Computing Natural Magnitudes in the Photometric Systems of Astronomical Plates using Gaia DR3 SEDs

TL;DR

The paper tackles the challenge of calibrating historical photographic plates by bridging spectral mismatches with modern catalogs. It replaces the traditional color-term approach with Gaia DR3 BP/RP XP spectra-based synthetic photometry to compute natural plate magnitudes, incorporating atmospheric reddening and optical transmission. The authors demonstrate that the conventional color-term method introduces sizable systematic errors (often >0.3 mag) and that the Gaia-based, SED-driven approach yields more consistent magnitudes, while also enabling discrimination of luminosity classes (dwarfs vs giants) and improved handling of air-mass effects. This approach promises more accurate cross-era photometry and can be integrated into tools like PyPlate to enhance calibration across plate archives such as APPLAUSE and DASCH, albeit contingent on comprehensive plate sensitivity data for all emulsions.

Abstract

Context. Accurate photometric calibration of astronomical photographic plates remains a fundamental challenge in astronomy, especially when bridging historical photographic data with modern observations due to the mismatch of spectral sensitivities of photographic plates and passbands of modern calibration catalogs. Aims. We intend to derive consistent natural magnitudes for celestial sources within the intrinsic photometric systems of astronomical photographic plates by using Gaia Data Release 3 (DR3) blue photometer (BP) and red photometer (RP) low-resolution spectral data and to show its superiority to former methods. Methods. We compiled spectral characteristic data for emulsions and filters applied in photometric observations using glass plates. The collected color sensitivities, modified by atmospheric reddening depending on the air mass, are then used to compute accurate natural magnitudes and fluxes of objects in the photographic plates through synthetic photometry, utilizing a catalog of Gaia spectral energy distributions (SEDs) over the wavelength range 330 nm to 1050 nm (XP spectra). This process uses GaiaXPy, a Python library designed to handle Gaia DR3 spectral data. These natural magnitudes are then compared with results from the color term method used to compile the data in existing photoplate archives. Results. Comparing the synthetic magnitudes with those existing in the Archives of Photographic PLates for Astronomical USE (APPLAUSE), we were able to reveal systematic errors of the existing data in the range of +/-0.3 mag and higher. In addition, the presented method allows for an accommodation of stars with similar color index but of different luminosity classes as well as an effective correction of atmospheric reddening at higher air masses, approximately 0.2 mag.

Figures (9)

• Figure 1: The color terms derived from a series of plates from two data releases of APPLAUSE archive for the randomly selected constant star UCAC4 600-107181. Left: Data release 3 using the Tycho-2 and APASS DR9 as a reference catalog. Right: Data release 4 using the Gaia EDR3 as a reference catalog.
• Figure 2: Time series of the constant star UCAC4 600-107181 (color index $B-V = 0.39\ \mathrm{mag}$) from 1 181 scans of 606 photographic plates, over 32 years, covering data from blue sensitive (pg, shown as blue dots) and yellow and red sensitive (pv, shown as red dots) plates from APPLAUSE DR3. Plates were classified based on known emulsions and reliable data (see appendix \ref{['sec:appendix:A']} for details about the plates used in this figure). Upper panel: Series of color terms determined by RMS fitting. Lower two panels: Calibrated $B$ magnitudes of the star, differing in that panel 3 includes 80 plates of the two emulsions Kodak IIa-O and Kodak 103a-D, exposed in the Lippert-Astrograph, only. The calibrated $V$ and $B$ magnitudes are identical except for a shift of the color index.
• Figure 3: Spectral sensitivities of various pg and pv plates coated with different emulsions are shown relative to the Gaia photometric system. The green areas represent the plate's emulsion color responses, which already include the atmospheric transmission data, using the air mass 1.5 to 0 ratio, while the dashed blue and solid red lines indicate Gaia's blue and red photometer sensitivities, respectively.
• Figure 4: The low-resolution spectrum of UCAC4 600-107181 is compared with two spectral sensitivity curves of Kodak emulsions with Schott filters for photographic photometry. The Gaia SED is shown with a black solid line (right axis). The blue-filled plot with the dashed line represents the spectral sensitivity of the Kodak IIa-O emulsion combined with the transmission of a 2 $\mathrm{mm}$ Schott GG13 (=385) filter, which closely matches the Johnson-Cousins B band (blue-sensitive plate). The red-filled plot with dotted line illustrates Kodak 103a-D emulsion plus a 2 $\mathrm{mm}$ Schott GG11 filter, which closely matches the Johnson-Cousins V band (visible-light-sensitive plate). The influence of atmospheric transmission was taken into account for both spectral sensitivity curves at zenith (air mass 1). All spectral curves here have been normalized to their individual maximum.
• Figure 5: Difference $\Delta = \mathrm{m}_\mathrm{synth} - m_\mathrm{cat}$ versus color index $\mathrm{color\_bv} = (B-V)$ for all in APPLAUSE DR3 identified, cleaned and with Gaia DR3 matched stars. Left: plate 12368, Großer Schmidt-Spiegel, Hamburg Observatory, 12.12.1968, Kodak IIa-O with filter GG13(=385), scan 14850, color term 0.83. Right: plate 11626, Großer Schmidt-Spiegel, Hamburg Observatory, 05.10.1964, Kodak 103a-D with filter GG11, scan 14057, color term 0.01. We used an air mass of 1.25 for the analysis in both cases. These plates in APPLAUSE have two scanned files with different color terms, producing two distributions. To keep this analysis figure simple, only one distribution for each such plate is shown.