Benchmarking differential reddening in front of globular clusters

TL;DR

Interstellar extinction and differential reddening hinder accurate stellar parameter determinations for globular clusters near the Galactic plane. The authors propose a cross-calibration approach that combines 3D reddening maps with Gaia DR3 RR Lyrae data to set absolute reddening zero-points for relative differential reddening maps, demonstrated on M9. They find that Bayestar 2019 provides a plausible foreground reddening ($E(B-V) ext{ in the range } [0.33,0.38]$ mag) while the SFD-based zeros tend to overestimate extinction, enabling a physically motivated reddening window of $0.33 \\le E(B-V) \\le 0.38$ mag. The method is extensible to other Gaia-accessible clusters and upcoming surveys like the Rubin Observatory, enabling more robust dereddening of globular cluster CMDs and improved stellar parameter determinations.

Abstract

Interstellar extinction is a major obstacle in determining accurate stellar parameters from photometry near the Galactic disk. It is especially true for globular clusters at low galactic latitudes, which suffer from significant amounts of, and spatially variable reddening. Although differential reddening maps are available for tens of clusters, establishing and validating the absolute zero point of relative maps is a challenge. In this study, we present a new approach to determine and evaluate absolute reddening zero-points for Galactic globular clusters by combining three-dimensional reddening maps with Gaia DR3 RR Lyrae data. As a first case study, we investigate the low-latitude globular cluster M9. We compare the Gaia photometry and color data of the cluster member RR Lyrae stars to field RR Lyrae stars with accurate parallaxes and whose photometric metallicities match that of M9, as well as to theoretical models. We calculate the dereddened Gaia colors for the M9 stars based on three zero points. We confirm that the original SFD map appears to be overcorrecting the reddening for at least some RR Lyrae stars, albeit not excessively. In contrast, the 3D Bayestar map and the recalibrated version of the SFD map provide physically plausible reddenings, which we accept as lower and upper limits for M9, respectively. Our results provide a physically motivated reddening range for M9, and outline a methodology that can be directly extended to other globular clusters that are accessible to the Gaia mission, and to other multicolor sky surveys, such as the Rubin Observatory.

Figures (4)

• Figure 1: Differential reddening maps for M9. Left: the recreated field of view using the three-dimensional Bayestar 2019 reddening map Green2019. Middle: the relative reddening map, as published by AlonsoGarcia2012. Right: we degraded the resolution of the AlonsoGarcia2012 map (AG12) to the same resolution as the Bayestar pixels, as can be seen here, then compared them to set the alternative absolute zero-point for the relative map.
• Figure 2: Bailey diagram for the Gaia DR3 SOS RR Lyrae catalog Clementini2023. It uses the recalculated photometric metallicities from Muraveva-2025 and also shows our filtered RRab (purple and red) and RRc (light blue) sample stars. The Oosterhoff separation is indicated by a dashed black line based on Luongo-2024. Purple dots mark Oosterhoff I-type RRab stars, while red dots correspond to Oosterhoff II.
• Figure 3: Phase folded light curves of the M9 RRL sample, based on Gaia epoch photometry and variability periods. Red, green and blue points correspond to $G_{\rm RP}$, $G$ and $G_{\rm BP}$ passbands, respectively. Note that for V2 and V10, the phase coverage of the red and blue bands are very poor, while for V2, V10, V18 and V19, the $G_{\rm BP}$ data are very close to, or above the $G$ band points. In both cases, they lead to incorrect estimates of the mean magnitude.
• Figure 4: The RRL CMD at the metallicity of M9. Small points represent the Gaia sample, with blue, purple and red corresponding to RRc, OoI RRab and OoII RRab stars, respectively. The origins of low-luminosity points are discussed in the text, and the dotted square corresponds to the observed RRL distribution of GCs Cruz_Reyes-2024. The solid and dashed lines indicate the Marconi2015 instability strip (IS) edges, for RRc (blue) and RRab (red) models. The dashed-dotted light blue and orange lines represent the outer edges of the RRL IS published by Marconi-2003 and DiCriscienzo-2004. Large star symbols are the M9 RR Lyrae stars for which we found the Gaia color data to be accurate. In the left and right panels we show these M9 stars using the SFD 1998 (empty), S&F 2011 (filled) and Bayestar 2019 reddening zero points, respectively. The Bayestar and the S&F 2011 values places the RR Lyrae stars within physically plausible limits, serving as reasonable bounds for the reddening in front of M9, whereas the SFD 1998 zero point shifts one of the RRab stars into the RRc instability strip, supporting the idea that it causes an overestimation.