A multi-wavelength study of the 2025 low state of the intermediate polar BG CMi

TL;DR

This study catalogs the first recorded low state of BG CMi through simultaneous XMM-Newton X-ray/optical data and supporting multi-wavelength observations, revealing a shift in accretion geometry. Timing analyses show a transition from spin-dominated optical variability to orbital-sideband-dominated signals and, in X-rays, a shift toward $2\Omega$ and other beat frequencies, consistent with disk-overflow accretion. Spectral modeling with cooling-flow emission and complex absorbers indicates a drastic drop in local absorption during the low state, while the unabsorbed X-ray flux remains similar or slightly higher, implying sustained WD accretion despite disk dissipation. The combined timing and spectral results argue for a partial disk mass loss leading to disk-overflow, with no evidence for a complete cessation of accretion or a full switch to stream-fed accretion, and they provide a cohesive timeline for the 2025 low state with implications for accretion mode switching in intermediate polars.

Abstract

We present multi-wavelength observations of the first recorded low state of the intermediate polar BG CMi. Optical monitoring of the source by members of the American Association of Variable Star Observers reveals a decrease of ~0.5 mag that lasted ~50 d in early 2025. During the low state the optical timing properties imply that BG CMi underwent a change in the accretion mode, as power at the spin frequency $ω$ dramatically dropped. An XMM-Newton observation revealed a substantial decrease in intrinsic absorption and a slight increase in intrinsic X-ray luminosity, compared to archival Suzaku data. Timing analysis of the X-ray light curves shows that power shifted from the orbital frequency $Ω$ (prominent in Suzaku data) to $2Ω$ in the low state XMM-Newton data, along with the strengthening of certain orbital sidebands. We suggest that BG CMi transitioned to disk-overflow accretion, where the white dwarf accreted matter via both a disk and a stream, the latter becoming more dominant during the low state due to a decrease in the mass and size of the disk.

Figures (10)

• Figure 1: Upper Panel: XMM-Newton/EPIC-pn 0.2--10 keV light curve of BG CMi with 100 s time resolution. Middle Panel: XMM-Newton/OM $V$-band light curve of BG CMi with 100 s time resolution. :Lower Panel: Ground-based optical light curves of BG CMi. Filters and telescopes are denoted in the legend. Some light curves have been offset for clarity. All light curves have been corrected to the solar system barycenter and thus times are given in Barycentric Modified Julian Date (BMJD).
• Figure 2: Upper Panel: Long-term AAVSO light curve of BG CMi from 2020 to 2025. Observations were taken in a clear filter and calibrated using a $V$-band zero-point ($CV$-band). Lower Panel: A zoomed in portion of the AAVSO light curve indicated by the red box in the upper panel and highlighting the low state of the source. Colors represent the epochs that we have broken the light curves down into for timing analysis. Though the first four epochs represent individual observing seasons, the low state is split into multiple epochs. In both panels, the purple line represents the time of the XMM-Newton observing campaign.
• Figure 3: Upper Panel: Lomb-Scargle periodogram of the XMM-Newton EPIC/pn 0.2--10 keV light curve of BG CMi. Middle Panel: Lomb-Scargle periodogram of the XMM-Newton EPIC/OM light curve of BG CMi. Lower Panel: Lomb-Scargle periodogram of the non-overlapping ground-based optical light curve of BG CMi. In all panels, vertical dashed lines represent known periodicities in the system, including the orbital and spin frequencies ($\Omega$ and $\omega$, respectively) and their multiples, as well as beat and sideband frequencies. $99.9\%$ significance thresholds for the Lomb-Scargle power are shown as horizontal dot-dashed lines in each panel.
• Figure 4: Lomb-Scargle periodograms of the AAVSO light curves of BG CMi. The colors match the epochs defined in Fig. \ref{['fig:AAVSO_LC']}, and the relevant MJD ranges are labeled in each panel. The vertical dashed lines represent known periodicities related to the system and the dot-dashed horizontal lines in each panel indicated the $99.9\%$ confidence level.
• Figure 5: Lomb-Scargle periodogram of the 2009 Apr 11 Suzaku 0.2--12 keV light curves of BG CMi. The vertical dashed lines represent known periodicities related to the system, with undetected periods shown in grey at the top of the figure. The dot-dashed horizontal line represents the $99.9\%$ confidence level.