N E W S
OF T H E N A T IO N A L A C A D E M Y OF SC IEN C ES OF THE R EPU BLIC OF K A Z A K H S T A N P H Y S IC O -M A T H E M A T IC A L S E R IE S
IS S N 1 9 91-346Х h ttp s://d o i.o r g /1 0 .3 2 0 1 4 /2 0 2 0 .2 5 1 8 -1 7 2 6 .5 5 V olu m e 3, N um ber 331 (2 0 2 0 ), 2 0 9 - 216
УДК 550.383
G .I. G o r d iy e n k o , A .F . Y a k o v e ts, Y u .G . L itv in o v
«Institute o f Ionosphere» JSC «NCSRT», Almaty, Kazakhstan, [email protected], [email protected], [email protected]
LONG - TERM VARIATIONS
IN THE F- AND D-REGIONS OF IONOSPHERE
Abstract. The long-term changes in the near-noon, diurnal, and midnight critical frequencies o f the ionosphere F2 layer (foF2) and near-noon minimum reflection frequencies (fmin) were studied using ground-based vertical radio sounding o f the ionosphere at Alma-Ata station [43.25N, 76.92E] for the period 1957-2017. The data on solar and geomagnetic activity were used as factors affecting the state o f the ionosphere.
It was taken into account that the minimum reflection frequency depends both on the absorption o f the probe signal in the lower ionosphere, and on the noise level and technical characteristics of the ionosonde. Therefore, it seems problematic to estimate the absolute value o f signal absorption in this way. However, as a qualitative characteristic, as an indicator o f the absorption level o f the probe signal, the parameter fmin can be used.
Arithmetic average values o f median foF2 for near noon (10-14 LT), near midnight (23-01 LT) hours and average diurnal values for this period o f ionosphere observation are considered as initial data. The variability o f the lower ionosphere was studied using fmin data for daylight hours. The monthly average values o f the solar radio emission flux F10.7 and the Ap index are considered as characteristics o f solar and geomagnetic activity. The geomagnetic (Ap) indexes, the near-noon, near-midnight, diurnal averaged foF2, and near-noon fmin variations are found to be in strong dependence from the solar activity; all o f them show a dominant pattern o f variation with a period o f ~34-36 years and linear negative trend. The correlation coefficient between the foF2 and F10.7 long-term variations is very high, up to 0. 99 that permit us to believe that the solar activity can be considered to be the main driver for the long-term variations in the ionospheric F2-region.
The geomagnetic (Ap) index, the near-noon, near-midnight, daily averaged foF2, and near-noon fmin variations are found to be in strong dependence from solar activity; all o f them show a dominant pattern o f variation with a period o f ~34-36 years and linear negative trend. The correlation coefficient between the foF2 and F10.7 long-term variations is very high, up to 0. 99, that permits us to believe that the solar activity can be considered as a main driver for the long-term variations in the ionospheric F2-region.
The long-term course o f fmin is similar to those found in F10.7, Ap, and foF2, i.e. the periodicity o f 34-36 years is also evident in the fmin variations. However, in contrast to the F2-layer parameters, the fmin variation clear demonstrates an upward (positive) linear trend that is opposite in sign to the trend found in F10.7. This means relatively high sensitivity o f the fmin values to the solar activity changes and significant influence o f other trend drivers on them, one o f which is the possible impact o f anthropogenic factors on the state o f the lower ionosphere.
Key words: midlatitude ionosphere; upper and lower ionosphere; D-region, F2-region, long-term trends.
In trod u ction . In our previous work w e investigated the long-term variations in the near-noon (10-14 LT) critical frequency o f the ionospheric F2-layer (foF 2) m easured at the m id-latitude station A lm a-A ta [43.25N , 76.92E ] over the period from 1957 to 20 1 2 [1]. The purpose o f this paper is to provide further analyses o f the F2-layer variations and study long-term trends (a long-term linear change) in the near- noon, near-m idnight and daily averaged F2-layer critical frequencies m easured at the A lm a-A ta station in m ore extended tim e period, from 1957 to 2 0 1 7 including the period o f very deep m inim um in solar activity observed in 2 0 0 8 -2 0 0 9 . In addition to the F2-layer parameters the m inim um frequency o f
to do the trend analyses in the ionospheric absorption. The fm in value depends, apart from the absorption o f radio w a v es in the ionosphere, on technical characteristics o f ion oson d es that does not g iv e us any opportunity to use this ionosphere parameter to determ ine the absolute valu es o f the absorption o f radio w aves. H ow ever, as a qualitative characteristic, as an indicator o f absorption o f radio w a v es, the fm in value is w id ely used ([2-3] and references therein).
For the present trend analyses, m onthly m edian foF 2 and fm in values routinely m easured at the A lm a- A ta station have been used. The arithmetic m eans o f the foF 2 valu es at the near-noon (1 0 -1 4 LT), near
m idnight (23-01 LT) hours, and daily m eans foF 2 are calculated for the analyses, the fm in values are taken on ly for the near-noon hours: foF210-14, foF 2 23-0 1, foF2av, and fm in n-13 correspondingly. The m onthly m ean solar flu x F 10.7 and geom agnetic in d ex A p are also used as the indexes for solar and geom agnetic activities (available at h tt://w w .sw pc.noaa.gov) to illustrate their long-term variations.
D e sc r ip tio n o f th e d a ta , an d o b se rv a tio n resu lts T re n d s in the io n o sp h e ric F -re g io n . A s an exam ple, m edian m ean values foF210-14, foF223-01, foF2d.av together w ith the corresponding data o f solar and geom agnetic activity is show n in figure 1a-e. The dots represent observed data, the thick lin es show the fitting curves. It should be also noted, the statistical processing o f the ionospheric data series requires their continuity that is not alw ays p ossib le because o f different reasons (ionosonde repair and other technical problem s).
The regression dependences b etw een the selected ionospheric parameters and F 10.7 (figure 1f-h) have been analyzed to define the m issin g data. Thick lin es represent the linear regression lines for these data;
dashed lin es correspond to the p olynom ial functions o f the second degree that better fit the g iv en F2-layer parameters (R 2polmomial>R2lmear) where R2 is the co efficien t o f determ ination that provides a m easure o f h ow w e ll the least-square curve fits the observational data, r is the correlation coefficien t. A ssu m in g the second-order polynom ial dependence w ith F 10.7 the m issin g foF 2 valu es have been defined from the regression equations to fill available gaps in the data sets. T w o things are evident in the figure, firstly the variations o f all parameters are strictly m odulated by c y clic 11-year variations o f solar activity and secon d ly the foF 2 data are m uch scattered relatively to the sm oothed lin es that can be attributed to various sources other than solar activity, including planetary w a v es and seasonal variations. A s an exam ple, the great part o f seasonal variations is evident in Figure 1e w here large deviations o f the foF223-01 values from the sm oothed line is caused by the fact that sum m er nighttim e foF 2 values are m uch higher than w inter ones. Since this intraseasonal variability has to be taken into consideration in the trend analyses, annual averages o f the parameters considered w ere calculated. Figure 2a-e dem onstrates tem poral variations o f the calculated annual averages (denoted by sym bol*) o f the F10.7*, Ap* indexes, and the F2-layer parameters (foF2*10-14, foF2*23-01, and foF2*av) for w h ole period o f observations; variations o f the ionospheric parameters w ith F10.7* are presented in Figure 2f-h.
The m ain feature o f the variations is their sim ilarity (figure 2a-e) and clo se connection b etw een the ionospheric parameters and solar activity (figure 2f-h). The coefficien ts o f determ ination R2 (the coefficien ts R 2 and derived regression equations are show n in the figure field s) are found to be very high, from 0 .9 7 to 0.99 it m eans that from 97% to 99% o f the annual foF 2 variations can be explained by their relationship w ith the 11-year cy cle o f solar activity.
A ccording to our previous w ork [1], the 11-year (132 m onths) running m ean values o f the annual averages o f the ionospheric parameters (foF2*10-14(132), foF2*23-01(132), and foF2*av(132) w ere calculated to obtained an independent picture o f long-term trends in the upper ionosphere; the 11-year sm oothing technique w as also applied to the F 10.7 and A p data sets (F 10.7*(132), A p*(132)); Figure 3 presents these calculated valu es and sh ow s that geom agn etic activity is strongly controlled by the solar activity, both
15/1/32 15/1/62 15/1/92
Date (dd/mm/yy)
15/1/22 15/1/32 15/1/62 15/1/92
_________________ Date (dd/mm/yy)
15/1/22
<4
16 14 12 10 8 6 4
f.
_ Y = 0.0422747 * X + 3.41567 . R2 = 0.788499
■ Jul. 1957 - Apr. 2018
15/1/62 15/1/92
Date (dd/mm/yy)
R2 = 0.81856 Y = 0.4840471 + 0.0895636 * X - 0.0001649 * pow(X,2)
i | i | i | i | I | i | i | i | I | i | i | i |
60 80 100 120 140 160 180 200 220 240 260 280 300
Solar flux (F10.7)
<4
,©
12 10 8 6 4 2
15/1/32 15/1/62 15/1/92
Date (dd/mm/yy)
15/1/22 60 80 100 120 140 160 180 200 220 240 260 280 300
Solar flux (F10.7)
12 10 8 6 4 2
Y = -0.0007 7 6 5 5 8 1 6 3 * X + 5.171121512 Jul. 1957 - Apr. 2018
1/1/32 1/1/62 1/1/92
Date (dd/mm/yy)
1/1/22 60 80 100 120 140 160 180 200 2 20 240 260 280 300
_______________ Solar flux (F10.7)_________________
Figure 1 - The long-term variations of the near-noon(b), daily mean (c) and near-midnight (d) monthly median foF2 values in Alma-Ata [43.250N, 76.920E] in 1957-2017 together with the corresponding variations of solar flux F10.7 (a)
and Ap index (c); solid dots - measured data, thick lines show the fitting lines, dashed lines - linear fits.
Monthly median near-noon (f), daily mean (g), and near-midnight (h) foF2 values versus monthly mean F10.7;
thick lines represent the linear regression lines for these data, dashed lines correspond to the polynomial functions of the second degree
15/1/32 15/1/62 15/1/92
________________ Date (dd/mm/yy)
15/1/32 15/1/62 15/1/92
________________ Date (dd/mm/yy)
b.
14
12
10
8
6
4
Jan. 1958 - Oct. 2017
\\ M
; i i \ л 1 11 »
? 1
A
—: # —j- -S- i_\ f I iP-
i i I I \ ; ; i 1 Г Г- -Л — Hi: г \ r - * : • V
M
U\I \ j \
Y = -0.002032832417 * X + 10.00657781
— 1—I— 1—I— 1— |— 1—|— 1— |— '— |—I— |—l— |—I— |
/32 15/1/62 15/1/92 15/1/
____________ Date (dd/mm/yy)_______________
16 14 12 10 8 6 4
Y = 0.0445755 * X + 3.1420578 R 2 = 0.968435
Jan. 1958 - Oct. 2017
R2 = 0.990587 Y = 0.4755632 + 0.088437 * X - 0.00015897 * pow(X,2)
“I— |— i— |— I— |— i— |— l— |— l— |— l— |— l— |— l— |— l— |— l— |— l— | 60 80 100 120 140 160 180 200 220 240 260 280 300
_______________Solar flux (F10.7*)________________
c.
Jan. 1958 - Oct. 2016
12
И 10 -
' t
l \ V/
Y = -0.001321675886 * X + 7.262293556
Y = 0.03028235 * X + 2.638842 R2 = 0.980738
Jan 1958 - O ct 2016
R2= 0.991508 Y = 1 .383659+0.050929 _ * X -7.483085E -005 * pow (X 2)
1
—I—1—|—1—|—1—|—I—|—I—|—I—|—I—|—I—|
1/1/32 1/1/62 1/1/92 1/1/22
___________________ D ate (dd/m m /yy)_____________________
—I—1—I—1—|—1—|—1—|—i—|—i—|—i—|—i—|—i—p —pn—]
60 80 100 120 140 160 180 200 220 240 260 280 300
Solar flux (F10.7*)
d.
4 -
h.
Y = 0.01751185959 * X + 2.480358595 R2 = 0.972788
Jan. 1958 - Oct. 2017
R2 = 0.974822 Y = 2.16364+ _0.022722 * X - 1.888E-005 * pow(X,2)
1/1/32 1/1/62 1/1/92
Date (dd/mm/yy)
1/1/22
l— |— \— \— i— \— \— |— \— \— i— \— \— \— i— \— \— |— \— p n — p i | 60 80 100 120 140 160 180 200 220 240 260 280 300
Solar flux (F10.7*)
10
8 8 —
6
to 6
4
10 —
8
6
2
Figure 2 - As the Figure 1 but the annual averages of the F10.7, Ap, and foF2 values.
a.
ГЛ
15/1/32 15/1/48 15/1/64 15/1/80 15/1/96 15/1/12
b. e.
<4
,©
15/1/32 15/1/48 15/1/64 15/1/80 15/1/96 15/1/12
Date (dd/mm/yy)
Figure 3 - The 132-month smoothed values of the annual averages of the ionospheric F2-layer parameters, F10.7, and Ap.
This period is sligh tly differs from our earlier finding [1] where the period w as estim ated to be in a range o f 30 -3 2 years. The extended data sets used in the study reveal the period more clearly. It should be m entioned that sim ilar period (3 1 -3 2 years) has been found in all the solar terrestrial parameters ([4-6], and references therein) and w as interpreted as the 35-year Brutcher clim atic periodicity. N o te, the negative trend (dashed line), the low ering in the F2-layer critical frequency, is also seen in figure 3. The long-term trends found in different ionospheric characteristics (figures 1-3) are g iv en in table.
Table
# Parameter Trend, MHz/year Parameter Trend, MHz/year Parameter Trend, MHz/year 1 foF210-14(med) -0.02822 foF2*10-14 -0.02439 foF2*10-14 (132) -0.01758 2 foF2d.av (med) -0.01884 foF2*d.av -0.01586 foF2*d.av (132) -0.01077 3 foF223-<>1 (med) -0.00932 foF2*23-01 -0.00756 foF2*23-01 (132) -0.00364
studied for the period 1963-2012 (the 11-year sm oothing technique that w as applied to the foF2* data sets reduced the available period for study to b etw een 1963 and 2 0 1 2 ), the coefficien ts o f determ ination (R 2) w ere found to be 0.96, 0.98, and 0.90 for near-noon, daily averaged and near-m idnight data correspondingly figure 4a.
Figure 4 - The regression dependence between foF2*(132) and F10.7*(132) for the periods 1963-2012, (a), 1972-1984 (b) and 1984-2008 (c)
H ow ever, a com bination o f tw o groups o f points w as evid en t in figure 4 (upper panel) that assum ed som e different dependence b etw een the parameters on different phases o f the 34-36-years cy cle (the period 1972-1984, and 1 9 8 4-2008), w hen the foF2*(132) values increase to their m axim um and then decrease to their m inim um values. E xam ples o f the regression dependences b etw een foF 2*(132) and F 10.7*(132) for the tw o tim e periods are show n in figure 4a,b. It is distinctly seen in both cases that 99%
o f the variations in the ionospheric parameters can be explained by linear dependence b etw een them and solar activity. So, the solar activity can be considered to be the m ain driver for the long-term variations in the ionospheric F2-region.
® иs 53
—"o
1/1/62 1/1/70 1/1/78 1/1/86 1/1/94 1/1/02 1/1/10 1/1/18
D a te (d d /m m /y y )
Figure 5 - The 132-month smoothed values of fmin and F10.7 where dashed lines represent their linear trends for the whole measuring interval
T re n d s in the D -re g io n . A sim ilar analysis, as that carried out for foF 2, has been carried out for the fm in data observed at the A lm a-A ta station in the years 1957-2017. A s it has been already m entioned, the ionospheric parameter fm in is used as a qualitative characteristic o f the ionospheric absorption in the D - region, and our interest “is there any trend in the fm in data, and i f “y e s ” w hat is its sign?” For this
lines (dashed lin es) and dem onstrates that the long-tim e course o f fm in (132) is sim ilar to th ose found for F 10.7 and foF 2 (figures 1-3) in the event that the periodicity o f 3 4 -3 6 years is also evident in the fm in (132) variation w h ich is found in variations o f the F2-layer parameters. H ow ever, figure 5 clear illustrates a stable upward (positive) trend in fm in (132) values opposite in sign to the trends observed in variations F 10.7 and foF2. It m eans relative high sensitivity o f the fm in values to the solar activity changes and significant influence o f other trend drivers on them that is very possib le the im pact o f anthropogenic factors on the state o f the low er ionosphere.
C o n c lu sio n . The F2-layer critical frequency (foF 2) and lo w e st frequency (fm in) observed at the m id
latitude ionospheric station A lm a-A ta [4 3 .2 5 N , 76.92E ] in the period 1957-2017 w ere used to study lo n g term trends in the upper (F2-layer) and low er (D -region) ionosphere.
The geom agn etic indexes (A p), the near-noon, near-m idnight, daily averaged foF 2 are found to be in strong dependence from the solar activity; all o f them sh ow a dom inant pattern o f variation w ith a period o f ~ 3 4 -3 6 years and linear negative trend.
The trend m agnitudes are found to be o f the sam e order independently o f the fact w hether the solar activity effects in the data sets are sm oothed or not; the trends are statistically significant and lie w ithin o f -0.018 to -0 .0 2 8 , -0.011 to -0 .0 1 9 , -0 .0 0 3 6 to -0.0093 M H z/year for near-noon, daily averaged and near
m idnight values correspondingly.
The long-term course o f fm in is sim ilar to those found in F 10.7, A p, and foF 2, i.e. the periodicity o f
~ 3 4 -3 6 years is also evident in the fm in variation. H ow ever, in contrast to the F2-layer parameters, the fm in variation clear dem onstrates an upward (p ositive) linear trend that is opposite in sign to the trend found in the F 10.7. This m eans relatively high sensitivity o f the fm in values to the solar activity changes and significant influence o f other trend drivers on them , one o f w h ich is the p ossib le im pact o f anthropogenic factors on the state o f the low er ionosphere.
T he w o r k w a s c a r r ie d o u t w ith a c c o rd a n c e G ra n t № A P 0 5 1 3 1 2 6 1 o f M in is tr y o f E d u c a tio n a n d S c ien ce RK.
А.Ф. Яковец, Г.И. Гордиенко, Ю .Г. Литвинов
ЕЖШС «Ионосфера институты» «¥ГЗТО» АК, Алматы, Казакстан
ИОНОСФ ЕРАНЬЩ F Ж ЭНЕ D АЙМ АГЫ НДАГЫ ¥З А Ц МЕРЗ1МД1 ВАРИАЦИЯЛАРЫ
Аннотация. Алматы станциясындагы ионосфераньщ жер Y ^ i тж радиозондауыньщ деректерi бойынша F2 Ионосфера кабатынын (foF2) жэне жер манындагы ен теменп шагылысу жишктершщ (fmin) узак мерзiмдi езгерiстерi зерттелдi [43.25 N, 76.92 E] 1957-2017 жж.
Сонымен катар, ен аз шагылысу ж и ш п теменп ионосферада зондирлеушi сигналдын жутылуына да, шу денгеш мен ионозондтын техникалык сипаттамаларына да байланысты. Сондыктан сигналдын жутылуынын абсолютпк мэнш осындай жолмен багалау проблемалы болып табылады. Алайда, колшатыр сигналынын жуту денгешнщ индикаторы ретшде сапалы сипаттама ретшде fmin параметрш пайдалануга болады.
Бастапкы деректер репнде fof2 медиандык fof2 (10-14 LT), терек жанындагы (23-01 LT) сагат арифметикалык орташа мэндерi жэне ионосферанын бакылауынын каралып отырган кезенiндегi орташа тэулiктiк мэндерi карастырылды. Теменгi ионосферанын езгергiштiгi кYндiзгi сагат Yшiн fmin деректерiн пайдалану аркылы зерттелген. ^ н радиосэулелендiру агынынын орташа айлык мэндерi F10. 7 жэне Ар индекс кYн жэне геомагнитпк белсендiлiктiн сипаттамалары ретiнде рассматрен.
^ н жэне геомагниттiк белсендiлiгi бойынша деректер ионосферанын жагдайына эсер ететiн факторлар ретiнде тартылды. Ионосфера аймагынын F2 параметрлерi, сондай-ак Ар геомагниттiк белсендшктщ индексi Т ~ 35 жыл мерзiмдiлiгiн жэне сызыктык терiс трендiн аныктай отырып, ^ н белсендiлiгiне катан тэуелдiлiктi сезшеда. FoF2 жэне F10.7 узак мерзiмдi вариация арасындагы детерминация коэффициенттерi 0.99 мэндерше жетедi, бул кYн белсендш п Ионосфера аймагынын F2 узак мерзiмдi езгергiштiгiнiн негiзгi драйверi болып табылады деп болжайды.
Бул нэтиже бiр жагынан кYн белсендiлiгiне жогары тэуелдш кп, екiншi жагынан баска факторлардын теменп ионосферанын жагдайына, антропогендш факторларга эсерш болжайды.
ТYЙiн сездер: орташалыкты ионосфера; жогаргы жэне теменгi ионосфера; D -облыс; F2-облыс; узак мерзiмдi трендтер.
ДТОО «Институт ионосферы» АО «НЦКИТ», Алматы, Казахстан ДОЛГО ВРЕМ ЕННЫ Е ВАРИ АЦИИ В ОБЛАСТИ F И D ИОНОСФ ЕРЫ
Аннотация. Исследованы долговременные изменения околополуденных, среднесуточных и полуночных критических частот слоя F2 ионосферы (foF2) и околополуденных минимальных частот отражения (fmin) по данным наземного вертикального радиозондирования ионосферы на станции Алма-Ата [43.25N, 76.92E] за период 1957-2017 гг.
Принималось во внимание, что минимальная частота отражения зависит как от поглощения зондирующего сигнала в нижней ионосфере, так и от уровня шумов и технических характеристик ионозонда.
Поэтому оценить абсолютное значение поглощения сигнала таким способом представляется проблематичным. Однако, как качественная характеристика, как индикатор уровня поглощения зондирующего сигнала, параметр fmin можно использовать.
В качестве исходных данных рассмотрены арифметически средние значения медианных foF2 для околополуденных (10-14 LT), околополуночных (23-01 LT) часов и среднесуточные значения за рассматриваемый период наблюдений ионосферы. Изменчивость нижней ионосферы исследована с использованием данных fmin для дневных часов. Среднемесячные значения потока радиоизлучения Солнца F10.7 и индекс Ар рассматрены как характеристики солнечной и геомагнитной активности.
Привлекались данные по солнечной и геомагнитной активности как факторы, оказывающие влияние на состояние ионосферы. Получено, что параметры F2 области ионосферы, а также индекс геомагнитной активности Ар испытывают строгую зависимость от солнечной активности, обнаруживая периодичность Т ~ 35 лет и линейный отрицательный тренд. Коэффициенты детерминации взаимосвязи долгопериодных вариаций foF2 и F10.7 достигают значений 0.99, это предполагает, что солнечная активность является основным драйвером долговременной изменчивости F2 области ионосферы.
Значения fmin, обнаруживая подобно параметрам F2 области ионосферы ~3 5-летнюю периодичность, демонстрируют положительный тренд в наблюдаемый период 1957-2017 гг. Данный результат предполагает с одной стороны высокую зависимость от солнечной активности, с другой стороны влияние других факторов на состояние нижней ионосферы, возможно факторов антропогенных.
Ключевые слова: среднеширотная ионосфера; верхняя и нижняя ионосфера; D -область; F2-область;
долговременные тренды.
Information about authors:
Gordienko Galina Ivanovna, Ph.D. mat. Sciences, ANS, Institute of the Ionosphere, JSC "NCSCIT", Almaty, [email protected]. 1) The author made a significant contribution to the research concept, obtaining data and interpretation, 2) the author wrote the first version of the article 3) the author approved the final version of the article before submitting it for publication. https://orcid.org/0000-0002-9337-206X;
Yakovets Arthur Fedorovich, Ph.D. mat. Sciences, VNS, Institute of the Ionosphere, JSC "NTSKIT", Almaty, [email protected]. 1) The author made a significant contribution to the research concept, data analysis and interpretation, 2) the author wrote the first version of the article, 3) the author approved the final version of the article before submitting it for publication. https://orcid.org/0000-0002-3374-9507;
Litvinov Yuri Georgievich, Ph.D. mat. Sci., SSS, Institute of the Ionosphere, NCSCIT JSC, Almaty, [email protected].
1) The author made a significant contribution to obtaining the data and their analysis, 2) the author made a critical review of the article for important intellectual content, 3) the author approved the final version of the article before submitting it for publication.
https://orcid.org/0000-0002-0600-0517
REFERENCES
[1] Gordiyenko G. I., Vodyannikov V. V., Yakovets A. F. and Litvinov Yu. G. (2014), Earth Planets and Space, 66 (125), 1-6, http://www.earth-planets-space.com/content/66/1/125.
[2] Lauter E. A., Entzian G. Winter anomaly 1980/81 as an example of stratomesospheric coupling//Phys. Solariterr. 1982.
Vol. 18. P. 83-90.
[3] Kokourov V. D. Minimum frequency of reflections as a climatic characteristic of the upper atmosphere// Geomagnetism and Aeronomy. 2003. Vol. 43(2). P.274-276 (in Russian).
[4] Clua de Gonzalez A. L, Gonzalez W. D, Dutra S. L. G., Tsurutani B. T. Periodic variation in the geomagnetic activity: a study based on the Ap index//J Geophys. Res. 1993. Vol. 98. P. 9215.
[5] Echer E., Rigozo N. R., Nordemann D. J. R., and Vieira L. E. A. (2004). Prediction of solar activity on the basis of
