Intermediate-Mass Stripped Stars in the Magellanic Clouds: Forward Modeling the Observed Population Discovered Via UV Excess

TL;DR

We address the poorly constrained population of intermediate-mass stripped stars in the Magellanic Clouds ($1-8\,M_\odot$) and develop a forward-modeling framework to interpret UV-excess sources from the SUMS survey. By combining binary population synthesis grids (anchored to spatially resolved star-formation histories) with a 3D dust extinction model and a realistic Swift-UVOT photometry pipeline, we quantify the completeness and purity of UV-excess detections and predict the true underlying population. Our forward model predicts $\sim1100$ and $\sim2600$ stripped stars in the SMC and LMC, respectively, with a mock-observed recovery rate of $\sim9-13\%$ (before footprint correction) and $\sim13-45\%$ of intrinsic UV-excess systems recovered after accounting for selection and crowding. We find that UV-excess selection preferentially picks stripped stars with compact-object companions and that MS contamination and crowding can significantly bias the observed sample; the results provide a quantitative framework to test binary evolution scenarios and to compare observed and simulated stripped-star demographics in a realistic way.

Abstract

Stripped stars are hot, helium-rich stars formed when binary interactions remove a star's hydrogen envelope. While low-mass ($\lesssim 1\,M_\odot$) and high-mass ($\gtrsim 8\,M_\odot$) stripped stars are well studied as hot subdwarfs and Wolf-Rayet stars, their intermediate-mass counterparts ($1-8 \,M_\odot$) have only recently been discovered. The Stripped-Star Ultraviolet Magellanic Cloud Survey (SUMS) identified UV-excess sources in the Magellanic Clouds using Swift-UVOT photometry and selected 820 photometric stripped-star candidates. However, the selection function, completeness, and purity of this sample remain poorly understood. We forward model the population of stripped stars in the Magellanic Clouds using a binary population synthesis model combined with spatially resolved star formation histories and simulated UV photometry. To assess survey sensitivity, we inject simulated sources into real Swift-UVOT images and reproduce the SUMS selection process, including crowding, extinction, and photometric quality cuts. We recover $\sim 250$ simulated stripped stars with masses $\gtrsim 1\,M_{\odot}$, which corresponds to a recover rate of $9-13\,\%$. The rest are missed due to dilution by luminous companions, crowding, and high extinction. The observed population is biased toward systems with low-mass companions formed by common envelope evolution and toward systems with compact object companions. Of the stripped stars which show UV excess, $25-45\,\%$ are identifiable by SUMS; higher-resolution data or improved reddening corrections are needed to detect the rest. We predict contamination of the observed stripped star candidates by main-sequence stars with spurious UV excess due to crowding and provide guidelines for selecting higher-purity subsamples. These results enable tests of binary evolution models and realistic comparison of observed and simulated stripped-star populations.

Figures (20)

• Figure 1: Observed stripped star candidates in the SMC (left) and LMC (right). Top: spatial distribution of the spectroscopically-confirmed sources [orange;][]drout_2023 and photometric candidates [gray;][]ludwig_2025. Bottom: one of the nine UV-optical CMDs used to assess UV excess. UV and optical photometry is de-reddened assuming a uniform extinction of $A_V = 0.22$ mag for the SMC and $A_V = 0.38$ mag for the LMC. Dashed line indicates the theoretical ZAMS.
• Figure 2: An example star formation history for an LMC subregion, where $t=0$ represents today. The dashed line indicates the star formation rate from Mazzi_2021. Star markers show the birth year of stripped stars alive today. The size of the star markers correspond to the masses of the stripped stars at $t = 0$.
• Figure 3: UV-optical CMD for a subset of synthetic stripped stars in the SMC. Dark blue markers indicate the stripped stars, light blue markers indicate the companions, and gray markers indicate the unresolved sources. The gray line connects each stripped star to its companion and unresolved source. The dashed line indicates the theoretical ZAMS. About 1/3 of the unresolved sources falls bluewards of the ZAMS and thus are UV-excess sources.
• Figure 4: Spatial distribution of the synthetic stripped-star binaries in the SMC (left) and LMC (right). Each box represents a VMC subregion that rubele_2018 and Mazzi_2021 derived an average star formation history for. Gray markers indicate young MS stars from MCPS. Blue/pink markers indicate simulated stripped-star binaries. The black line shows the footprint of the SUMaC images used to perform UV photometry in SUMS.
• Figure 5: Average extinction curves for the SMC Gordon_2024 and LMC Gordon_2003. The gray lines show the transmission curves for the optical ($U$, $B$, $V$, $I$) and UV ($UVW2$, $UVM2$, $UVW1$) bandpasses.