Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.izmiran.rssi.ru/~obridko/papers/0128.pdf Дата изменения: Tue Nov 17 10:38:30 2015 Дата индексирования: Sun Apr 10 01:08:52 2016 Кодировка: Поисковые слова: spacecraft

ISSN 0016 7932, Geomagnetism and Aeronomy, 2013, Vol. 53, No. 7, pp. 813­817. © Pleiades Publishing, Ltd., 2013.

Long Term Variations of Geomagnetic Activity and their Solar Sources1
B. Kirova, V. N. Obridkob, K. Georgievaa, E. V. Nepomnyashtayab, and B. D. Shelting
a b

Space Research ant Technologies Institute ­ BAS, Sofia, Bulgaria b IZMIRAN e mail: bkirov@space.bas.bg

Abstract--Geomagnetic activity in each phase of the solar cycle consists of 3 parts: (1) a "floor" below which the geomagnetic activity cannot fall even in the absence of sunspots, related to moderate graduate com mencement storms; (2) sunspot related activity due to sudden commencement storms caused by coronal mass ejections; (3) graduate commencement storms due to high speed solar wind from solar coronal holes. We find that the changes in the "floor" depend on the global magnetic moment of the Sun, and on the other side, from the height of the "floor" we can judge about the amplitude of the sunspot cycle. DOI: 10.1134/S0016793213070128
1

1. INTRODUCTION As early as in 1852 it was noted that geomagnetic disturbances are related to solar activity (Sabine, 1852), and in 1982 it was found that geomagnetic activity is caused by two types of solar agents, the first one related to sunspots and caused by coronal mass ejections (CMEs), and the other one­not related to sunspots and caused by high speed solar wind from solar coronal holes (Feynman, 1982). CME caused geomagnetic disturbances have a maximum in sunspot maximum, and coronal holes related disturbances-- in the descending phase of sunspots. In the present study we aim to determine how these two components of geomagnetic activity vary from cycle to cycle, and the variations of which solar agents cause these changes.

2. COMPONENTS OF GEOMAGNETIC ACTIVITY Feynman (1982) noted that if a geomagnetic activ ity index (e.g. aa index) is plotted as a function of the sunspot number R (Fig. 1), all aa index values lie above a line given by the equation aaR = a0 + b.R. Fey nman (1982) suggested that aaR is the geomagnetic activity caused by sunspot related solar activity, and the aa values above this line represent the geomagnetic activ ity caused by high speed solar wind: aaP = aa ­ aaR. (Feynman, 1982) determined that aaR = 5.38 + 0.12.R. We calculated the values of a0 and b in consec utive periods of around 30 years, each including 3 sun spot cycles (cycles 9­11, 10­12 and so on). The cal culated values of the coefficients demonstrate a clear quasi secular cycle (Fig. 2).
1

Actually, geomagnetic activity can be divided into 3 rather than 2 components. The first one is the "floor", equal to the a0 coefficient which represents the geo magnetic activity in the absence of sunspots. It is prac tically determined by the activity in the cycle mini mum and varies smoothly from cycle to cycle. The sec ond component is the geomagnetic activity caused by sunspot related solar activity which is described by the straight line aaT = b.R so that aaR = a0 + aaT. The slope b of this line also changes cyclically. The third component aaP (the value above aaR) is caused by high speed solar wind. It can be seen that when the coeffi cient a0 is big (high "floor" of the geomagnetic activ ity), the geomagnetic activity is almost independent of the sunspot number (small coefficient b), and when the "floor" is low, the geomagnetic activity quickly grows with growing sunspot number. Figure 3 demonstrates that both the "floor" of the geomagnetic activity a0 and the slope b don't depend on the phase of the sunspot cycle (upward or down ward branch), but are different in different intervals. Besides, the scatter of values above aaR in any interval is much bigger during the downward branch which has to be expected because of the strong influence of high speed solar wind streams on the geomagnetic activity in this phase of the sunspot cycle (Georgieva, K., Kirov, B, 2005). The question now is what solar agents cause the nonzero values of the geomagnetic activity "floor", and what factors lead to the changes in the coefficients a0 and b. 3. TYPES OF GEOMAGNETIC STORMS The geomagnetic storms differ in intensity as well as in characteristics. The intensity of the storm is

The article is published in the original.

813


814 aa/ak index 40 35 30 25 20 15 10 5 0 1878­1912

KIROV et al. aa/ak index 40 35 30 25 20 15 10 5 0 1954­1985 k

d

20 40 60 80 100 120 140 160 180 200 Sunspot number

20 40 60 80 100 120 140 160 180 200 Sunspot number

Fig. 1. Dependence of aa index on R in the period 1878­1912 (left) and 1954­1985 (right).

determined by the parameters D, H and Z--the mag netic declination, change of the horizontal and verti cal components of the Earth's magnetic field, respec tively, measured in nT (table), and the storm can be with a sudden or gradual commencement. In order to determine the origin of different storms, (Shelting, Obridko, 2011) compared the occurrence frequency of all, strong, moderate and weak storms with sudden and with gradual commencement to the number of sunspots. They found that the occurrence frequency of the sudden commencement storms correlates with the sunspot number, with correlation coefficient 0.872 +/­ 0.06. In contrast, the correlation coeffi cient of the occurrence frequency of the gradual com mencement storms and the sunspot number is practi cally zero (0.014 +/­ 0.13), and their maximum is shifted by 1­3 years after the sunspot maximum. Fur
b 0.14 0.12 0.10 0.08 0.06 0.04 0.02
b(L) a(R)

ther analysis shows that there is high correlation among storms of different intensity inside each of the groups, and absolutely no correlation between the two groups of storms. From there the authors make the conclusion that these two types of storms have differ ent origin. Sudden commencement storms are caused by coronal mass ejections (CMEs). As the solar coro nal mass ejections are associated with active areas, i.e., with local magnetic fields, a high correlation is observed between the occurrence frequency of sudden commencement storms and the number of sunspots. Gradual commencement storms are caused by high speed solar wind. It originates from solar coronal holes whose maximum is on the downward branch of the sunspot cycle. This leads to two maxima in geomag netic activity in the course of the sunspot cycle, the second one being due mainly to the occurrence of many gradual commencement storms (Tsurutani, et al. 2006). 4. GEOMAGNETIC ACTIVITY "FLOOR" The geomagnetic activity "floor" is the geomag netic activity in the absence of any sunspots. Practi cally it is determined by the geomagnetic activity in the sunspot minimum. As seen in Fig. 4, during the whole investigated period, the geomagnetic activity with aa index between 10 and 30 is caused by gradual commencement storms, but this is especially well pro nounced in sunspot minimum periods. We can conclude that the geomagnetic activity "floor" is determined by non sunspot related solar activity. Figure 5 demonstrates that the variations in the geomagnetic activity floor follow, with the excep tion of cycle 13, the variations in the number of 3 hour intervals with aa index between 10 and 30 in the min imum of the respective cycle. Moreover, from cycle 14 to 21 (around 1985), an increase is observed in both the number of 3 hour intervals with 10 < aa < 30, and of the geomagnetic activity floor, after which they both
Vol. 53 No. 7 2013

a 18 16 14 12 10 8 6

Fig. 2. Cyclic variations of the coefficients a0 and b.

1844­1877 1856­1888 1867­1900 1878­1912 1889­1922 1901­1932 1913­1943 1926­1953 1933­1963 1944­1976 1954­1985 1964­1996 1976­2005 Year

0

4

GEOMAGNETISM AND AERONOMY


LONG TERM VARIATIONS OF GEOMAGNETIC ACTIVITY aa/ak geomagnetik index 40 Upward branch, 1844­1964 35 30 25 20 15 10 5 0 40 35 30 25 20 15 10 5 0 20 40 60 80 100 120 140 160 180 200 20 40 60 80 100 120 140 160 180 200 Downward branch, 1844­1964

815

40 35 30 25 20 15 10 5 0 20 40 60 80 100 120 140 160 180 200 Downward branch, 1865­2006 Upward branch, 1865­2006

40 35 30 25 20 15 10 5 0

20 40 60 80 100 120 140 160 180 200 Sunspot number

Fig. 3. Dependence of aa on R in the upward (upper panel) and in the downward branch (lower panel) of the sunspot cycle in the period 1844­1964 (left) and in the period 1865­2006 (right).

begin decreasing. The correlation between the two quantities is 0.85 with p < 0.01. It is seen that the geomagnetic activity floor in each cycle is determined by the geomagnetic activity in the range 10 < aa < 30 in the cycle minimum, and as far as this geomagnetic activity, especially in the cycle mini mum, is related as shown above to gradual commence ment storms, the floor is determined by the high speed solar wind streams causing gradual commencement storms. Therefore, the reason for the changing geo magnetic activity floor is the changing high speed solar wind reaching the Earth during sunspot minimum. It is interesting to note that at the same time as the increase in the floor changed to decrease (cycle 21),
Table Storm intensity Minor Moderate Strong Severe D 100­139 140­200 201­290 291
Vol. 53 No. 7

the increase in the solar global magnetic moment changed to a sharp decrease (Obridko, Shelting, 2009)--(Fig. 6). It seems that the change in the solar global magnetic moment is reflected in the geomag netic activity caused by the high speed solar wind dur ing the sunspot minimum through its effects on the thickness of the heliospheric current sheet (Simon and Legrand, 1987). 5. COEFFICIENT b (SLOPE) The sunspot related geoeffective solar agents are the CMEs. It is known that the geoeffectiveness of CMEs var little in the course of the sunspot cycle. The geoeffectivness of magnetic clouds (a subclass of

H 80­125 126­200 201­270 271
2013

Z 40­90 31­140 141­250 251

GEOMAGNETISM AND AERONOMY


816 60 50 40 30 20 10 0 1950

KIROV et al. Number (10 < aa < 30) 1400 1300 1200 1100 1000 900 800 700 600 500 400 11 12 13 14 15 16 17 18 19 20 21 22 23 Cycle No a0 20 18 16 14 12 10 8 6 4 24

1960

1970

1980

1990

2000

2010

Fig. 4. Annual number of weak sudden commencement storms (solid line) and gradual commencement storms (dotted line).

Fig. 5. Number of 3 hour intervals with aa between 10 and 30 (solid line) and the geomagnetic activity floor a0 (dotted line), in consecutive sunspot cycle minima.

CMEs with high and smoothly rotating magnetic field) does change significantly, but their number is negligible compared to the total number of CMEs except around sunspot minimum when their geoeffec tiveness is low. Therefore, the varying contribution of CMEs to geomagnetic activity from year to year depends on their varying number. (Georgieva and Kirov, 2005). Figure 7 presents the ratio of CMEs to the number of sunspots for the period 1996­2012. For the number of CMEs, the SOHO/LASCO CME cat alog is used (http://cdaw.gsfc.nasa.gov/CME_list/) as it provides a fairly homogenous data set. It can be seen that the ratio is almost constant dur ing the greater part of the sunspot cycle, with the exception of the periods of sunspot minima, especially in the minimum between cycles 23 and 24 which is known to be very peculiar. In 2007, 2008, and 2009 we had total annuals of 20, 70, and 60 CMEs, respectively, with the average annual number of sunspots between 3
MU 60

and 7. A small change in the correlation between the sunspot number and the number of CMEs is observed in the previous sunspot minimum also, but it is only due to CMEs in a few single months. We can conclude that the increasing geomagnetic activity with increas ing sunspot number (the coefficient b) is due to the increasing number of CMEs with increasing number of sunspots. 6. FORECASTING THE FOLLOWING CYCLE Figure 8 demonstrates that the sunspot number in the cycle maximum is related to the geomagnetic activity floor in the same cycle. The correlation between the two variables is 0.792 with p = 0.001. From this it follows that the characteristics of the whole cycle are set already at its beginning, and mea suring the geomagnetic activity we can forecast to a great extend its maximum. On the other hand, given that the evolution of the Sun's magnetic moment is indicative of the changes in a0, we can evaluate in advance the direction of change of the geomagnetic activity "floor". 7. SUMMARY AND CONCLUSIONS

40

20

0 1918 1928 1936 1948 1958 1968 1978 1988 1996 2008 Years
Fig. 6. Evolution of the solar magnetic moment from 1918 to 2006.

The geomagnetic activity consists of three compo nents: (1) a0--the geomagnetic activity "floor", theo retically equal to the activity at zero number of sun spots. Practically this activity is determined by the gradual commencement geomagnetic disturbances with 10 < aa < 30 in the beginning of the sunspot cycle; (2) aaT--the geomagnetic activity caused by CMEs whose number linearly increases with increasing num ber of sunspots aaT = b.R, so that aaR = a0 + b.R; (3) aaP--the values of aa above aaR, caused by high speed solar wind streams from solar coronal holes. We
Vol. 53 No. 7 2013

GEOMAGNETISM AND AERONOMY


LONG TERM VARIATIONS OF GEOMAGNETIC ACTIVITY Ratio (number of SMEs/sunspot number) 120 100 80 60 40 20 0 ­20 1996 1998 2012 2000 2003 2005 2007 2010 1997 1999 2001 2004 2006 2008 2011 Years
Fig. 7. Ratio of the number of CMEs to the sunspot number (monthly averages).

817

have found that the variations of a0 have a quasi secu lar cycle, with the direction of variation following the variation of the global magnetic moment of the Sun. On the other hand, we have found that a0 determined by the geomagnetic activity at the beginning of a cycle is directly related to the sunspot maximum in the same cycle. Therefore, if we know the direction of change (increasing or decreasing) of the global magnetic moment, we can forecast the increase or decrease of a0 in the next cycle, and from there--the increase or decrease of the sunspot maximum. Moreover, during
a0 20 18 16 14 12 10 8 6 4 Support number 200 180 160 140 120 100 80 60 40 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Solar cycle

the sunspot minimum we can quite accurately forecast the following sunspot minimum. REFERENCES
Feynman, J., Geomagnetic and solar wind cycles, 1900­ 1975, J. Geophys. Res., 1982, vol. 87, pp. 6153­6162. Georgieva, K. and Kirov, B., Helicity of magnetic clouds and solar cycle variations of their geoeffectiveness, Proc. IAU Symposium 226 Coronal and Stellar Mass Ejections, Oxford, 2005, pp. 470­472. Obridko, V.N. and Shelting, B.D., Anomalies in the evolu tion of global and large scale solar magnetic fields as the precursors of several upcoming low solar cycles, Astron. Lett., 2009, vol. 35, no. 3, pp. 38­44. Sabine, E., On periodical laws discoverable in the mean effects of the larger magnetic disturbances, Phil. Trans. R. Soc. London, 1852, vol. 142, pp. 103­124. doi 10.1098/rstl.1852.0009 Shelting, B.D. and Obridko, V.N., Sudden and gradual commencement magnetic storms as indices of solar activity, Russian Annual Conference on Solar Physics, Pulkovo, 2011. Simon, P.A. and Legrand, J.P., Some solar cycle phenom ena related to the geomagnetic activity from 1868 to 1980. III. Quiet days, fluctuating activity or the solar equatorial belt as the main origin of the solar wind flow ing in the ecliptic plane, Astron. Astrophys., 1987, vol. 182, pp. 329­336. Tsurutani, B.T., Gonzalez, W.D., Gonzalez, A.L.C., et al., Corotating solar wind streams and recurrent geomag netic activity: A review, J. Geophys. Res., 2006, vol. 111, pp. 11 107­11 132.
2013

Fig. 8. Geomagnetic activity floor (solid line) and sunspot number in the cycle maximum (dotted line). GEOMAGNETISM AND AERONOMY Vol. 53 No. 7