shift could not be determined, default was to apply a time-dependent correction estimated from long-term behavior of post-SM4 observations of standard star BD+28°4211.

Next, different layers of co-addition and splicing were applied to order-merged 1D spectra of each object.

STAGE ZERO — Sub-exposures, if any, of a basic observation were combined. Individual spectra aligned in velocity by cross-correlating against sub-exposure with highest apparent throughput, determined by comparing total net count rates integrated over central zone of echelle format. (With narrow slits used extensively in ASTRAL Hot-Stars, effective throughput can vary up to ∼30% during single orbit owing to "telescope breathing.") Next, sub-exposures were interpolated onto wavelength scale of reference spectrum; scaled to reference in flux density according to net count rates; then co-added, weighting by total net counts, but taking into account bad pixels and gaps. Resulting files are "o-type" and have rootname of original observation, e.g., obkk52040.

STAGE ONE/TWO — Same-setting exposures of an object taken in different orbits of visit, or in different visits, were combined. Hybrid of Stages ONE (same slit) and TWO (different apertures) described in StarCAT, to take advantage of uniform observing strategy of ASTRAL. As in Stage ZERO, cross-correlation alignments were relative to exposure exhibiting highest apparent throughput. Also, individual observations were scaled to reference exposure according to flux ratios determined by global average over high S/N intervals. Again, weighting was by total net counts, which allows for different integration times in exposure set (or use of different apertures in different visits). Resulting files are "E-type," and have appended aperture code and MJD date range to reflect diversity of constituent exposures, e.g., "E140M-1425_0.2×0.2_55543-55554." If two or more different slits were used, aperture code was changed to "MULTIaperture," e.g., "E230H-2513_MULTIaperture_52507-57585."

STAGE THREE — Different wavelength segments of object were spliced to produce, ideally, seamless spectrum covering full FUV + NUV range. As in other steps, wavelengths of adjacent segments were aligned by cross-correlation. "Bootstrapping" calibration procedure took advantage of intentional broadly overlapping spectral coverage to refine velocity zero point and absolute flux scale (by pair-wise comparisons of overlap regions in velocity and flux). Resulting file is "U-type," e.g., "UVSUM_1X_55543-55554." "UVSUM" part signals a multi-wavelength splice; middle numeral indicates particular grouping of spectra spliced; adjacent letter tells whether all were medium resolution ("M"), all were high resolution ("H"), or mixed ("X"; latter is case for all but one ASTRAL Cool Stars, but several Hot Stars are purely M and some purely H); and trailing date range summarizes minimum and maximum starting MJDs of spliced group.

In splicing procedure, general philosophy was to minimize overlap regions between observations of different resolutions, to extent possible given desire to include enough overlap to determine accurate flux ratios, as well as at least one suitable cross-correlation feature. Exception was made for frequent combination E140H-1291 + E140M-1425, to boost S/N in key FUV interval below 1350 Å. In this instance, entire overlap zone (1150–1350 Å) was co-added. In all cases of mixed resolution, higher resolution spectrum was filtered with lower resolution line-spread-function, and vice versa. Thus, spectral resolution in mixed resolution overlap zone is convolution of the two lsf's, and is lower than lower of the two. Photometric errors were adjusted for the filtering, and co-addition weighting was according to smoothed version of respective inverse squared errors (in flux density units). Dual-filtering procedure avoids awkward lsf resulting from simply adding two mixed-resolution overlaps (as was done in StarCAT). New parameter "RESOL" (λ/Δλ) was added to U-type FITS files to keep track of places where mixed-resolution spectra were spliced. Finally, for Cool Stars targets, an E230M-2707 taken for flux calibration purposes was spliced only over small interval at longwavelength end of merged NUV H-res spectrum.


If highest resolution is desired, say for ISM studies, avoid top-level spliced (U-type) spectrum below 1350 Å, and instead consult E140H-1291 co-add one level down (if available).

There remain spurious defects in STIS spectra, especially noticeable in Hot Stars targets with long stretches of smooth continuum. Most appear as weak, narrow "emission lines," mainly in NUV. They apparently result from flat-field issues. Occasionally defects occur in FUV where local areas of sensitivity degradation on detector were not flagged: these appear as broad (few Å) shallow absorptions. Also, new flag (EPSILON=450) marks places between adjacent echelle settings where difference between co-added fluxes was larger than 3 times average photometric noise. Flag introduced to help identify possible processing glitches.

Because of necessity to de-tilt and rescale many sub-exposures, owing to "breathing" effects, ASTRAL Hot Stars material not well suited for temporal studies of continuum variations (0.2″×0.2″ Cool Stars o-type exposures should be better in time-domain regard). Also, fact that exposures of each setting were scaled to observation with maximum throughput could introduce bias, if source had been variable over span of STIS visits. Thus, absolute fluxes of ASTRAL spectra should be viewed cautiously; although relative flux distributions should be reliable.

ASTRAL Cool Stars

A few words concerning the eight late-type targets. They cover wide range of spectral types, F–M, mainly giants and supergiants. All had been observed by IUE, and all previously by HST, usually GHRS but two by STIS. However, especially for GHRS, spectral coverage was incomplete, resolution was not as high as STIS H modes, or S/N did not meet goals. To round out original ASTRAL Cool Stars, nearby, bright late-type dwarfs Alpha Centauri A (G2 V) and B (K1 V) were added (from ongoing joint Chandra/HST program to track coronal activity cycles). Key characteristics:

Beta Cassiopeia (Caph: F2 IV)— Most extreme "X-ray deficient" case in group of already anomalous fast-spinning Hertzsprung gap giants. These display powerful FUV emissions, but surprisingly under-luminous X-ray coronae. Beta Cas, like Procyon (see next), falls at edge of convection: essential ingredient (together with rotation, which Beta Cas has in abundance, but Procyon not) for Dynamo generation of magnetic fields, with their consequent effects on high-energy atmospheric processes. Important link in magnetic evolution of Hertzsprung gap giants: 'cool' 1 MK corona versus 'hot' 10 MK for later G0 IIIs.

Alpha Canis Minoris A (Procyon: F5 IV-V)— Nearby, bright, warmer analog of low-activity Sun. Important cool-corona object (2 MK). Chandra transmission grating spectrum mid-way between solar-like Alpha Cen A and more active K-type companion Alpha Cen B.

Alpha Centauri (Rigel Kentaurus) A (G2 V) and B (K1 V)— Nearest sunlike stars, only 1.3 pc away. Binary system with 80-yr period. Alpha Cen A is near twin of Sun in fundamental properties, including age and coronal activity; B is smaller, cooler, less luminous, but coronally more active. Important comparisons to more evolved stars of original Cool Stars sample. AB observed by STIS on roughly semi-annual basis since 2010 in joint program with Chandra.

Alpha Aquarii (Sadalmelik: G2 Ib)— "Hybrid chromosphere" supergiant in class originally noted by L. Hartmann and colleagues in early 1980s: harboring cool massive winds, imprinting blueshifted circumstellar absorptions on Mg II; but also displaying hot FUV lines like C IV, combination usually avoided in "non-coronal" giants like Arcturus (Alpha Boo: K1 III) and Aldebaran (Alpha Tau: K5 III). Weak coronal X-ray source detected by Chandra. Important comparison to sibling Beta Aqr (G0 Ib), another certified hybrid star, which previously was observed by STIS (and Chandra).

Beta Draconis (Rastaban: G2 Iab)— Yellow supergiants Beta Dra and Alpha Aqr, although superficially similar in spectral type and luminosity, are strikingly different at high energies: former is strong X-ray source with bright FUV emissions; latter is cool-wind dominated with suppressed FUV emissions and barely detected corona. Pivotal pair for understanding dichotomy between coronally active and quiet supergiants.

Beta Geminorum (Pollux: K0 IIIb)— Early-K giant with solar-like coronal properties; key comparison to non-coronal red giants mentioned above. Important contrast to equally puzzling class of active helium-core-burning "clump giants" like Iota Cap (G8 III) and Beta Cet (K0 III), both previously observed by STIS. One of few giant stars with detected weak magnetic field and suspected planetary companion.

Gamma Draconis (Etamin: K5 III)— Another hybrid chromosphere star, showing weaker fluoresced molecular lines (CO and H2) than archetype red giant Arcturus. Faint Chandra X-ray source, but stronger than Arcturus. Important link to more active hybrids like Alpha Aqr (above).

Gamma Crucis (Gacrux: M3.5 III)— Classic M giant exhibiting complex atmosphere and wind, similar to more exotic cool supergiant Betelgeuse (next) in surface temperature, equally extreme non-coronal object, but simpler, cleaner UV spectrum (narrower, less blended chromospheric and wind emission lines), yet significant mass outflow. Bridge to warmer non-coronal K giants like Gamma Dra.

Alpha Orionis (Betelgeuse: M2 Iab)— Iconic windy red supergiant, with clumpy surface convection and mysterious distant cold circumstellar shell, prominent in FUV absorptions of CO. Extreme object in terms of low surface temperature, high visual luminosity, and lack of coronal signatures. Again, important player in story of hot-corona/cool-wind transition.

ASTRAL Hot Stars

The twenty-one Hot Stars are too numerous to describe individually. Brief characteristics are noted in associated Table. Specific objects were chosen to cover full range of spectral types, early-O to early-A; main-sequence and evolved stars; normal plus chemically peculiar subtypes; fast and slow rotators; magnetic exotica; as well as nearby objects relevant to ISM studies. Practical consideration was UV-bright early-type stars can trigger over-light conditions on STIS's highly sensitive MAMA cameras. Pair of neutral density filters is available in STIS slit wheel, but initial ND step is 2 (100× attenuation); very inefficient for targets barely violating bright limits. To work around limitation, Cycle 19 calibration program ("Bridging STIS's Neutral Density Desert") validated three 31″×0.05″ intermediate-ND slits (ND=0.6, 1.0, and 1.3) that were "available but not supported." Adding these slits to STIS tool kit greatly increased number of observable objects to meet scientific objectives (still requiring 6–12 orbits per target to achieve high S/N). Case of Vega was deemed specially important (zero point of photometric scale; ultra-fast rotator seen pole-on; possible Chandra X-ray source), despite falling partly into inefficient category, even with intermediate ND slits. Vega accordingly was allocated additional orbits to boost S/N for under-performing echelle settings.

Metal Deficient Turn-Off Stars

Four metal-poor turn-off stars from GO-14161 (R. Peterson, PI) were added to ASTRAL, as examples of low-metallicity objects in contrast to more normal compositions of main sample. These (relatively faint) F–G dwarfs were recorded exclusively in the 5 NUV prime high-res settings with CENWAVE>2000 Å. About a dozen orbits, spread over several visits, were devoted to each target. Unfortunately, these metal-poor stars were too faint for FUV echelles. Nevertheless, NUV co-added, spliced spectra are of high quality.

Calibration White Dwarfs

Several hot White Dwarfs are HST standard stars. Most important for STIS are G191-B2B (photometric calibration, order-dependent blaze correction curves) and BD+28°4211 (long-term sensitivity changes, slit throughputs). G191-B2B was observed in all 44 STIS echelle settings near beginning of STIS operations, then again shortly after Servicing Mission 4 (2009, when STIS was restored to operating condition after long hibernation following electrical failure in 2004). BD+28°4211 has been recorded on regular basis (every few months) during STIS operations since beginning, in three primary medium-res echelle settings (M-1425, M-1978, M-2707), and two prime high-res settings (H-1416, H-2263). Co-adding and splicing these extensive data collections yield spectra of unusually high S/N and quality. However, because ASTRAL-modified CALSTIS was based mainly on post-SM4 calibration material, in some cases pre-SM4 echellegrams did not process well. Consequently, G191-B2B and BD+28°4211 datasets with processing defects were culled out. As result, not all of original spectra were included in co-added, spliced WD spectra here, but nevertheless final tracings are very high quality, exceeding most other ASTRAL stars.

Acknowledgments. Author expresses appreciation to the ASTRAL Co-Investigators, especially STIS colleagues at STScI who helped materially in planning and executing the complex observing programs. Based on observations with NASA/ESA Hubble Space Telescope, obtained from Mikulsky Archive at Space Telescope Science Institute, operated by Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-26555. Support for ASTRAL through grants STScI HST-GO-12278.01-A and HST-GO-13346.01-A. Project made use of public databases hosted by SIMBAD, maintained by CDS, Strasbourg, France.