[Paper Review] Statistical properties of superflares on solar-type stars with Kepler data

Yuta Notsu, Hiroyuki Maehara+6 more|arXiv (Cornell University)|Jul 31, 2016

Stellar, planetary, and galactic studies1 references4 citations

TL;DR

This study analyzes Kepler long- and short-cadence light curves to investigate superflares on solar-type stars, finding that superflare energy distribution follows a power-law with index -1.5 for energies between 10^33 and 10^36 erg. The flare frequency increases with shorter rotation periods, but the maximum flare energy per star shows no clear correlation with rotation, indicating superflares can occur on Sun-like stars, albeit rarely.

ABSTRACT

Superflares are flares that release total energy 10$\sim$10$^{4}$ times greater than that of the biggest solar flares with energy of $\sim$10$^{32}$ erg. We searched superflares on solar-type stars (G-type main sequence stars) using the Kepler 30-min (long) and 1-min (short) cadence data. We found more than 1500 superflares on 279 stars from 30-min cadence data (Q0-6) and 187 superflares on 23 stars from 1-min cadence data (Q0-17). The bolometric energy of detected superflares ranges from the order of 10$^{32}$ erg to 10$^{36}$ erg. Using these data, we found that the occurrence frequency ($dN/dE$) of superflares is expressed as a power-law function of flare energy ($E$) with the index of -1.5 for $10^{33}

Motivation & Objective

• To determine whether superflares—flares 10–10,000× more energetic than the largest solar flares—can occur on solar-type stars (G-type main-sequence stars) using Kepler photometric data.
• To investigate the statistical properties of superflares, including energy distribution, occurrence frequency, and dependence on stellar rotation.
• To examine whether the presence of close-in giant planets (e.g., hot Jupiters) correlates with superflare occurrence, challenging prior hypotheses about planetary influence on stellar magnetic activity.
• To assess the physical mechanism behind superflares by linking flare energy to magnetic energy stored in starspots and testing the Alfvén timescale model.

Proposed method

• Analyzed Kepler 30-minute and 1-minute cadence light curves from Quarters 0–6 (long-cadence) and 0–17 (short-cadence) to detect superflares on G-type main-sequence stars.
• Defined solar-type stars by effective temperature (5300–6300 K) and surface gravity (log g ≥ 4.0), using revised stellar parameters to reduce systematic bias.
• Calculated bolometric flare energy from flux increases in light curves, with energies ranging from ~10^32 to 10^36 erg.
• Fitted the flare energy distribution to a power-law function: dN/dE ∝ E^−α, with α = 1.5 ± 0.1 for 10^33 < E < 10^36 erg.
• Correlated flare frequency and maximum flare energy with stellar rotation period derived from quasi-periodic brightness variations.
• Tested the Alfvén timescale model by examining the scaling of flare duration with energy: τ ∝ E^0.39±0.03.

Experimental results

Research questions

• RQ1 Do superflares with energies exceeding 10^33 erg occur on solar-type stars, and what is their statistical energy distribution?
• RQ2 How does the occurrence frequency of superflares depend on the rotation period of solar-type stars?
• RQ3 Is there a correlation between the maximum observed flare energy and the rotation period of the host star?
• RQ4 Can the duration of superflares be explained by the Alfvén timescale, and what is the scaling relation with energy?
• RQ5 Is the presence of close-in giant planets (e.g., hot Jupiters) a necessary condition for superflare occurrence on solar-type stars?

Key findings

• More than 1,500 superflares were detected on 279 solar-type stars using 30-minute cadence data, and 187 superflares on 23 stars using 1-minute cadence data.
• The energy distribution of superflares follows a power-law with index -1.5 for energies between 10^33 and 10^36 erg, consistent with a universal scaling law.
• Flare frequency increases with shorter rotation periods, indicating that faster-rotating stars exhibit more frequent superflares.
• The maximum flare energy observed per star shows no significant correlation with rotation period, suggesting that magnetic energy storage capacity is not strongly rotation-dependent.
• Flare duration scales with energy as τ ∝ E^0.39±0.03, consistent with the Alfvén timescale model for magnetic reconnection.
• No significant excess of superflare stars hosting close-in giant planets was found—only 1.1% of superflare stars host confirmed or candidate planets, compared to 2.1% in the general solar-type star sample—indicating hot Jupiters are not a necessary condition for superflares.