Influence of solar flare's location and heliospheric current sheet on the associated shock's arrival at Earth

Page created by Teresa Taylor
 
CONTINUE READING
JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 112, A06107, doi:10.1029/2006JA012205, 2007
  Click
  Here
    for
 Full
Article

Influence of solar flare’s location and heliospheric
current sheet on the associated shock’s arrival at Earth
Xinhua Zhao,1,2 Xueshang Feng,2 and Chin-Chun Wu3
Received 3 December 2006; revised 9 March 2007; accepted 15 March 2007; published 5 June 2007.

[1] We study the source locations of 130 solar flare-type II radio burst events with the
associated interplanetary shocks observed by L1 spacecraft (type A events) and
217 flare-type II events without such shocks observed at L1 (type B events) during
February 1997–August 2002. In particular, we investigate the relative positions between
the flare sources, the heliospheric current sheet (HCS), and the Earth. We found the
following results: (1) Solar flares are usually distributed within [S30°, N30°] in
heliographic latitude and [S30°, N30°]  [E10°, W30°] is the predominant source region
on the solar disk that includes the majority of geoeffective solar flares. (2) The shocks
with the associated flares located near the HCS would have a lower probably of
reaching the Earth. For the Earth-encountered shocks, their initial speeds are distinctly
higher when their associated flares are located near the HCS. (3) The angular distance
from the flare source to the Earth (defined as Y below) also contributes to the probability
of the associated shock being observed at the Earth. The shock arrival probability
decreases with the increment of Y and the mean initial shock speed increases with Y for
those Earth-encountered shocks. (4) The so-called ‘‘same-opposite side effect’’ of the
HCS is confirmed to exist. That is, the shocks whose associated flares are located on the
same side of the HCS as the Earth (called as ‘‘same side events’’) have a greater chance
of reaching the Earth than those shocks with their associated flares on the opposite
side (‘‘opposite side events’’). Here for the first time, a comprehensive sample of solar
transient events of both arriving and nonarriving ones (at Earth) is used to testify to the
same-opposite side effect. These results would be valuable in understanding the
solar-terrestrial relations, and helpful for space weather prediction.
Citation: Zhao, X., X. Feng, and C.-C. Wu (2007), Influence of solar flare’s location and heliospheric current sheet on the associated
shock’s arrival at Earth, J. Geophys. Res., 112, A06107, doi:10.1029/2006JA012205.

1. Introduction                                                           produce corresponding geoeffects [e.g., Cane et al., 2000;
                                                                          Park et al., 2002; Wang et al., 2002; Cid et al., 2004;
  [2] It is widely accepted nowadays that nonrecurrent                    Yermolaev et al., 2005; Zhao et al., 2006]. Many factors
interplanetary (IP) and geomagnetic disturbances come                     would affect the probability of a solar transient encountering
from solar transient activities such as solar flares, erupting            the Earth, such as its source location, strength, angular
prominences, type II radio bursts, coronal mass ejections                 width, propagation direction, speed, trajectory, surrounding
(CMEs) and so on. Arrival at the Earth of these transient                 coronal/interplanetary plasma and magnetic environment,
disturbances usually indicates the possible initiation of                 and also its interaction with other transients and/or back-
geomagnetic storms if sufficiently long and sufficiently                  ground solar wind as being pointed out by Cane et al.
large magnitude southward components of the interplane-                   [2000], Wang et al. [2004, 2006], and Zhao et al. [2006].
tary magnetic field exist following the interplanetary dis-                 [3] The source location of a solar transient is perhaps one
turbances [Dryer, 1994]. On the other hand, statistical                   of the most important factors. Webb et al. [2000] found that
studies have demonstrated that not all the solar transients               halo CMEs associated with surface activity within 0.5 Rs of
happening at the frontside of the Sun hit the Earth and                   the Sun center appear to be an excellent indicator of
                                                                          increased geoactivity several days later. Cane et al. [2000]
    1
      Solar-Interplanetary-Geomagnetic Weather Group, State Key Labora-
                                                                          pointed out that most of the interplanetary ejecta observed at
tory for Space Weather, Center for Space Science and Applied Research,    Earth are associated with halo CMEs within [E40°, W40°]
Chinese Academy of Sciences, Beijing, China.                              in longitude. Especially, for the transients that produce
    2
      College of Earth Sciences, Graduate University of the Chinese       strong geomagnetic disturbance at Earth, their solar source
Academy of Sciences, Beijing, China.                                      distribution has the east-west asymmetry in longitude. The
    3
      Center for Space Plasma and Aeronomic Research, University of
Alabama in Huntsville, Alabama, USA.                                      great magnetic storms are usually caused by the solar
                                                                          transients coming from the Western Hemisphere [e.g., Wei
Copyright 2007 by the American Geophysical Union.                         et al., 1985; Wei and Deng, 1987; Wang et al., 2002; Cane
0148-0227/07/2006JA012205$09.00

                                                                   A06107                                                         1 of 9
A06107                   ZHAO ET AL.: INFLUENCE OF SOLAR FLARE LOCATION AND THE HCS                                       A06107

and Richardson, 2003; Zhang et al., 2003; Zhao et al.,
2006]. Zhao et al. [2006] used a large data set to statistically
study the characteristics of solar flares associated with IP
shock or nonshock events at L1 and found that the most
probable location for flares associated with IP shocks at
Earth is W20° in longitude. Our results also indicated that
the flare-associated shocks originating from the longitudinal
range of [E10°, W30°] would have higher probability to
reach the Earth.
  [4] The effect of the heliospheric current sheet (HCS) on
the transmission and geoeffects of solar transients is another
interesting topic. Wei et al. [1990, 1991] studied the source
distribution of 277 flare-shock events on the solar synoptic
maps and found that the same side events whose sources are
located on the same side of the HCS as Earth are more
prone to arrive at Earth than the opposite side events, which
is derived from the fact that the number of the same side
events is much larger than that of the opposite side events
among the total events. The idea of Wei et al. [1990, 1991]
has been demonstrated and applied by others [Dryer, 1994;
                                                                   Figure 1. Heliographic latitude as a function of time for
Odstrcil et al., 1996; Tokumaru et al., 2000; Zhao and
                                                                   the whole 347 solar flares.
Feng, 2005].
  [5] In this paper, we follow the previous work of Zhao
et al. [2006] and continue to search for the difference            presence of an associated interplanetary shock observed at
between two kinds of solar disturbances, i.e., those asso-         L1 spacecraft about 1 – 5 days later, which are marked as
ciated with/without IP counterparts observed at near Earth         type A events. The other 217 nonshock events without the
space. We lay emphasis on the latitudinal distribution this        associated IP shock observed at the near Earth space are
time. Especially we take into account the location and shape       marked as type B. The whole data set covers an uninter-
of the HCS, and calculate quantitatively the angular distance      rupted period of about 5 years during the rising and
from flare source to the HCS. This latitudinal regulation          maximum phase of solar cycle 23, and they can serve as
combined with the longitudinal restrictions obtained by            an unbiased database for statistically studying the solar/
Zhao et al. [2006] would give a more explicit depiction            interplanetary transient disturbance and their geoeffects.
of the impact of solar sources on the presence of the              The detailed information about these events can be found
associated IP disturbances at Earth. We utilize the same           in space weather ‘‘Fearless Forecast’’ of Fry et al. [2003]
data set as Zhao et al. [2006] and use solar flare as a proxy      and McKenna-Lawlor et al. [2006].
for solar transient activity. It is well known that after a long
debate on the controversial ‘‘solar flare myth’’ controversy,
people begin to know that CMEs and flares are closely              3. Statistical Analysis
correlated solar transient phenomena regarded as the solar         3.1. Heliographic Latitude
source of the IP disturbances. CMEs and flares are probably          [7] The heliographic coordinate system is widely used in
two phenomena in one process just as recognized by many            describing the location of a point on the solar surface.
previous studies [e.g., Dryer, 1996; Harrison, 1996; Zhao          Figure 1 gives the heliographic latitude variation versus
et al., 2006].                                                     occurrence time for these 347 solar flares including both
                                                                   type A and type B events. It can be seen from this figure
2. Data Set                                                        that solar flares do not appear randomly on solar surface but
                                                                   are concentrated in two latitude bands on either side of the
  [6] We use the data set of 347 solar flare-type II radio         equator. At solar activity minimum years, solar flares are
burst events during February 1997 – August 2002 collected          mainly concentrated near ±30°; as solar activity increases,
by Zhao et al. [2006]. The near Sun observations of                flares emerge closer and closer to the equator in the course
these events include: (1) the start time of the event denoted      of a cycle and reach about ±15° at solar maximum years of
by the beginning of the metric type II burst; (2) solar            2001 and 2002. The distribution straddles the solar equator,
activity location provided by optical Ha flare observations;       and looks like two wings of a butterfly. Such a latitudinal
(3) initial coronal shock speed estimated by the metric radio      distribution characteristic is similar to the well-known
type II temporal drift in plasma frequency with the use of a       butterfly diagram of sunspots, suggesting that this might
coronal density model; (4) piston driving time suggested by        be an intrinsic feature of solar active regions. This is also
the proxy duration of the GOES spacecraft’s soft X-ray             consistent with previous studies [e.g., Maunder, 1904;
enhancement due to the flare; (5) flare intensity defined by       Byrne et al., 1964; Knoska and Krivsky, 1978; Li et al.,
the maximum flux of X-rays (of wavelength 0.1 to 0.8 nm)           1998; Mari and Popescu, 2004].
during the flaring process; and (6) background solar wind            [8] Figure 2 gives the event number histogram of solar
speed. The type II radio burst is interpreted as the signature     flares along their heliographic latitude for type A (Figure 2a)
of shock wave initiation in the solar corona. Out of the 347       and type B (Figure 2b). The x coordinate in this figure
events in our sample, there are 130 shock events with              represents the latitude bins of flares, and the y coordinate

                                                              2 of 9
A06107                   ZHAO ET AL.: INFLUENCE OF SOLAR FLARE LOCATION AND THE HCS                                      A06107

         Figure 2. The event number frequency of solar flares plotted versus their heliographic latitude in
         interval of 10° for (a) type A and (b) type B. The event number on each latitude bin is also shown.

represents the event numbers on each latitude bin of 10°          activity conditions and topology of the HCS during the
from 40° to 40°. The event numbers on each latitude bin          time the data were taken.
of 10° are also displayed. It can be seen from this figure that
solar flares do not occupy the whole solar disk in latitude       3.2. The Relative Locations of the Flare Source, the
direction. Most of the flares are distributed within the          HCS, and the Earth
latitude range of ±30° for both type A and type B events            [10] The above discussed heliographic latitude is in fact
and rare flares appear in high latitude region over ±30° as       the angular distance from flare source to the solar equator.
shown by the dotted lines in Figures 2a and 2b. Especially,       In order to depict the relative positions between the flare
the histogram distribution presents a double-peak charac-         source, the heliospheric current sheet (HCS) and the Earth
teristic with the maximum located near ±(10° –20°). For the       quantitatively, we define new parameters of lS and Y. Here
130 solar flares associated with interplanetary shocks at L1      lS is the angle from the flare source to the HCS, thus
of type A, 125 of them are located within the latitude range      denotes the latitude relative to the HCS (abbreviated as
of ±30°, accounting for 96.2% (125/130); similarly, 208 of        LRHCS) of flares; while Y stands for the angular distance
the 217 nonshock flares are located within the latitude range     from source to the Earth (abbreviated as ADFSE). We use
of ±30°, accounting for 95.9% (208/217). Therefore the            one event as an example to show how to obtain these two
latitude range [30°, 30°] would include the majority of          parameters. The type II radio burst of this event started at
solar sources of flares associated with/without interplanetary    03:18 UT on 8 August 1998 with an associated M3.0 solar
shocks. Considering the conclusion about longitude restric-       flare located at N30°E09°; the estimated initial shock speed
tion by Zhao et al. [2006], we can see that [S30°, N30°]         is 600 km/s, and the corresponding IP shock arrived at the
[E10°, W30°] is the predominant source region on solar            L1 spacecraft, ACE, at 22:40 UT on 11 August. Figure 3a
disk from which nonrecurrent geoeffective solar disturban-        gives the spread of the flare source on the contour of the
ces most likely originate. This is a closed rectangular region    solar source surface magnetic field for CR 1939 located at
including the disk center of the Sun. Solar transient dis-        2.5 solar radii from the Wilcox Solar Observatory (http://
turbances originating from this region would have higher          wso.stanford.edu/). This Carrington rotation started at
probability to reach the Earth and cause relatively intense       09:44 UT on 1 August 1998 and ended at 16:20 UT on
geomagnetic activity. Although the region boundary is             28 August, containing the time of this event. The thick red
derived from statistical study and is only a coarse estima-       line in this figure is the HCS at 2.5 Rs. On the basis of the
tion, this region can be used as a helpful guide to roughly       start time and flare location, we find the point S for the
judge a solar transient’s geoeffectiveness by only basing on      source location in Figure 3a. P is the foot of the perpen-
its source location.                                              dicular from the point S onto the HCS, meaning that the
  [9] For type A events in Figure 2a, 71 flares occurred in       spherical distance from S to P is the shortest arc. E denotes
the Northern Hemisphere but only 59 flares in the Southern       the Earth location at the start time of the event, and H is its
Hemisphere. The number of northern flares is 20% 7159            foot of the perpendicular on the HCS. The LRHCS of the
                                                          59
higher than that of southern flares. While for the type B         flare, i.e., lS, is ffSOP as shown in Figure 3b. Figure 3b is
events in Figure 2b, 107 flares occurred in the Northern          the three-dimensional display of Figure 3a. The spherical
Hemisphere and 110 flares in the Southern Hemisphere;             surface in Figure 3b is the source surface at 2.5 solar radii,
their event numbers are nearly equivalent. The reason for         and the gridding denotes the heliographic longitude and
the preponderance of the northern flares among the total          latitude projected onto the source surface. The shape and
shock-arriving events may be correlated with the solar            location of the HCS, point S, P, E in Figure 3b are the same

                                                             3 of 9
A06107                   ZHAO ET AL.: INFLUENCE OF SOLAR FLARE LOCATION AND THE HCS                                      A06107

         Figure 3. (a) The spread of solar flare source on the source surface magnetic field located at 2.5 solar
         radii, (b) Its three-dimensional display. The red solid line stands for the HCS at solar source surface;
         S and E denote the flare source and Earth position projected onto the source surface, P and H are
         their perpendicular feet, respectively; O is the center of the Sun.

as those in Figure 3a. The LRHCS of solar flare lS can be       graphic latitude, lS is defined to be positive when the flare
obtained from the following equation:                           lies to the north of the HCS and negative on the other hand.
                                                                In this figure, Ni (i = 1; 2; . . . ; 5) stands for the number of
           lS ¼ arccosðcosðqS  qP Þ cosðfS  fP ÞÞ       ð1Þ   type A events with their lSs in the bins of [50°, 30°],
                                                                [30°, 10°],. . ., [30°, 50°], respectively; and ni stands for
Here (qS; fS) and (qP; fP) denote the heliospheric longitude    the corresponding type B event number. Then the event
and latitude of S and P in Figure 3a. Another parameter Y is    percentage of type A (Pi) and type B (pi) can be obtained as
angle ffSOE, that is:                                            follows:

            Y ¼ arccosðcosðqS  qE Þ cosðfS  fE ÞÞ       ð2Þ                                      Ni
                                                                                          Pi ¼                               ð3Þ
                                                                                                 Ni þ ni
and (qE; fE) stands for the heliospheric longitude and
latitude of E in Figure 3a. The Y, defined in this way,
                                                                                                        ni
presents the angular span from the solar transient’s erupting                        pi ¼ 1  Pi ¼                           ð4Þ
normal to the Sun-Earth direction. It will be seen below that                                        Ni þ ni
this parameter Y can have an effect on the propagation/
arrival of the associated IP shock. From Figure 3a or 3b,         [12] Therefore Pi and pi, derived in this way, reflect the
8 August 1998 flare is located on the opposite side of          relative percentage of type A and type B events within fixed
the HCS as the Earth. Then this event is called a ‘‘opposite    lS interval. The solid line in Figure 4 denotes the percent-
side’’ event. Otherwise, the event will be called a ‘‘same      age of type A events (Pi) every 20° along lS; and the dashed
side’’ event. For the 347 events in our data set, we compute    line denotes the corresponding percentage of type B events
lS and Y for each event, and divide it into the same side or    (pi). The event numbers of Ni and ni in each bin are shown
opposite side category. In the following, we try to             on the solid/dashed lines together with Vi and vi (in units of
investigate the contribution of these new defined parameters    km/s), which represents the mean value of initial shock
to the associated shock’s arrival probability at the Earth.     speed for these Ni type A events (Vi) or ni type B events (vi).
These new parameters include lS, Y, and whether the flare       Figure 4 indicates that the event percentage is not averagely
and the Earth are on the same side or on opposite side of the   distributed along solar flare’s lS. The percentage of type A
HCS. They are discussed in the following three subsections,     events decreases gradually when lS approaches zero, while
respectively.                                                   that of type B events increases. The percentage of type A is
3.2.1. Latitude Relative to the Heliospheric Current            minimum at [10°, 10°] of lS. This means that the shocks
Sheet (ls)                                                      originating near the HCS have a smaller probability of
  [11] Figure 4 gives the event number percentage distri-       reaching the Earth. The HCS, as we know, is the magnetic
bution of type A and type B along lS. Similar to helio-         field reversal region embedded within the heliospheric

                                                           4 of 9
A06107                       ZHAO ET AL.: INFLUENCE OF SOLAR FLARE LOCATION AND THE HCS                                           A06107

          Figure 4. The percentage distribution along solar flare’s latitude relative to the HCS (lS) for type A
          (solid line) and type B (dashed line) events. Ni and ni are the event number of type A and type B within
          the fixed lS interval; Vi and vi are their mean initial shock speed, respectively.

plasma sheet (HPS) with high density and low solar wind                    3.2.2. Angular Distance From Flare Source to the
speed. Therefore the interaction of solar transient disturban-             Earth (Y)
ces with the HCS/HPS would greatly change the propaga-                       [14] The event number percentage distribution along Y,
tion process and signatures of these solar disturbances [e.g.,             i.e., the angular distance from the flare source to the Earth,
Wu and Dryer, 1997; Smith et al., 1998; Hu and Jia, 2001].                 is shown in Figure 5. In this figure, Ni (i = 1; 2;. . .; 6) is
That is to say, the shocks near the HCS would have larger                  the type A event number with Y in the bins of [0°, 20°],
probability of interacting with the HCS and then to weaken                 [20°, 40°], . . . , [100°, 120°] respectively; and ni is the
or even dissipate before reaching the Earth.                               corresponding type B event number. The event percentage
  [13] Figure 4 also shows that the mean value of initial                  of type A (Pi) and type B (pi) are derived according to
shock speed of type A events is larger than that of type B                 equations (3) and (4) in the same way. The event number of
events at all the five lS intervals, i.e., Vi > vi for i = 1; 2; . . . ;   Ni and ni together with their mean initial shock speed of Vi
5. The initial shock speed, as we know, can be regarded as a               and vi are also labeled on the solid and dashed lines as we
proxy for the energy of solar transient events and used as                 have done in Figure 4. Figure 5 indicates that the percentage
input for predicting shock arrival time in the prevalent                   of type A event decreases with the increment of Y, while
‘‘fearless forecast’’ models. Especially, the ISPM model                   that of type B increases. This reveals the potential contri-
supposed that the total energy of a solar transient event is               bution of Y to the IP shock’s arrival ability. As a reasonable
expected to be proportional to the kinetic energy flux, i.e.,              approximation, the flare source can be regarded as the
E / V3s [Smith and Dryer, 1990, 1995]. This means that the                 originating place of the associated shock on solar surface;
energy of type A shock-associated events is higher than                    then the line through the Sun center and the flare source (OS
that of type B nonshock events in statistical sense. That is               in Figure 3b) is assumed to be the shock’s erupting normal,
to say, the shocks of more energy would have larger                        which can be taken as the main propagation direction of the
probability to arrive at Earth before decaying into ordinary               shock in interplanetary space; also, the shock intensity is the
MHD waves during their passage. This is consistent with                    strongest along the main propagation direction; therefore,
the work of Zhao et al. [2006], where the X-ray peak flux                  large Y event corresponds to the large deviation from the
is used as an indicator for the total energy released in the               shock’s main propagation direction to the Sun-Earth line;
flaring process. Another thing to mention is that for the                  the flank of a shock is relatively weak, then the satellites at
type A events originating near the HCS (lS within [10°,                   L1 would have a higher probability to miss the shock in this
10°]), the mean value of their initial shock speed (V3 =                   case. This is why the shocks associated with solar flares
1012 km/s) is larger than the nearby value of V2 and V4.                   erupting far away from the Sun-Earth direction would have
Therefore we can say that only the higher speed shocks                     less probability to arrive at Earth. It can also be seen from
have possibility to arrive at Earth if the shock originates                Figure 5 that the percentage of type A events is larger than
near the HCS. Stated differently, this result again indicates              that of type B events when Y            40°, but the type A
the low probability of arrival for the shocks near the HCS,                percentage decreases dramatically when Y exceeds 40°.
even for relatively high-speed shocks.                                     As for the initial shock speed, the mean value of type A
                                                                           events (Vi) is higher than that of the corresponding type B
                                                                           events (vi) on each Y interval, which is consistent with

                                                                      5 of 9
A06107                    ZHAO ET AL.: INFLUENCE OF SOLAR FLARE LOCATION AND THE HCS                                      A06107

         Figure 5. The percentage distribution along the angular distance from flare source to Earth (Y) for type
         A (solid line) and type B (dashed line) events. Ni, ni, Vi, and vi are defined the same as those in Figure 4.

Figure 4; on the other hand, the mean shock speed of type A         number is about three times as large as that of the opposite
events, i.e., Vi, increases with the increment of Y. This           side event. However, all these conclusions are only based on
implies that the larger the angular distance is, the more           the solar activities for which the related interplanetary
energy is needed for the IP shocks to reach Earth. But the          counterparts (IP shock or ICME) were observed at Earth.
extremely high value of V6 may not have the general                 The solar events with no such IP manifestations at Earth
representative as the event number of N6 is only one. For           were not discussed in those works. On the other hand, Hu
the type B events, their mean shock speed vi do not change          and Jia [2001] presented a one-dimensional analysis and
evidently along Y except for the last one, v6. While, the type      2.5-dimensional MHD simulations on the interplanetary
B event number in the last interval is also little, i.e., n6 = 4.   shock interaction with the HCS and its associated structures;
For the same reason, the abnormity of v6 does not have              they found that the HCS alone is transparent for any
much significance in statistical sense.                             interplanetary shocks. Therefore a more reliable witness
3.2.3. The Same Side-Opposite Side Effect                           about the same-opposite side effect should be based on a
  [15] As the HCS is the most important structure in an             comprehensive and unbiased sample of solar transient
undisturbed interplanetary medium, its effect on the prop-          activities both with and without the associated IP counter-
agation of solar disturbances and corresponding geoeffects          parts at Earth.
has attracted extensive attention both in data analysis and           [16] Our data set gives 347 solar flare-type II radio
simulation. Much of this kind of work is pursued in terms of        burst events covering an uninterrupted time period during
the relative positions between the source of solar transients       February 1997 –August 2002. For the total events, 212 of
and the observer (i.e., the Earth) with respect to the HCS in       them are the same side events with flares located on the
order to investigate the effect of the HCS on the propagation       same side of the HCS as the Earth, and the same side events
properties of these transient disturbances. This approach is        account for 61% of the total; the other 135 events belong to
appropriate because of lack of sufficient observations at the       the opposite side ones, which account for 39%. We can see
larger (>1 AU) distances in the interplanetary medium.              that the same side percentage is larger than the opposite side
Henning et al. [1985] pointed out that solar flares located         percentage during the total events including both shock-
on the same side of the HCS as the Earth would likely               associated and nonshock-associated flares. However, the
produce larger geomagnetic disturbances than flares located         same-opposite side percentages are not equal for type A
on the opposite side. Wei et al. [1990, 1991] studied the           and B events. For the 130 type A events, 94 of them are the
source distribution of 277 flare-shock events under a newly         same side events and their same side percentage is 72%,
established flare-HCS coordinate system. They found that            larger than 61% of the total events. While for the 217 type B
the same side event number was larger than the opposite             events, only 118 of them are the same side ones; the same
side event number and also that the strong shocks and               side percentage of type B is 54%, lower than 61%. The
corresponding geomagnetic storms were usually associated            comparison between our same-opposite side percentages
with the same side events. This indicates that the HCS              obtained here and those published in previous papers are
might have the ‘‘impeding’’ effect on the trans-propagation         listed in Table 1. We can see from Table 1 that the same side
of shocks in interplanetary space. Recently, the source             percentage of type A events (72%) is fairly consistent with
distribution of nearly 100 CME-ICME conjugated events               previous results [Wei et al., 1990, 1991; Zhao and Feng,
have been investigated [Zhao and Feng, 2005; Feng and               2005; Feng and Zhao, 2006] where only solar events with
Zhao, 2006], and it is found that the same side event               IP manifestations at Earth were discussed. But the previous

                                                               6 of 9
A06107                        ZHAO ET AL.: INFLUENCE OF SOLAR FLARE LOCATION AND THE HCS                                                  A06107

           Table 1. Summary and Comparison of the Same Side (S.S.) and Opposite Side (O.S.) Percentages Between the
           Previously Published Results and This Paper Results (Denoted as *)
                 Event Kind           Event Number            S.S. Percentage,         O.S. Percentage              Published Paper
                               a
           Flare-Disturbances               135                      56                       44                Henning et al. [1985]
           Flare-IP Shockb                  277                      71                       29                Wei et al. [1990, 1991]
           CME-ICMEs                        80                       70                       30                Zhao and Feng [2005]
           CME-ICMEs                        100                      74                       26                Feng and Zhao [2006]
           Flare-Type II                    347                      61                       39                            *
           Flare-IP Shockc                  130                      72                       28                            *
           Flare-Nonshockd                  217                      54                       46                            *
             a
              Here ‘‘disturbance: refers to solar wind disturbance or Dst disturbance associated with a particular flare.
             b
               The ‘‘IP shock’’ refers to the interplanetary shocks observed by spacecrafts at near Earth space.
             c
              This is the type A events in the paper; and ‘‘IP shock’’ refers to the interplanetary shocks observed by L1 spacecrafts.
             d
               This is the type B events in the paper for which no interplanetary shocks are observed by L1 spacecrafts.

papers exclude those solar events without IP disturbances.                 depends on the relative locations between the flare source,
Here our analysis is based on solar transient events with and              the HCS, and the Earth. Then the angular distance from the
without IP disturbances and shows the opposite side per-                   flare source to the Earth location (Y) is the most related
centage of those nonshock-flares (type B) to be higher than                parameter to this kind of sorting. Generally speaking, the
that of the shock-associated flares (type A) revealing the                 flare has a high chance of lying on the opposite side of the
underlying ‘‘impeding’’ effect of the HCS on the trans-                    HCS from the Earth for the large Y events. Specific
propagation of IP shocks when the flare source is located on               calculations show that the averaged Y for these 212 same
the opposite side of the HCS as the Earth.                                 side events is 44°, less than the averaged value of 58° for
  [17] To explore further the effect of the HCS on the                     those 135 opposite side events. Unfortunately, Y has evident
arrival of the shock at the Earth, we compare the shock                    contribution to the associated shock’s arrival probability as
arrival probabilities for the same side (S.S.) and opposite                discussed previously. In order to exclude the contribution of
side (O.S.) events in a more direct way. The total same side               Y, the same-opposite side effect of the HCS is explored
event number is 212, and 94 of them have the associated                    using event with equivalent angular distances of Y. Figure 6
shock being observed at Earth (accounting for 44.3%).                      gives the percentage distribution of type A (solid line) and
While among the total 135 opposite side events, only 36                    type B (dashed line) events along Y, where the same side
of their shocks arrived at Earth (26.7%). However, we                      and opposite side events are treated separately. The same
should keep in mind that this same-opposite side sorting                   side distribution is displayed on the right-hand half of the

         Figure 6. The distribution of type A (solid line) and type B (dashed line) percentage along Y for the
         same side and opposite side events. This figure is similar to Figure 5, but different from it in dealing with
         the same-opposite side events separately, as shown in the right and left half panel, respectively. NSi (i = 1;
         2; 3) denotes the type A event number of the same side events in Y intervals of [0°, 40°], [40°, 80°], [80°,
         120°] and VSi denotes their mean initial shock speed corresponding; nSi and vSi are those for the type B
         events of the same side events. The definitions of NOi, VOi, nOi, and vOi are in the same way but for the
         opposite side events.

                                                                     7 of 9
A06107                    ZHAO ET AL.: INFLUENCE OF SOLAR FLARE LOCATION AND THE HCS                                                     A06107

figure, while the opposite side distribution is on the left-        the same side events still exists among the same-opposite
hand half of the figure. The x coordinate is symmetrical            side events with the same angular distance of Y. In this
about the dotted line in the middle. In the right half of the       context, the ‘‘impeding’’ of the HCS turns into the exclusive
figure, NSi (i = 1; 2; 3) is the type A event number of the         explanation if other factors are neglected.
same side events with their Y in intervals of [0°, 40°], [40°,
80°], [80°, 120°], respectively; VSi is the mean initial shock        [19] Acknowledgments. We are appreciated to the Wilcox Solar
speed of them; nSi and vSi are those for the type B events of       Observatory (http://wso.stanford.edu/) for its free data policy. This work
                                                                    is jointly supported by National Natural Science Foundation of China
the same side events. In the left half, the definitions of NOi,     (40621003, 40536029, 40336053, and 40523006), 973 program under
VOi, nOi, and vOi are used for the opposite side events. We         grant 2006CB806304 and the CAS International Partnership Program for
can see from Figure 6 that the percentage distribution is not       Creative Research Teams.
symmetrical about the central dotted line. The solid line             [20] Zuyin Pu thanks Elizabeth Lucek and another reviewer for their
                                                                    assistance in evaluating this paper.
within each Y bin in the right half is higher than that within
the relevant Y bin in the left half; while the dashed line          References
distribution is on the contrary as required. This means that        Byrne, F. N., M. A. Ellison, and J. H. Reid (1964), A survey of solar flare
for a group of solar flares with similar angular distances            phenomena, Space Sci. Rev., 3, 319.
(including both same side and opposite side), the IP shocks         Cane, H. V., I. G. Richardson, and O. C. St. Cyr (2000), Coronal mass
                                                                      ejections, interplanetary ejecta and geomagnetic storms, Geophys. Res.
associated with the same side flares would have larger                Lett., 27(21), 3591.
probability to reach Earth than those associated with the           Cane, H. V., and I. G. Richardson (2003), Interplanetary coronal mass
opposite side flares. This is straightforward evidence for the        ejections in the near-Earth solar wind during 1996 – 2002, J. Geophys.
                                                                      Res., 108(A4), 1156, doi:10.1029/2002JA009817.
effect of the HCS on the shock’s trans-propatation in               Cid, C., M. A. Hidalgo, E. Saiz, Y. Cerrato, and J. Sequeiros (2004),
interplanetary medium as the contribution of Y is eliminat-           Sources of intense geomagnetic storms over the rise of solar cycle 23,
ed. As for the mean shock speed, the type A event speed is            Sol. Phys., 223, 231.
larger than that of type B for both same side and opposite          Dryer, M. (1994), Interplanetary studies: Propagation of disturbances
                                                                      between the sun and the magnetosphere, Space Sci. Rev., 67, 363.
side events, i.e., VSi > vSi and VOi > vOi for i = 1; 2; 3; while   Dryer, M. (1996), Comments on the origins of coronal mass ejections, Sol.
among the type A events, the speed of the opposite side               Phys., 169, 421.
events is larger than that of the same side events, i.e., VO1 >     Feng, X. S., and X. H. Zhao (2006), Geoeffective analysis of CMEs under
                                                                      current sheet magnetic coordinates, Astrophys. Space Sci., doi:10.1007/
VS1, VO2 > VS2 and only VO3 < VS3 where the event numbers             s10509-006-9039-6.
NS3 and NO3 are merely 6. This implies that for the events          Fry, C. D., M. Dryer, Z. Smith, W. Sun, C. S. Deehr, and S.-I. Akasofu
with the same angular distances of Y, the opposite side               (2003), Forecasting solar wind structures and shock arrival times using an
                                                                      ensemble of models, J. Geophys. Res., 108(A2), 1070, doi:10.1029/
shocks need more energy (proportional to V3 as assumed) to            2002JA009474.
transport through the HCS and then arrive at Earth, dem-            Harrison, R. A. (1996), Coronal magnetic storms: A new perspective on
onstrating another aspect of the potential ‘‘impeding’’ effect        flares and the ‘solar flare myth’ debate, Sol. Phys., 166(2), 441.
of the HCS on shock propagation.                                    Henning, H. M., P. H. Scherrer, and J. T. Hoeksema (1985), The influence
                                                                      of the heliopheric current sheet and angular separation on flare-
                                                                      accelerated solar wind, J. Geophys. Res., 90, 11,055.
4. Summary                                                          Hu, Y. Q., and X. Z. Jia (2001), Interplanetary shock interaction with the
                                                                      heliospheric current sheet and its associated structures, J. Geophys. Res.,
  [18 ] Here are investigated the source properties of                106(A12), 29,299.
                                                                    Knoska, S., and L. Krivsky (1978), Time-latitude occurrence of flares in
130 flare-shock events (type A) and 217 flare-nonshock                solar cycle No 20 /1965-1976/, Astronomical Institutes of Czechoslovakia,
events (type B) during an uninterrupted period of February            Bulletin, 29(6), 352.
1997 – August 2002. It is found that most of these flares           Li, K. J., B. Schmieder, and Q. S. Li (1998), Statistical analysis of the X-ray
                                                                      flares (M 1) during the maximum period of solar cycle 22, Astron.
(including type A and B) are distributed within [S30°,                Astrophys. Suppl., 131, 99.
N30°] in heliographic latitude, and that [S30°, N30°]              Mari, G., and M. D. Popescu (2004), Solar flare cycles, Rom. Rep. Phys.,
[E10°, W30°] is the predominant source region on solar                56(2), 141.
disk, including the majority of geoeffective solar flares. We       Maunder, E. W. (1904), Note on the distribution sun-spots in heliographic
                                                                      latitude, 1874 – 1902, Mon. Not. R. Astron. Soc., 64, 747.
also probe the distribution of flare source relative to the         McKenna-Lawlor, S. M. P., M. Dryer, M. D. Kartalev, Z. Smith, C. D. Fry,
HCS. The statistical results indicate that the shocks asso-           W. Sun, C. S. Deehr, K. Kecskemety, and K. Kudela (2006), Real-
ciated with flares near the HCS have a lower probability of           time predictions of the arrival at the Earth of flare-generated shocks
                                                                      during solar cycle 23, J. Geophys. Res., 111, A11103, doi:10.1029/
reaching the Earth. This might be caused by the interaction           2005JA011162.
between the shock and the HCS that weakens the intensity            Odstrcil, D., M. Dryer, and Z. Smith (1996), Propagation of an interplane-
of the shock as a result. The angular (azimuthal) distance            tary shock along the heliospheric plasma sheet, J. Geophys. Res.,
                                                                      101(A9), 19,973.
from the flare source to the Earth (Y) also has an effect on        Park, Y. D., Y.-J. Moon, S. Iraida Kim, and H. S. Yun (2002), Delay times
the associated shock’s arrival at the Earth, with the shocks          between geoeffective solar disturbances and geomagnetic indices, Astro-
associated with large Y flares having a smaller probability           phys. Space Sci., 279, 343.
of reaching the Earth. For the Earth-encountered shocks,            Smith, Z., and M. Dryer (1990), MHD study of temporal and spatial evolu-
                                                                      tion of simulated interplanetary shocks in the ecliptic plane within 1 AU,
their initial speeds increase with the increment of Y in a            Sol. Phys., 129, 387.
statistical sense. Furthermore, for the first time, the same-       Smith, Z., and M. Dryer (1995), The interplanetary shock propagation
opposite side effect of the HCS is confirmed by using an              model: A model for prediction solar-flare-caused geomagnetic sudden
                                                                      impulses based on the 2 1/2 D MHD numerical simulation results
unprejudiced set of both arriving and nonarriving (at the             from the Interplanetary Global Model (2D IGM), NOAA Tech. Memo.,
Earth) solar transients. The shocks associated with solar             ERL/SEL-89.
flares located on the same side of the HCS as the Earth have        Smith, Z., D. Odstrcil, and M. Dryer (1998), A 2.5-dimensional MHD
                                                                      parametric study of interplanetary shock interactions with the helio-
a greater chance to reach the Earth than those shocks with            spheric current sheet/heliospheric plasma sheet, J. Geophys. Res.,
the associated flares on the opposite side. This superiority of       103(A9), 20,581.

                                                               8 of 9
A06107                         ZHAO ET AL.: INFLUENCE OF SOLAR FLARE LOCATION AND THE HCS                                                         A06107

Tokumaru, M., M. Kojima, K. Fujiki, et al. (2000), Three-dimensional               associated interplanetary shock waves in the flare-heliospheric current
  propagation of interplanetary disturbances detected with radio scintilla-        sheet coordinate system, Acta Geophys. Sin., 34, 133.
  tion measurements at 327 MHz, J. Geophys. Res., 105(A5), 10,435.               Wu, C. C., and M. Dryer (1997), Three-dimensional MHD simulation of
Wang, Y. M., P. Z. Ye, S. Wang, G. P. Zhou, and J. X. Wang (2002), A               interplanetary magnetic field changes at 1 AU caused by a simulated
  statistical study on the geoeffectiveness of Earth-directed coronal mass         solar disturbance and a tilted heliospheric current/plasma sheet, Sol.
  ejections from March 1997 to December 2000, J. Geophys. Res.,                    Phys., 173, 391.
  107(A11), 1340, doi:10.1029/2002JA009244.                                      Yermolaev, Y. I., M. Y. Yermolaev, G. N. Zastenker, L. M. Zelenyi, A. A.
Wang, Y. M., C. L. Shen, S. Wang, and P. Z. Ye (2004), Deflection of               Petrukovich, and J.-A. Sauvaud (2005), Statistical studies of geomagnetic
  coronal mass ejection in the interplanetary medium, Sol. Phys., 222, 329.        storm dependencies on solar and interplanetary events: A review, Planet.
Wang, Y. M., X. H. Xue, C. L. Shen, P. Z. Ye, S. Wang, and J. Zhang                Space Sci., 53, 189.
  (2006), Impact of major coronal mass ejections on geospace during 2005         Zhang, J., K. P. Dere, R. A. Howard, and V. Bothmer (2003), Identification
  September 7 – 13, Astrophys. J., 646, 625.                                       of solar sources of major geomagnetic storms between 1996 and 2000,
Webb, D. F., E. W. Cliver, N. U. Crooker, and B. J. Thompson (2000),               Astrophys. J., 582, 520.
  Relationship of halo coronal mass ejections, magnetic clouds, and mag-         Zhao, X. H., and X. S. Feng (2005), A tentative study for the prediction of
  netic storms, J. Geophys. Res., 105(A4), 7491.                                   the CME related geomagnetic storm intensity and its transit time, Sci.
Wei, F. S., and X. D. Deng (1987), North-south asymmetrical propagation            China E, 48(6), 648.
  of flare-shocks, Chin. J. Geophys., 30(5), 443.                                Zhao, X. H., X. S. Feng, and C.-Chun Wu (2006), Characteristics of solar
Wei, F. S., G. Yang, and G. L. Zhang (1985), Three-dimensional propaga-            flares associated with interplanetary shock or nonshock events at Earth,
  tion characteristics of flare-shocks based on IPS observations, Acta Astro-      J. Geophys. Res., 111, A09103, doi:10.1029/2006JA011784.
  phys. Sin., 5(3), 214.
Wei, F. S., J. H. Zhang, and S. P. Huang (1990), Study of the propagation        
  characteristics of flare-associated interplanetary shock waves in the flare-      X. Feng and X. Zhao, SIGMA Weather Group, State Key Laboratory for
  heliospheric current sheet coordinate system, Acta Geophys. Sin., 33,          Space Weather, Center for Space Science and Applied Research, Chinese
  125.                                                                           Academy of Sciences, P. O. Box 8701, Beijing 100080, China.
Wei, F. S., S. Q. Liu, and J. H. Zhang (1991), Study of distribution char-       (fengx@spaceweather.ac.cn; xhzhao@spaceweather.ac.cn)
  acteristics of the geomagnetic disturbances corresponding to the flare-           C.-C. Wu, Center for Space Plasma and Aeronomic Research, University
                                                                                 of Alabama in Huntsville, Alabama 35899, USA.

                                                                            9 of 9
You can also read