Документ взят из кэша поисковой машины. Адрес оригинального документа : http://hea-www.harvard.edu/AstroStat/SolStat2012/jack_ireland.pdf Дата изменения: Fri Jun 22 22:45:41 2012 Дата индексирования: Tue Oct 2 04:18:23 2012 Кодировка: Поисковые слова: с р р с с п п р п п с с с р р р р р п п р р р п п р

Finding oscillatory regions in SDO/AIA EUV data
J. Ireland, C. A.Young ADNET Systems, Inc. at NASA GSFC Funded by a NASA ROSES 2008 HGI award.


Solar structures oscillate
· ·
many different oscillation periods have been identified - lots of frequency space to examine. concentrate on 3 and 5 minute wavebands.


Scientific return
·
Diagnostic potential: if you are sure that the oscillation can be identified with a predicted wave mode, then observations can be used to measure coronal properties, e.g. coronal field strength. Potential energy source that heats the corona.

·


Measurements
· ·
Frequencies. Occurrence locations, longevity, recurrence rate at a given location, conditions for required for an oscillation to occur.


SDO/AIA data
· · · · ·
16 Mpx/image. 10 wavebands. One image in each waveband every 12 seconds continuously. Overlapping timeranges imply 24 analyses per active region. ~ 4 x 10 FFTs per day.
9

Identify oscillating structures.


Approach
· · ·
Look for 3 and 5 minute period oscillations only. Examine active regions only (smaller number of time-series examined). Identification of oscillating individual pixels, then segmentation into significant groups


Data
· · · ·
Use SPoCA active region detections. Download cutouts from AIA cutout service in 171 е and 193 е. Remove solar rotation, sum 2 в 2. One hour duration = 300 samples.


Analysis
· ·
Oscillation model in each pixel
dj = A cos( tj )+ B sin( tj )

Calculate (Bayesian) probability p() that the timeseries dj observed at times tj= j.t (1jN) supports a frequency .
p( ) 1 - 2C ( ) Nd
2

· ·



1-N/2

N 1 C ( ) = | dj e N j =1

i tj 2

|

Integrate over pre-defined frequency ranges. Use the Fourier frequencies k = k.2/(N.t).


Analysis
· · ·
Group together pixels that have a high probability of oscillating within the frequency range. Measure local coherence properties of these groups. Keep groups of highly coherent pixels.


Analysis
· · · ·
Active region size ~ 240 x 240 arcsec2 with 2в2 spatial summing, ~ 104-5 px. Approximately 1-10 active regions per day. Analyze durations of 1 hour of data only. Two wave bands analyzed (out of a possible 6).


Results


Results


Results


Conclusions
· ·
Not many coherently oscillating groups of pixels in171 е and 193 е. Consistent with the suggestion most oscillatory signals are not wave-like. knowledge of the underlying physical structure (where are the loops? where is the moss?) to give a definitive answer.

· But.... - need