Young M dwarfs flare activity model: Towards better exoplanetary atmospheric characterisation

TL;DR

The study addresses how energy is deposited during flares on young M dwarfs and how this impacts exoplanetary atmospheres. It develops the YMDF framework by combining RADYN-based radiative-hydrodynamic models with RH1.5D radiative transfer to produce time-dependent, panchromatic flare spectra, represented as a linear combination of two atmosphere models parameterized by hat{X}_{1} and hat{X}_{2}. The results show that YMDF can reproduce the TESS continuum rise and FUV-A emission and that flare energy distributions in simulated populations are well described by a broken power law with α_1 ≈ 1.5 and α_2 ≈ 2, aligning with multi-wavelength observations of AU Mic and related stars. This framework enables generation of synthetic, time-resolved spectra for planetary-atmosphere chemistry and escape studies, with future extensions to broaden the atomic/molecular networks and to capture more complex flare morphologies across a wider range of stellar types.

Abstract

Context. Stellar flares can significantly influence the atmospheres and habitability of orbiting exoplanets, especially around young and active M dwarfs. Understanding the temporally and spectrally resolved activity of such stars is essential for assessing their impact on planetary environments. Aims. We aim to examine in detail state-of-the-art concepts of flare models to identify what is missing in our understanding of energy deposition during the flare event. By comparing synthetic and observed flare spectra, we seek to determine the modelling frameworks best suited to represent flare energetics and spectral far-ultraviolet features while providing a foundation for investigating flare impacts on exoplanet atmospheres. Methods. In this work, we built the Young M Dwarfs Flare (YMDF) model utilising the combination of radiative-hydrodynamic (RHD) stellar atmosphere models with a high and low-energy electron beam and corresponding synthetic observables. These models are based on physical principles and were validated with solar and stellar observations. Results. The newly developed YMDF model reproduces the observed continuum rise in both the TESS photometric band and the FUV-A spectral range. Furthermore, the flare distributions generated within this framework show consistency with those observed in our sample of stars. Conclusions. We have developed the YMDF model as a tool to reproduce the time-dependent spectra of flaring young M dwarfs, providing a physically motivated description of their spectral and temporal evolution during flare events.

Figures (11)

• Figure 1: Coefficients of the fitted models for flares in young M dwarfs. The values of $\hat{X}_1$ and $\hat{X}_2$ are plotted on the y- and x-axes, respectively. Each point represents an individual flare event and is colour-coded by the flare energy released in the TESS (left) or the FUV bandpass (right). Model fitting was performed using multiple setups: H6 and H6CIII for the TESS data, and COScut1 and COScut2 for the HST-COS data. Linear regressions fitted to the flare events for each star are shown as dashed coloured lines. To guide the eye, in the panels, the H6, H6CIII, COScut1 and COScut2 ratios are shown as red dotted, black dotted, grey dash-dotted and grey dotted lines, respectively.
• Figure 2: The light curves of several flares observed in TESS and HST-COS. The upper row shows TESS flux light curves (e$^-$ s${^{-1}}$) for flares denoted as AU Mic T2, HIP 107345 T1 and 2MASS J02365171-5203036 T1 (the observation details are stated in Table \ref{['tab:102']} in Appendix \ref{['sec:appendixTBF']}) as dashed lines with red spheres for 20-second observation stamps and black error bars with capped ends. Note that errors can be smaller than the sphere size. Three temporal flare model MCMC fitting results are shown: Mendoza (light blue), Davenport (light coral), and Feinstein (teal) models fitted to the observed data. The H6 and H6CIII models, that were inversely calculated using observed EDs for these particular flares shown as solid blue and green lines with data-point spheres. In the upper row, the H6 model is hidden by the H6CIII due to overlap as their flux values are very similar. The lower row shows FUV HST-COS flux observations for flares denoted as AU Mic F1, AU Mic F2, and Karmn J07446+035 F1, plotted with consistent styles and model fits as the upper row. The details of observations for these flares can be found in Table \ref{['tab:101']} in Appendix \ref{['sec:appendixcos']}.
• Figure 3: Surface flux density observed and modelled. The m2F12-37-2.5 (dark green) and cF13-500-3 (light green) stellar atmosphere models are plotted at coefficients $\hat{{X}}_{{1}}$=1 and $\hat{{X}}_{{2}}$ = 1 (the theoretical assumption of the entire stellar surface exhibiting flaring activity). For AU Mic during a flare, the red solid and orange dashed lines show the fitted model spectrum at flare maximum using coefficients from the example fit (the flare data are stated in Table \ref{['tab:102']} in Appendix \ref{['sec:appendixTBF']}). The inverse H6 model at flare maximum, pre-flare and the 'fiducial flare' model from 2018ApJ...867...71L with a comparable ED is plotted as blue dash-dotted, light blue dashed, and purple dashed lines, respectively. Blackbody curves for 3850 K, 6000 K, 8000 K, and 10,000 K are included in black solid, dotted, dash-dotted and dashed lines. The panchromatic AU Mic flux from 2022AJ....164..110F is plotted in light grey. Insets highlight the FUV (lower left) and TESS (lower right) wavelength ranges.
• Figure 4: Flare spectra in FUV HST-COS. Green circles and grey diamonds indicate the continuum flux at the flare peak and preflare, respectively, with uncertainties displayed as corresponding light-coloured lines; only the upper error bars are shown for visibility, as the errors are symmetric. The best-fit model to the continuum fluxes, derived using the COScut1 spectral setup, is plotted as a light red line. The blue, green, and light green dashed lines correspond to the H6 and H6CIII setups at the peak, and pre-flare H6, respectively. The light grey, light orange and purple dashed lines represent: the sum of the extracted one-dimensional spectra from all exposures in the orbit, the smoothed panchromatic spectrum of AU Mic from 2022AJ....164..110F, and the 'fiducial flare' model from 2018ApJ...867...71L at a similar ED, respectively.
• Figure 5: Cumulative FFDs (scatter) of simulated ED distributions for broken power law relation (blue spheres) with $\alpha_1$=1.39 and $\alpha_2$=1.8 found from the observed distribution of AU Mic white light flares. The intrinsic observed AU Mic's FFD is plotted as red spheres (adopted from [Fig. 5, left panel]2025AA...700A..53M). Simulated ED distributions for single power laws with $\alpha$=1.39 and $\alpha$=1.8 as light blue and lime spheres, respectively. The resulting distributions are from one example run with 1000 random samples, only plotted if the sample meets p>0.01 in the K-S test. The grey and blue dashed guides correspond to power law coefficients $\alpha$=2.0 and $\alpha$=1.5, respectively, for a range of $\beta$ coefficients.