Observations of ‘‘zebra’’ pattern in cm-range with spatial resolution

Advances in Space Research 35 (2005) 1789–1794
Observations of ‘‘zebra’’ pattern in cm-range with spatial resolution
Alexander T. Altyntsev a, Alexey A. Kuznetsov
Natalya S. Meshalkina a, Yihua Yan
Department of Radioastrophysics, Institute of Solar-Terrestrial Physics SB RAS, Lermontov Street 126, Irkutsk 664033, Russia
National Astronomical Observatories, Beijing, China
Received 28 September 2004; received in revised form 6 January 2005; accepted 10 January 2005
We present the results of the first observations of the solar microwave burst with fine spectral structure of zebra type at the
frequency about 5.7 GHz. The burst has been detected simultaneously by the Siberian Solar Radio Telescope and by the spectropolarimeter of the National Astronomical Observatory of China. Zebra pattern consisted of three parallel stripes with complex
frequency drift. The degree of circular polarization of emission reached 100%, the polarization sense corresponded to the extraordinary wave (X-mode). We have determined the plasma parameters in the emission source: plasma density about 1011 cm3,
magnetic field strength 60–80 G. We argue that in the given event the most probable mechanism of the zebra pattern generation
is non-linear coupling of harmonics of Bernstein modes.
2005 COSPAR. Published by Elsevier Ltd. All rights reserved.
Keywords: Solar flares; Microwave emission; Zebra pattern; Non-linear processes
1. Introduction
Observations of radio bursts with fine spectral structure of zebra type, seen as a number of parallel bright
stripes in the dynamic spectrum, give unique means
for the local measurements of the magnetic field in the
solar corona. In the meter and decimeter waves, the
bursts with zebra pattern are observed rather frequently.
Until now, the highest frequency at which a zebra pattern has been observed is 3.8 GHz (Chernov et al.,
2003). There are several approaches to the interpretation
of zebra-like fine structure (Rosenberg, 1972; Aurass
et al., 2003; Zlotnik et al., 2003; Chernov, 1976; LaBelle
et al., 2003.
We present the results of the first observations of the
zebra pattern burst at the frequency about 5.7 GHz. The
burst occurred during the solar flare on January 05, 2003
Corresponding author. Tel.: +7 3952 437299.
E-mail addresses: [email protected] (A.T.
[email protected] (A.A. Kuznetsov).
in AR 0243. It was observed simultaneously by the Siberian Solar Radiotelescope (SSRT) and by the spectropolarimeter of the National Astronomical Observatory of
China (NAOC).
2. Observation techniques
The dynamic spectra of zebra structures were received by the Solar Radio Broadband Fast Dynamic
Spectrometers (5.2–7.6 GHz) at the Huairou Solar
Observing Station of NAOC. The reception band of
the NAOC spectropolarimeter individual frequency
channel is 20 MHz, and the temporal resolution is
5.9 ms.
The spatial characteristics of the microwave sources
were recorded by the SSRT. The SSRT is a crossed
radio interferometer, consisting of two lines of antennas,
the east-west (EW), and the north-south (NS), operating
in the 5.67–5.79 GHz frequency range. Radio maps of
the solar disk are recorded at intervals of 3–5 min. The
0273-1177/$30 2005 COSPAR. Published by Elsevier Ltd. All rights reserved.
A.T. Altyntsev et al. / Advances in Space Research 35 (2005) 1789–1794
investigations of the fine temporal structure of flare
bursts are based on the data, recorded by the EW and
NS arrays, which provide one-dimensional images
(scans) of the solar disk every 14 ms simultaneously.
The SSRT beam width is down to 1500 and depends on
the array direction and the observation local time.
We used also the data from Nobeyama spectropolarimeters, SOHO/MDI magnetograms, SOHO/EIT
images in the soft ultraviolet radiation.
3. Observations
The radio emission profiles of the flare are shown in
Fig. 1. The background microwave burst has a relatively
low polarization degree. The spike-like pulses with duration of several seconds were observed at the frequencies
below 5.7 GHz. At 3.8 GHz the right-hand polarized
pulses were registered for 12 min.
We managed to detect two events with zebra pattern
in the dynamic spectrum. Zebra structures were observed 3 min prior to the background burst maximum.
One can see the highly polarized bursts at 3.8 GHz at
the same time. The NAOC dynamic spectra in 2.6–
3.8 GHz frequency range show that these bursts were
short-duration wide-band pulses, without zebra
The brighter zebra pattern is shown in Fig. 2. The
dynamic spectrum contains three bright stripes that
had no frequency drift firstly. Then, the drift towards
lower frequencies began, and the duration of this
(decreasing) branch is about 2.5 s. One can notice the
second (increasing) branch with duration up to 2 s,
although this part of the burst is expressed more
poorly. Thus, the spectrum has the U-like shape. The
mean frequency of the event can be estimated as
5.6 GHz. The frequency gap between adjacent bright
stripes is about 0.16 GHz and this value is the same
during the whole burst. The instantaneous bandwidth
of stripes is about 0.06 GHz. We note, that all the zebra stripes started their drift simultaneously, with the
accuracy within 50 ms, which shows that the emission
of all stripes has the same source. The zebra structure
is recorded by both instruments in the right handed
mode only. Thus, the zebra pattern circular polarization degree reaches 100%.
Another observed zebra pattern consisted of four
stripes without frequency drift, the frequency separation
Fig. 1. The temporal profiles of the radio emission fluxes of right (continuous curve) and left (dotted curve) handed circular polarization, recorded by
the Nobeyama and Huairou spectropolarimeters (magnitude in sfu) during the flare on January 05, 2003. Vertical lines mark the zebra pattern times.
A.T. Altyntsev et al. / Advances in Space Research 35 (2005) 1789–1794
Fig. 2. The dynamic spectrum with zebra pattern (at the top, the darker areas correspond to the emission higher intensity) and the time profiles,
recorded by the NAOC spectropolarimeters and the SSRT linear arrays at the same frequencies.
between emission stripes was 0.22 GHz. Like in the first
event, the fine structures were observed only in the rightpolarized channel. Unfortunately, the determination of
source spatial parameters in this case is less accurate,
so we will not discuss it further.
The magnetogram of the flare region is shown in Fig.
3 (left panel). The magnitude of the field in the leading
spot of the S-polarity reaches 875 G; the spot of the
N-polarity is displaced 4500 to the north-east. The contours in Fig. 3 show the emission brightness in the UV
(195 Å) at the flare peak. The UV contours show a loop
connecting the two sites with the magnetic fields of
opposite direction (sources 1 and 2). The distance between them is about 3500 . The crossing of the straight
lines points to the brightness centroid of the zebra
source, which is located near the north-east loop foot,
where the magnetic field value (at the photosphere) is
about 150 G of the N-polarity (source 1). Thus, the zebra emission polarization corresponds to the extraordinary wave (X-mode). This conclusion is confirmed by
the coincidence of the polarization signs of the zebra
pattern and the background burst. The background
microwave burst is most probably caused by the gyro-
synchrotron emission mechanism, which produces preferably extraordinary wave.
The structure of the microwave sources at 5.7 GHz is
shown in Fig. 3 (right panel). There is a correspondence
of sources 1 and 2 to the sources with the right and left
polarization, accordingly. The brightness center of the
background flare burst is close to source 1. The same
structure of the background burst is seen in the NoRH
radio maps at 17 GHz. The straight dashed lines (right
panel) define the bands around the straight lines shown
on the magnetogram (left panel). Their widths are equal
to the half widths of the corresponding interferometer
beams, and can be considered as the upper estimate of
the error in the source location. Thus, the zebra source
is located within the parallelogram formed by the
straight dashed lines. The overlapping with the magnetogram shows that the magnetic field inside the parallelogram has N-polarity. In Fig. 3, we also show the
magnetic field lines, calculated in the potential field
approximation using the MDI magnetogram. The calculation method was developed by Rudenko (2001).
In Fig. 4 the one-dimensional distributions of the
radio brightness (scans), recorded by the EW array,
A.T. Altyntsev et al. / Advances in Space Research 35 (2005) 1789–1794
Fig. 3. Left: The magnetogram (halftone) and the UV emission map (contour) of AR 0243. The crossed straight lines show the directions of scanning
by SSRT interferometers. Zebra pattern source is situated at the point of intersection of these lines. Some reconstructed magnetic field lines are also
shown. The source of zebra pattern most probably lies on the line drawn as thick. Right: The structure of the microwave emission at 5.7 GHz. The
brightness temperature in intensity is shown by grayscale. The dashed lines correspond to the contours in polarization. The straight dashed lines
mark the bands on the solar disk, the emission from which is recorded during the zebra pattern observation.
Fig. 4. The one-dimensional distributions (intensity vs. coordinate) of the microwave emission (scans) and polarization degree, recorded by the
SSRT/EW array. Top panel shows the scans of the background burst in different polarizations and the scans of the different stripes of zebra emission;
bottom panel shows the burst polarization degree. The background burst scan is determined as the scan at the 06:06:12.274 moment, when the zebra
stripes went out the SSRT receiving band. The zebra scans are obtained by subtracting the background scan from the scans, corresponding to the
zebra brightening.
A.T. Altyntsev et al. / Advances in Space Research 35 (2005) 1789–1794
are shown in the intensity and polarization. The relative
scans of the zebra source (with subtracted background)
are shown for the two consecutive moments, corresponding to recording the zebra pattern different stripes.
It is evident, that the different stripe sources spatially
coincide, and they are displaced a few arcsec to the east
relative to the brightness center of the background
burst. The zebra source size is less than 1000 or
7000 km, and is significantly smaller than the background burst size. At the background burst polarization
scan the positions of 130 and 16500 correspond to the
loop footpoints, and one can see different polarization
signs at these points. Note that we show the relative zebra scans only for the right polarized channel. The leftpolarized emission level after background subtraction is
reduced to zero, without any fine structures.
Using the GOES X-ray observations, we can estimate
the plasma temperature in the microwave emission
source as 1.1 · 107 K, and the plasma density (from
the integral emission measure) as n . 7.2 · 1010 cm3.
This means that the emission is generated at the frequency close to the double plasma frequency.
In our opinion, the model of non-linear interaction of
Bernstein waves is the most suitable to explain the polarization, spectral and temporal parameters of the observed zebra pattern. In this case, the frequency
interval between adjacent bright stripes is almost equal
to the electron cyclotron frequency Dm = (eB)/(2pmc) in
the emission source. Therefore, the observations of a zebra pattern provide a direct method for magnetic field
magnitude measurement. In the event under consideration, we managed to detect the zebra patterns twice
with frequency intervals of 220 and 160 MHz (Fig. 2),
respectively. These values correspond to the magnetic
field strength B of 60–80 G in the zebra source.
One of the magnetic lines connecting sources 1 and 2
is marked by the thick line in Fig. 3 (left panel). Its
length is 130,000 km and its height is about 45,000 km.
It is remarkable that the point of this line with the magnetic field value of 60 G is close to the visible position of
the zebra source. So, we believe that this site, marked by
the cross in Fig. 3 (left panel), corresponds to the zebra
source location. The height of this point above the photosphere is about 14,000 km.
We also note that there were three fluctuations of
microwave burst intensity with period about 170 s during the flare (Fig. 1). These fluctuations can be caused
by the MHD waves that propagate along the loop with
Alfven velocity.
4. Interpretation
There are several approaches to interpretation of the
bursts with the fine structure of zebra type, which were
developed with reference to bursts in the meter and deci-
meter ranges. In the observed event, the two factors are
of main importance when determining the emission
100% polarization degree, corresponding to extraordinary wave.
Small source size.
The small source size follows not only from the direct
measurements, but also from the high synchronism of
the frequency drifts of different stripes. Indeed, it is seen
in Fig. 2 that zebra stripes have no frequency drift
firstly. Then, at 06:06:11.6, all zebra stripes start their
drift simultaneously, with possible delay not more than
50 ms (note that here, we discuss the onset of frequency
drift, not the onset of the emission itself). The change of
emission frequency requires the change of source parameters (such as plasma density and magnetic field); this
change must be simultaneous in all parts of the emission
source. In the magnetized plasma the time scales of variation of plasma density and magnetic field are determined by the Alfven velocity (for the MHD waves
which are the fastest disturbances affecting the above
mentioned plasma parameters). Thus for plasma density
1011 cm3, magnetic field B 6 200 G and time delay
Dt 6 50 ms we obtain source size r < 100 km.
In our opinion, among the mechanisms that have
been proposed for an interpretation of zebra patterns,
the process of non-linear coalescence of Bernstein
waves (Zheleznyakov and Zlotnik, 1975a,b; Mollwo
and Sauer, 1977) seems to be the most suitable. It suggests, that all emission stripes are generated in the
same compact source. In the observed microwave zebra pattern (Fig. 2) the ratio of frequency interval between adjacent zebra stripes to the emission frequency
is Dx/x . 1/35, which corresponds to the coalescence
of cyclotron harmonics with numbers about 17–18.
In this process, the polarization corresponds to the
extraordinary wave (Mollwo and Sauer, 1977). The
numerical modeling carried out by Haruki and Sakai
(2001) shows that the polarization degree of the electromagnetic waves (in the X-mode sense) can reach
100% for coupling of Bernstein modes. The investigation by Willes (1999) shows that the efficiency of the
non-linear interaction is sufficient to provide radio
emission with the observed intensity; it is also shown
that the polarization sign depends on the specific conditions, but in most cases it corresponds to the
extraordinary wave. The polarization degree can be
as high as 60% in the X-mode sense. However, these
investigations correspond to the conditions that somewhat differ from the conditions in the observed flare
loop. Thus, we can conclude that the emission mechanism of the observed microwave zebra pattern is not
clearly determined yet: the coalescence of Bernstein
modes is the most probable model of the existing
A.T. Altyntsev et al. / Advances in Space Research 35 (2005) 1789–1794
ones, but the further investigation of this process is
necessary (and, possibly, a new mechanism has to be
This work was supported by the RFBR (02-02-39030,
03-02-16229, 04-02-39003 and 04-02-31012), Russia
Department of Education (E02-3.2-489), Integratsiya
project (I0208), MOST (G2000078403), NSFC
(10225313 and 10333030), and CAS.
Aurass, H., Klein, K.-L., Zlotnik, E.Ya., Zaitsev, V.V. Solar type IV
burst spectral fine structures. I. Observations. Astron. Astrophys.
410, 1001–1010, 2003.
Chernov, G.P. Microstructure in continuous emission of type IV meter
bursts. Modulation of continuous emission by wave packets of
whistlers. Soviet Astron. 20, 582–589, 1976.
Chernov, G.P., Yan, Y.H., Fu, Q.J. A superfine structure in solar
microwave bursts. Astron. Astrophys. 406, 1071–1081, 2003.
Haruki, T., Sakai, J.-I. Electromagnetic wave emission from a
dynamical current sheet with pinching and the coalescence of
magnetic islands in solar flare plasmas. Astrophys. J. 552, L175–
L179, 2001.
LaBelle, J., Treumann, R.A., Yoon, P.H., Karlicky, M. A model of
zebra emission in solar type IV radio bursts. Astrophys. J. 593,
1195–1207, 2003.
Mollwo, L., Sauer, K. A model explaining type IV continuum bursts
by coherent non-linear interaction of Bernstein waves. Sol. Phys.
51, 435–458, 1977.
Rosenberg, H. A possibly direct measurement of coronal magnetic
field strengths. Sol. Phys. 25, 188–196, 1972.
Rudenko, G.V. Extrapolation of the solar magnetic field within the
potential-field approximation from full-disk magnetograms. Sol.
Phys. 198, 5–30, 2001.
Willes, A.J. Polarization of multiple-frequency band solar spike bursts.
Sol. Phys. 186, 319–336, 1999.
Zheleznyakov, V.V., Zlotnik, E.Ya. Cyclotron wave instability in the
corona and origin of solar radio emission with fine structure. I.
Bernstein modes and plasma waves in a hybrid band. Sol. Phys. 43,
431–451, 1975a.
Zheleznyakov, V.V., Zlotnik, E.Ya. Cyclotron wave instability in the
corona and origin of solar radio emission with fine structure. III.
Origin of zebra-pattern. Sol. Phys. 44, 461–470, 1975b.
Zlotnik, E.Ya., Zaitsev, V.V., Aurass, H., Mann, G., Hofmann, A.
Solar type IV burst spectral fine structures. II. Source model.
Astron. Astrophys. 410, 1011–1022, 2003.