Even though flares evolve on timescales as short as seconds, most optical flare monitoring is carried out at cadences of minutes. I am currently carrying out the first large survey of white light flares with resolved peaks using the new 20 sec cadence TESS mode. TESS is observing 460 of the nearest and brightest flare stars across the sky as part of my GO programs 3174 and 4132. In a first paper, I discovered degeneracies present at 2 min cadence in the morphologies of 3792 flares can be resolved at 20 sec cadence, and quantified the debated occurrence rates of short-period pulsations in flares. We also estimate the UV surface fluences of temperate planets during the peaks of the largest flares. We find a third of these flares reach the 90% lethal dose threshold of the radiation hardy bacterium D. Radiodurans in just 20 seconds.
The TESS Extended Mission observed 460 of the nearest and brightest flare stars at 20 sec cadence to resolve the rapid changes in emission
How rapidly does flare emission evolve?
Many of these flares show structures not resolvable at lower cadence. While flares are generally thought to have impulsive or “spiky” peaks, we observed a number of “flat top” flares. Flat top flares have a period of constant emission during the peak. and likely impact planetary atmospheres differently than do impulsive peaks. Subsequent studies down to one second cadence (e.g. Tomo-e Gozen; Aizawa+2022) have confirmed flat top flares are very common at these cadences, sparking new investigations in the field to characterize the emission mechanisms and habitability impacts of these fast flares.
Flare light curves are usually assumed to be highly-peaked based on templates constructed from data obtained at minute cadence. However, the 20 sec TESS data shows many flares to have flat tops, or periods of constant emission during the peak.
Even though flares evolve on timescales as short as seconds, most optical flare monitoring is carried out at cadences of minutes. I am currently carrying out the first large survey of white light flares with resolved peaks using the new 20 sec cadence TESS mode. TESS is observing 460 of the nearest and brightest flare stars across the sky as part of my GO programs 3174 and 4132. In a first paper, I discovered degeneracies present at 2 min cadence in the morphologies of 3792 flares can be resolved at 20 sec cadence, and quantified the debated occurrence rates of short-period pulsations in flares. We also estimate the UV surface fluences of temperate planets during the peaks of the largest flares. We find a third of these flares reach the 90% lethal dose threshold of the radiation hardy bacterium D. Radiodurans in just 20 seconds.
Two minute cadence flares (orange) observed by TESS are each revealed to have unresolved structure only visible at 20 second cadence (grey).
The TESS Extended Mission observed 460 of the nearest and brightest flare stars at 20 sec cadence to resolve the rapid changes in emission
How rapidly does flare emission evolve?
Many of these flares show structures not resolvable at lower cadence. While flares are generally thought to have impulsive or “spiky” peaks, we observed a number of “flat top” flares. Flat top flares have a period of constant emission during the peak. and likely impact planetary atmospheres differently than do impulsive peaks. Subsequent studies down to one second cadence (e.g. Tomo-e Gozen; Aizawa+2022) have confirmed flat top flares are very common at these cadences, sparking new investigations in the field to characterize the emission mechanisms and habitability impacts of these fast flares.
Flare light curves are usually assumed to be highly-peaked based on templates constructed from data obtained at minute cadence. However, the 20 sec TESS data shows many flares to have flat tops, or periods of constant emission during the peak.
The TESS mission has vastly increased the number of planets available for detailed characterization. At the time of writing, more than 5000 TESS Objects of Interest (TOIs) have now been detected. TOIs are candidate or confirmed transiting planets. Exoplanet characterization efforts require a careful treatment of starspots and flares to avoid false positives in atmospheric line or bio-signature detections and to probe non-equilibrium atmospheric states resulting from high rates of flaring.
Previously, no comprehensive survey of the flare rates of the TOI catalog existed. Most of the sky has now been observed continuously for at least 2 months during the TESS Prime and Extended Mission, sufficient to place strong upper limits on the flare rates of low mass TOIs. I carried out the first comprehensive search for flares for all non-retired TOIs with 2 min light curves. Non-retired TOIs are those that have not been classified as false positives by the TESS mission. We use injection testing to obtain upper limits on the flare rates of all non-flaring TOIs in order to determine “worst case scenarios" for atmospheric characterization.
My survey discovered 93 planets orbit flare stars. I found 25% of terrestrial M-dwarf planets suitable for transmission spectroscopy with JWST orbit flare stars. Smaller planets orbit flare stars more frequently than do larger planets due to planet detection biases. However, none of the terrestrial planet hosts appear to flare at sufficient levels for 99.9% atmospheric ozone depletion. Terrestrial planets in the habitable zone in our sample included TOI 700 d, LHS 1140 b, TOI 715.01, and TOI 4312.01, each among the most exciting planets for follow-up efforts.
Smaller TESS planets orbit flare stars more frequently than do larger planets, a result of the planet detection biases of TESS and increased activity of small stars.
Which planets discovered by TESS orbit flare stars?
Skymap illustration of the flare rates/upper limits of all two minute cadence TOIs at the time of publication. Upper limits are obtained with flare injection and recovery testing. The colorbar gives the number of large flares each year.
While the multi-wavelength properties of large stellar flares have received much recent attention, small flares have not received the same degree of attention despite their high frequency and connection to space weather. I led a paper on the first stellar flare observed by ALMA and Chandra, a flare so small its multi-wavelength properties can be directly compared to M-X solar flares. The Sun emits 170 X1-class flares during its 11 yr cycle, providing a unique opportunity to compare the pan-chromatic properties of solar and stellar flares at this energy for the first time. The flare was observed in the millimeter, optical, and soft X-ray.
I found the flare's coronal temperature and particle density measured from the soft X-ray are typical of solar flares, but smaller than in many stellar flares. However, the millimeter luminosity is 1000X higher than comparable X1 solar flares, indicating more intense particle acceleration in M-dwarfs. I am now obtaining simultaneous ALMA and Chandra data of a young Proxima Cen analog to find out if these trends hold more broadly. Simulations of 5-100 GHz flux densities resulting from optically thin gyrosynchrotron emission find a dependence on the high energy electron density and cutoff energy (Wu et al. 2019). Millimeter emission currently remains one of the few direct probes of the accelerated particle environment in M-dwarf flares. By constraining the electron environments of a larger sample of millimeter flares, it may be possible to determine if they are more energetic than those of typical solar flares.
We reproduce the soft X-ray to millimeter scaling relationship from Krucker et al. (2013) and overlay our flare in red. We extrapolate the likely position of larger stellar flares previously observed with ALMA by assuming the ratio of SXR and millimeter emission from our flare.
Are flares from small stars similar to solar flares?
Left panel: The first flare detected with Chandra and ALMA, alongside coordinated optical photometry. Right panel: While among the smallest stellar flares, the flare is an X1 class solar analog. The Chandra spectrum shows its coronal temperature and emission measure (particle density) are typical of solar flares (triangles) rather than other stellar flares (circles).
More than forty large flares were detected simultaneously with TESS, allowing me to measure color-temperatures as high as 40,000 K. A 9000 K blackbody is assumed for white light flares from empirical fits to the spectra of smaller events (Kowalski et al. 2013). Hot superflares emit 10X more UV radiation than expected from the 9000 K blackbody, under-predicting the effects of flares on atmospheres and the survival of micro-organisms (e.g. Tilley et al. 2019, Abrevaya et al. 2020) that assume 9000 K. Although unexpected, such high temperatures are now supported by radiative-hydrodynamic flare modeling (Kowalski et al. 2022).
Left panel: A superflare observed by Evryscope and TESS. Middle panel: Temperature evolution of the left panel flare. The flare reached 42,000 K and would have produced 20X the UV radiation expected from a single optical bandpass and 9000 K blackbody. Right panel: although a cool flare and a hot flare have the same g’ band energy, the UV emission differs significantly.
Do hotter flares douse planets in more UV radiation?
For my thesis at UNC Chapel Hill, I measured the occurrence rate of M-dwarf superflares across the sky with the Evryscope array of small telescopes. Superflares are events of more than 10^33 erg, 10-1000X more energetic than the largest flares from our Sun. Evryscope is an array of 61 mm telescopes, together imaging 8150 square degrees on the sky at 2 minute cadence down to g’=16 in a dark sky. I created the Evryscope Flare (EvryFlare) Survey to discover how often extreme flare events impact planets orbiting M-dwarfs, leveraging Evryscope and TESS data of the brightest low mass stars. My survey of 4069 of the nearest and brightest low mass stars in the sky detected 413 superflares from 285 stars spanning a range of masses and ages, with more flares from younger stars. Evryscope is uniquely suited to determine the long-term flare rates of nearby stars due to the high cadence and multi-year span of observations. TESS contributes the smaller events that fall below Evryscope’s detection capabilities from the ground.
I found 15±2% of the stars around which TESS may discover temperate planets are currently emitting flares large enough to significantly affect the potential habitability of those planets. I observed 17 stars that may deplete an Earth-like atmosphere via repeated flaring at current levels, including one superflare with sufficient energy to photo-dissociate all ozone in an Earth-like atmosphere in a single event.
The largest flares detected by Evryscope from planet-search targets. The energy of the top left flare could fully disrupt the chemical equilibrium of an Earth-like atmosphere.
The Evryscope Flare (EvryFlare) Survey
Proxima Cen, the nearest star to our solar system, hosts a likely-rocky planet (Proxima b) in its habitable zone (where surface water can be liquid). I led an ApJ Letter reporting a “superflare” from the star on March 18 2016, captured with the Evryscope array of small telescopes. During this flare, Proxima Cen increased in brightness by a factor of ~100X and irradiated its terrestrial planet Proxima b with potentially-lethal doses of UV radiation. A recent study by Ximena Abrevaya et al. has tested the survival of common micro-organisms in lab experiments mimicking conditions during the superflare. They found approximately 1 in 10,000 survived the flare.
We also observed 23 other large flares on Proxima and use these to demonstrate Proxima b experiences 2-5 superflares in a year, in agreement with a previous estimate by James Davenport using the MOST satellite. In our paper published in the Astrophysical Journal Letters we used this superflare rate to estimate the habitability of Proxima b. We found a 90% loss to an Earth-like planet’s ozone layer in 5 years and estimated a complete collapse of the ozone layer on geologically- short timescales.
The Proxima superflare captured in a loop of Evryscope images.
The Proxima superflare, observed by Evryscope in March 2016, during which the star increased in brightness by ~100X and irradiated the potentially-habitable planet Proxima b with lethal doses of UV radiation
Evryscope detection of the Proxima Superflare