P-24 Mutual Information of Short-Term Memory of Stellar Flares
Presenter Status
Student
Second Presenter Status
Professor of Engineering
Third Presenter Status
Adjunct Faculty Research Assistant to Engineering
Fourth Presenter Status
Principal Professional Staff Physicist, Johns Hopkins University Applied Physics Laboratory
Preferred Session
Poster Session
Location
Buller Hall Hallways
Start Date
22-10-2021 2:00 PM
End Date
22-10-2021 3:00 PM
Presentation Abstract
Stars radiate energy stored in their magnetic fields in the form of stellar flares, and the dynamics of a system that causes flare events can be well-described by the sequence of times between flare events, known as waiting times. From the stellar intensity data collected by the Kepler satellite, the waiting time sequences of different stars is created by finding sudden elevations in the light curve above a certain threshold. The mutual information, which measures the information within a data set, is then calculated for the original data as well as for surrogate data sets created by random permutations of waiting times within regions with constant flaring rates. Comparing the mutual information of the actual waiting times against the mutual information of the surrogate waiting times, we find that there is a significant elevation in the mutual information of the original data set, and thus our information theory analysis indicates a dependence between successive flares. This increased mutual information is due to the clustering of flares, evidenced by comparing the cumulative distribution function (CDF of successive flares in the original data set with the CDF of flares in the surrogate sequences.
P-24 Mutual Information of Short-Term Memory of Stellar Flares
Buller Hall Hallways
Stars radiate energy stored in their magnetic fields in the form of stellar flares, and the dynamics of a system that causes flare events can be well-described by the sequence of times between flare events, known as waiting times. From the stellar intensity data collected by the Kepler satellite, the waiting time sequences of different stars is created by finding sudden elevations in the light curve above a certain threshold. The mutual information, which measures the information within a data set, is then calculated for the original data as well as for surrogate data sets created by random permutations of waiting times within regions with constant flaring rates. Comparing the mutual information of the actual waiting times against the mutual information of the surrogate waiting times, we find that there is a significant elevation in the mutual information of the original data set, and thus our information theory analysis indicates a dependence between successive flares. This increased mutual information is due to the clustering of flares, evidenced by comparing the cumulative distribution function (CDF of successive flares in the original data set with the CDF of flares in the surrogate sequences.