No Arabic abstract
The Kepler spacecraft observed a total of only four AM Herculis cataclysmic variable stars during its lifetime. We analyze the short-cadence K2 light curve of one of those systems, Tau 4 (RX J0502.8+1624), which underwent a serendipitous jump from a low-accretion state into a high state during the final days of the observation. Apart from one brief flare, there was no evidence of accretion during the 70 d of observations of the low state. As Tau 4 transitioned into a high state, the resumption of accretion was very gradual, taking approximately six days (~90 binary orbits). We supplement Tau 4s K2 light curve with time-resolved spectroscopy obtained in both high and low states of accretion. High-excitation lines, such as He II 468.6 nm, were extraordinarily weak, even when the system was actively accreting. This strongly suggests the absence of an accretion shock, placing Tau 4 in the bombardment regime predicted for AM Herculis systems with low accretion rates. In both the high-state and low-state spectra, Zeeman absorption features from the white dwarfs photosphere are present and reveal a surface-averaged field strength of $15pm2$ MG. Remarkably, the high-state spectra also show Zeeman-split emission lines produced in a region with a field strength of $12pm1$ MG. Zeeman emission has not been previously reported in an AM Herculis system, and we propose that the phenomenon is caused by a temperature inversion in the WDs atmosphere near the accretion region.
We present photometry of the intermediate polar FO Aquarii obtained as part of the K2 mission using the Kepler space telescope. The amplitude spectrum of the data confirms the orbital period of 4.8508(4) h, and the shape of the light curve is consistent with the outer edge of the accretion disk being eclipsed when folded on this period. The average flux of FO Aquarii changed during the observations, suggesting a change in the mass accretion rate. There is no evidence in the amplitude spectrum of a longer period that would suggest disk precession. The amplitude spectrum also shows the white dwarf spin period of 1254.3401(4) s, the beat period of 1351.329(2) s, and 31 other spin and orbital harmonics. The detected period is longer than the last reported period of 1254.284(16) s, suggesting that FO Aqr is now spinning down, and has a positive $dot{P}$. There is no detectable variation in the spin period over the course of the K2 observations, but the phase of the spin cycle is correlated with the system brightness. We also find the amplitude of the beat signal is correlated with the system brightness.
We present an analysis of K2 short cadence data of 34 M dwarfs which have spectral types in the range M0 - L1. Of these stars, 31 showed flares with a duration between $sim$10-90 min. Using distances obtained from Gaia DR2 parallaxes, we determined the energy of the flares to be in the range $sim1.2times10^{29}-6times10^{34}$ erg. In agreement with previous studies we find rapidly rotating stars tend to show more flares, with evidence for a decline in activity in stars with rotation periods longer than $sim$10 days. The rotational modulation seen in M dwarf stars is widely considered to result from a starspot which rotates in and out of view. Flux minimum is therefore the rotation phase where we view the main starspot close to the stellar disk center. Surprisingly, having determined the rotational phase of each flare in our study we find none show any preference for rotational phase. We outline three scenarios which could account for this unexpected finding. The relationship between rotation phase and flare rate will be explored further using data from wide surveys such as NGTS and TESS.
We analyse the K2 short cadence data of the intermediate polar FO Aqr and provide accurate and updated orbital and spin periodicities. We additionally find small spin period changes as a function of orbital phase of ~0.02 seconds translating to velocities of ~ a few km/s. The obtained orbital-folded velocity profile displays two clear maxima and minima, and cannot be explained by the radial velocity of the orbiting white dwarf. Instead we propose that the observed velocities are the sum of the radial velocities of both the white dwarf and of the stellar surface facing the white dwarf which reprocesses the WD spin pulses. This combination can explain the observed low velocities in FO Aqr. However asymmetries in the orbital configuration are required to explain the double peaked velocity profile. One possible scenario would invoke binary eccentricity. We thus developed a simple binary model to explain and fit our observations, and find a small binary eccentricity of e=0.03. Although small, persistent eccentricity in a close interacting binary would induce enhanced mass transfer occurring preferentially at periastron passages. We thus discuss alternative scenarios where other asymmetries might explain our observations assuming circular orbits. Since FO Aqr is the first system where the combined radial velocities of both the WD and secondary surface have been measured, it is possible that other mass-transferring binaries also display similar velocity curves when observed with Kepler. These will provide additional valuable tests to either confirm or rule out small eccentricities in similar systems.
Using light curves obtained by the K2 mission, we study the relation between stellar rotation and magnetic activity with special focus on stellar flares. Our sample comprises 56 bright and nearby M dwarfs observed by K2 during campaigns C0-C18 in long- and short-cadence mode. We derive rotation periods for 46 M dwarfs and measure photometric activity indicators such as amplitude of the rotational signal, standard deviation of the light curves, and the basic flare properties (flare rate, flare energy, flare duration, and flare amplitude). We found 1662 short-cadence flares, 363 of which have a long-cadence counterpart with flare energies of up to $5.6cdotp10^{34}$erg. The flare amplitude, duration, and frequency derived from the short-cadence light curves differ significantly from those derived from the long-cadence data. The analysis of the short-cadence light curves results in a flare rate that is 4.6 times higher than the long-cadence data. We confirm the abrupt change in activity level in the rotation-activity relation at a critical period of ~10d when photometric activity diagnostics are used. This change is most drastic in the flare duration and frequency for short-cadence data. Our flare studies revealed that the highest flare rates are not found among the fastest rotators and that stars with the highest flare rates do not show the most energetic flares. We found that the superflare frequency ($Egeq5cdotp10^{34}$erg) for the fast-rotating M stars is twice higher than for solar like stars in the same period range. By fitting the cumulative FFD, we derived a power-law index of $alpha=1.84 pm 0.14$, consistent with previous M dwarf studies and the value found for the Sun.
We present the results of a search for stellar flares from stars neighbouring the target sources in the Kepler short cadence data. These flares have been discarded as contaminants in previous surveys and therefore provide an unexplored resource of flare events, in particular high energy events from faint stars. We have measured M dwarf flare energies up to 1.5$times$10^35 erg, pushing the limit for flare energies measured using Kepler data. We have used our sample to study theflaring activity of wide binaries, finding that the lower mass counterpart in a wide binary flares more often at a given energy. Of the 4430 flares detected in our original search, 298 came from a neighbouring star, a rate of 6.7$pm$0.4 per cent for the Kepler short cadence lightcurves. We have used our sample to estimate a 5.8$pm$0.1 per cent rate of false positive flare events in studies using TESS short cadence data.