1: \documentclass[12pt,preprint]{aastex}

2: %%\documentclass[manuscript]{aastex}

3: 

4: %\newcommand{\vdag}{(v)^\dagger}

5: %\newcommand{\myemail}{skywalker@galaxy.far.far.away}

6: 

7: %\shorttitle{Solar Cycle prediction}

8: %\shortauthors{Cameron and Sch\"ussler}

9: 

10: \begin{document}

11: 

12: %\title{Properties of sunspot area/number time series and solar cycle

13: %prediction}

14: 

15: \title{A robust correlation between growth rate and amplitude of solar

16:   cycles: consequences for prediction methods}

17: 

18: \author{R. Cameron}\email{cameron@mps.mpg.de}

19: 

20: \and

21: 

22: \author{M. Sch\"ussler}\email{schuessler@mps.mpg.de}

23: 

24: \affil{Max-Planck-Institut f\"ur Sonnensystemforschung, 

25:    37191 Katlenburg-Lindau, Germany}

26: 

27: 

28: 

29: \begin{abstract}

30: We consider the statistical relationship between the growth rate of

31: activity in the early phase of a solar cycle with its subsequent

32: amplitude on the basis of four datasets of global activity indices (Wolf

33: sunspot number, group sunspot number, sunspot area, and 10.7-cm radio

34: flux). In all cases, a significant correlation is found: stronger cycles

35: tend to rise faster. Owing to the overlapping of sunspot cycles, this

36: correlation leads to an amplitude-dependent shift of the solar minimum

37: epoch. We show that this effect explains the correlations underlying

38: various so-called precursor methods for the prediction of solar cycle

39: amplitudes and also affects the prediction tool of Dikpati et al. (2006)

40: based upon a dynamo model.  Inferences as to the nature of the solar

41: dynamo mechanism resulting from predictive schemes which (directly or

42: indirectly) use the timing of solar minima should therefore be treated

43: with caution.

44: 

45: \end{abstract}

46: 

47: 

48: \keywords{Sun:Activity}

49: 

50: \section{Introduction} 

51: Solar activity is the driver of space weather, which has practical

52: consequences for human activities in space. This is one reason why the

53: search for methods to predict its (short-term and long-term) future

54: levels has found much interest in the literature. Further motivation

55: arises from the potential implications for understanding the {\em

56: origin} of solar activity: a reliable method of predicting the amplitude

57: of future solar cycles could provide a constraint on dynamo models. On

58: the other hand, the converse proposition is not necessarily true, as has

59: been pointed out by \citet{Bushby:Tobias:2007}: the nonlinear dynamics

60: of the dynamo might be such as to make mid- to long-term prediction

61: impossible even if an almost perfect physical understanding of the

62: dynamo mechanism is achieved.

63: 

64: In many cases, recipes for prediction are inferred from correlations

65: found in historical records of measured quantities, which are (directly

66: or indirectly) related to solar activity. As illustrated by

67: \citet[][cf. his Fig. 14.2]{Wilson:1994} and \citet[][cf. their

68: Fig.~6]{Lantos:Richard:1998}, the success of most methods in actually

69: {\em predicting} the unknown amplitude of a future cycle is rather

70: disappointing. Nevertheless, the correlations between several

71: `precursors', i.e., quantities measured during the descending or minimum

72: phase of a cycle and the amplitude of the subsequent cycle

73: \citep[e.g.,][]{Hathaway:etal:1999, Schatten:2003} might have a

74: non-random origin and thus call for a physical explanation. This could

75: have implications for dynamo models.

76: 

77: In this paper, we consider the effect of the overlapping of solar cycles

78: in combination with their asymmetric shape on the correlations between

79: precursors and following cycle amplitudes. In this connection, the

80: important aspect of the asymmetry is the difference of the

81: amplitude-dependent ascent rate in the early cycle phase (related to the

82: so-called Waldmeier effect) compared to the decay rate near the end of a

83: cycle. Since sunspot cycles overlap for typically 2 to 3 years

84: \citep{Harvey:1992a}, this asymmetry affects the timing of the activity

85: minima, which are pivotal epochs for most precursor methods. We show

86: that these effects can explain the correlations upon which such methods

87: are based, without necessarily implying a direct physical connection

88: between the precursor quantity and the following cycle. We also show

89: that the essence of the Waldmeier effect, i.e., that stronger cycles

90: tend to show a faster rise of activity levels during their ascending

91: phase than weaker cycles, is a robust property present in all activity

92: indices.  On this basis, we explain how cycle asymmetry and cycle

93: overlapping may also affect the dynamo-based prediction method of

94: \citet{Dikpati:Gilman:2006}, in spite of recent claims to the contrary

95: \citep{Dikpati:etal:2008}.

96: 

97: 

98: \section{Overlapping asymmetric cycles and precursors}

99: 

100: Precursor methods are based upon the existence of a correlation between

101: some physical quantity measured during the descending or minimum phase

102: of a cycle and the amplitude of the following cycle. The correlation is

103: established by considering historical records of data, the longest of

104: which is the record of sunspot numbers. Assuming a non-random origin of

105: such a correlation, there are three possible explanations for

106: its existence:

107: %

108: \begin{enumerate}

109: \item The precursor is a feature of the {\em old} cycle that represents

110:   or is related to an input quantity for the dynamo process, so that its

111:   magnitude directly affects the amplitude of the next cycle. For

112:   example, it has been suggested that the strength of the polar field

113:   during solar minimum is a measure of the poloidal field from which the

114:   toroidal field for the next cycle is generated

115:   \citep[e.g.,][]{Schatten:etal:1978, Choudhuri:etal:2007}.

116: \item The precursor quantity represents an early manifestation of the

117:   {\em new} cycle, which already affects the high latitudes of the Sun but

118:   does not yet produce sunspots. The notion of such an `extended solar

119:   cycle' \citep{Wilson:etal:1988} is supported by observations of

120:   coronal activity \citep{Altrock:1997}, ephemeral magnetic regions

121:   \citep{Harvey:1994a}, and zonal flows \citep{Howe:etal:2006,

122:   Altrock:etal:2008}.

123: \item If the definition of the precursor quantity directly or indirectly

124:   depends on the timing of the solar minimum, the amplitude-dependent

125:   shift of the minimum due to the overlap of asymmetric cycles can lead

126:   to a correlation of the precursor with the amplitude of the new cycle

127:   \citep[][see their Fig.~10]{Cameron:Schuessler:2007}. This does not

128:   necessarily involve a physical connection between the precursor and

129:   the dynamo process, in principle even permitting the prediction of

130:   random cycle amplitudes.

131:   

132: \end{enumerate}

133: %

134: The third possibility arises from the observation that information about

135: the height of the maximum of a cycle is already contained in its early

136: rise phase \citep{Waldmeier:1935,Waldmeier:1955}. In fact, the shapes of

137: most historical solar cycles can reasonably well be described by simple

138: functions containing a few parameters

139: \citep[e.g.,][]{Hathaway:etal:1994,Li:1999}, so that an estimate of the

140: further development of a cycle is possible a few years after sunspot

141: minimum \citep{Waldmeier:1936, Elling:Schwentek:1992}. Since cycles

142: overlap, i.e., the rise of the new cycle starts already when the decay

143: of the old cycle is still ongoing, the timing of the cycle minimum (the

144: epoch when the sum of the activities of both cycles is minimal) and its

145: height are affected by the amplitude of the new cycle: the minimum

146: occurs earlier and is higher for a subsequent strong cycle than for a

147: weak cycle. In fact, stronger cycles tend to be preceded by shorter

148: cycles \citep[earlier minima,][]{Solanki:etal:2002b} with higher minimum

149: activity levels. Both quantities can be considered as precursors with

150: correlation coefficients with the amplitude of the next cycle of about

151: 0.7 \citep{Hathaway:etal:1999} .

152: 

153: \citet{Cameron:Schuessler:2007} have noted that the amplitude-dependent

154: shift of the minimum epoch is enhanced by the amplitude-dependent

155: {\em asymmetry} of solar cycles: strong cycles tend to grow faster in activity

156: than weak cycles, while the decay rate in the late descending phase is

157: largely independent of the cycle amplitude.  This is related to, but not

158: identical to, the so-called `Waldmeier effect' \citep{Waldmeier:1935}

159: which is often stated in the form: the time between cycle minimum and

160: maximum is shorter for stronger cycles

161: \cite[e.g.,][]{Hathaway:etal:2002}. However, the crucial quantity for

162: the minimum shift is not the time between minimum and maximum but the

163: {\em steepness} of the activity rise (decay) in the initial (late) phase

164: of the cycle, which can be determined independently of the timing of the

165: activity minima and maxima by considering the slope of the activity

166: curve. We show in the next section that the correlation between initial

167: growth rate and cycle amplitude is a very robust feature shown by all

168: available global activity indices.

169: 

170: \section{The robust amplitude-dependent cycle asymmetry}

171: 

172: \citet{Hathaway:etal:2002} found that the `classical' Waldmeier effect,

173: i.e., the correlation between the rise time (time interval between

174: minimum and maximum) and the cycle amplitude, is by a factor of about

175: two weaker for the group sunspot numbers \citep{Hoyt:Schatten:1998} than

176: for the Wolf sunspot numbers. For both datasets, the historically

177: strongest cycles (numbers 18, 19, 21, and 22) do not fit the linear

178: relationship very well, showing somewhat long rise times. This weakening

179: of the correlation can be understood by the shift in the timing of the

180: minima: stronger, rapidly growing cycles lead to early preceding minima,

181: thus increasing the time between minimum and maximum. This effect

182: becomes even more pronounced in the analysis of sunspot area data by

183: \citet{Dikpati:etal:2008} owing to the fact that the sunspot areas for

184: cycles 21, 22 and 23 reach a slightly higher second `Gnevyshev maximum'

185: \citep{Gnevyshev:1967} a few years after the first maximum coinciding

186: with the sunspot number maximum. It is not surprising that determining

187: the rise time in terms of these later maxima destroys the correlation

188: with the cycle amplitude.

189: 

190: As we have pointed out in the previous section, the relevant quantity

191: for the minimum shift of overlapping cycles is not the time interval

192: between minimum and maximum but the {\em steepness} of the activity

193: growth (decay) in the initial (late) phase of the cycle.  An inspection

194: of Fig.~5 of \citet{Hathaway:etal:2002} and of Fig.~1 of

195: \citet{Dikpati:etal:2008} already suggests that the different datasets

196: do not differ strongly in this respect.  In order to give a quantitative

197: account, we need to measure the steepness without reference to the

198: epochs of maximum and minimum; in particular, the latter is already

199: affected by the correlations that we wish to study. For instance,

200: \citet{Lantos:2000} considered the slope of the sunspot number curve at

201: the inflexion point during the ascending part of cycles 9 to 22 and

202: found a correlation coefficient of $r=0.88$ with the cycle amplitude,

203: while the (anti)correlation with the rise time from minimum to maximum

204: only yields $\vert r\vert=0.61$. In our case, we need to consider

205: specifically the early rise and late decay phases, so that we estimate

206: the rise and decay rates by determining the time required for the

207: activity index under consideration to cover a fixed interval of

208: values. This interval is chosen such that it is not significantly

209: affected by the cycle overlap. The rise and decay rates are then defined

210: as the ratios of the value intervals and the corresponding rise and

211: decay times, respectively.

212: 

213: We have analyzed four datasets: monthly Wolf sunspot

214: numbers\footnote{http://sidc.oma.be} (SN) for cycles 7 to 23, monthly

215: group sunspot

216: numbers\footnote{http://www.ngdc.noaa.gov/stp/SOLAR/ftpsunspotnumber.htm}

217: \citep[GSN,][]{Hoyt:Schatten:1998} for cycles 7 to 22, sunspot areas

218: \citep[SAR,][]{Balmaceda:etal:2005} for cycles 12 to 23, and 10.7-cm

219: solar radio

220: flux\footnote{http://www.drao.nrc.ca/icarus/www/sol\_home.shtml} (SRF)

221: for cycles 19 to 23. While we have used the full available data sets for

222: SRF and SAR, the SN and GSN data are only considered from cycle 7 on

223: because a) we wish to compare the results for the different datasets

224: during roughly the same period of time, and b) for very low-amplitude

225: cycles like those during the Dalton minimum and the first cycles after

226: the Maunder minimum we cannot use the same intervals for the

227: determination of the rise and decay times.  Fig.~\ref{fig_data} shows

228: the datasets, also indicating the intervals chosen for the determination

229: of growth and decay rates: 30--50 (SN, GSN), 300--600~microhemispheres

230: (SAR), and 900--1100~sfu (SRF). All data have been smoothed by a Gaussian

231: with a FWHM of 1 year.

232: 

233: \clearpage

234: \begin{figure}

235: \epsscale{1.0}

236: \plottwo{f1a.ps}{f1b.ps}

237: \plottwo{f1c.ps}{f1d.ps}

238: \caption{Data sets used in this study: Wolf sunspot number (SN, top

239:  left), group sunspot number (GSN, top right), sunspot area (SAR, bottom

240:  left), and 10.7-cm solar radio flux (SRF, bottom right). SN and GSN are

241:  considered from cycle no. 7 onward (after the Dalton minimum). The

242:  dashed horizontal lines indicate the intervals used for the

243:  determination of the rise and decay rates.}

244: \label{fig_data}

245: \end{figure}

246: 

247: \begin{figure}

248: \epsscale{.9}

249: \plottwo{f2a.ps}{f2b.ps}

250: \plottwo{f2c.ps}{f2d.ps}

251: \plottwo{f2e.ps}{f2f.ps}

252: \plottwo{f2g.ps}{f2h.ps}

253: \caption{Scatter diagrams of rise rates (left panels) and decay rates

254: (right panels) versus cycle amplitude for four sets of solar activity

255: indices: Wolf sunspot number (SN, top row), group sunspot number (GSN,

256: second row), sunspot area (SAR, third row, cycles 21--23 indicated by

257: triangles, two of which nearly coincide on the left-hand panel), and

258: 10.7-cm solar radio flux (bottom row). The corresponding linear

259: correlation coefficients are given. }

260: \label{fig_corr}

261: \end{figure}

262: \clearpage

263: 

264: Scatter diagrams of the resulting rise and decay rates with the cycle

265: amplitudes are shown in Fig.~\ref{fig_corr}, together with the

266: corresponding correlation coefficients. The figure demonstrates that

267: high correlation coefficients ($r=0.83\dots0.89$) between rise rate and

268: cycle amplitude exist for all activity indices. Assuming a Gaussian

269: distribution, the corresponding significance levels are between 95\%

270: (SRF) and over 99\% (SN, GSN, SAR). In particular, the group sunspot

271: numbers and the sunspot areas exhibit an even higher correlation than

272: the Wolf sunspot numbers. Note that cycles 21--23 (indicated by

273: triangles in the third row of Fig.~\ref{fig_corr}), which, owing to

274: their higher {\bf Gnevyshev} maxima in the SAR data, largely destroyed the

275: correlation of the `classical' Waldmeier effect in the analysis of 

276: \citet{Dikpati:etal:2008}, nicely fit the correlation for the

277: growth rate.  On the other hand, the decay rates show no

278: significant correlation with the cycle amplitude, so that we find an

279: {\em amplitude-dependent asymmetry} between the rise rates early in a cycle

280: and the decay rates in the late phase. This asymmetry acts to make

281: the behavior near sunspot minimum rather insensitive to the amplitude of

282: the old cycle but sensitive to that of the new.

283: 

284: We have checked how strongly the correlation coefficients depend on the

285: chosen intervals for the determination of the growth and decay rates, and

286: found them to be rather insensitive. For instance, in the case of the growth

287: (decay) rates for the SN data, we have $r=0.79\,(0.05)$ for the interval

288: $[25,45]$, $r=0.82\,(0.22)$ for $[30,50]$ (the interval chosen for

289: Fig.~\ref{fig_corr}), $r=0.85\,(0.42)$ for $[35,55]$, and

290: $r=0.88\,(0.54)$ for $[40,60]$. Similar results are found for the other

291: datasets.

292: 

293: %We note that we find similar, albeit somewhat less significant, results

294: %when considering separately the sunspot area data for the northern and

295: %southern hemispheres (using the range 225--450 microhemispheres for

296: %determining the growth and decay rates): $r_{\rm N,rise}=0.7$, $r_{\rm

297: %N,decay}=0$, $r_{\rm S,rise}=0.29$, $r_{\rm S,decay}=0$.

298: 

299: 

300: \section{Implications for precursor methods}

301: 

302: The overlapping of cycles with an amplitude-correlated asymmetry leads

303: to a dependence of the minimum epoch on the amplitude of the following

304: cycle: since the initial activity of a large-amplitude cycle increases

305: faster while the decline in the decay phase is not or less

306: amplitude-dependent, the minimum level of activity between overlapping

307: cycles occurs the earlier the higher the amplitude of the following

308: cycle. Such a minimum shift affects the value of any precursor quantity

309: which is defined (directly or indirectly) with regard to the minimum

310: epoch; in the extreme, it may completely explain the predictive power of

311: the precursor. \citet{Cameron:Schuessler:2007} have demonstrated this

312: for the case of a very simply defined precursor, namely, the activity

313: level (such as sunspot number or area), three years before minimum. The

314: earlier the minimum occurs (i.e., the stronger the new cycles is), the

315: higher is the precursor value, simply because it is measured at an

316: earlier time when the activity of the old cycle had not declined as

317: much. This explains why the method can, in principle, even predict

318: cycles with random amplitudes. It is important to note, however, that

319: such methods always require that the minimum epoch {\em is already

320: known}. For a real prediction (as opposed to a postdiction of past

321: cycles) such a method can be applied only some time into the new cycle,

322: when proper averaging can be performed to define the minimum time in

323: view of the strong fluctuations of activity around minimum

324: \citep[cf.][]{Harvey:White:1999b}.

325: 

326: Precursors that are defined (directly or indirectly) in relation to the

327: timing of the minimum are affected by the amplitude-dependent minimum

328: shift, which thus may (partly or fully) explain their correlation with

329: the amplitude of the next cycle (i.e., their predictive power). Among

330: the various precursor methods affected we mention just a few examples:

331: geomagnetic activity before and around solar minimum

332: \citep[e.g.,][]{Ohl:1966, Legrand:Simon:1981, Lantos:Richard:1998,

333: Kane:2007}, sunspot activity around minimum \citep{Hathaway:etal:1999},

334: length of the preceding sunspot cycle \citep{Kane:2008}, skewness of the

335: preceding cycle \citep{Ramaswamy:1977,Lantos:2006}, as well as

336: properties of the decay phase of the preceding cycle

337: \citep{Podladchikova:etal:2008}.

338: 

339: 

340: \section{Implications for the model of Dikpati et al.}

341: 

342: \citet{Cameron:Schuessler:2007} suggested that at least part of the

343: predictive skill of the model of \citet{Dikpati:etal:2006} and

344: \citet{Dikpati:Gilman:2006}, henceforth referred to as the DGT model,

345: may result from the overlap of asymmetric cycles in the sunspot area

346: data from which their source term is derived. We have shown above that

347: the underlying effect, i.e., the amplitude-dependent asymmetry between

348: the growth and decay phases, is present in all global activity datasets,

349: including the sunspot area data used in the DGT model.

350: 

351: The effect of this crosstalk between cycles on the DGT model results

352: from the assumption of a fixed speed and range of latitude drift of the

353: source term in the course of a cycle. The shortening of the cycle length

354: (defined as the time between minima) by the shift of the minimum

355: preceding a strong cycle then leads to larger source amplitudes (sunspot

356: areas) assigned to the low-latitude part of the artificial butterfly

357: diagram used in the DGT model. As a consequence, the magnetic flux

358: crossing the equator as the potentially relevant quantity for the dynamo

359: amplitude of the next cycle is higher, while the opposite is true for a

360: weak following cycle. In this way, the correlation in the data survives

361: the artificial stretching/compressing of cycles carried out in the DGT

362: model in order to obtain a constant cycle period. In fact,

363: \citet{Dikpati:etal:2008b} found that, in their model, the flux crossing

364: the equator during a cycle is significantly correlated ($r=0.76$) with

365: the amplitude of the next cycle. This is in accordance with Fig. 8a in

366: \citet{Dikpati:Gilman:2006}, which demonstrates that the memory in their

367: model does not extend much longer than one cycle, but it is in contrast

368: to the claim of \citet{Dikpati:etal:2008b} that it takes 17-21 years for

369: the poloidal surface field to reach the tachocline and generate the new

370: toroidal field, so that 2-3 preceding cycles contribute to the toroidal

371: flux of a cycle. {\bf A stronger diffusive coupling between surface and

372: tachocline than indicated by the explicit value of their magnetic

373: diffusivity (for example, due to numerical diffusion) could possibly

374: explain this discrepancy \citep[cf.][]{Yeates:etal:2008}.}

375: 

376: By disregarding the observed sunspot latitudes in favor of an imposed

377: fixed latitude drift, the DGT model becomes affected by the

378: amplitude-dependent minimum shift. The extent to which this contributes

379: to the predictive skill of the model could be easily tested by replacing

380: the actual sunspot area data as input by a synthetic dataset with random

381: cycle amplitudes (without any memory), but keeping the correlation

382: resulting from the minimum shift of overlapping cycles.

383: \citet{Cameron:Schuessler:2007} used this approach to demonstrate the

384: effect on precursor methods. The amount of predictive skill shown by the

385: DGT model for such an input will directly indicate how strongly this

386: correlation affects the prediction.

387: 

388: \section{Discussion}

389: 

390: The shift of the solar minimum epoch due to the

391: amplitude-dependent asymmetry between the growth and decay phases of

392: solar cycles can explain the rather high correlation coefficients

393: between various precursor quantities and the strength of the subsequent

394: cycle \citep[e.g.,][]{Hathaway:etal:1999}. In this connection, we should

395: note that even a statistically highly significant correlation does not

396: automatically imply a high skill of the precursor for the actual {\em

397: prediction} of a future cycle:

398: 

399: \begin{enumerate}

400: \item A typical value of, say, $r=0.8$ for a `good' precursor leaves

401:   about one third of the variance in amplitude unexplained by the

402:   correlation, so that the prediction is prone to considerable

403:   statistical uncertainty. Even the value of $r=0.97$ shown by the method

404:   of \citet[][see Hathaway et al. 1999]{Thompson:1993} did not prevent

405:   an utterly inaccurate prediction for cycle 23 (30\% too high). 

406: \item The data used to determine the correlations in most cases cover

407:   only part of the known range of variability in the solar

408:   cycle. Periods such as the Maunder and Dalton minima indicate that

409:   nonlinear and stochastic effects may severely limit the predictability

410:   since it is unclear how strongly these effects are represented in the

411:   observed surface flux \citep{Bushby:Tobias:2007}. `Tuning' of free

412:   parameters in prediction schemes can lead to seemingly high

413:   correlations for past cycles while the predictive skill for future

414:   cycles in fact is much lower.

415: \end{enumerate}

416: 

417: The dependence of many precursor methods on knowing the minimum epoch

418: further limits their practical usefulness: the large relative

419: fluctuations of the activity indices around sunspot minimum make it

420: necessary to average the data (over 1 year, for instance) in order to

421: obtain a proper definition of the minimum epoch. Consequently, a

422: sensible prediction can only be made one or two years after

423: the minimum has passed, so that there is no big advantage compared 

424: to methods that use the information provided by the early phases of the

425: new cycle itself, like the steepness of the activity rise

426: \citep{Waldmeier:1936, Elling:Schwentek:1992}.

427: 

428: The relation between the shift of the minimum time and the amplitude

429: of the next cycle due to cycle overlapping is strongest for a dynamo where

430: the time interval between the onsets of activity of subsequent cycles is 

431: constant. Fluctuations in the timing of the onset of activity introduces 

432: noise into the relation between minimum shift and activity and thus reduces 

433: the predictive skill of precursor methods.  Quantifying such fluctuations

434: in the solar cycle data is difficult because the length of the available 

435: datasets is not sufficient to reliably determine the phase fluctuations and 

436: drifts \citep{Gough:1978,Hoyng:1996} and to decide on the existence or 

437: otherwise of a solar `clock' \citep[cf.][]{Dicke:1978, Dicke:1988}. 

438: However the expected shifts due to the overlapping of the cycles and 

439: the amplitude dependent asymmetry produce variations in the cycle length 

440: (as measured from minimum to minimum) which are consistent with those 

441: observed since the Dalton minimum. This means the effect we have outlined  

442: is relevant.  In passing, we also note that in the case of 

443: advection-dominated dynamo models the cycle period is determined by the 

444: meridional flow, {\bf which supports phase stability 

445: \citep{Charbonneau:Dikpati:2000, Charbonneau:2005}.}

446: 

447: 

448: For our understanding of the solar dynamo, it would certainly be useful

449: to clarify to which extent the skill of some prediction methods results

450: 1) from correlations in the input data (such as discussed here), 2) from

451: capturing early high-latitude manifestations of an extended new cycle,

452: and 3) from capturing mechanisms that connect properties of the old

453: cycle to the strength of the new cycle (such as a dynamo model with a

454: memory of at least one cycle). It is reasonable to suppose that there

455: is some memory in the solar cycle i.e., that the solar cycle amplitudes do not

456: constitute a purely random sequence. Potentially important information

457: could be gleaned from identifying a quantity that unequivocally

458: represents this memory and is not `contaminated' by early information

459: leaking in from the new cycle. Clarifying whether the memory extends

460: over more than one cycle would help to decide between

461: diffusion-dominated and advection-dominated dynamo models

462: \citep{Yeates:etal:2008}.

463: 

464: \section{Conclusions}

465: 

466: We have confirmed a highly significant correlation between the growth

467: rate of activity during the early phase of a solar cycle and its maximum

468: amplitude for all global activity indices, i.e., Wolf and group sunspot

469: numbers, total sunspot area, and 10.7-cm radio flux. On the other hand,

470: there is no significant correlation between the decay rate in the late

471: cycle phase and the cycle amplitude.

472: 

473: Owing to the overlapping of individual cycles, this asymmetry leads to

474: an amplitude-dependent shift of the minimum epoch, thus explaining

475: (fully or partly) the predictive power of precursor methods which

476: (directly or indirectly) use the timing of the activity minimum as a

477: pivotal point. The resulting correlation in the sunspot area data

478: probably also affects the predictions with the dynamo-based model of

479: \citet{Dikpati:Gilman:2006}.

480: 

481: For our understanding of the origin of the solar magnetic field, it is

482: important to disentangle the effects of `real' physical precursors,

483: i.e., properties of the old cycle directly affecting the flux generation

484: for the next cycle or early high-latitude manifestations of the new

485: cycle, from apparent precursors, which derive their predictive power

486: from the the amplitude-dependent shift of the minimum epoch. 

487: 

488: %%\acknowledgments 

489: 

490: 

491: \bibliographystyle{apj}

492: %\bibliography{apj-jour,papref}

493: %\bibliography{joushort,papref}

494: %\bibliography{x.bbl}

495: 

496: 

497: %\ifnum 2<1

498: \begin{thebibliography}{46}

499: 

500: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi

501: 

502: \bibitem[{{Altrock} {et~al.}(2008){Altrock}, {Howe}, \&

503:   {Ulrich}}]{Altrock:etal:2008}

504: {Altrock}, R., {Howe}, R., \& {Ulrich}, R. 2008, in ASP Conf. Ser., Vol. 383,

505:   Subsurface and Atmospheric Influences on Solar Activity, ed. R.~{Howe}, R.~W.

506:   {Komm}, K.~S. {Balasubramaniam}, \& G.~J.~D. {Petrie}, 335

507: 

508: \bibitem[{{Altrock}(1997)}]{Altrock:1997}

509: {Altrock}, R.~C. 1997, \solphys, 170, 411

510: 

511: \bibitem[{{Balmaceda} {et~al.}(2005){Balmaceda}, {Solanki}, \&

512:   {Krivova}}]{Balmaceda:etal:2005}

513: {Balmaceda}, L., {Solanki}, S.~K., \& {Krivova}, N. 2005, Memorie della Societa

514:   Astronomica Italiana, 76, 929

515: 

516: \bibitem[{{Bushby} \& {Tobias}(2007)}]{Bushby:Tobias:2007}

517: {Bushby}, P.~J. \& {Tobias}, S.~M. 2007, \apj, 661, 1289

518: 

519: \bibitem[{{Cameron} \& {Sch{\"u}ssler}(2007)}]{Cameron:Schuessler:2007}

520: {Cameron}, R. \& {Sch{\"u}ssler}, M. 2007, \apj, 659, 801

521: 

522: \bibitem[{{Charbonneau}(2005)}]{Charbonneau:2005}

523: {Charbonneau}, P. 2005, Living Reviews in Solar Physics, LRSP-2005-2,

524:   http://solarphysics.livingreviews.org

525: 

526: \bibitem[{{Charbonneau} \& {Dikpati}(2000)}]{Charbonneau:Dikpati:2000}

527: {Charbonneau}, P. \& {Dikpati}, M. 2000, \apj, 543, 1027

528: 

529: \bibitem[{{Choudhuri} {et~al.}(2007){Choudhuri}, {Chatterjee}, \&

530:   {Jiang}}]{Choudhuri:etal:2007}

531: {Choudhuri}, A.~R., {Chatterjee}, P., \& {Jiang}, J. 2007, Phys. Rev. Lett.,

532:   98, 131103

533: 

534: \bibitem[{{Dicke}(1978)}]{Dicke:1978}

535: {Dicke}, R.~H. 1978, \nat, 276, 676

536: 

537: \bibitem[{{Dicke}(1988)}]{Dicke:1988}

538: ---. 1988, \solphys, 115, 171

539: 

540: \bibitem[{{Dikpati} {et~al.}(2006){Dikpati}, {de Toma}, \&

541:   {Gilman}}]{Dikpati:etal:2006}

542: {Dikpati}, M., {de Toma}, G., \& {Gilman}, P.~A. 2006, Geophys. Res. Lett., 33,

543:   5102

544: 

545: \bibitem[{{Dikpati} {et~al.}(2008{\natexlab{a}}){Dikpati}, {de Toma}, \&

546:   {Gilman}}]{Dikpati:etal:2008b}

547: ---. 2008{\natexlab{a}}, \apj, 675, 920

548: 

549: \bibitem[{{Dikpati} \& {Gilman}(2006)}]{Dikpati:Gilman:2006}

550: {Dikpati}, M. \& {Gilman}, P.~A. 2006, Astrophys. J, 649, 498

551: 

552: \bibitem[{{Dikpati} {et~al.}(2008{\natexlab{b}}){Dikpati}, {Gilman}, \& {de

553:   Toma}}]{Dikpati:etal:2008}

554: {Dikpati}, M., {Gilman}, P.~A., \& {de Toma}, G. 2008{\natexlab{b}}, \apjl,

555:   673, L99

556: 

557: \bibitem[{{Elling} \& {Schwentek}(1992)}]{Elling:Schwentek:1992}

558: {Elling}, W. \& {Schwentek}, H. 1992, \solphys, 137, 155

559: 

560: \bibitem[{{Gnevyshev}(1967)}]{Gnevyshev:1967}

561: {Gnevyshev}, M.~N. 1967, \solphys, 1, 107

562: 

563: \bibitem[{{Gough}(1978)}]{Gough:1978}

564: {Gough}, D. 1978, in Pleins Feux sur la Physique Solaire, ed. S.~{Dumont} \&

565:   J.~{Roesch} (CNRS, Paris), 81

566: 

567: \bibitem[{{Harvey}(1992)}]{Harvey:1992a}

568: {Harvey}, K.~L. 1992, in The Solar Cycle, ed. K.~L. Harvey (San Francisco:

569:   Astronomical Society of the Pacific, ASP Conf. Series Vol. 27), 335

570: 

571: \bibitem[{{Harvey}(1994)}]{Harvey:1994a}

572: {Harvey}, K.~L. 1994, in Solar Surface Magnetism, ed. R.~J. {Rutten} \& C.~J.

573:   {Schrijver} (Dordrecht: Kluwer), 347

574: 

575: \bibitem[{{Harvey} \& {White}(1999)}]{Harvey:White:1999b}

576: {Harvey}, K.~L. \& {White}, O.~R. 1999, \jgr, 104, 19759

577: 

578: \bibitem[{{Hathaway} {et~al.}(1994){Hathaway}, {Wilson}, \&

579:   {Reichmann}}]{Hathaway:etal:1994}

580: {Hathaway}, D.~H., {Wilson}, R.~M., \& {Reichmann}, E.~J. 1994, Sol. Phys.,

581:   151, 177

582: 

583: \bibitem[{{Hathaway} {et~al.}(1999){Hathaway}, {Wilson}, \&

584:   {Reichmann}}]{Hathaway:etal:1999}

585: ---. 1999, J. Geophys. Res., 104, 22375

586: 

587: \bibitem[{{Hathaway} {et~al.}(2002){Hathaway}, {Wilson}, \&

588:   {Reichmann}}]{Hathaway:etal:2002}

589: ---. 2002, Sol. Phys., 211, 357

590: 

591: \bibitem[{{Howe} {et~al.}(2006){Howe}, {Komm}, {Hill}, {Ulrich}, {Haber},

592:   {Hindman}, {Schou}, \& {Thompson}}]{Howe:etal:2006}

593: {Howe}, R., {Komm}, R., {Hill}, F., {Ulrich}, R., {Haber}, D.~A., {Hindman},

594:   B.~W., {Schou}, J., \& {Thompson}, M.~J. 2006, \solphys, 235, 1

595: 

596: \bibitem[{{Hoyng}(1996)}]{Hoyng:1996}

597: {Hoyng}, P. 1996, Sol. Phys., 169, 253

598: 

599: \bibitem[{{Hoyt} \& {Schatten}(1998)}]{Hoyt:Schatten:1998}

600: {Hoyt}, D.~V. \& {Schatten}, K.~H. 1998, Sol. Phys., 179, 189

601: 

602: \bibitem[{{Kane}(2007)}]{Kane:2007}

603: {Kane}, R.~P. 2007, \solphys, 243, 205

604: 

605: \bibitem[{{Kane}(2008)}]{Kane:2008}

606: ---. 2008, \solphys, 248, 203

607: 

608: \bibitem[{{Lantos}(2000)}]{Lantos:2000}

609: {Lantos}, P. 2000, \solphys, 196, 221

610: 

611: \bibitem[{{Lantos}(2006)}]{Lantos:2006}

612: ---. 2006, Sol. Phys., 236, 199

613: 

614: \bibitem[{{Lantos} \& {Richard}(1998)}]{Lantos:Richard:1998}

615: {Lantos}, P. \& {Richard}, O. 1998, Sol. Phys., 182, 231

616: 

617: \bibitem[{{Legrand} \& {Simon}(1981)}]{Legrand:Simon:1981}

618: {Legrand}, J.~P. \& {Simon}, P.~A. 1981, Sol. Phys., 70, 173

619: 

620: \bibitem[{{Li}(1999)}]{Li:1999}

621: {Li}, K. 1999, \aap, 345, 1006

622: 

623: \bibitem[{{Ohl}(1966)}]{Ohl:1966}

624: {Ohl}, A.~I. 1966, Soln. Dann., 12, 84

625: 

626: \bibitem[{{Podladchikova} {et~al.}(2008){Podladchikova}, {Lefebvre}, \& {van

627:   der Linden}}]{Podladchikova:etal:2008}

628: {Podladchikova}, T., {Lefebvre}, B., \& {van der Linden}, R. 2008, Journal of

629:   Atmospheric and Terrestrial Physics, 70, 277

630: 

631: \bibitem[{{Ramaswamy}(1977)}]{Ramaswamy:1977}

632: {Ramaswamy}, G. 1977, \nat, 265, 713

633: 

634: \bibitem[{{Schatten}(2003)}]{Schatten:2003}

635: {Schatten}, K.~H. 2003, Adv. Space Res., 32, 451

636: 

637: \bibitem[{{Schatten} {et~al.}(1978){Schatten}, {Scherrer}, {Svalgaard}, \&

638:   {Wilcox}}]{Schatten:etal:1978}

639: {Schatten}, K.~H., {Scherrer}, P.~H., {Svalgaard}, L., \& {Wilcox}, J.~M. 1978,

640:   Geophys. Res. Lett., 5, 411

641: 

642: \bibitem[{{Solanki} {et~al.}(2002){Solanki}, {Krivova}, {Sch{\" u}ssler}, \&

643:   {Fligge}}]{Solanki:etal:2002b}

644: {Solanki}, S.~K., {Krivova}, N.~A., {Sch{\" u}ssler}, M., \& {Fligge}, M. 2002,

645:   \aap, 396, 1029

646: 

647: \bibitem[{{Thompson}(1993)}]{Thompson:1993}

648: {Thompson}, R.~J. 1993, Sol. Phys., 148, 383

649: 

650: \bibitem[{{Waldmeier}(1935)}]{Waldmeier:1935}

651: {Waldmeier}, M. 1935, Mitt. Eidgen. Sternw. Z{\"u}rich, 14, 105

652: 

653: \bibitem[{{Waldmeier}(1936)}]{Waldmeier:1936}

654: ---. 1936, Astron. Nachr., 259, 267

655: 

656: \bibitem[{{Waldmeier}(1955)}]{Waldmeier:1955}

657: ---. 1955, {Ergebnisse und Probleme der Sonnenforschung.} (Leipzig, Geest {\&}

658:   Portig)

659: 

660: \bibitem[{{Wilson}(1994)}]{Wilson:1994}

661: {Wilson}, P.~R. 1994, {Solar and stellar activity cycles} (Cambridge

662:   Astrophysics Series, Cambridge University Press)

663: 

664: \bibitem[{{Wilson} {et~al.}(1988){Wilson}, {Altrock}, {Harvey}, {Martin}, \&

665:   {Snodgrass}}]{Wilson:etal:1988}

666: {Wilson}, P.~R., {Altrock}, R.~C., {Harvey}, K.~L., {Martin}, S.~F., \&

667:   {Snodgrass}, H.~B. 1988, \nat, 333, 748

668: 

669: \bibitem[{{Yeates} {et~al.}(2008){Yeates}, {Nandy}, \&

670:   {Mackay}}]{Yeates:etal:2008}

671: {Yeates}, A.~R., {Nandy}, D., \& {Mackay}, D.~H. 2008, \apj, 673, 544

672: 

673: 

674: \end{thebibliography}

675: %\fi

676: 

677: \clearpage

678: 

679: 

680: \end{document}

681: 

682: