Published February 25, 2021 | Version v1
Poster Open

Understanding Flare Statistics of Young Stars with TESS and Machine Learning

  • 1. University of Chicago
  • 2. University of New South Wales
  • 3. NASA HQ
  • 4. Fermi National Accelerator Laboratory, Kavli Institute for Cosmological Physics, University of Chicago
  • 5. Massachusetts Institute of Technology
  • 6. The University of Texas at Austin
  • 7. NASA Goddard Space Flight Center


The evolution of young stellar magnetic fields is not well understood. By studying observable proxies, such as stellar flares and starspots, we can begin to better understand such mechanisms. All-sky photometric time-series missions have allowed for the monitoring of thousands of young stars to understand the evolution of stellar activity. We have developed a convolutional neural network, stella, specifically trained to find flares in TESS short-cadence data. We applied the network to 3200 young stars to evaluate flare rates as a function of age and spectral type. We also measured rotation periods for 1500 of our targets and found that flares of all amplitudes are present across all spot phases, suggesting high spot coverage across the entire surface. Additionally, flare rates and amplitudes decrease for stars older than 50 Myr across all temperatures hotter than 4000 K, while cooler stars show no evolution across 800 Myr. Cooler stars also show higher flare rates and amplitudes across all ages, particularly those that are fully convective. 



Files (3.5 MB)

Name Size Download all
3.3 MB Preview Download
177.1 kB Preview Download