Statistical Signatures of Nanoflare Activity. II. A Nanoflare Explanation for Periodic Brightenings in Flare Stars Observed by NGTS

Abstract

Several studies have documented periodic and quasi-periodic signals from the time series of dMe flare stars and other stellar sources. Such periodic signals, observed within quiescent phases (i.e., devoid of larger-scale microflare or flare activity), range in a period from 1 to 1000 s and hence have been tentatively linked to ubiquitous p-mode oscillations generated in the convective layers of the star. As such, most interpretations for the observed periodicities have been framed in terms of magnetohydrodynamic wave behavior. However, we propose that a series of continuous nanoflares, based upon a power-law distribution, can provide a similar periodic signal in the associated time series. Adapting previous statistical analyses of solar nanoflare signals, we find the first statistical evidence for stellar nanoflare signals embedded within the noise envelope of M-type stellar lightcurves. Employing data collected by the Next Generation Transit Survey (NGTS), we find evidence for stellar nanoflare activity demonstrating a flaring power-law index of 3.25 ± 0.20, alongside a decay timescale of 200 ± 100 s. We also find that synthetic time series, consistent with the observations of dMe flare star lightcurves, are capable of producing quasi-periodic signals in the same frequency range as p-mode signals, despite being purely composed of impulsive signatures. Phenomena traditionally considered a consequence of wave behavior may be described by a number of high-frequency but discrete nanoflare energy events. This new physical interpretation presents a novel diagnostic capability, by linking observed periodic signals to given nanoflare model conditions.

NGTS Observations of Enhanced Nanoflare Activity in Fully Convective Stars Linked to Plasma Resistivity

“…We present two example histograms of statistical intensity fluctuations in Figure 1 The signatures of the M4V star shown in Figure 1 are consistent with previous positive stellar nanoflare identifications 8 . The distinct increase of visible nanoflare signatures within the expected regime of full convection is a promising sign that the enhanced nanoflare rates are related to the underlying convective nature of the star.…”

“…Interestingly, the M4V stars display evidence for longer e-folding timescales and larger power-law indices when compared to their M2.5V and M3V counterparts. As discussed above, this is likely an explanation for the relatively constant ζ values throughout the spectral sub-type range, whereby the statistical effects of higher power-law indices are negated by the slower decay timescales associated with those stars, which is consistent with previous solar and stellar case studies 8,14 .…”

“…To ensure consistency with previous stellar nanoflare investigations, the filtering steps employed were identical to those used in 8 , with the filtered lightcurves subsequently cropped to 97 060 datapoints each to match the number statistics from the previous study. This allows a direct comparison to be made with the work of 8 , since the previously published nanoflare simulations could be re-used due to identical number statistics, filtering techniques, desired α (power-law index) and τ (e-folding time) ranges, in addition to specific NGTS-modelled noise characteristics.…”

“…The same normalisation steps applied to the observational data were applied to the modelled timeseries, i.e., 3 σ threshold 'flare' subtraction, low order polynomial detrending, and the z-score normalisation. These Monte-Carlo models were first developed for previous stellar nanoflare analysis 8 , and fully described therein. The nanoflare modelling techniques were originally adapted from 2-D Solar nanoflare modelling code, which have been documented extensively 14 .…”

“…The periodic signatures manifest as features present in Fourier spectral energy densities, namely spectral slopes, an associated 'turning-point' where the spectral slope begins, and clear peaks in the range 1 − 10 mHz. These techniques do not rely on the individual identification of nanoflare events, so can identify nanoflare activity in seemingly quiescent time series 8 .…”