Konus-Wind and Helicon-Coronas-F Observations of Solar Flares
V. D. Pal'shin, Yu. E. Charikov, R. L. Aptekar, S. V. Golenetskii, A. A. Kokomov, D. S. Svinkin, Z. Ya. Sokolova, M. V. Ulanov, D. D. Frederiks, A. E. Tsvetkova
aa r X i v : . [ a s t r o - ph . S R ] D ec Konus-
Wind and Helicon-
Coronas-F
Observations of Solar Flares
V. D. Pal’shin , , Yu. E. Charikov , , , R. L. Aptekar , S. V. Golenetskii , A. A. Kokomov ,D. S. Svinkin , Z. Ya. Sokolova , M. V. Ulanov , , D. D. Frederiks , and A. E. Tsvetkova ABSTRACT
Results of solar flare observations obtained in the Konus-
Wind experiment fromNovember, 1994 to December, 2013 and in the Helicon-
Coronas-F experiment duringits operation from 2001 to 2005, are presented. For the periods indicated Konus-
Wind detected in the trigger mode 834 solar flares, and Helicon-
Coronas-F detected morethan 300 solar flares.A description of the instruments and data processing techniques are given. Asan example, the analysis of the spectral evolution of the flares SOL2012-11-08T02:19(M1.7) and SOL2002-03-10T01:34 (C5.1) is made with the Konus-
Wind data and theflare SOL2003-10-26T06:11 (X1.2) is analyzed in the 2.223 MeV deuterium line withthe Helicon-
Coronas-F data.
Subject headings:
Sun: X-rays – Sun: gamma-rays – Sun: flares
1. INTRODUCTION
Studying the dynamics of the energy spectra of X-ray and gamma-ray emission of solar flaresis a powerful tool for the investigation of charged particle acceleration and the diagnostics of flare Ioffe Physical Technical Institute, St. Petersburg, Russian Federation [email protected] St. Petersburg State Polytechnical University, St. Petersburg, Russian Federation Central (Pulkovo) Astronomical Observatory, Russian Academy of Sciences, St. Petersburg, Russian Federation
RHESSI spectrometer (Lin et al. 2002), which has excellent energy andangular resolution. Operating since 2008 onboard the
Fermi observatory the Gamma-Ray BurstMonitor (GBM) (Meegan et al. 2009) is also carrying out observations of solar flares in a range from ∼ ∼
40 MeV. The Russian gamma-ray spectrometer Konus, which successfully operatesonboard the U.S.
Wind spacecraft, has a number of advantages when compared with
RHESSI andGBM, operating on near-Earth spacecraft. The Konus-
Wind experiment has been continuouslycarried out since November 1994 in the favorable conditions of interplanetary space far from theEarth’s magnetosphere, which ensures no interference from Earth’s radiation belts nor occultingby the Earth. Owing to such an orbit a useful observation time is ∼ ∼
18 keV to ∼
15 MeV under extremely stable background (in the absence of powerful streams ofcharged particles from the Sun). In addition, the effective area of the spectrometer in the range >
100 keV may be more than the effective area of the
RHESSI detector (for certain configurations)and resolution in the MeV range is better than that of the GBM detectors. During the operationof Konus-
Wind and other series of the Konus experiments a unique database on solar flares hasbeen accumulated. The aim of this paper is to describe the obtained data, the methods of theirprocessing, as well as to present the results of data analysis through examples on specific flares.
2. INSTRUMENTS2.1. Konus-
Wind
The Russian-American experiment Konus-
Wind is designed to study the temporal and spectralcharacteristics of gamma-ray bursts, soft gamma repeaters and other transient phenomena in a wideenergy range from ∼
18 keV to ∼
15 MeV. The instrument consists of two spectrometric gamma-raydetectors NaI(Tl) (13 cm in diameter, 7.5 cm in height). Their axes are directed to the south and 3 –the north ecliptic pole, respectively, that provides a continuous monitoring of the entire celestialsphere. Each detector has an effective area of ≃ −
160 cm , depending on the energy and incidentangle of a gamma quantum. Resolution of the spectrometer is about 8% (FWHM) at the 662 keVline, the sensitivity of ∼ (1 − × − erg cm − . Detectors operate in two modes: “Background”and “Burst” (trigger mode). In the “Background” mode a counting rate is measured in threeenergy channels G1, G2, G3 with the nominal ranges 10–50, 50–200, and 200–750 keV ( ∼ ∼ ∼ Coronas-F
The Helicon gamma-ray spectrometer was one of the instruments onboard the
Coronas-F solar space observatory (Oraevskii et al. 2002), which had been in operation from August 2001 toDecember 2005 in a near-Earth low-eccentricity polar orbit (inclination 82 . ◦
5, distance from theEarth 500–550 km). The spacecraft was stabilized by rotation with respect to the axis directedtoward the Sun within 10 ′ . The spectrometer consisted of two detectors, similar to those of Konus- Wind , one of which was oriented toward the Sun and the other viewed the antisolar hemisphere.The “Burst” mode was similar to the Konus-
Wind trigger mode; in the “Background” mode a timehistory was measured in 8 energy channels covering the 10–200 keV range with a time resolutionof 1 s and 256-channel spectra were measured in the 0.2–10 MeV range with accumulation time of33.6 s. Data output to the onboard memory was made without interruption in measurements.
3. OBSERVATIONS OF SOLAR FLARES IN KONUS-WIND EXPERIMENT
During the period from its launch in November 1994 until the end of 2013 Konus-
Wind hadregistered in the “Burst” mode 834 solar flares with the GOES classification: 113 X-class, 454M-class, 262 C-class and 5 B-class. Figure 1 shows the distribution of the number of flares overthe years. It is seen that the largest number of flares occurred in years of maximum solar activity2000–2002, while in 2008 and 2009 corresponding to the minimum of solar activity, there were nosufficiently intense flares to cause a trigger.Due to the nature of the trigger algorithm (see Section 2.1), the transition to the “Burst” modetakes place at appearing in a flare of high-energy radiation with quite a rapid increase in intensity,wherein a smoother and softer initial rise phase is skipped (and registered only in the backgroundrecord). Besides, the duration of many intense flares exceeds the duration of measurements in thetrigger mode. As a result, for a significant proportion of long and intense flares, spectral data andtime profiles of high resolution are available only for part of a flare. 5 –
Wind solar flares
Spectral analysis of the Konus-
Wind data is performed using the software package XSPEC(Arnaud 1996), for this purpose spectra and a detector response matrix are converted into fits-format. The incident angle on the detector is 90 ◦ for solar flares, this entails the strong absorptionof soft radiation when passing through the aluminum container 2 mm thick housing the scintillator.Despite this, the intensity of radiation from powerful flares in the soft range may cause the overflowof the G1 channel counter and significant distortion of spectra.In the standard analysis of Konus- Wind data a spectrum, averaged over an interval of ∼ >
40 keV range, where the contribution of the thermal component of radiation is usuallysmall, flare spectra are often well described by a simple power-law model: dN/dE ∝ E − γ . Anexample of such a spectrum measured by Konus- Wind is shown in Fig. 2.
Figure 3 shows a solar flare recorded by Konus-
Wind on November 8, 2012 (class M1.7). Theleft panel shows the time profile of the flare in three energy channels and two hardness ratios: G2/G1 6 –and G3/G2, produced from background data; on the right – time profile with a resolution of 1.024 s,the evolution of the power-law spectral index and the energy flux in the 40–1000 keV range obtainedfrom “Burst” mode data. It can be seen that the flare starts with a fairly soft emission, followed bya powerful hard pulse (which caused the trigger event) during which a correlation between intensityand hardness is clearly seen, then the intensity of the radiation gradually decreases, but its hardnessremains almost constant with the spectral index of ∼ Figure 4 (similar to Figure 3) shows a solar flare detected by Konus-
Wind on March 10, 2002(class C5.1). The flare is quite unusual – the duration of the main pulse is only about 15 s, withoutany significant emission before the pulse even in the soft channel, and the spectrum of the emissionis extremely hard with an index of 2.4 at the maximum intensity. In ∼
70 s after the main pulse amuch weaker pulse follows that is only visible in the channel G1 (in this case the pulse has greaterintensity in the GOES channels, as shown in the left panel of Fig. 4). This flare belongs to a classof so-called “early impulsive flares” (Sui et al. 2006). For such flares a hard radiation peak precedesa soft radiation peak, indicating weak plasma preheating, while the images obtained by
RHESSI ,typically exhibit two-footpoint morphology of the source for the hard peak and looptop source forthe delayed soft radiation peak; the spectrum of the early impulsive peak is nonthermal and veryhard and the spectrum is thermal in the delayed soft radiation pulse (see, for example, Su et al.(2008)). In this flare the delay between the soft and hard peaks is ∼
20 s.
4. OBSERVATIONS OF SOLAR FLARES IN HELICON-CORONAS-FEXPERIMENT
Useful exposure for the Coronas-F observations was severely limited by passing through theradiation belts at high latitudes and through the South Atlantic Anomaly (SAA). In total, for 4years of the experiment 300 flares had been registered in the trigger mode: 4 X-class, 54 M-class, 7 –163 C-class and 69 B-class (flares were counted providing their trigger fired on the portions of orbitwith a modest background level, i.e. at low and moderate latitudes away from the SAA). Theflares were registered by the detector whose axis was oriented to the Sun, herewith, the radiationpassed through an entrance beryllium window that provided low absorption in the soft part ofthe spectrum. Measurements of 256-channel spectra in the “Background” mode (see section 2.2)allowed registering the deuterium line 2.223 MeV and studying its evolution on the timescale of33.6 s in strong long flares.
Figure 5 shows a solar flare detected on October 26, 2003 (class X1.2). The flare is interestingbecause after ∼
40 min after its start a powerful pulse of hard emission was detected, which wasnot accompanied by any increase in the intensity of soft X-rays (Zimovets & Struminsky 2012).The top panels of Fig. 5 show the time profile derived from the data of the Konus-
Wind standbysystem (see section 2.1) and from the anti-coincidence shield (ACS) of the SPI telescope onboardthe
INTEGRAL observatory. The bottom panel shows the emission intensity evolution in thedeuterium line, derived from the Helicon data. It can be seen that the line intensity is followingthe intensity of the powerful hard X-rays pulse (with a peak at ∼ K present in the detector materials. The deuterium line width is determined bythe resolution of the instrument that amounts to 130 keV at 2.223 MeV ( ≃ − to the photon flux (photons cm − s − ), the measured intensity was divided by the 8 –detector effective area in the full absorption peak, which is equal to 32 cm at 2.223 MeV. The linefluence measured during the hard pulse (906 s since 07:27:16 UT), was 22.1 ± − ,which is comparable with the fluences measured with the GRS gamma-ray spectrometer onboardthe Solar Maximum Mission in the 2.223 MeV line for the most intense solar flares (Vestrand et al.1999).
5. CONCLUSIONS
In the Konus-
Wind experiment, a large database of solar flares has been accumulated since1994 upto the present (834 flares in the “Burst” mode). These data allow us to investigate thedynamics of hard X-ray and gamma-ray emission spectra in different phases of flares on timescalesvarying from 0.256 to 8.2 s (but the initial rise phase is often skipped), as well as the temporalevolution and the evolution of the hardness ratio on the timescales from 2 to 256 ms in the “Burst”mode and on the timescale of 2.994 s in the “Background” mode.The Helicon Coronas-F data allow studying the dynamics of the intensity of the 2.223 MeVdeuterium line and its relationship to the intensity of hard X-ray and gamma radiation.A list of the Konus-
Wind
REFERENCES
Aptekar, R., Frederiks, D., Golenetskii, S., et al. 1995, Space Sci. Rev. 71, 265Arnaud, K. A. 1996, in ASP Conf. Ser. 101, Astronomical Data Analysis Software and Systems V,ed. G. Jacoby & J. Barnes (San Francisco, CA: ASP), 17Fletcher, L., Dennis, B.R., Hudson, H.S., et al. 2011, Space Sci. Rev., 159, 19 9 –Lin, R.P., Dennis, B.R., Hurford, G.J., et al. 2002, Sol. Phys., 210, 3Mazets, E.P., Aptekar, R.L., Golenetskii, S.V., et al. 2012, JETP Letters, 96, 544Meegan, C., Lichti, G., Bhat, P., et al. 2009, ApJ, 702, 791Oraevskii, V.N., Sobelman, I.I., Zhitnik, I.A., Kuznetsov, V.D. 2002, Phys. Usp., 45, 886Su, Y., Gan, W.Q., & Li, Y. P. 2008, AdSpR, 41, 988Sui, L., Holman, G.D., & Dennis, B.R. 2006, ApJ, 645, L157Vestrand, W.T., Share, G.H., Murphy, R.J., et al. 1999, ApJS, 180, 409Zimovets, I., & Struminsky, A. 2012, Sol. Phys., 281, 749
This preprint was prepared with the AAS L A TEX macros v5.2.
10 – N u m be r o f f l a r e s Year Total class B class C class M class X
Fig. 1.— Distribution of solar flares detected by Konus-
Wind in the trigger mode over time. 11 – −3 no r m a li ze d c oun t s s − k e V − data and folded model
100 100050 200 500−202 χ Energy (keV)
Fig. 2.— Spectrum of the flare SOL2002-03-10T01:34 (see section 3.3, measured by Konus-
Wind during the interval 0–5.632 s. The spectrum is well described by a power-law model with the index γ = 2 . ± .
30 ( χ =46/55 dof); significant emission is seen up to ∼ G3 (300-1160 keV) c oun t s / . s G1 (18-70 keV)0.00.10.20.3 G / G -250 -200 -150 -100 -50 0 50 100 150 200 2500.000.050.10 T - T (s) G / G c oun t s / . s P ho t on I nde x T-T (s) F l u x ( - e r g / c m / s ) Fig. 3.— Solar Flare on November 8, 2012 (class M1.7). T =8365.302 s UT (02:19:25.302). 13 – G3 (300-1160 keV)
G2 (70-300 keV) c oun t s / . s G1 (18-70 keV) G / G -150 -100 -50 0 50 100 150 200 2500.000.050.10 T - T (s) G / G GO ES f l u x ( - e r g c m - s - ) GOES (1-8 Å ) c oun t s / . s P ho t on I nde x T-T (s) F l u x ( - e r g / c m / s ) Fig. 4.— Solar Flare on March 10, 2002 (class C5.1) – “early impulsive flare”. T =5693.874 s UT(01:34:53.874). 14 – c oun t s / . s c oun t s / s pho t on s / c m / s Konus-Wind (70-300 keV)Helicon-Coronas-F (2.223 MeV)
Fig. 5.— Time profile of the solar flare on October 26, 2003 drawn from the Konus-
Wind and
INTEGRAL
SPI-ACS data and the evolution of the 2.223 MeV line intensity derived from theHelicon data. 15 – sp9(268.44-301.99 s) - c oun t s s - k e V - sp10(301.99-335.55 s) E (MeV) sp11(335.55-369.10 s) - c oun t s s - k e V - E (MeV) sp12(369.10-402.66 s)
Fig. 6.— Four consecutive spectra measured by Helicon during the flare SOL2003-10-26T06:11(continuum subtracted). The spectrum accumulation times are given relative to the Konus-
Wind trigger time T =22285.091 s UT (06:11:25.091). The spectrum shows the 2.223 MeV neutroncapture line and the strong 1.460 MeV line caused by the decay of40