J/ApJ/833/284 Quasi-periodic pulsations in solar flares (Inglis+, 2016)
A large-scale search for evidence of quasi-periodic pulsations in solar flares.
Inglis A.R., Ireland J., Dennis B.R., Hayes L., Gallagher P.
<Astrophys. J., 833, 284-284 (2016)>
=2016ApJ...833..284I 2016ApJ...833..284I (SIMBAD/NED BibCode)
ADC_Keywords: Sun ; Stars, flare ; X-ray sources ; Gamma rays
Keywords: methods: statistical; Sun: corona; Sun: flares; Sun: oscillations;
Sun: X-rays, gamma rays
Abstract:
The nature of quasi-periodic pulsations (QPP) in solar flares is
poorly constrained, and critically the general prevalence of such
signals in solar flares is unknown. Therefore, we perform a
large-scale search for evidence of signals consistent with QPP in
solar flares, focusing on the 1-300s timescale. We analyze 675 M- and
X-class flares observed by the Geostationary Operational Environmental
Satellite (GOES) series in 1-8 A soft X-rays between 2011 February 1
and 2015 December 31. Additionally, over the same era we analyze
Fermi/Gamma-ray Burst Monitor (GBM) 15-25keV X-ray data for each of
these flares associated with a Fermi/GBM solar flare trigger, a total
of 261 events. Using a model comparison method, we determine whether
there is evidence for a substantial enhancement in the Fourier power
spectrum that may be consistent with a QPP signature, based on three
tested models; a power-law plus a constant, a broken power-law plus
constant, and a power-law-plus-constant with an additional QPP
signature component. From this, we determine that ∼30% of GOES events
and ∼8% of Fermi/GBM events show strong signatures consistent with
classical interpretations of QPP. For the remaining events either two
or more tested models cannot be strongly distinguished from each
other, or the events are well-described by single power-law or broken
power-law Fourier power spectra. For both instruments, a preferred
characteristic timescale of ∼5-30s was found in the QPP-like events,
with no dependence on flare magnitude in either GOES or GBM data. We
also show that individual events in the sample show similar
characteristic timescales in both GBM and GOES data sets. We discuss
the implications of these results for our understanding of solar
flares and possible QPP mechanisms.
Description:
We have used data from the Geostationary Operational Environmental
Satellite (GOES) instrument series, and from Fermi/Gamma-ray Burst
Monitor (GBM). For this reason, we choose the interval 2011 February 1
- 2015 December 31, as it not only coincides with the availability of
GOES-15 satellite data, but also includes regular solar observations
by GBM.
GOES satellites are equipped with solar X-ray detectors that record
the incident flux in the 0.5-4Å and 1-8Å wavelength ranges.
Solar X-ray data from the most recent satellite, GOES-15, has been
available since 2010 at a nominal 2s cadence. To access the GOES
catalog, we use the Heliophysics Event Knowledgebase (HEK).
Fermi/GBM operates in the 8keV-40MeV range and regularly observes
emission from solar flares, with a solar duty cycle of ∼60%, similar
to the solar-dedicated Reuven Ramaty High Energy Solar Spectroscopic
Imager (RHESSI). To accumulate the database of Fermi/GBM events, we
use the GBM trigger catalog produced by the instrument team, selecting
all events marked as flares.
File Summary:
--------------------------------------------------------------------------------
FileName Lrecl Records Explanations
--------------------------------------------------------------------------------
ReadMe 80 . This file
table3.dat 104 703 List of studied GOES events and analysis result
parameters
table4.dat 104 297 List of studied Fermi/GBM events and analysis
result parameters
--------------------------------------------------------------------------------
See also:
J/A+A/304/563 : Cool X-ray flares of Sun with GOES (Phillips+, 1995)
J/ApJ/757/94 : Solar flares observed with GOES and AIA (Aschwanden, 2012)
J/ApJ/797/50 : Global energetics of solar flares. I. (Aschwanden+, 2014)
J/ApJ/831/105 : Global energetics of solar flares. IV. CME (Aschwanden, 2016)
J/ApJ/845/36 : Complex network for solar active regions (Daei+, 2017)
http://www.lmsal.com/hek/ : Heliophysics Events Knowledgebase (HEK) home page
http://fermi.gsfc.nasa.gov/ssc/data/access/gbm/ : GBM data products page
Byte-by-byte Description of file: table[34].dat
--------------------------------------------------------------------------------
Bytes Format Units Label Explanations
--------------------------------------------------------------------------------
1- 4 I4 yr Obs.Y [2011/2015] Year of the observation
5- 6 I2 month Obs.M Month of the observation
7- 8 I2 d Obs.D Day of the observation
10- 13 A4 --- Class GOES classification (M1.0 to X6.9)
15- 16 I2 h StObs.h Hour of observation start
17- 18 I2 min StObs.m Minute of observation start
19- 20 I2 s StObs.s Second of observation start
22- 23 I2 h EndObs.h Hour of observation end
24- 25 I2 min EndObs.m Minute of observation end
26- 27 I2 s EndObs.s Second of observation end
29- 33 F5.1 --- dBIC0-1 [-21/462] The Δ Bayesian Information
Criterion between models S0 and S1 (1)
35- 39 F5.1 --- dBIC0-2 [-17/324] The Δ Bayesian Information
Criterion between models S0 and S2 (1)
41- 46 F6.1 --- dBIC2-1 [-167/288] The Δ Bayesian Information
Criterion between models S2 and S1 (1)
48- 50 A3 --- Strong1 Model S1 strongly favored?
52- 55 F4.2 --- chi2-0 [0/6] The χ2 of S0 model
57- 61 F5.3 --- p-val0 [0/1] The p-value of S0 model
63- 66 F4.2 --- chi2-1 [0/6] The χ2 of S1 model
68- 72 F5.3 --- p-val1 [0/1] The p-value of S1 model
74- 77 F4.2 --- chi2-2 [0/6] The χ2 of S2 model
79- 83 F5.3 --- p-val2 [0/1] The p-value of S2 model
85- 89 F5.1 s Period [4.5/299]? Period of best model
91- 95 F5.3 --- Width [0.05/0.3]? Width of peak in log-frequency
space
97-104 A8 --- Flags Additional flags (2)
--------------------------------------------------------------------------------
Note (1): Model are defined in Section 3 as:
S0 = single power-law plus constant model see Equation (2):
S0(f)=A0f-α0+C0 ;
S1 = power-law-plus-constant model with an additional localized enhancement;
see Equation (3);
S2 = broken-power law model; see Equation (4).
Note (2): Flag as follows:
B0 = bad fit to model S0 (p-val0 < 0.01);
B1 = bad fit to model S1 (p-val1 < 0.01);
B2 = bad fit to model S2 (p-val2 < 0.01);
S = short data series (N < 200);
D = discontinuity in data series.
--------------------------------------------------------------------------------
History:
From electronic version of the journal
(End) Prepared by [AAS], Emmanuelle Perret [CDS] 05-Apr-2018