CALSPEC

Access to the latest CALSPEC files

CALSPEC contains the composite stellar spectra that are flux standards on the HST system. All files are in machine independent binary FITS table format. Information about the pedigree of a given spectrum is in the header of the FITS file, which can be read with the IRAF hedit task, by mrdfits.pro in IDL, or with the Python astropy.io.fits package. Table 1 below summarizes this set of standard star spectra. Columns 2-4 give the spectral type, V, and B-V magnitudes of the stars. Column 5 is the computer compatible file name with the plus and minus signs converted to underscores or omitted. Thus, the actual CALSPEC file name is the prefix in column 5, plus one of the suffixes in columns 6 or 7, plus ".fits". For example, a standard that has STIS data is "bd_28d4211_stis_005.fits". The naming convention in column 7 includes STIS, NIC, and WFC to indicate some contribution from any of the three HST spectrophotometers. Column 8 has the amount of variability found by Mullally et al. (2022), including upper limits. More documentation on some of these stars, including coordinates, finding charts and spectral types are in Turnshek et al. (1990), Bohlin, Colina & Finley (1995), and Colina & Bohlin (1997). 

The BOSZ grid for main-sequence (MS) stars (Bohlin et al. 2017) provides models for each G-A type MS standard, while a separate file includes the STIS and NICMOS fluxes concatenated with a BOSZ R=500 model for the extrapolation to 32 microns (col. 7). In 2023, Szabolcs Meszaros updated his BOSZ models because of improvements of the model atmosphere spectral synthesis codes and linelists (Meszaros et al. 2024). The original models are now replaced by updated calculations with the old R=300,000 model (col. 6) now replaced with R=200,000 resolution models that extend to a shorter wavelength limit of 500A. Starting in 2022 April, the BOSZ LTE models are replaced with R=5000 Hubeny NLTE models for stars hotter than 15,000K (Bohlin et al. 2022). Beginning in 2019, WFC3 IR grism fluxes (Bohlin & Deustua 2019) supplement the NICMOS spectral energy distributions (SEDs). Coordinates are in Table 2. The final two columns of Table 2 provide the complete long names needed for a successful Simbad search.

High-fidelity, non-BOSZ models at resolutions of less than R=200,000 are available for a few of the WD stars. The modeled IR fluxes for 13 stars are compared to Spitzer IRAC photometry in Bohlin et al. (2011). NLTE models for the three fundamental primary WD standards GD71, GD153, and G191B2B are updated by Bohlin, et al. (2020), where the pure H I models of Rauch (TMAP) are used for GD71 and GD153, while G191B2B has a Hubeny (TLUSTY) metal line-blanketed model. Wavelength dependent changes are <1% from 0.167-10 microns, and there is also a wavelength independent increase of 0.87% for all SEDs. A driving force for the 2020 update is greatly improved agreement to <1% at 0.15-30 microns between the TMAP and TLUSTY models. As a result of the changes to the primary standards, the flux calibration is revised, and all stars with STIS spectra and their models are updated as of 2020 March.

The observational spectra from column 7 can be compared with the column 6 models. The model calculations cover the long wavelength limits of 2.5 microns for NICMOS, 1.7 microns for WFC3, one micron for STIS and ACS, and ~30 microns for some JWST or Spitzer filters. The composite flux standard alpha_lyr_stis consists of a special Kurucz 9550K model from 900-1152A (Kurucz 2003), IUE data from 1152-1675A, STIS CCD fluxes from 1675-10200A (Bohlin & Gilliland 2004a), and the 9550K model longward of 10200A. The STIS and ACS observations of the SDSS standard BD+17 4708 are documented in Bohlin & Gilliland (2004b) and are extended to 2.5 microns with NICMOS data, while Bohlin & Landolt (2015) present the evidence for variability of BD+17 4708. All observed wavelengths are corrected to the heliocentric vacuum reference frame. Model wavelengths are corrected to the same vacuum heliocentric standard, if the radial velocity appears in Simbad. In summary, the files of choice for calibration of photometry and low resolution spectra are the *stis* files. For calibration of high resolution spectrometers, the R=200,000 *mod* files may be helpful.

CALSPEC also contains the ultraviolet to near-infrared absolute flux distribution of the Sun to 2.7 microns (filename: sun_reference_stis_002.fits). The solar reference spectrum combines absolute flux measurements from space and the ground with a model spectrum for the near-infrared (Colina, Bohlin, & Castelli 1996). The absolute solar flux measurements of Thuillier (2003) from 1990-23975A, the measurements of Coddington et al. (2021), and the solar model from the website of R. Kurucz are additional choices for a solar reference spectrum. The Kurucz model for the Sun (sun_mod_001.fits) at a spectral resolution of R=5000 is in column 6 of Table 1b, where the model has been normalized to the SED of Thuillier (2003) at 6800-7700A.

Also in the CALSPEC data directory are a combined FUSE and STIS high resolution spectrum of G191B2B (Rauch et al. 2013) g191b2b_stisfuse_001.fits and the covariance matrix WDcovar_002.fits for the uncertainty of the HST flux system as a function of wavelength. As explained in more detail by Bohlin et al. (2014), the uncertainty is relative to the reference wavelength of 5556A (5557.5 in vacuum), where this relative uncertainty is zero and the absolute uncertainty is 0.5% (Bohlin 2014, Bohlin et al. 2020).

Table 1a: Files of the Flux¹ Standards with Complete STIS Coverage

Access to the latest CALSPEC files

Table 1b: Files of the Flux¹ Standards with Incomplete STIS Coverage

* WARNING. Some wavelength coverage from sources of lower precision, such as IUE, FOS, or Oke.
** The STIS low resolution modes cover the 1140-10200A range, the NICMOS grism coverage is 0.8-2.5 micron, 
and the two WFC3 grisms cover 0.8-1.7 micron. See the *.fits headers for the exact source ranges.

(1)  The unit of flux in all files is erg s-1 cm-2 A-1.
(2)  Dust ring (Bohlin et al, 2011). Variable at +/- 0.01 mag (Pancino et al. 2012).
(3)  BD+17 4708 is variable (Bohlin & Landolt 2015).
(4)  BD+28 4211 has a companion at 2.8arcsec (Landolt & Uomoto (2007).
(5)  Dust ring (Bohlin et al, 2011).
(6)  Transiting Planet.
(7)  No single model fits well in the sense that an unresolved cool companion would explain the problem.
(8)  HZ43 and HZ3B are separated by ~3arcsec (BDC).
(9)  P041C has an M companion 0.57arcsec away (Gilliland & Rajan 2011)

Access to the latest CALSPEC files

Table 2:  Coordinates, Radial Velocity (km/s), Proper Motion (PM), and Alternate Star Names

Access to the latest CALSPEC files

History 

2024 June Update

In the previous delivery, the BOSZ model files have air wavelengths, instead of the proper vacuum wavelengths. This error is corrected, and the revised models are from Meszaros, S, et al. (2024). This CALSPEC delivery updates both the STIS and the model files for 55 stars.

2023 September Update

2023 April Update

George Rieke provided precision Ks photometry that improved the constraints on fitting BOSZ LTE models for seven stars (18sco, hd106252, hd205905, hd37962, ngc2506g31, p177d, and snap2.) Because of instabilities of the fits due to partial degeneracy of effective temperature and reddening, two stars required priors of E(B-V)=0.02 for 18sco and E(B-V)=0.08 for ngc2506g31. The Rieke Ks values for 16CygB, HD159222, and P330E are consistent to better than 1% with previous fits to their HST spectrophotometry.

2023 February Update

2022 October Update

The CTE correction for the STIS CCD data is updated (Bohlin & Lockwood 2022); and there are 18 new stars of which only two have models. There are a total of 69 new and revised stellar spectra files (*stis*), plus new or revised model files (*mod*) for 30 of the 69 stars. An existing SED is revised if there are substantial changes of >0.5% in the stellar flux. With the new formulation for the STIS CTE correction the STIS, ACS, and WFC3 flux measures are in accord at the 1% level, not only for the primary standards and other stars in the neighborhood of V=13 and brighter, but also now for fainter stars in the V=16 range. 

2022 April Update

2021 March Update

There are new pure helium Koester models for LDS749B and WD0308-565; both the model and the observational SEDs are new for these two stars. The 19 SEDs with WFC3 IR grism data are updated. The columns for FOS+Oke and IUE+Oke are deleted, and those SEDs are removed from current_calspec, because of lower precision than the robust set of modern HST data. Those demoted SEDs remain in the full CALSPEC directory.

2020 April Update

Three stars with more STIS observations and 12 new stars in the NGC6681 globular cluster are delivered. WD1057+719 now has full STIS coverage and is promoted from Table 1b to 1a. A preferred faint standard, WD1657+343 has a robust set of STIS observations along with WFC3 and NICMOS spectra, while WD1057+719 is deficient in G750L with only one exposure. The faint WD0308-565 has only two repeat verification observations, except for G230L with the desired minimum of three observations. The 12 faint NGC6681 stars are in a crowded field, and some have a few percent long-wavelength contamination from neighbors. These 12 stars appear in the finding chart of Proffitt, et al, (2003), Avila et al. (2019) and correspond to those labels a-m with the omission of star d. The Table 1a B and V values are actually HST/WFC3 F438W and F606W photometry from here.

2020 March Update

The most recent SEDs calculated by the TLUSTY and TMAP NLTE model atmosphere codes for the primary WDs G191B2B, GD153, and GD71 show improved agreement to 1% from 0.15 to 30 microns, in comparison to the previous 1% consistency only from 0.2 to 5 microns. These new NLTE models of hot WDs now provide the flux standards from the FUV to the mid-IR (Bohlin, Hubeny, and Rauch 2020, AJ, in press). The wavelength dependent change in the HST flux scale exceeds 1% only shortward of 0.167 and longward of 10 microns but does reach 2% at the short wavelength limit of STIS at 0.115 micron. In addition, there is a gray increase of 0.87% in all fluxes due to the increase of the 5556A (air) flux of Vega from 3.44 to 3.47 e-9 erg s^-1 cm^-2 A^-1.

2019 September Update

To improve the photometric precision, the standard extraction height is increased for the STIS G230LB and G430L gratings, making the default height 11 pixels for all five CCD modes. New STIS observations of Table 1b stars in Cycle 26 have promoted eta UMa, Feige 110, Feige 34, HD93521, HZ21, HZ4, and HZ44 from Table 1b to Table 1a. New stars for JWST calibration are 109 Vir, 16 CygB, 18 Sco, delta UMi, eta1 Dor, HD101452, HD115169, HD128998, HD142331, HD167060, HD2811, and HD55677. Models for 12 low-metallicity, Southern stars from program 12813 now have R=300,000 BOSZ models. There is one new WD, while Vega and Sirius change by 1-2%, because the correction for these highly saturated data is found to change with time (Bohlin, et al. 2019).

2019 January Update

New WFC3 IR grism SEDs are incorporated in revised SEDs for 17 stars and in two new WDs WD1327_083 and WD2341_322 (Bohlin & Deustua 2019 subm.) Two new stars with only STIS data are added: BD_11D3759 (Maíz Apellániz & Weiler 2018) and a faint WD SDSS132811 for ACS/SBC calibration.

2017 November Update

The R=5000 resolution solar model from the Kurucz website is added as sun_mod_001.fits.

2017 August Update

The STIS G430L region of GRW+70 5824 is upgraded from the poor single observation to an average of multiple visits. Because of the multitude of UV observation for monitoring the changing STIS sensitivity, GRW+70 5824 is now among the best choices of standard stars. The sparse sample of cool stars in CALSPEC motivated new STIS observations of KF08T3 (K1.5III) for combining with the existing NICMOS SED to make the new kf08t3_stisnic_001. The models for the JWST calibration are updated to include the computed model continuum.

2017 February Update

Thirty-two main sequence stars for JWST calibration are re-delivered after fitting BOSZ models to the STIS & NICMOS SEDs (Bohlin et al. 2017). The R=500 model provides the extrapolation of the data to 32 microns, while the pure R=300,000 model is also available here; both models are normalized to the STIS flux distribution at 6000-9000A.

2015 April Update

The wavelength coverage of P330E is extended to 2000A, and Table 2 is updated.

2015 January Update

The G750L STIS spectra require a wider extraction width of 11 pixels, resulting in flux changes of up to ~2% (Bohlin & Proffitt 2015).

2014 March Update

A STIS flux distribution for the primary IR standard Sirius (Bohlin 2014) is added to CALSPEC with an IUE extension to shorter wavelengths and with a special Kurucz model beyond 1 micron. The specially tailored Kurucz models for both Sirius and Vega, as normalized to STIS, are also now included. Bohlin (2014) reconciled the absolute visible and IR flux measures, resulting in a 5556A flux reduction from 3.46 to 3.44 e-9 erg s^-1 cm^-2 A^-1. Thus, all of the STIS and NICMOS SEDs are reduced by 0.6% at all wavelengths.

2013 December Update

Added coordinates for Reach et al. (2005) stars to Table 2.

2013 November Update

The low dispersion (R~500) CALSPEC database is expanded from 60 to 93 members with 19 new SEDs for JWST flux calibration (programs 12428 and 12682), 13 new southern standards (12813), and one partial flux distribution from the MAST HST archive. The new stars are measured with STIS and have preliminary extensions to 40 microns from model atmosphere fits. The basis for the wavelength dependence of the fluxes is switched from Hubeny pure hydrogen to Rauch model atmosphere calculations, and all dependence on ground based V magnitudes is removed (Bohlin, Gordon, & Tremblay 2014). This delivery encompasses an update for 48 stars with STIS or NICMOS spectrophotometry, the 32 new standards with complete STIS coverage, and eta UMa with archival STIS G230LB data. In addition, 8 high fidelity stellar models are new or updated. For the first time, high dispersion observations of G191B2B and an uncertainty covariance matrix are included.

2011 December Update

Calibration standard WD0308-565 derived from STIS data was added to the list of CALSPEC data.

2010 February Update

The 2010 February CALSPEC delivery included six revised models, 22 revised standard star flux distributions, and four new SEDs from STIS observations. HZ43 was demoted from its primary standard category some years ago because of its bright cool companion. The revisions to the flux distributions are all generally less than 1%. The main change was driven by the discovery of a small error of ~0.57% in the gain 4 to gain 1 ratio for the STIS CCD (Goudfroiij, et al. 2009). The weighting for the STIS sensitivity is changed from the total counts for each observation to equal weight for each of the three primary standard WDs, regardless of the number of observations. A few stars have additional STIS observations included in their observed average SED. The NICMOS dispersion relations have been updated (Pirzkal, Bohlin, & Thatte 2009). For seven G stars (Bohlin 2010), the observed flux distributions have been extended from the NICMOS long wavelength limit of ~2.5 microns to 40 microns, using the best fitting Castelli & Kurucz (2004) model SED.

2009 January Update

The 2009 January CALSPEC delivery encompassed six new models and 22 standard star flux distributions. The switch to TLUSTY203 caused a small change in the continuum fluxes with a 0.3% increase in the 1300A/10000A flux ratio. In a narrow region around 4000A between H-delta and H-epsilon, the new models are up to ~0.9% fainter. The correction of a V bandpass error made all stars fainter by a constant 0.35%, independent of wavelength. Pure hydrogen NLTE models are provided for the first time for the faint stars WD1057+719 and WD1657+343; and these models include a small amount of interstellar reddening (Bohlin 2007). There is one new standard, HD209458, delivered to CALSPEC for the first time. The G230LB STIS spectra of G191B2B have replaced the FOS fluxes at 1700-3850A, and new NICMOS observations are included for many stars.

UPDATES: 1996Feb22, 1999Jul19, 2001Jun5, 2004Feb27, 2006Sep1, 2007Oct LDS749B (Bohlin & Koester 2008), 2008Jul Bohlin & Cohen (2008), 2009Jan, 2010Feb, 2011Dec WD0308-565, 2013Nov, 2013Dec, 2014Mar, 2015Jan, 2015Apr, 2017Feb, 2017Aug, 2017Nov, 2019Jan, 2019Sep, 2020Mar, 2020Apr, 2021Mar, 2022Apr, 2022Oct, 2023Feb, 2023Apr, 2023Sep..

PLEASE report errors or send comments to bohlin@stsci.edu. 

Thanks to Pierre Maxted for a set of corrections.

This page made use of the SIMBAD database, operated at CDS, Strasbourg, France.

References

Avila, R. J., et al. 2019, Instrument Science Report, ACS 2019-05, (Baltimore:STScI)
Bohlin, R. C., & Lockwood, S. 2022, Instrument Science Report, STIS 2022-07, (Baltimore:STScI)
Bohlin, R. C., Krick, J. E., Gordon, K. D., and Hubeny, I. 2022, AJ, 164, 10
Bohlin, R. C., Hubeny, I., & Rauch, T. 2020, AJ, 160, 21
Bohlin. R. C., & Deustua, S. E. 2019, AJ, 157, 229.
Bohlin, R. C., Deustua, S. E., and de Rosa 2019, AJ, 158, 211
Bohlin, R. C., Meszaros, Sz., Fleming, S., Gordon, K. D., Koekemoer, A. M., & Kovacs, J. 2017, AJ, 153, 234
Bohlin, R. C., & Landolt, A. U. 2015, AJ, 149, 122
Bohlin, R. C., & Proffitt, C. R. 2015, Instrument Science Report, STIS 2015-01, (Baltimore:STScI)
Bohlin, R. C. 2014, AJ, 147, 127
Bohlin, R. C., Gordon, K. D., & Tremblay, P.-E. 2014, PASP, 126, 711
Bohlin, R. C., et al. 2011, AJ, 141, 173
Bohlin, R. C. 2010, AJ, 139, 1515
Bohlin, R. C., & Cohen, M. 2008, AJ, 136, 1171
Bohlin, R. C., & Koester, D. 2008, AJ, 135, 1092
Bohlin, R. C. 2007, in The Future of Photometric, Spectrophotometric, and Polarimetric Standardization, ASP Conf. Series, 
    Vol. 364, p. 315 ed. C. Sterken
Bohlin, R. C., & Gilliland, R. L. 2004a, AJ, 127, 3508 
Bohlin, R. C., & Gilliland, R. L. 2004b, AJ, 128, 3053
Bohlin, R. C., Dickinson, M. E., & Calzetti, D. 2001, AJ, 122, 2118 (BDC)
Bohlin, R.C. 2000, AJ, 120, 437
Bohlin, R.C. 1996, AJ, 111, 1743
Bohlin, R.C., Colina, L., & Finley, D.S. 1995, AJ, 110, 1316
Castelli, F., & Kurucz, R. 2004, IAU Symp. No. 210, Modeling of Stellar Atmospheres, eds. N. Piskunov, W. Weiss, 
    & D. Gray 2003, poster A20; also arXiv:0405.087 [astro-ph
Coddington, O. M., Richard, E. C., Harber, D. et al. 2021, Geophys. Res. Let., 48, 9170
Cohen, M., Megeath, S.~T., Hammersley, P.~L., Martin-Luis, F., & Stauffer, J. 2003, AJ, 125, 2645
Colina, L., & Bohlin, R. 1994, AJ, 108, 1931
Colina, L., & Bohlin, R. 1997, AJ, 113, 1138
Colina, L., Bohlin, R.C., & Castelli, F. 1996, AJ, 112, 307
Gilliland, R., & Rajan, A. 2011, Instrument Science Report, WFC3 2011-03, (Baltimore:STScI)
Gordon, K. D., et al. 2023, ApJ, 950, 86
Goudfrooij, P., Wolfe, M. A., Bohlin, R. C., Proffitt, C. R., & Lennon, D. J. 2009, Instrument Science Report, 
    STIS 2009-02, (Baltimore:STScI)
Kurucz, R. 2003, and all references to Kurucz website, http://kurucz.harvard.edu/
Landolt, A., & Uomoto, A. 2007, AJ, 133, 768
Maíz Apellániz, J., & Weiler, M. 2018, A&A, 619, 180
Megessier, C. 1995, A&A, 296, 77
Meszaros, S, et al. 2024, A&A, in press
Mullally, S. E., et al. 2022, AJ, 163, 136
Narayan, G., et al. 2019, ApJS, 241, 20
Oke, J.B. 1990, AJ, 99, 1621
Pancino, E., et al. 2012, MNRAS, 426, 1767
Pirzkal, N., Bohlin, R.,  & Thatte D. 2009, Instrument Science Report, NICMOS 2009-006, (Baltimore:STScI)
Proffitt, C. R., Brown, T. M., Mobasher, B., & Davies, J. 2003, Instrument Science Report, STIS 2003-01, 
    (Baltimore:STScI)
Rauch, T., Werner, K., Bohlin, R., & Kruk, J. 2013, A&A, 560, A106
Reach, W. T., et al. 2005, PASP, 117, 978
Rieke, G. H., et al. 2008, AJ, 135, 2245 
Thuillier, G., Herse, M., Labs, D., Foujols, T., Peetermans, W., Gillotay, D., Simon, P., & Mandel, H. 2003, 
    Sol. Phys., 214, 1
Turnshek, D.A., Bohlin, R.C., Williamson, R., Lupie, O., Koornneef, J., & Morgan D. 1990, AJ, 99, 1243

Back to the Astronomical Catalogs Page

Last modified, June 2024