Content uploaded by J.-E. Solheim
Author content
All content in this area was uploaded by J.-E. Solheim on Jan 11, 2014
Content may be subject to copyright.
Pattern Recogn. Phys., 1, 159–164, 2013
www.pattern-recogn-phys.net/1/159/2013/
doi:10.5194/prp-1-159-2013
©Author(s) 2013. CC Attribution 3.0 License.
Open Access
Pattern Recognition
in Physics
The sunspot cycle length – modulated by planets?
J.-E. Solheim
formerly at: Department of Physics and Technology, Universtity of Tromsø, Norway
Correspondence to: J.-E. Solheim (janesol@online.no)
Received: 1 October 2013 – Revised: 8 November 2013 – Accepted: 12 November 2013 – Published: 4 December 2013
Abstract. The Schwabe frequency band of the sunspot record since 1700 has an average period of 11.06yr
and contains four major cycles, with periods of 9.97, 10.66, 11.01 and 11.83yr. Analysis of the O–C residuals
of the timing of solar cycle minima reveals that the solar cycle length is modulated by a secular period of about
190yr and the Gleissberg period of about 86 yr.
Based on a simple harmonic model with these periods, we predict that the solar cycle length will in average
be longer during the 21st century. Cycle 24 may be about 12yr long, while cycles 25 and 26 are estimated
to be about 9 and 11yr long. The following cycle is estimated to be 14yr long. In all periods during the last
1000yr, when the solar cycle length has increased due to the 190yr cycle, a deep minimum of solar activity
has occurred. This is expected to re-occur in the first part of this century.
The coherent modulation of the solar cycle length over a period of 400yr is a strong argument for an external
tidal forcing by the planets Venus, Earth, Jupiter and Saturn, as expressed in a spin-orbit coupling model.
1 Introduction
A possible relation between solar activity as manifested by
sunspots and the Earth’s climate has been discussed many
times since William Herschel (1801) speculated on a possi-
ble connection. In recent times Reid (1987) showed, based
on data on globally averaged sea surface temperature (SST),
that the solar irradiance may have varied in phase with the
80–90yr cycle represented by an envelope of the 11yr solar
activity cycle, called the Gleissberg cycle.
Friis-Christensen and Lassen (1991) investigated the re-
lation between the sunspot numbers and Northern Hemi-
sphere land temperature, and found similar variations, but
with the temperature variations leading the sunspot numbers.
They then discovered that using the solar cycle length (SCL)
as an indicator of solar activity in the sense that a shorter
cycle means higher activity, they could much better corre-
late with the NH land temperature variations. It was also
demonstrated (Friis-Christensen and Lassen, 1991; Hoyt and
Schatten, 1993; Larssen and Friis-Christensen, 1995) that the
correlation between SCL and climate has probably been in
operation for centuries. A statistical study of 69 tree ring
sets, covering more than 594yr, demonstrated that wider tree
rings (better growth conditions) were associated with shorter
sunspot cycles (Zhou and Butler, 1998).
In their study, Friis-Christensen and Lassen (1991) used
a smoothed mean value for the SCL with the length of five
solar cycles weighted 1-2-2-2-1. In a follow-up paper, Re-
ichel et al. (2001) concluded that the right cause-and-effect
ordering, in the sense of Granger causality, is present be-
tween the smoothed SCL and the cycle mean temperature
anomaly for the Northern Hemisphere land air temperature
in the 20th century at the 99% significance level. This sug-
gests that there may exist a physical mechanism linking so-
lar activity to climate variations. However, at the turn of the
century, a discrepancy between the SCL and NH land series
developed (Thejll and Lassen, 2000; Thejll, 2009), because
the short cycle 22 was followed by a much longer cycle 23,
without sign of cooling.
Recognizing that averaged temperature series from differ-
ent meteorological stations of variable quality and changing
locations may contain errors and partially unknown phenom-
ena derived from the averaging procedure, Butler (1994) pro-
posed instead to use long series of high quality from sin-
gle stations. He showed that this improved the correlation
when used for temperature series for Armagh, which corre-
lates strongly with the NH land temperature.
Archibald (2008) was the first to realize that the length of
the previous sunspot cycle (PSCL) has a predictive power
Published by Copernicus Publications.
160 J.-E. Solheim: The sunspot cycle length – modulated by planets?
for the temperature in the next sunspot cycle for certain lo-
cations, if the raw (unsmoothed) value for the SCL is used.
Based on the estimated longer SC23 than SC22, he predicted
cooling during SC24 for some selected locations. A system-
atic study of the correlation for locations around the North
Atlantic was published by Solheim et al. (2012). They found
that maximum correlation was obtained with an 8–12yr lag
for locations around and in the North Atlantic, and found that
a correlation with a lag of one solar cycle could explain 25 to
72 per cent of the temperature variance in that region. This
one cycle lag could therefore be used for forecasting the tem-
perature in the next solar cycle. Based on SC23 being consid-
erably longer than SC22, they forecast a temperature decline
during SC24 for the sites investigated.
In order to forecast the development of SCL for longer pe-
riods, it is necessary to investigate the long-term variability
of the SCLs. This was done for the first time by Fairbridge
and Hameed (1983), who found that the phase differences re-
peated after 16 sunspot cycles, or 178 yr, if they used minima
as the start time for a cycle.
This was followed up by Richards et al. (2009), who used
median trace analyses of the SCL and power spectrum anal-
ysis of the O–C residuals (as explained in Eq. 1). They found
that the solar cycle length is controlled by periods of 188
and 87yr. They concluded that the length of the solar cycle
should increase gradually the next ≈75yr. They did not dis-
cuss the origin of their determined periods.
Regarding the 11yr sunspot period, many scientists have
noticed the bimodal structure of the distribution of solar cy-
cle length. According to analysis by Scafetta (2012), the
sunspot length probability distribution consists of three pe-
riods of about 9.98, 10.9 and 11.86yr. The side periods ap-
pear to be closely related to the spring period of Jupiter and
Saturn, which has a range between 9.5 and 10.5yr with a
median length of 9.93yr, and the sidereal period of Jupiter
(about 11.86yr). Scafetta (2012) proposed that the central
cycle period is associated with a quasi 11yr solar dynamo
cycle, which is forced by the two cyclical side attractors with
periods of 9.93 and 11.86 yr. He also suggested that the secu-
lar variations of the solar cycle amplitude and length are beat
periods of the three solar cycle periods, and that it is possible
to describe the secular variations of the sunspot cycle with
these beat periods.
Scafetta’s analysis covered the period 1755–2008 (solar
cycles 1–23). In the following we will investigate the solar
cycles for the longer period 1700–2010, and we will also in-
vestigate the O–C residuals all the way back to 1610 to search
for period combinations or harmonics. Based on a simple
harmonic model we will estimate the length of the next solar
cycles. Finally we will discuss if the modulation of the SCL
may be controlled by the planets, as proposed by Scafetta
(2012) and Wilson et al. (2008).
2 Data and methods
Yearly average sunspot numbers were downloaded
from the Solar Influences Data Center (SIDC). The
length and time of solar cycles were downloaded
from http://www.ngdc.noaa.gov/stp/space-weather/
solar-data/solar-indices/sunspot-numbers/cycle-data/table_
cycle-dates_maximum-minimum.txt.
For the analysis of the sunspot number time series I have
used the Period04 analysis package (Lenz and Breger, 2005),
downloaded from the Period04 website at http://www.astro.
univie.ac.at/dsn/dsn/Period04/. This program performs least
square fitting of a number of frequencies, where initial fre-
quencies may be determined by Fourier transform (FT) or
given as input. Error analysis is done by an analytical for-
mula (Breger et al., 1999) assuming an ideal case, or with
a least square error calculation. The largest of the obtained
errors is used.
The O–C technique for investigation of secular modulation
of the SCL is described in detail in Richards et al. (2009).
We follow their description and use the downloaded set of
SCLs determined between the minima, and construct the O–
C residuals cycle by cycle using the formula:
(O−C)i=(ti−t0)−(Ni×P0),(1)
where tiis the end time of cycle no. Ni,P0is the reference
period investigated, andCi=t0+Ni×P0.
3 Results
3.1 The 11yr cycle
The solar cycle length variation with time since 1610 is
shown in Fig. 1. We notice large variations in the 17th and
18th centuries, but with a generally shorter length from about
1850. The data set covers a total of 36 cycles, and the mean
length is 11.06±1.5 yr. In Fig. 2 we show the distribution of
the SCL between solar minima. The median value is between
10.7 and 11.0yr, but there are no observations in this range.
This clearly indicates a double or multiple bell distribution.
The resulting periodogram of the sunspot numbers from
1700–2010 is shown in Fig. 3. We find, as did Scafetta
(2012), a dominating band with periods 10–12yr, where
we identify four peaks: P1=9.97 ±0.02, P2=10.66±0.02,
P3=11.010±0.001 and P4=11.83±0.02 yr. The errors are
determined by an analytical formula (Breger et al., 1999).
There is also a triplet of periods in an 8.5 yr band, and a triplet
around 5.5yr. The latter is most likely higher harmonics of
three peaks in the 11yr band.
The long period of 53±0.6 yr is best explained as a 4th
subharmonics of P2(5 ×10.66 =53.3), and the long period
of 100±15 yr may be related to the known Gleissberg period
of 87yr.
Pattern Recogn. Phys., 1, 159–164, 2013 www.pattern-recogn-phys.net/1/159/2013/
J.-E. Solheim: The sunspot cycle length – modulated by planets? 161
8
10
12
14
16
Length(yr)
8
10
12
14
16
1650 1700 1750 1800 1850 1900 1950 2000
Yea r
Solar Cycle length
Figure 1. The solar cycle length (SCL) from 1610 as downloaded
from the National Geophysical Data Center (NGDC). We observe
that the SCL was longer than the mean of 11.06yr in most of the
19th century and shorter than the mean in most in the 20th century.
0
1
2
3
4
5
6
7
8
No
78910111213141516
Length (Year)
Sunspot Cycle Length between minima
Figure 2. The distribution of the solar cycle lengths in bins of 0.5 yr
width. The distribution covers 36 cycles from 1610 to 2008.
3.2 Long-term modulation of the length of the solar
cycle
We use the average period P=11.06yr as our reference pe-
riod and obtain the O–C residuals as shown in Fig. 4, where
the O–C residuals are given as a function of the cycle no.
As the starting point for cycle −13 we use 1610.8 with an
O–C=−0.95. The residuals give us a picture of the long-
term trends in SCL. We observe that the residuals increase
most of the time between SC4 and SC14 (1775–1900), be-
cause the SCL is then nearly always longer than 11yr (see
also Fig. 1). Then we enter a period with shorter periods, and
a warming Earth. The question is now if that will continue.
To investigate what controls the length of the solar cycle,
we calculate a periodogram of the residual O–C data string,
and get the amplitude spectrum shown in Fig. 5.
The spectrum consists of two dominating periods: 190±9
and 85.6±2yr. Periods shorter than 50 years are harmonics
of the two main periods. There is also a period of the order
440yr, which explains why the peak around 1900 is higher
5
10
15
20
25
30
35
Amplitude
5
10
15
20
25
30
35
0.05 0.1 0.15 0.2 0.25
Frequency (Year-1)
Periodogram of Sunspot Numbers 1700-2010
100
53
21.3
11. 83
11. 01
10.66
9.97
8.47
5.5
Figure 3. Amplitude spectrum of the yearly average sunspot num-
bers 1700–2010.
-4
-2
0
2
4
O-C (yrs)
-4
-2
0
2
4
-15 -10 -5 0 5 10 15 20 25
Sunspot Cycle No
O-C for Sunspot Cycle Minima P = 11.06 yrs
1610 1712 1810 1900 2015
Figure 4. O–C residuals for the length of the solar cycle compared
with the average period of 11.06yr. The curve is increasing for
SCL>11.06 yr.
than the peak around 1700. A similar result was obtained by
Richards et al. (2009), who identified a Gleissberg period of
86.5±12.5 yr and a secular period of 188±38 yr. In their anal-
ysis they use SCLs based on both solar maxima and minima.
In Fig. 6 we show the O–C residuals with the strongest
controlling period ≈190yr and its subharmonic at ≈440 yr.
This dominant cycle is the reason for an increasing period
length in the 19th century and a decreasing length in the 20th
century. We can therefore expect increasing SCLs in the 21st
century.
Adding the Gleissberg cycle and three of the harmonics
gives the fit shown in Fig. 7, where we may also obtain an
estimate of near future SCLs. Times of minima can be esti-
mated from the following equation:
tmin =1755.5+11.06×Ni+(O−C)est,(2)
where (O–C)est is the estimated O–C value determined with
the harmonic model as shown in Fig. 7 (red curve). For the
www.pattern-recogn-phys.net/1/159/2013/ Pattern Recogn. Phys., 1, 159–164, 2013
162 J.-E. Solheim: The sunspot cycle length – modulated by planets?
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
Amplitude
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
0.0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Frequency (Cycles-1)
Periodogram of O-C residuals
190
86 47-42 38
Figure 5. Amplitude spectrum of O–C residuals of the SCL mea-
sured between minima.
-4
-2
0
2
4
O-C (yrs)
-4
-2
0
2
4
-15 -10 -5 0 5 10 15 20 25 30
Sunspot Cycle No
O-C for Sunspot Cycle Minima P = 11.06 yrs
Figure 6. O–C residuals for SCL minima, with a simulation based
on the dominating periods of 190 and 440 yr.
next minimum after SC24, Eq. (2) gives 2020.9, since the
(O–C)est then is close to zero.
4 Discussion
We have shown that the solar cycle length since 1600 is con-
trolled by stable oscillations, which provide an average cycle
length of 11.06yr. The cycle length is modulated by 3 long
periods of ≈440, ≈190 and ≈86yr, and some of their har-
monics. If the dominating period of ≈190 yr is followed back
in time, it is found (Richards et al., 2009) that all known solar
deep minima during the last 1000yr (the Oort, Wolf, Spörer,
Maunder and Dalton minima) are close to the minimum or on
the rising branch of this oscillation. We can therefore expect
another grand minimum during the first part of this century.
Looking more closely at the model simulations in Fig. 7,
we estimate the length of SC24 ≈12 yr, SCL25 ≈9yr, SCL26
≈11yr and SCL27 ≈14 yr. The forecast for the time of the
next minimum (2020.9) can be compared with the forecast
-4
-2
0
2
4
O-C (yrs)
-4
-2
0
2
4
-15 -10 -5 0 5 10 15 20 25 30
Sunspot Cycle No
O-C for Sunspot Cycle Minima P = 11.06 yrs
Figure 7. O–C residuals for SCL minima, with a simulation based
on 6 harmonics with periods 440, 190, 86, 48, 43, and 38 yr.
based on a mathematical model (Salvador, 2013), which es-
timates the end of solar cycle 24 in 2018.
It has for some time been discussed if the solar cycle length
is controlled by an internal or external clock. Dicke (1978)
argued that the phase of the solar cycle appears to be cou-
pled to an internal clock, because shorter cycles are usually
followed by longer cycles, as if the Sun remembers the cor-
rect phase. Another view (Huyong, 1996) is that the memory
effect can be explained by mean field theory, which predicts
coherent changes in frequency and amplitude of a dynamo
wave. However, it is admitted by solar physicists that present
solar dynamo theories, although able to describe the peri-
odicities and the polarity reversal of solar activity well, are
not yet able to quantitatively explain the 11 and 22yr cycles,
nor the other observed quasi-cycles (de Jager and Versteegh,
2005). The remarkable resemblance between planetary tidal
forcing periods and observed solar quasi-periods is a strong
argument for a planetary tidal forcing on the solar activity.
Regarding the splitting of the 11yr solar cycle band into
4 distinct peaks, the most remarkable is the strongest peak
P=11.010±0.001 yr. A period so close to 11 Earth years
has a great chance to be related to the Earth’s orbit. Wilson
(2013) explains that the Venus–Earth–Sun periodic align-
ments create a tidal bulge, which for a period of 11.07yr is
speeded up by Jupiter’s movement, and the next 11.07yr are
slowed down by the same. This is called the VEJ tidal-torque
coupling model, and explains both the average Schwabe and
Hales cycles. These tidal forces work to increase or decrease
the solar rotation rate in the convective layers where the solar
dynamo is situated (Wilson, 2013).
Among the other three periods in the 11yr band, 9.97 yr
is close to the Jupiter–Saturn spring tide period of 9.93yr,
which is half of the Jupiter–Saturn heliocentric conjunc-
tion period of 19.86yr. It should be noticed that the spring
tide period of Jupiter/Saturn varies between 9.5 and 10.5yr
(Scafetta, 2012). The period of 11.83yr is close to Jupiter’s
orbital period of 11.86yr. Scafetta (2012) proposes that the
solar cycle period ≈11.0yr is generated by the two side at-
tractors controlled by the two giant planets. We have found
another sunspot period at 10.66yr, which also may be a
Pattern Recogn. Phys., 1, 159–164, 2013 www.pattern-recogn-phys.net/1/159/2013/
J.-E. Solheim: The sunspot cycle length – modulated by planets? 163
dynamo period. Both these periods are strongly forced, since
they have higher harmonics of 5.5 and 5.25yr, and one sub-
harmonic of 21.3yr.
By our O–C analysis we find, as did Richards et al. (2009),
that the SCL is modulated by a secular period of 190±9 yr
in addition to a period of 86±2 yr, which most likely is the
Gleissberg period. The long period is close to the Jose cycle
of 178.7yr, which is the period of recurrent pattern of the
movement of the Sun around the barycenter of the solar sys-
tem (Jose, 1965). Fairbridge and Hameed (1985) found phase
coherence of solar cycle minima over two 176yr cycles, or
16 Schwabe periods. Our 190yr period is also close to a pe-
riod of 208 yr, which is found in cosmic ray observations and
in cosmogenic isotopes, and explained by tidal torque on the
Sun by the planets (Abreu et al., 2012).
However, a far better match with the 190 yr period is found
by introducing a so-called Gear Effect, which modulates the
tangential torque applied by the alignments of Venus and
Earth to the Jupiter–Sun–Saturn system as explained by Wil-
son (2013). He shows that prograde and retrograde torque os-
cillate in a quasi bidecadal period controlled by the 19.859 yr
synodic period of Jupiter and Saturn. Figure 13 in Wilson
(2013) shows the angel between the center of mass of the
Jupiter, Sun and Saturn system and Venus/Earth from 1013
to 2015. If we compare this with our Fig. 6, we find an excel-
lent match between periods and phases, indicating a strong
link between the modulation of the solar cycle length and
the torque effect proposed by Wilson (2013). The modula-
tion period can be calculated as the beat period between the
Hale-like period of 22.137yr and the Jupiter–Saturn synodic
period of 19.859 yr. The result is a beat period of 192.98 yr or
193±2 yr, when the orbital variations are included (Wilson,
2013). By also introducing the Gear Effect to the VEJ-tidal
torque model, he can also explain an 88.1yr Gleissberg cy-
cle.
Finally, it may be instructive to compare our predictions
of the next solar cycle lengths with a prediction made by de
Jager and Duhau (2009), based on the dynamo model that is
constructed from the relationship between the polodial and
torodial magnetic cycles. They conclude that the polar cycle
24 will be similar to polar cycle 12, which means that the
maxima of sunspot cycles 23 and 24 will be quite similar to
those of the cycle pair 11 and 12. They further conclude that
a short Dalton minimum will occur, lasting a maximum of 3
cycles (SC24-26), whereafter a grand minimum will follow,
starting with cycle 27. They predict the maximum sunspots
of SC24 to be 68±17 with a maximum at 2014.5±0.5, but
do not predict the length.
At the moment we are close to the solar maximum of
SC24, but have 7 more years to the next minimum, accord-
ing to our forecast. During that period we will observe if the
cooling forecast for the North Atlantic region will take place,
and if this will also keep the global temperature in hiatus, as
it has been since the start of SC23.
5 Conclusions
We have shown that the Schwabe frequency band of the
sunspot record since 1700 has an average period of 11.06yr
and contains four major cycles, with periods of 9.97, 10.66,
11.01 and 11.83 yr. Analysis of the O–C residuals of the tim-
ing of solar cycle minima reveals that the solar cycle length is
modulated by a secular period of about 190yr and a Gleiss-
berg period of about 86yr. Our result is a confirmation of
earlier phase studies by Fairbridge and Hameed (1983) and
Richards et al. (2009).
Based on a simple harmonic model with these periods, we
predict that the solar cycle length will increase during the
21st century. Cycle 24 may be about 12yr long, while cycles
25 and 26 are estimated to be about 9 and 11yr long. The
following cycle 27 will be much longer. In all periods when
the solar cycle length has increased due to the 190yr cycle
during the last 1000yr, a deep minimum of solar activity has
occurred. This is also to be expected in the early part of this.
The coherent modulation of the solar cycle length over a
period of 400yr is a strong argument for an external forcing
by the planets Venus, Earth, Jupiter and Saturn, expressed in
the spin-orbit coupling model as proposed by Wilson (2013).
Excellent phase coherence with this model is a strong
added argument for this interpretation.
Acknowledgements. The author acknowledges the use of
sunspot numbers and times of minima from the National Geophys-
ical Data Center. He also thanks the Vienna astroseismological
group for the excellent Period04 program package, and two referees
with helpful advice for improving this publication.
Edited by: N.-A. Mörner
Reviewed by: H. Yndestad and H. Jelbring
References
Abreau, J. A., Beer, J., Ferriz-Mas, A., McCracken, K. G., and
Steinhilber, F.: Is there a planetary influence on solar activity?
Astron. Astrophys., 548, 9 pp., 2012.
Archibald, D.: Solar cycle 24: Implications for the United
States, in: International Conference on Climate Change (www.
davidarchibald.info), 2008.
Breger, M., Handler, G., Garrido, R, Audard, N., Zima, W., Paparó,
M., Beichbuchner, F., Li, Zhi-ping, Jiang, Shi-yang, Liu, Zong-
li, Zhou, Ai-ying, Pikall, H., Stankov, A., Guzik, J. A., Sperl, M.,
Krzesinski, J., Ogloza, W., Pajdosz, G., Zola, S., Thomassen, T.,
Solheim, J.-E., Serkowitsch, E., Reegen, P., Rumpf, T., Schmal-
wieser, A., and Montgomery, M. H.: 30+frequencies for the delta
Scuti variable 4 Canum Venaticorum: results of the 1996 multi-
site campaign, Astron. Astrophys., 349, 225–235, 1999.
Butler, C. J.: Maximum and minimum temperatures at Armagh Ob-
servatory, 1844–1992, and the length of the sunspot cycle, Sol.
Phys., 152, 35–42, 1994.
de Jager, C. and Versteegh, J. M.: Do planetary motions drive solar
variability?, Sol. Phys., 229, 175–179, 2005.
www.pattern-recogn-phys.net/1/159/2013/ Pattern Recogn. Phys., 1, 159–164, 2013
164 J.-E. Solheim: The sunspot cycle length – modulated by planets?
de Jager, C. and Duhau, S.: Forecasting the parameters of sunspot
cycle 24 and beyond, J. Atmos. Sol.-Terr. Phy., 71, 239–245,
2009.
Dicke, R. H.: Is there a chronometer hidden deep in the Sun?, Na-
ture, 276, 676–680, 1978.
Fairbridge, R. W. and Hameed, S.: Phase coherence of solar cy-
cle minima over two 178-year periods, Astron. J., 88, 867–869,
1983.
Friis-Chrisetensen, E. and Lassen, K.: Length of the solar cycle:
an indicator of the solar activity closely associated with climate,
Science, 254, 698–700, 1991.
Herschel, W.: Observations tending to investigate the nature of the
Sun, in order to find the causes or symptoms of its variable emis-
sion of light and heat: With remarks on the use that may possibly
be drawn from solar observations, Philos. T. R. Soc. Lond., 91,
265–318, 1801.
Hoyng, P.: Is the solar cycle timed by a clock?, Sol. Phys., 169,
253–264, 1996.
Hoyt, D. V. and Schatten, H. K.: A discussion of plausible solar
irrandiance variations, 1700–1992, J. Geophys. Res., 98, 18895–
18906, 1993.
Jose, P. D.: Sun’s Motion and Sunspots, Astron. J., 70, 193–200,
1965.
Larssen, K. and Friis-Christensen, E.: Variability of the solar cycle
length during the past five centuries and the apparent association
with terrestrial climate, J. Atmos. Sol.-Terr. Phy., 57, 835–845,
1995.
Lenz, P. and Breger, M.: Period04 User Guide, Communications in
Asteroseismology, 146, 53–136, 2005.
Reichel, R., Thejll, P., and Lassen, K.: The cause-and-effect rela-
tionship of solar cycle length and the Northern Hemisphere air
surface temperature, J. Geophys. Res., 106, 15635–15641, 2001.
Reid, G. C.: Influence of solar variability on global sea surface tem-
peratures, Nature 329, 142–143, 1987.
Richards, M. T., Rogers, M. L., and Richard, D. St. P.: Long-Term
variability in the Length of the Solar Cycle, Publ. Astron. Soc.
Pac., 121, 797–809, 2009.
Salvador, R. J.: A mathematical model of the sunspot cycle for the
past 1000 yr, Pattern Recogn. Phys., in preparation, 2013.
Scafetta, N.: Multi-scale harmonic model for solar and climate
cyclical variation throughout the Holocene based on Jupiter-
Saturn tidal frequencies plus the 11.year solar dynamo cycle, J.
Atmos. Sol.-Terr. Phy., 80, 296–311, 2012.
Solheim, J.-E., Stordahl, K., and Humlum, O.: The long sunspot
cycle 23 predicts a significant temperature decrease in cycle 24,
J. Atmos. Sol.-Terr. Phy., 80, 267–284, 2012.
Thejll, P. and Lassen, K.: Solar forcing of the Northern Hemisphere
land air temperature: new data, J. Atmos. Sol.-Terr. Phy., 62,
1207–1213, 2000.
Thejll, P.: Update of the Solar Cycle Length Curve, and the Rela-
tionship to the Global Mean Temperature, Danish Climate Centre
Report 09-01, 2009.
Wilson, I. R. G., Carter, B. D., and Waite, I. A.: Does a Spin-Orbit
Coupling Between the Sun and the Jovian Planets Govern the
Solar Cycle?, Publ. Astron. Soc. Aust., 25, 8–95, 2008.
Wilson I. R. G.: The Venus-Earth-Jupiter Spin-Orbit Coupling
Model, Pattern Recogn. Phys., in preparation, 2013.
Zhou, K. and Butler, C. J.: A statistical study of the relationship be-
tween the solar cycle length and tree-ring index values, J. Atmos.
Sol.-Terr. Phy., 60, 1711–1718, 1998.
Pattern Recogn. Phys., 1, 159–164, 2013 www.pattern-recogn-phys.net/1/159/2013/