Astronomy and Astrophysics – Astrophysics
Scientific paper
Dec 2010
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2010agufmsh43b1820b&link_type=abstract
American Geophysical Union, Fall Meeting 2010, abstract #SH43B-1820
Astronomy and Astrophysics
Astrophysics
[7519] Solar Physics, Astrophysics, And Astronomy / Flares
Scientific paper
Principal Component Analysis was applied to selected sequences of Hydrogen-alpha images of solar active regions observed by the Improved Solar Optical Observation Network (ISOON) telescope at Sacramento Peak, NM at one-minute intervals. Resulting eigenvalues were used to determine the explained variance by the corresponding eigenvectors with elements for each ISOON measurement time. The subset of eigenvectors accounting for 99.9% of the variance were used as the source of the variable data in an application of multivariate discriminant analysis using the Fisher’s Linear Discriminant (FLD) algorithm applied to two groups. Independently obtained GOES x-ray flux measurements at one-minute intervals for the same measurement times were used to determine which group (non-flaring or flaring) the measurement time was assigned to, using a flux threshold based on the background x-ray flux from the day prior to the day of analysis. The product of the FLD algorithm applied to the eigenvectors from several selected ISOON measurement sequences was the discriminant vector (DV) consisting of coefficients for each of the subset eigenvector elements. The DV was then applied to same subset of eigenvector elements for an independent ISOON sequence to diagnose the value of the FLD function, which was used to determine the probability of flaring (0 - no likelihood of flaring, 1 - full likelihood of flaring) at each independent sequence measurement time. Results showed that for many of the image times, the diagnosed flaring probability was able to indicate the presence of flaring at the correct times of the image sequence. This flare/no flare diagnostic scheme, if successfully projected forward in time, might serve as a basis for short-term flare probability predictions.
Balasubramaniam K. S.
Norquist Donald C.
No associations
LandOfFree
Development of a Statistical Diagnostic Scheme for Flare Probability does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Development of a Statistical Diagnostic Scheme for Flare Probability, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Development of a Statistical Diagnostic Scheme for Flare Probability will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1510748