1: \documentclass[12pt,preprint]{aastex}
2: \usepackage{graphicx}
3: \usepackage{amsmath,amssymb}
4: \usepackage{natbib}
5: \usepackage{multirow}
6:
7:
8: \shorttitle{Advective vs Diffusive Solar Convection Zones}
9: \shortauthors{Yeates et al.}
10:
11: \newcommand{\pdiff}[2]{\frac{\partial #1}{\partial #2}}
12: \newcommand{\ppdiff}[2]{\frac{\partial^2 #1}{\partial #2^2}}
13: \newcommand{\pddiff}[3]{\frac{\partial^2 #1}{\partial #2 \partial #3}}
14: \newcommand{\pdlin}[2]{\partial #1/\partial #2}
15: \renewcommand{\eqref}[1]{(\ref{eqn:#1})}
16: \newcommand{\mpsec}{\,\textrm{m}\,\textrm{s}^{-1}}
17: \newcommand{\cmsqpsec}{\,\textrm{cm}^2\,\textrm{s}^{-1}}
18: \newcommand{\permet}{\,\textrm{m}^{-1}}
19: \newcommand{\met}{\,\textrm{m}}
20: \newcommand{\erf}{\,\textrm{erf}}
21: \newcommand{\bpol}{\mathbf{B}_{\textrm{p}}}
22: \newcommand{\mx}{\,\textrm{Mx}}
23:
24:
25: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
26:
27: \begin{document}
28:
29: \title{Exploring the Physical Basis of Solar Cycle Predictions:\\
30: Flux Transport Dynamics and
31: Persistence of Memory in Advection versus Diffusion Dominated Solar Convection Zones}
32:
33: \author{Anthony~R.~Yeates}
34: \affil{School of Mathematics and Statistics, University of St Andrews, St Andrews, KY16 9SS, UK}
35: \email{anthony@mcs.st-and.ac.uk}
36: \author{Dibyendu Nandy}
37: \affil{Department of Physics, Montana State University, Bozeman, MT 59717, USA}
38: \email{nandi@mithra.physics.montana.edu}
39: \and
40: \author{Duncan~H.~Mackay}
41: \affil{School of Mathematics and Statistics, University of St Andrews, St Andrews, KY16 9SS, UK}
42: \email{duncan@mcs.st-and.ac.uk}
43:
44: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
45:
46: \begin{abstract}
47:
48: The predictability, or lack thereof, of the solar cycle is governed by numerous
49: separate physical processes that act in unison in the interior of the Sun. Magnetic
50: flux transport and the finite time delay it introduces, specifically in the so-called
51: Babcock-Leighton models of the solar cycle with spatially segregated source regions for
52: the $\alpha$ and $\Omega$-effects, play a crucial rule in this predictability. Through
53: dynamo simulations with such a model, we study the physical basis of solar cycle
54: predictions by examining two contrasting regimes, one dominated by diffusive magnetic
55: flux transport in the solar convection zone, the other dominated by advective flux
56: transport by meridional circulation. Our analysis shows that diffusion plays an
57: important role in flux transport, even when the solar cycle period is governed by the
58: meridional flow speed. We further examine the persistence of memory of past cycles in
59: the advection and diffusion dominated regimes through stochastically forced dynamo
60: simulations. We find that in the advection-dominated regime, this memory persists for
61: up to three cycles, whereas in the diffusion-dominated regime, this memory persists for
62: mainly one cycle. This indicates that solar cycle predictions based on these two
63: different regimes would have to rely on fundamentally different inputs -- which may be
64: the cause of conflicting predictions. Our simulations also show that the observed solar
65: cycle amplitude-period relationship arises more naturally in the diffusion dominated
66: regime, thereby supporting those dynamo models in which diffusive flux transport plays
67: a dominant role in the solar convection zone.
68: \end{abstract}
69:
70: \keywords{Sun: activity --- Sun: interior --- Sun: magnetic fields}
71:
72: \section{Introduction}
73: %==============================================================================
74:
75: Direct observations of sunspot numbers over 400 years, as well as proxy data for much
76: longer timescales \citep{beer2000}, show that both the amplitude and the duration of
77: the solar magnetic cycle vary from one cycle to the next. The importance of this
78: phenomenon lies in the contribution of varying levels of solar activity to long-term
79: climate change, and to short-term space weather \citep{nandy2007}. While there is now a
80: concensus that the Sun's magnetic field is generated by a hydromagnetic dynamo
81: \citep{ossendrijver2003,charbonneau2005}, the origin of fluctuations in the basic cycle
82: is yet to be conclusively determined. Several different mechanisms have been proposed,
83: including nonlinear effects \citep{tobias1997, beer1998, knobloch1998, kueker1999,
84: wilmotsmith2005}, stochastic forcing
85: \citep{choudhuri1992,hoyng1993,ossendrijver1996,charbonneau2000,mininni2002}, and
86: time-delay dynamics \citep{yoshimura1978,durney2000,charbonneau2005b,wilmotsmith2006}.
87: A coupled, equally important, but ill-understood issue is how the memory of these
88: fluctuations, whatever may be its origin, carries over from one cycle to another
89: mediated via flux transport processes within the solar convection zone (SCZ). A
90: unified understanding of all these disparate processes lays the physical foundation for
91: the predictability (or lack-thereof) of future solar activity. These considerations
92: motivate the current study.
93:
94: The main flux transport processes in the SCZ involve magnetic buoyancy (timescale on
95: the order of months), meridional circulation, diffusion and downward flux-pumping
96: (timescales relatively larger). Because magnetic buoyancy, i.e., the buoyant rise of
97: magnetic flux tubes, acts on timescales much shorter than the solar cycle timescale,
98: the fluctuations that it produces are also short-lived in comparison. Our focus here is
99: on longer-term fluctuations, on the order of the solar cycle period, that may lead
100: to predictive capabilities.
101:
102: Through an analysis of observational data, \citet{hathaway2003} have shown that the
103: solar cycle amplitude and duration are correlated with the equatorward drift velocity
104: of the sunspot belts during the cycle. They associate this drift velocity with the deep
105: meridional counterflow that must exist to balance the poleward flows that are observed
106: at the surface (\citealp{hathaway1996}, \citeyear{hathaway2005}; \citealp{miesch2005}).
107: The results show a significant negative correlation between the drift velocity and the
108: cycle duration, so that the drift is faster in shorter cycles, consistent with the
109: interpretation of meridional circulation as the timekeeper of the solar cycle
110: (\citealp{nandy2004}; but see also \citealp{schuessler2004}). In addition
111: \citet{hathaway2003} identified positive correlations between the drift velocity of
112: cycle $n$ and the amplitudes of both cycles $n$ and $n+2$. While the two-cycle time lag
113: was a new result, the positive correlation between circulation speed and amplitude of
114: the same cycle is supported by several earlier studies. In their surface flux transport
115: model, \citet{wang2002b} needed a varying meridional flow, faster in higher-amplitude
116: cycles, to sustain regular reversals in the Sun's polar field. They cited observational
117: evidence from polar faculae counts \citep{sheeley1991}, which peaked early for two of
118: the stronger cycles, coinciding with poleward surges of magnetic flux. Furthermore,
119: observations show a statistically-significant negative correlation between peak sunspot
120: number and the duration of cycles 1 to 22 (Figure 1c of \citealp{charbonneau2000}; see
121: also \citealp{solanki2002}). Such a negative correlation between cycle amplitude and
122: duration is also found in the models of \citet{hoyng1993} and \citet{charbonneau2000}.
123: Taken with the inverse relation between cycle duration and circulation speed, this is
124: again suggestive of a positive correlation between circulation speed and cycle
125: amplitude.
126:
127: Meridional circulation plays an important role in a certain class of theoretical solar
128: cycle models often referred to as ``flux-transport'', ``advection-dominated,'' or
129: ``circulation-dominated'' dynamo models (see, e.g., the review by Nandy 2004). Such
130: models have gained popularity in recent years owing to their success in reproducing
131: various observed features of the solar cycle
132: \citep{choudhuri1995,durney1995,dikpati1999,kueker2001,bonanno2002,
133: nandy2001,nandy2002,chatterjee2004}. In these models, a single-cell meridional
134: circulation in each hemisphere (which is observed at the solar surface) is invoked to
135: transport poloidal field, first poleward at near-surface layers and then down to the
136: tachocline where toroidal field is generated. Subsequently, the return flow in the
137: circulation advects this toroidal field belt equatorward through a region at the base
138: of the SCZ which is characterized by low diffusivity. From this deep toroidal field
139: belt, destabilized flux tubes rise to the surface due to magnetic buoyancy, producing
140: sunspots \citep{parker1955}. We may point out here that the name
141: ``flux-transport dynamo'' is somewhat inappropriate to classify a circulation or
142: advection-dominated dynamo (where the diffusion timescale is much larger than the
143: circulation timescale throughout the dynamo domain). Our results indicate that
144: diffusive flux-transport in the SCZ could play a dominant role in dynamos even when the
145: cycle period is governed by meridional circulation speed, pointing out that
146: flux-transport is a shared process. So, henceforth, by ``flux-transport'' dynamo, we
147: imply a dynamo where the transport of magnetic field is shared by magnetic buoyancy,
148: meridional circulation, and diffusion.
149:
150: Flux-transport dynamos offer the possibility of prediction because of their inherent
151: memory. This arises specifically when the dynamo source regions for poloidal field
152: production (the traditional $\alpha$-effect) and toroidal field generation (the
153: $\Omega$-effect) are spatially segregated. A brief discussion on important timescales
154: (we identify three significant ones) in the dynamo process is merited here. The first
155: is governed by the buoyant rise of toroidal flux tubes from the $\Omega$-effect layer
156: to the $\alpha$-effect layer to generate the poloidal field; since this is a fast
157: process on the order of months, no significant memory is introduced here. The second
158: involves the transport of poloidal field back into the $\Omega$-effect layer (either by
159: circulation or diffusion). This could be a slow process where significant memory is
160: introduced which is dominated by the fastest of the competing processes (advection
161: versus diffusion). The third timescale relates to the slow equatorward transport of the
162: toroidal field belt through the base of the SCZ, which sets the period of the sunspot
163: cycle. In this class of dynamo models, with meridional circulation and low
164: diffusivities in the tachocline (at the base of the SCZ), the third timescale is almost
165: invariably determined by the circulation speed. It is the second timescale above, with
166: competing effects of diffusive flux transport and advective flux transport, that becomes
167: important in the context of the persistence of memory. In the advection-dominated,
168: stochastically fluctuating model of \citet{charbonneau2000}, this second timescale
169: (governed by advection of poloidal field due to meridional circulation) was about 17
170: years, so that the polar field at the end of cycle $n$ correlated strongest with the
171: toroidal field of cycle $n+2$ rather than that of cycle $n+1$. The length of memory of
172: any particular flux-transport dynamo model is unfortunately dependent on the internal
173: meridional flow profile, and on other chosen properties of the convection zone which
174: are not yet well-determined observationally. A particular problem is the strength of
175: diffusivity in the convection zone, which strongly affects the mode of operation of the
176: dynamo.
177:
178: Even if one assumes that these flux-transport dynamos capture enough of the realistic
179: physics of the SCZ to make predictions of future solar activity, these predictions are
180: critically dependent on the relative role of diffusion and advection in the SCZ.
181: \citet{dikpati2006b}, in their highly {\it advection-dominated} model, show that bands
182: of latitudinal field from three previous cycles remain ``lined up in the meridional
183: circulation conveyor belt''. They suggest that poloidal fields from cycles $n-3$,
184: $n-2$, and $n-1$ combine to produce the toroidal field of cycle $n$. Based on an
185: assumed proxy for the solar poloidal fields (sunspot area), this leads them to predict
186: that Cycle 24 will be about $50\%$ stronger than Cycle 23 \citep{dikpati2006}. In stark
187: contrast, \citet{choudhuri2007}, using a flux-transport dynamo model with {\it
188: diffusion-dominated} SCZ, and using as inputs the observed strength of the solar dipole
189: moment (as a proxy for the poloidal field), predict that Cycle 24 will be about $35\%$
190: \emph{weaker} than Cycle 23. \citet{choudhuri2007} argue that the main contribution to
191: the toroidal field of cycle $n$, comes only from the polar field of cycle $n-1$ (see
192: also \citealp{jiang2007} for further details of this model).
193:
194: The conflicting predictions from these two solar dynamo models presumably result from
195: the difference in the memory (i.e., survival) of past cycle fields in these models and
196: could be to some extent influenced by the different inputs they use as proxies for the
197: solar poloidal field. We also hypothesize that stronger diffusion in the
198: \citet{choudhuri2007} model destroys polar field faster, and that flux transport by
199: diffusion across the SCZ in this model short-circuits the meridional circulation
200: conveyor belt, thereby shortening the memory of previous cycles. We perform a detailed
201: analysis to test these ideas. To begin with, we consider a wider parameter space in the
202: present paper, where we study the effect of varying meridional circulation speed and
203: SCZ diffusivity on the amplitude and period of the solar cycle. In these simulations,
204: we keep all other parameters the same, allowing a direct comparison between advection-dominated and diffusion-dominated SCZ regimes -- which has previously been clouded by
205: other differences between models. Then we introduce stochastic fluctuations in the
206: model $\alpha$-effect to self-consistently generate cycle-amplitude variations -- as a
207: completely theoretical construct towards studying cycle-to-cycle variations, in
208: contrast to using diverse observed proxies for time-varying poloidal fields.
209: Subsequently, we perform a comparative analysis of the persistence of memory in this
210: stochastically forced dynamo model in both the advective and diffusive flux-transport
211: dominated regimes. Therefore, in spirit, this paper deals with the underlying physics
212: of solar cycle predictability, and is not concerned with making a prediction itself.
213: The layout of this paper is as follows. The main features of the model are summarised
214: in Section \ref{sec:model}, and the results of the parameter-space study are presented
215: in Section \ref{sec:results}. These results are interpreted in Section
216: \ref{sec:regimes}. In Section \ref{sec:timedelays} we analyze the persistence of memory
217: in the advection versus diffusion dominated regimes. We conclude in Section
218: \ref{sec:conclusion} with a discussion on the relevance of this work in the context of
219: developing predictive capabilities for the solar activity cycle.
220:
221: \section{The Model} \label{sec:model}
222: %==============================================================================
223:
224: We use the solar dynamo code \emph{Surya}, which has been studied extensively in
225: different contexts (e.g. \citealp{nandy2002}, \citealp{chatterjee2004},
226: \citealp{chatterjee2006}), and is made available to the public on request. The major
227: ingredients of the code include an analytic fit to the helioseismically-determined
228: differential rotation profile, a single-cell meridional circulation in the SCZ,
229: different diffusivities for the toroidal and poloidal fields, a buoyancy algorithm to
230: model radial transport of magnetic flux, and a Babcock-Leighton (BL;
231: \citealp{babcock1961}, \citealp{leighton1969}) type $\alpha$-effect localized near the surface layer (signifying the
232: generation of poloidal field due to the evolution of
233: tilted bipolar sunspot pairs under surface flux transport). The code solves the
234: kinematic mean-field dynamo equations for an axisymmetric magnetic field, which may be
235: expressed in spherical coordinates $(r,\theta,\phi)$ as
236: \begin{equation}
237: \mathbf{B} = B(r,\theta)\mathbf{e}_{\phi} + \bpol,
238: \end{equation}
239: where $B(r,\theta)$ and $\bpol=\nabla\times\left [ A(r,\theta)\mathbf{e}_{\phi} \right ]$
240: correspond to the toroidal and poloidal components respectively. The mean-field MHD
241: induction equation (see e.g. \citealp{moffatt1978}) then leads to the following
242: standard equations for the $\alpha$-$\Omega$ dynamo problem:
243: \begin{align}
244: \pdiff{A}{t} + \frac{1}{s}\left( \mathbf{v}\cdot\nabla \right)\left(sA\right) &= \eta_p\left(\nabla^2 - \frac{1}{s^2}\right)A + \alpha B, \label{eqn:Aevol}\\
245: \pdiff{B}{t} + \frac{1}{r}\left( \pdiff{}{r}\left( r v_r B \right) + \pdiff{}{\theta}\left( v_\theta B\right) \right) &= \eta_t \left(\nabla^2 - \frac{1}{s^2} \right)B \nonumber \\
246: &+ s\left( \bpol \cdot \nabla \right) \Omega + \frac{1}{r}\frac{d\eta_t}{dr}\pdiff{}{r}\left( rB \right). \label{eqn:Bevol}
247: \end{align}
248: Here $s=r\sin\theta$, and we specify the meridional flow $\mathbf{v}$, the internal
249: angular velocity $\Omega$, the diffusivities $\eta_p$ and $\eta_t$, and the coefficient
250: $\alpha$ for the BL $\alpha$-effect which describes the generation of poloidal field at the solar
251: surface from the decay of bipolar sunspots. Note that although for modelling purposes
252: the BL $\alpha$-effect is mathematically similar to the traditional mean-field
253: $\alpha$-effect due to small-scale helical turbulence, the former is fundamentally
254: different. The BL $\alpha$-effect acts on much larger spatial (on the order of active
255: regions or greater) and temporal (surface flux transport) scales, and is quenched at
256: much stronger field strengths ($10^5 \mathrm{G}$). The profiles of $\Omega$ and $\alpha$ were
257: described in \citet{chatterjee2004} and will not be repeated here. We will, however,
258: describe the meridional circulation and diffusivity profiles in more detail.
259:
260: \subsection{Meridional Circulation}
261: %------------------------------------------------------------------------------
262:
263: The meridional circulation is defined in terms of a streamfunction $\psi(r,\theta)$,
264: giving the velocity by
265: \begin{equation}
266: \rho\mathbf{v}=\nabla\times\left [\psi(r,\theta)\mathbf{e}_\phi\right ],
267: \end{equation}
268: where we assume the density stratification
269: \begin{equation}
270: \rho=C\left(R_\odot/r - 0.95\right)^{3/2}.
271: \end{equation}
272: The streamfunction is given by
273: \begin{align}
274: \psi r\sin\theta &= \psi_0\left(r-R_p\right)\sin\left(\frac{\pi(r-R_p)}{R_\odot-R_p}\right) \nonumber\\
275: & \qquad \times
276: \left\{1-\textrm{e}^{-\beta_1\theta^\epsilon}\right\}\left\{1-\textrm{e}^{\beta_2(\theta-\pi/2)}\right\}\textrm{e}^{-\left((r-r_0)/\Gamma\right)^2},
277: \end{align}
278: where $\beta_1=1.5\times 10^{-8}\permet$, $\beta_2=1.8\times
279: 10^{-8}\permet$, $\epsilon=2.0000001$, $\Gamma=3.47\times 10^{8}\met$, and
280: $r_0=(R_\odot-R_b)/4$. Here $R_\odot=6.96\times 10^{8}\met$ is the radius of the Sun,
281: $R_b=0.55R_\odot$ is the bottom of the simulation domain, and $R_p=0.61R_\odot$ is the
282: penetration depth of the meridional circulation. We combine the arbitrary constants $C$
283: and $\psi_0$ in the parameter $v_0=-\psi_0/(0.95C)$ which gives, approximately, the
284: flow speed near the surface at mid-latitudes. It is this parameter $v_0$ which we vary
285: to change the circulation speed in this study.
286:
287: The circulation profile is illustrated in Figure \ref{fig:merid} for $v_0=25\mpsec$.
288: The dots are plotted at yearly intervals for particles moving along the streamlines
289: shown.
290: % fig:merid
291:
292:
293: \subsection{Diffusion}
294: %------------------------------------------------------------------------------
295:
296: We use different diffusivities for the toroidal and poloidal fields, defined as
297: follows:
298: \begin{align}
299: \eta_t(r) &= \eta_{RZ} + \frac{\eta_1-\eta_{RZ}}{2}\left[1+\erf\left(\frac{r-r'_{BCZ}}{d_t}\right)\right]\nonumber\\
300: & \qquad\qquad\qquad +\frac{\eta_0-\eta_1}{2}\left[1+\erf\left(\frac{r-r_{TCZ}}{d_t}\right)\right],\\
301: \eta_p(r) &= \eta_{RZ} + \frac{\eta_2-\eta_{RZ}}{2}\left[1+\erf\left(\frac{r-r_{BCZ}}{d_t}\right)\right] \nonumber\\
302: & \qquad\qquad\qquad +
303: \frac{\eta_0-\eta_2}{2}\left[1+\erf\left(\frac{r-r_{TCZ}}{d_t}\right)\right].
304: \end{align}
305: Here $d_t=0.025R_\odot$, $r_{BCZ}=0.7R_\odot$, $r'_{BCZ}=0.72R_\odot$, and
306: $r_{TCZ}=0.975R_\odot$. In the radiative core we choose a low diffusivity, namely
307: $\eta_{RZ}=2.2\times 10^8\cmsqpsec$, representing molecular diffusivity only since
308: there is no turbulent convection. We always choose $\eta_1<\eta_2$ so that the toroidal
309: field diffusivity $\eta_t$ in the convection zone is lower than the poloidal field
310: diffusivity $\eta_p$. This is to model the suppression of turbulent diffusivity by
311: strong magnetic fields, as toroidal field tends to be strong and concentrated in
312: localised flux tubes and therefore subject to less diffusion \citep{choudhuri2003},
313: whereas poloidal field is weaker and subject to more diffusion. At the surface both
314: diffusivities increase to a high value (of the order of $10^{12}\cmsqpsec$), in line
315: with surface flux transport models and observational estimates. Typical profiles are
316: illustrated in Figure \ref{fig:eta}.
317: % fig:eta
318:
319:
320: \subsection{Numerical Domain and Boundary Conditions}
321: %-----------------------------------------------------------------------------------
322:
323: We solve Equations \eqref{Aevol} and \eqref{Bevol} in a meridional plane $0.55 R_\odot
324: < r < R_\odot$, $0 < \theta < \pi/2$, representing the Northern hemisphere. This is
325: divided into a spherical grid of 128 by 128 cells, uniformly spaced in $r$ and
326: $\theta$. We use the same boundary conditions as \citet{chatterjee2004}, except that we
327: consider only the Northern hemisphere and set $B=0$ and $\pdlin{A}{\theta}=0$ at the
328: equator ($\theta=\pi/2$), thereby forcing the solution to have dipolar parity.
329:
330: \subsection{Example Solutions}
331: %-----------------------------------------------------------------------------------
332:
333: Example solutions for two different runs are shown in time-latitude plots in Figures
334: \ref{fig:fields}(a) and (b), where the black lines denote contours of toroidal field
335: strength at the base of the convection zone ($r=0.71R_\odot$). This corresponds to the
336: solar butterfly diagram, with the strongest field located at the active latitudes and
337: migrating equatorward during each cycle. The background shading shows the strength of
338: the radial field at the solar surface ($r=R_\odot$), which peaks at the pole several years after the toroidal field maxima (of the same sign) at low latitudes. The two solutions in Figure \ref{fig:fields} characterize the diffusion-dominated SCZ (Figure \ref{fig:fields}a) and advection-dominated SCZ (Figure \ref{fig:fields}b) regimes of the dynamo. In Figure \ref{fig:fields}(b) the toroidal field shows a poleward branch at high latitude which is absent in Figure \ref{fig:fields}(a), and also a stronger radial polar field at the surface. The cause of these differences between the two regimes will become clear in Section \ref{sec:regimes}.
339:
340: % fig:fields
341:
342:
343: \section{Results} \label{sec:results}
344: %==============================================================================
345:
346: We have carried out a parameter-space study to investigate how the cycle duration and
347: amplitude in our model depend on the speed of meridional circulation and on the
348: diffusivity in the convection zone. In each run of the code the parameters are held
349: constant in time, but they are varied between different runs. Specifically, we vary the
350: parameter $v_0$, which gives the maximum circulation speed, and also $\eta_2$, which
351: affects the diffusive decay and transport of the \emph{poloidal} field in the
352: convection zone, but not the toroidal field. In all runs we keep a surface diffusivity
353: of $\eta_0 = 2.0 \times 10^{12}\cmsqpsec$, and a toroidal field diffusivity of $\eta_1
354: = 0.04 \times 10^{12}\cmsqpsec$ in the convection zone. These choices approximate the
355: fact that turbulent diffusion is expected to be more efficient in the decay and
356: dispersal of the weaker poloidal field, but less so for the stronger toroidal field;
357: the latter suppresses the convective motions that give rise to turbulent diffusivity in
358: the first place. The $\alpha$-effect coefficient is not varied, but is set to $\alpha_0
359: = 30 \mpsec$ for each run. This particular value was chosen to ensure that periodic
360: solutions could be obtained for a wide range of the parameters $v_0$ and $\eta_2$. Each
361: run was started from an arbitrary initial state, and then evolved until initial
362: transients had disappeared, leaving a steady periodic dynamo solution. Such a periodic
363: solution was found to exist only within a certain range of $v_0$ for each value of
364: $\eta_2$. The cycle duration and amplitude were then measured from the periodic
365: solutions, in the cases where such a solution was found.
366:
367: We define the cycle duration and amplitude by considering the time evolution of the
368: toroidal magnetic flux $\Phi_\mathrm{tor}$ in a certain region around the base of the
369: convection zone. Specifically, the toroidal field $B(r,\theta)$ is integrated at each
370: time step over a region $r=0.677 R_\odot$ to $0.726 R_\odot$, $\theta=45^\circ$ to
371: $80^\circ$ (i.e., over the tachocline and latitudes $45^\circ$ to $10^\circ$). This
372: magnetic flux $\Phi_\mathrm{tor}$ should be proportional to the active region magnetic
373: flux at the solar surface, under the assumption that more toroidal flux at the base of
374: the convection zone leads to more buoyant eruptions. In a steady dynamo solution the
375: flux $\Phi_\mathrm{tor}$ varies in strictly periodic manner, with its maximum amplitude
376: giving the ``cycle amplitude''. We define the ``cycle duration'' to be the interval
377: between successive peaks of $|\Phi_\mathrm{tor}|$ (half of the full dynamo period).
378: This is therefore equivalent to the standard definition of the 11-year solar activity
379: cycle, but of course the simulated periods are different.
380:
381: The resulting cycle duration and cycle amplitude are plotted as functions of the
382: circulation speed $v_0$ in Figures \ref{fig:dur_v0} and \ref{fig:amp_v0} respectively.
383: In these figures each curve corresponds to a different value of the diffusivity
384: $\eta_2$. The range of speeds covered by each curve indicates the range for which the
385: code relaxed to a steady periodic dynamo solution, up to a maximum of $v_0=38\mpsec$.
386: In Figure \ref{fig:amp_diff} the cycle amplitude is plotted as a function of $\eta_2$,
387: and in this case each curve corresponds to a different circulation speed $v_0$.
388:
389: % fig:dur_v0
390: % fig:amp_v0
391: % fig:amp_diff
392:
393:
394: \subsection{Dependence of Cycle Period on Meridional Circulation and Diffusion}
395: %-----------------------------------------------------------------------------------------------------------------------------------------
396:
397: Figure \ref{fig:dur_v0} shows a clear inverse dependence of the cycle duration on the
398: meridional circulation speed $v_0$, with faster circulation leading to shorter cycles.
399: A least-squares fit for the curve with $\eta_2=0.5\times 10^{12}\cmsqpsec$ gives the
400: dependence of the cycle period on meridional flow speed
401: \begin{equation}
402: T = 217.716\,v_0^{-0.885},
403: \end{equation}
404: where $T$ is in years and $v_0$ is in metres per second. This agrees with the $T \sim
405: v_0^{-0.89}$ found by \citet{dikpati1999}, and this inverse relation is a
406: well-established result for Babcock-Leighton dynamo models. In these models the
407: circulation, and specifically the equatorward counterflow at the bottom of the
408: convection zone, is the primary determinant of the cycle period \citep{nandy2004}.
409:
410: The cycle duration is only weakly dependent on the diffusivity $\eta_2$. A power-law
411: fit for the curve with $\eta_2=2.0\times 10^{12}\cmsqpsec$ gives $T =
412: 150.745\,v_0^{-0.756}$ years. The lower power of $v_0$ here indicates that for
413: higher-diffusivity solutions the cycle duration is slightly less dependent on
414: circulation speed, presumably because flux transport by diffusive dispersal starts
415: becoming important. Also, it is evident from Figure \ref{fig:dur_v0} that, at lower
416: circulation speeds, there is a maximum diffusivity for which a periodic solution can
417: exist. If there is too much diffusion at a low circulation speed, then the poloidal
418: field will decay too much during its transport from high to low latitudes, thus
419: generating insufficient toroidal field to sustain a periodic dynamo process. The
420: essential difference between advective and diffusive flux transport is that the latter
421: also reduces field strength during transport, due to diffusive decay.
422:
423:
424: \subsection{Dependence of Cycle Amplitude on Meridional Circulation and Diffusion}
425: %-----------------------------------------------------------------------------------------------------------------------------------------
426:
427: Now we turn to the dependence of cycle amplitude on the speed of meridional
428: circulation. This is shown in Figure \ref{fig:amp_v0}, where each curve corresponds to
429: a different diffusivity $\eta_2$ according to the legend. Rather than being monotonic,
430: the cycle amplitude first increases with $v_0$ for low $v_0$ and then decreases with
431: $v_0$ for large $v_0$, with a turnover at some value of $v_0$ in between. The location
432: of this turnover shifts to higher speeds as the diffusivity is increased.
433:
434: The dependence of cycle amplitude on diffusivity at any given circulation speed is not
435: entirely clear on Figure \ref{fig:amp_v0}, but is evident in Figure \ref{fig:amp_diff},
436: where cycle amplitude is plotted against diffusivity $\eta_2$. In this figure each
437: curve corresponds to a different value of the circulation speed $v_0$. We see a similar behavior in that the cycle amplitude first increases with $\eta_2$ for
438: low $\eta_2$ and then decreases with $\eta_2$ for high $\eta_2$, with a turnover in
439: between. If the circulation speed is increased, then the value of $\eta_2$
440: corresponding to this turnover also increases.
441:
442: The behaviour of the cycle amplitude in our model, as illustrated in Figures
443: \ref{fig:amp_v0} and \ref{fig:amp_diff}, is more complex than expected \emph{a priori}.
444: Rather than a simple linear dependence on the circulation speed $v_0$, there is a
445: turnover in cycle amplitude, at a speed which changes depending on the diffusivity in
446: the convection zone. In the next section we investigate the cause of this behaviour in
447: the model.
448:
449: \section{Advection versus Diffusion Dominated Solar Convection Zones} \label{sec:regimes}
450: %==============================================================================
451:
452: The turnover of cycle amplitude as depicted in Figure \ref{fig:amp_v0} occurs at a
453: higher circulation speed $v_0$ as the diffusivity $\eta_2$ is increased. The asterisks
454: (joined by a thin line) in Figure \ref{fig:regimes}(a) show the location of this
455: turnover as a function of $\eta_2$. We may think of this line in the $(\eta_2,v_0)$
456: plane as the dividing line between two distinct regimes of the dynamo, which we call
457: \emph{advection-dominated} and \emph{diffusion-dominated}. The advection-dominated
458: regime corresponds to high circulation speed and low diffusivity, while the
459: diffusion-dominated regime corresponds to low circulation speed and high diffusivity. A
460: shift from one of these regimes to another affects flux-transport dynamics in a way
461: that results in contrasting dependence of cycle amplitude on the governing parameters.
462: Consider how the cycle amplitude varies with $v_0$ for a fixed value of $\eta_2$,
463: corresponding to a curve on Figure \ref{fig:amp_v0}. In the diffusion-dominated regime,
464: a higher circulation speed means less time for diffusive decay of the poloidal field
465: during its transport through the convection zone, leading to more generation of
466: toroidal field and hence a higher cycle amplitude. In the advection-dominated regime, a
467: higher circulation speed leads to a lower cycle amplitude because there is less time to
468: amplify toroidal field in the tachocline (through which magnetic fields are swept
469: through at a faster speed). It is the balance between these conflicting influences that
470: leads to a turnover in cycle amplitude at some intermediate circulation speed.
471:
472: % fig:regimes
473:
474:
475: The bold line in Figure \ref{fig:regimes}(a) shows the transition point between the two
476: regimes that may be inferred from a simple balance between circulation and diffusion
477: timescales $\tau_{\textrm{C}}$ and $\tau_{\textrm{D}}$. For a given circulation speed
478: $v_0$, we define the circulation timescale $\tau_{\textrm{C}}$ as the time taken for
479: meridional circulation to advect poloidal fields from $r=0.95R_\odot, \theta=45^\circ$
480: to the location where the strongest toroidal field is formed at the tachocline
481: ($\theta=60^\circ$). The diffusion timescale is defined as
482: $\tau_{\textrm{D}}=L^2/\eta_2$, where $L=0.285R_\odot$ is the radial distance across the
483: convection zone from the same starting point. The two timescales are compared in Figure
484: \ref{fig:regimes}(b), where each horizontal line gives the circulation time
485: $\tau_{\textrm{C}}$ for a different speed $v_0$, and the bold curve gives
486: $\tau_{\textrm{D}}$ as a function of $\eta_2$. The crossing points of horizontal lines
487: with this curve give the transition points between the advection dominated
488: ($\tau_{\textrm{C}} < \tau_{\textrm{D}}$) and diffusion dominated ($\tau_{\textrm{C}} >
489: \tau_{\textrm{D}}$) regimes from these simple theoretical considerations -- which are in
490: good agreement with the simulated transition points.
491:
492:
493: \subsection{Magnetic Field Evolution in Advection versus Diffusion Dominated Regimes}
494: %-----------------------------------------------------------------------------------------------------------------------------------------
495:
496: We now compare the poloidal and toroidal field evolution in the two regimes. Figure
497: \ref{fig:adpoloidal} shows the poloidal field lines for two runs, at different times
498: through the cycle, starting from one cycle minimum and finishing at the next cycle
499: minimum, so that the fields reverse in sign. The left-hand column is taken from a run
500: with $v_0=20\mpsec$ and $\eta_2=0.5\times 10^{12}\cmsqpsec$, which is in the
501: advection-dominated regime. The right-hand column is from a diffusion-dominated run
502: with the same $v_0$ but with $\eta_2=2.0\times 10^{12}\cmsqpsec$. Figure
503: \ref{fig:adtoroidal} shows the evolution of the toroidal field for the same two runs.
504:
505: % figLadpoloidal
506: % fig:adtoroidal
507:
508:
509: The key difference between the two regimes is the rate at which the poloidal field is
510: able to diffuse through the convection zone, after it is generated at the surface by
511: the Babcock-Leighton $\alpha$-effect. This is seen clearly by comparing the poloidal
512: field evolution between 8 and 12 years in the two runs (Figures \ref{fig:adpoloidal}c/h
513: and d/i). In the diffusion-dominated run the new clockwise poloidal field diffuses
514: directly down to the tachocline at all latitudes between these two times. However, in
515: the advection-dominated run the new poloidal field does not reach the tachocline until
516: the end of the cycle (16 years), and reaches the high latitudes before it reaches the
517: tachocline; i.e., in this case, the field evolution follows the meridional circulation
518: conveyor belt. There is still significant anticlockwise poloidal field remaining below
519: the tachocline from the previous cycle, and even a lower band of clockwise field from
520: the cycle before that. In the diffusion-dominated case there is only a weak band of
521: anticlockwise field remaining from the previous cycle at solar minimum.
522:
523: This suggests that in the advection-dominated regime, poloidal fields from cycles $n$,
524: $n-1$, and $n-2$ combine to produce the toroidal field for cycle $n+1$, while in the
525: diffusion-dominated regime it is produced primarily from cycle $n$ poloidal field, with
526: a small contribution from cycle $n-1$.
527:
528: The main difference in toroidal field evolution seen in Figure \ref{fig:adtoroidal} is
529: during the rising phase of the cycle, seen at 4 and 8 years (panels b/g and c/h). In
530: the advection-dominated regime there are two separate regions of toroidal field
531: generation, one region at high latitudes from poloidal field which has been advected
532: poleward by the meridional circulation, and a second region at lower latitudes arising
533: from direct diffusion of poloidal field across the convection zone. In the
534: diffusion-dominated case there is no strong generation of toroidal field at high
535: latitudes. In this case the strongest field generation occurs at mid to low latitudes,
536: with direct diffusion presumably being the primary means of transporting poloidal field
537: to the base of the convection zone (i.e., the meridional circulation conveyor belt is
538: ``short-circuited''). We point our however that, contrary to usual expectations,
539: diffusive flux transport still plays a role in the advection-dominated case, and it is
540: responsible for the complex dependence of cycle amplitude on diffusivity in the
541: advection-dominated regime; we explore this issue below.
542:
543: \subsection{The Role of Diffusive Flux Transport}
544: %-----------------------------------------------------------------------------------------------------------------------------------------
545:
546: We have thus far identified the turnover in cycle amplitude to lie at the transition
547: point between advection-dominated and diffusion-dominated regimes of the dynamo. Within
548: the umbrella of this model, this maximum in the amplitude is understood to be a balance
549: between the time available for toroidal field amplification and the time available for
550: poloidal field decay. Figure \ref{fig:amp_diff} shows how the cycle amplitude varies
551: with the diffusivity $\eta_2$ for a fixed circulation speed $v_0$. We see that there is
552: a turnover in cycle amplitude for some value of $\eta_2$, with lower diffusivities
553: corresponding to the advection-dominated regime, and higher diffusivities to the
554: diffusion-dominated regime. In the diffusion-dominated regime, which is only reached
555: for lower speeds $v_0$ in Figure \ref{fig:amp_diff}, the cycle amplitude decreases with
556: increasing diffusivity. This is expected due to increased cancellation and decay of the
557: poloidal field. However, in the advection-dominated regime, the cycle amplitude
558: \emph{increases} as the diffusivity $\eta_2$ is increased. This initially seems
559: counter-intuitive, but we show here that it is caused by the influence of direct
560: diffusive flux transport of poloidal field across the convection zone.
561:
562: This direct diffusion (from the solar surface to the base of the SCZ) was visible in
563: the poloidal field evolution plots shown in the previous section; we now demonstrate
564: its quantitative effect as $\eta_2$ is varied, by comparing the poloidal field strength
565: $\left|\bpol\right|$ at the base of the convection zone ($r=0.715R_\odot$) with that at
566: the solar surface ($r=R_\odot$). We take the ratio $\left|\bpol(\textrm{base})\right| /
567: \left| \bpol(\textrm{top})\right|$, using the peak value of $\left| \bpol\right|$ at
568: each location during the solar cycle. This ratio is plotted in Figure
569: \ref{fig:polratio}, measured at latitudes $30^\circ$ and $60^\circ$, and for two
570: different circulation speeds. Thin lines correspond to $v_0=30\mpsec$, where the dynamo
571: is in the advection-dominated regime for the whole range of $\eta_2$ shown. Thick lines
572: correspond to $v_0=20\mpsec$, for which the dynamo changes between the two regimes at
573: about $\eta_2=1.1\times 10^{12}\cmsqpsec$.
574:
575: % fig:polratio
576:
577:
578: Consider first the behaviour at $30^\circ$ latitude (solid lines in Figure
579: \ref{fig:polratio}). Here the curves for both $v_0$ have positive slope, which implies
580: that a greater proportion of poloidal field from the surface reaches the bottom of the
581: convection zone as $\eta_2$ is increased. Thus direct diffusive flux transport at lower
582: latitudes always acts to increase the amount of poloidal field reaching the base of the
583: convection zone. Nearer to the pole, at $60^\circ$ latitude (dashed lines in Figure
584: \ref{fig:polratio}), the behaviour is different. Here the ratio decreases as $\eta_2$
585: is increased, both for the curve with $v_0=30\mpsec$ and in the diffusion-dominated
586: regime for $v_0=20\mpsec$. In the advection-dominated regime for $v_0=20\mpsec$
587: however, the ratio first increases with $\eta_2$. This suggests a more complex relation
588: between the surface and tachocline poloidal fields at high latitude, with competing
589: influence from both diffusive and advective flux transport. This is expected because
590: the downflow in the circulation is located at high latitudes.
591:
592: The analysis presented in this section supports the idea that direct diffusive
593: transport of poloidal field across the convection zone, especially around mid-latitudes, is responsible for the trend of increasing cycle amplitude with increasing
594: diffusivity, found in the advection-dominated regime. Although such diffusive transport
595: acts to increase cycle amplitude in both regimes, diffusion also causes the poloidal
596: field that is being transported by meridional circulation to decay, cancelling with
597: field from the previous cycle that is stored below the tachocline. Thus diffusion also
598: has a negative effect on cycle amplitude. It is this negative effect which dominates at
599: higher diffusivities, forcing the dynamo into the diffusion-dominated regime where
600: cycle amplitude decreases with increasing diffusivity.
601:
602: \section{Persistence of Memory: Cycle-to-Cycle Correlations in Advection versus
603: Diffusion Dominated Regimes in a Stochastically Forced Dynamo} \label{sec:timedelays}
604: %==============================================================================
605:
606: It is expected that the memory of a flux-transport dynamo is much longer in the
607: advection-dominated regime than in the diffusion-dominated regime, and solar cycle
608: predictions have been based on this expectation \citep{dikpati2006b,jiang2007}.
609: However, a detailed comparative analysis of persistence of memory in these different
610: regimes under the umbrella of the same model had not been previously performed. Our
611: analysis in the previous section has brought us closer to understanding the flux
612: transport dynamics (in these two regimes) that is the physical basis for any memory
613: mechanism. In this section, we consider how the persistence of this memory differs
614: between the two regimes, by looking at the correlation between peak polar and toroidal
615: fields of subsequent cycles. Since the simulations considered earlier relaxed to a
616: regular periodic cycle, we cannot use these to study correlations between different
617: cycles. Therefore, we now introduce self-consistent fluctuations in the cycle
618: properties by means of a stochastically varying $\alpha$-effect, and explore the
619: resulting correlations between different cycles.
620:
621: \subsection{Stochastic Fluctuations}
622: %-------------------------------------------------------------------------------
623: We introduce fluctuations in the model by varying the coefficient $\alpha_0$ of the
624: $\alpha$-effect (see \citealp{chatterjee2004} for the full expression). We set
625: \begin{equation}
626: \alpha_{0} = \alpha_{\textrm{base}} + \alpha_{\textrm{fluc}}\,\sigma(t; \tau_{\textrm{cor}}),
627: \end{equation}
628: where $\alpha_{\textrm{base}}=30\mpsec$ is the mean value,
629: $\alpha_{\textrm{fluc}}=30\mpsec$ gives the maximum amplitude of the fluctuations
630: (corresponding to the 200\% level), and $\sigma$ is a uniform random deviate selected
631: from the interval $[-1,1]$, with a new value after each coherence time
632: $\tau_{\textrm{cor}}$. Although for our purposes this is essentially a device for
633: changing the cycle properties from one cycle to the next, there is a strong physical
634: basis for stochastic variations in $\alpha$ which have been invoked in several previous
635: studies \citep{choudhuri1992,hoyng1993,ossendrijver1996,charbonneau2000,mininni2002}.
636: Our model uses a Babcock-Leighton $\alpha$-effect where poloidal field is generated at
637: the surface from the decay of tilted active regions \citep{babcock1961,leighton1969}.
638: Thus stochastic variations in the $\alpha$ coefficient are natural, because it arises
639: from the cumulative effect of a finite number of discrete flux emergence events (active
640: region eruptions with varying degrees of tilt).
641:
642: To compare the two regimes we consider two runs, both with $\eta_{2}=1.0\times
643: 10^{12}\cmsqpsec$. The circulation speed $v_0$ is kept constant throughout each run,
644: and only the $\alpha$ effect is varied. Run 1 has $v_0=15\mpsec$, so is
645: diffusion-dominated, while run 2 has $v_0=26\mpsec$ and is advection-dominated. The
646: coherence time $\tau_{\textrm{cor}}$ is set to $2.3$ years in run 1 and $1.5$ years in
647: run 2, so as to keep the ratio of the former to the cycle duration roughly the same in
648: each case. We note that although the exact value of the coherence time is not important
649: for our study (and is introduced just as a means to enable sufficient fluctuations),
650: the timescale -- on the order of a year -- is chosen to reflect that the BL $\alpha$-effect
651: is a result of surface flux transport processes (diffusion, meridional circulation and
652: differential rotation) which can take up to a year to generate a net radial (component
653: of the poloidal) field from multiple flux emergence events \citep{mackay2004}.
654:
655: \subsection{Correlation Analysis}
656: %-------------------------------------------------------------------------------
657: In this section we compare the peak surface radial flux $\Phi_r$ for cycle $n$ with the
658: peak toroidal flux $\Phi_{\textrm{tor}}$ for cycles $n$, $n+1$, $n+2$, and $n+3$. The
659: toroidal flux is defined as before by integrating $B(r,\theta)$ over the region
660: $r=0.677R_\odot$ to $0.726R_\odot$, $\theta=45^\circ$ to $80^\circ$. The radial flux
661: $\Phi_r$ is found by integrating $B_r(R_\odot,\theta)$ over the solar surface between
662: $\theta=1^\circ$ to $20^\circ$, (i.e., latitudes $70^\circ$ to $89^\circ$). Note that
663: the peak toroidal flux precedes the peak surface radial flux for the same cycle, which
664: has the same sign. The poloidal field then produces the toroidal field for cycle $n+1$
665: with the opposite sign. We measure the correlation of the surface radial flux for cycle
666: $n$ with the toroidal flux of different cycles, comparing the absolute value of each
667: total signed flux.
668:
669: Both runs were computed for a total of 275 cycles with fluctuating $\alpha_0$, so as to
670: produce meaningful statistics for each of the dynamo regimes. The results are
671: illustrated in Figures \ref{fig:cordiff} and \ref{fig:coradv} as scatter-plots of
672: $\Phi_\textrm{tor}$ for different cycles against $\Phi_r(n)$. The (non-parametric)
673: Spearman's rank correlation coefficient $r_\textrm{s}$ is given above each plot, along
674: with its significance level. The correlation coefficients are summarised in Table
675: \ref{tab:correlations}, where the Pearson's linear correlation coefficient
676: $r_\textrm{p}$ is also given for comparison. Although the latter is less reliable, as
677: it assumes a linear relation, it agrees well with $r_\textrm{s}$ in each case.
678:
679: % fig:cordiff
680: % fig:coradv
681:
682: % table
683:
684: The results show a clear difference between the two regimes. The advection-dominated
685: regime shows significant correlations at all 4 time delays, apparently suggesting that
686: the memory of past poloidal field survives for at least 3 cycles; however, more on this
687: later. The diffusion-dominated regime has a strong correlation only between $\Phi_{r}(n)$
688: and $\Phi_{\textrm{tor}}(n+1)$, suggesting that the dominant memory relates to just a one
689: cycle time-lag, although very weak correlations are also found with
690: $\Phi_{\textrm{tor}}(n)$ and $\Phi_{\textrm{tor}}(n+3)$.
691:
692: In both regimes the strongest relation is the positive correlation between $\Phi_{r}(n)$
693: and $\Phi_{\textrm{tor}}(n+1)$. This is to be expected as this is the more deterministic
694: phase of the cycle -- where toroidal fields (of cycle $n+1$, say) are inducted from the
695: older cycle $n$ poloidal field via the relatively steady differential rotation. Note
696: however that the two fluxes do not have to be directly coupled, in that the two fluxes
697: may be positively correlated because they are both created from the mid-latitude
698: poloidal field of cycle $n$ (generated by the $\alpha$-effect). The polar flux
699: $\Phi_{r}(n)$ arises through poleward meridional transport of the cycle $n$ poloidal
700: field, while the toroidal flux $\Phi_{\textrm{tor}}(n+1)$ is generated from cycle $n$
701: poloidal field that is diffusively transported across the convection zone. This is
702: particularly the case in the diffusion-dominated regime. Nonetheless, even this
703: indirect scenario suggests that the strongest correlation should in fact be between
704: the cycle $n$ poloidal field and cycle $n+1$ toroidal field in this class of
705: $\alpha$-$\Omega$ dynamo models.
706:
707: The other phase of the cycle, in which the poloidal field is generated by the
708: $\alpha$-effect, is inherently more random due to the fluctuating $\alpha$-effect in
709: these runs. Nevertheless, there is a strong positive correlation between
710: $\Phi_{\textrm{tor}}(n)$ and $\Phi_{r}(n)$ in the advection-dominated regime, while this
711: correlation is largely absent in the diffusion-dominated regime. This we attribute to
712: the relatively stronger role of advective flux transport in the advection-dominated
713: regime -- which implies that a larger fraction of the original toroidal flux that has
714: buoyantly erupted is transported to the polar regions by the circulation. In effect
715: therefore, the advection-dominated regime allows correlations to propagate \emph{in
716: both phases of the cycle}, whereas the diffusion-dominated case allows correlations to
717: propagate only in the poloidal-to-toroidal phase. The other correlation is broken in
718: the diffusion-dominated regime because the advection is short-circuited by direct
719: diffusion, which transports flux downwards and equatorward -- where it is cancelled by
720: oppositely signed flux from the other hemisphere. This explains how the correlations
721: can survive for multiple cycles in the advection-dominated regime, but not in the
722: diffusion-dominated regime.
723:
724: \section{Conclusion} \label{sec:conclusion}
725: %==============================================================================
726:
727: Significant uncertainties remain in our understanding of the physics of the solar
728: dynamo mechanism, implying that prediction of future solar activity based on physical
729: models is a challenging task. Here we have demonstrated how a flux-transport dynamo
730: model behaves differently in advection and diffusion dominated regimes. Such
731: differences, amongst others, have previously led to conflicting predictions of the
732: amplitude of Cycle 24. \citet{dikpati2006} use an advection-dominated model to predict a
733: much stronger cycle than Cycle 23, whereas \citet{choudhuri2007} use a
734: diffusion-dominated model to predict a much weaker cycle. The latter prediction is
735: somewhat similar in spirit to the precursor methods \citep{schatten2005,
736: svalgaard2005}, which use the polar field at cycle minimum to predict the amplitude of
737: the following cycle. Owing to the lack of observations of conditions inside the
738: convection zone, opinions differ as to whether the real solar dynamo is weakly or
739: strongly diffusive (e.g. \citealp{dikpati2006b}; \citealp{jiang2007}).
740:
741: We find that for low circulation speeds $v_0$ (in the diffusion-dominated regime), the
742: cycle amplitude is an increasing function of $v_0$, as in the observations of
743: \citet{hathaway2003}. However, the amplitude curve has a turnover point and is a
744: decreasing function of $v_0$ at higher $v_0$ (in the advection-dominated regime),
745: opposite to the observed correlation. When the diffusivity in the convection zone is
746: increased, the location of this turnover moves to a higher $v_0$. Our extensive
747: analysis shows that this turnover corresponds to the transition between the diffusion
748: and advection dominated regimes. In the diffusion-dominated regime, faster circulation
749: means less time for decay of the poloidal field, leading to a higher cycle amplitude,
750: whereas in the advection-dominated regime diffusive decay is less important and a
751: faster circulation means less time to induct toroidal field, thus generating a lower
752: cycle amplitude. If the observed statistics of the past $12$ cycles as reported by
753: \citet{hathaway2003} reflect a true underlying trend, then our results imply that the
754: solar dynamo is in fact working in a regime which is dominated by diffusive flux
755: transport in the main body of the SCZ (although the cycle period is still governed by
756: the slow meridional circulation counterflow at the base of the SCZ). This conclusion
757: supports the analysis of \citet{jiang2007}.
758:
759: Through a correlation analysis in a stochastically forced version of our model, we have
760: also explored the persistence of memory in the solar cycle for both the diffusion-dominated and advection-dominated regimes. It is this memory mechanism which is
761: understood to lead to predictive capabilities in $\alpha$-$\Omega$ dynamo models with
762: spatially segregated source regions for the $\alpha$ and $\Omega$ effects. This
763: understanding is based on the finite time delay required for flux transport to
764: communicate between these different source regions. We find that the polar field of
765: cycle $n$ correlates strongest with the amplitude (toroidal flux) of cycle $n+1$ in
766: both the regimes. In the diffusion-dominated regime this is the only significant
767: correlation, indicative of a one-cycle memory only. However, in the advection-dominated
768: case, there are also significant correlations with the amplitude of cycles $n$, $n+2$,
769: and $n+3$. In contrast to the correlations that we infer, \citet{charbonneau2000} found
770: that the strongest correlation in their advection dominated model was with a two-cycle
771: time lag. Since such correlations lead to predictive capabilities, and obviously seem
772: to be model and parameter-dependent as suggested by our results, such a correlation
773: analysis should be the first step towards any prediction, the latter being based on the
774: former. In hindsight, however, both \citet{dikpati2006} -- who use an advection-dominated model and inputs from multiple previous cycles, and \citet{choudhuri2007} --
775: who use a diffusion-dominated model and input from only the past cycle to predict the
776: next cycle, appear to be have made the correct choices within their modelling
777: assumptions.
778:
779: Note that the memory mechanism in our advection-dominated case appears to have a
780: different cause than that implied by \citet{dikpati2006b}, who invoke the survival of
781: multiple old-cycle polar fields feeding into a new cycle toroidal field. All of
782: the surviving correlations in our advection-dominated regime (Figure \ref{fig:coradv})
783: are positive; they do not alternate in sign. This alternation in sign would be expected if
784: bands of multiple older cycle poloidal field survive in the tachocline -- odd and even
785: cycle poloidal fields would obviously contribute oppositely because of their
786: alternating signs. In that case we would expect the absolute value of $\Phi_{r}(n)$ to
787: correlate positively with $\Phi_{\textrm{tor}}(n+1)$ and $\Phi_{\textrm{tor}}(n+3)$ and so
788: on, but negatively with $\Phi_{\textrm{tor}}(n+2)$ and $\Phi_{\textrm{tor}}(n+4)$ and so on,
789: as evident in the results of \citet[][Figure~9; after accounting for the fact that they use signed magnetic fields]{charbonneau2000}. Rather, in the advection-dominated regime
790: of our model, the correlations appear to persist simply because fluctuations in field
791: strength are passed on in both the poloidal-to-toroidal and toroidal-to-poloidal phases
792: of the cycle, as evidenced by the correlation between amplitude and polar flux of cycle
793: $n$. In a recent analysis, \citet{cameron2007} find that the predictive skill of a
794: surface flux transport model -- similar in spirit to the advection-dominated dynamo
795: model of \citet{dikpati2006} -- is contained in the input information of sunspot areas
796: in the declining phase of the cycle. They argue that memory of multiple past cycles, in
797: the form of surviving bands of poloidal field (its surface manifestations in their
798: case), need not be the only reason behind the predictive capability of the
799: advection-dominated dynamo model of \citet{dikpati2006}. Our analysis of the
800: advection-dominated regime supports this suggestion of \citet{cameron2007}.
801:
802: Coming back to the diffusion-dominated regime, our comparative analysis indicates that
803: in this case, the memory of past cycles is governed by downward diffusion of poloidal
804: field into the tachocline -- which primarily results in a one-cycle memory. The fact
805: that diffusion is an efficient means for transporting flux is often ignored, especially
806: in this era of advection-dominated models; however, we find that diffusive flux
807: transport is quite efficient. The identification of this one cycle memory in our
808: stochastically forced model contradicts \citet{dikpati2006b} -- who claim that
809: prediction is not possible in this regime. As long as the source regions are spatially
810: segregated, and one of the source effects is observable and the other deterministic, flux
811: transport $\alpha$-$\Omega$ dynamos will inherently have predictive skills no matter
812: what physical process (i.e., circulation, or diffusion, or downward flux-pumping) is
813: invoked to couple the two regions. We may also point out that in the context of cycle-to-cycle correlations, downward flux pumping \citep{tobias2001} would have the same
814: effect as diffusion in that it also acts to short-circuit the meridional circulation
815: conveyor belt. So although downward flux pumping differs from diffusive transport
816: because in the latter case the fields may reduce in strength due to decay, the overall
817: persistence of memory is expected to be similar if diffusive flux transport was
818: replaced or complemented by downward flux pumping.
819:
820: In summary, our analysis has served both to explore the diffusion dominated and
821: advection dominated regimes within the framework of a BL type dynamo, and to
822: demonstrate how the memory of the dynamo may be different in these two regimes. Based
823: on our analysis we assert that diffusive flux transport in the SCZ plays an important
824: role in flux transport dynamics, even if the dynamo cycle period is governed by the
825: meridional flow speed. In fact, the observed solar cycle amplitude-period dependence
826: may arise more naturally in the diffusion-dominated regime, as discussed earlier. Taken
827: together therefore, we may conclude that diffusive flux transport is a significant
828: physical process in the dynamo mechanism and this process leads primarily to a
829: one-cycle memory which may form the physical basis for solar cycle predictions, if
830: other physical mechanisms involved in the complete dynamo chain of events are well
831: understood. Separate, detailed examinations of these other related physical
832: mechanisms will be performed in the future.
833:
834: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
835: \acknowledgements
836: This research was funded by NASA Living With a Star grant NNG05GE47G. We
837: also acknowledge support from the UK STFC and the Solar Physics NSF-REU program
838: at Montana State University.
839:
840: \pagebreak
841: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
842: \begin{thebibliography}{}
843:
844: \bibitem[\protect\citeauthoryear{Babcock}{1961}]{babcock1961}
845: Babcock, H. W. 1961, ApJ, 133, 572
846:
847: \bibitem[\protect\citeauthoryear{Beer}{2000}]{beer2000}
848: Beer, J. 2000, Space Sci. Rev., 94, 53
849:
850: \bibitem[\protect\citeauthoryear{Beer et al.}{1998}]{beer1998}
851: Beer, J., Tobias, S., \& Weiss, N. 1998, Sol. Phys., 181, 237
852:
853: \bibitem[\protect\citeauthoryear{Bonanno et al.}{2002}]{bonanno2002}
854: Bonanno, A., Elstner, D., R\"udiger, G., \& Belvedere, G. 2002, A\&A, 390, 673
855:
856: \bibitem[\protect\citeauthoryear{Cameron \& Sch\"ussler}{2007}]{cameron2007}
857: Cameron, R., \& Sch\"ussler, M. 2007, ApJ, 659, 801
858:
859: \bibitem[\protect\citeauthoryear{Charbonneau}{2005}]{charbonneau2005}
860: Charbonneau, P. 2005, Living Rev. Sol. Phys., 2, 2,
861: \url{http://solarphysics.livingreviews.org/Articles/lrsp-2005-2/}
862:
863: \bibitem[\protect\citeauthoryear{Charbonneau \& Dikpati}{2000}]{charbonneau2000}
864: Charbonneau, P., \& Dikpati, M. 2000, ApJ, 543, 1027
865:
866: \bibitem[\protect\citeauthoryear{Charbonneau et al.}{2005}]{charbonneau2005b}
867: Charbonneau, P., St-Jean, C., \& Zacharias, P. 2005, ApJ, 619, 613
868:
869: \bibitem[\protect\citeauthoryear{Chatterjee \& Choudhuri}{2006}]{chatterjee2006}
870: Chatterjee, P., \& Choudhuri, A. R. 2006, Sol. Phys., 239, 29
871:
872: \bibitem[\protect\citeauthoryear{Chatterjee et al.}{2004}]{chatterjee2004}
873: Chatterjee, P., Nandy, D., \& Choudhuri, A. R. 2004, A\&A, 427, 1019
874:
875: \bibitem[\protect\citeauthoryear{Choudhuri}{1992}]{choudhuri1992}
876: Choudhuri, A. R. 1992, A\&A, 253, 277
877:
878: \bibitem[\protect\citeauthoryear{Choudhuri}{2003}]{choudhuri2003}
879: Choudhuri, A. R. 2003, Sol. Phys., 215, 31
880:
881: \bibitem[\protect\citeauthoryear{Choudhuri et al.}{2007}]{choudhuri2007}
882: Choudhuri, A. R., Chatterjee, P., \& Jiang, J. 2007, Phys. Rev. Lett., 98, 131103
883: \bibitem[\protect\citeauthoryear{Choudhuri et al.}{1995}]{choudhuri1995}
884: Choudhuri, A. R., Sch\"ussler, M., \& Dikpati, M. 1995, A\&A, 303, L29
885:
886: \bibitem[\protect\citeauthoryear{Dikpati \& Charbonneau}{1999}]{dikpati1999}
887: Dikpati, M., \& Charbonneau, P. 1999, ApJ, 518, 508
888:
889: \bibitem[\protect\citeauthoryear{Dikpati et al.}{2006}]{dikpati2006}
890: Dikpati, M., de Toma, G., \& Gilman, P. A. 2006, Geophys. Res. Lett., 33, 5102
891:
892: \bibitem[\protect\citeauthoryear{Dikpati \& Gilman}{2006}]{dikpati2006b}
893: Dikpati, M., \& Gilman, P. A. 2006, ApJ, 649, 498
894:
895: \bibitem[\protect\citeauthoryear{Durney}{1995}]{durney1995}
896: Durney, B. R. 1995, Sol. Phys., 160, 213
897:
898: \bibitem[\protect\citeauthoryear{Durney}{2000}]{durney2000}
899: Durney, B. R. 2000, Sol. Phys., 196, 421
900:
901: \bibitem[\protect\citeauthoryear{Hathaway}{1996}]{hathaway1996}
902: Hathaway, D. H. 1996, ApJ, 460, 1027
903:
904: \bibitem[\protect\citeauthoryear{Hathaway}{2005}]{hathaway2005}
905: Hathaway, D. H. 2005, in ASP Conf. Ser. 346, Large-scale Structures and their Role in
906: Solar Activity, ed. K. Sankarasubramanian, M. Penn, \& A. Pevtsov (San Francisco, ASP),
907: 19
908:
909: \bibitem[\protect\citeauthoryear{Hathaway et al.}{2003}]{hathaway2003}
910: Hathaway, D. H., Nandy, D., Wilson, R. M., \& Reichmann, E. J. 2003, ApJ, 589, 665
911:
912: \bibitem[\protect\citeauthoryear{Hoyng}{1993}]{hoyng1993}
913: Hoyng, P. 1993, A\&A, 272, 321
914:
915: \bibitem[\protect\citeauthoryear{Jiang et al.}{2007}]{jiang2007}
916: Jiang, J., Chatterjee, P., \& Choudhuri, A.R. 2007, MNRAS, submitted
917:
918: \bibitem[\protect\citeauthoryear{Knobloch et al.}{1998}]{knobloch1998}
919: Knobloch, E., Tobias, S. M., \& Weiss, N. O. 1998, MNRAS, 297, 1123
920:
921: \bibitem[\protect\citeauthoryear{K\"uker et al.}{1999}]{kueker1999}
922: K\"uker, M., Arlt, R., \& R\"udiger, G. 1999, A\&A, 343, 977
923:
924: \bibitem[\protect\citeauthoryear{K\"uker et al.}{2001}]{kueker2001}
925: K\"uker, M., R\"udiger, G., \& Schultz, M. 2001, A\&A, 374, 301
926:
927: \bibitem[\protect\citeauthoryear{Leighton}{1969}]{leighton1969}
928: Leighton, R. B. 1969, ApJ, 156, 1
929:
930: \bibitem[\protect\citeauthoryear{Mackay et al.}{2004}]{mackay2004}
931: Mackay, D. H., Jardine, M., Collier Cameron, A., Donati, J.-F, \& Hussain, G. A. J.
932: 2004, MNRAS, 354, 737
933:
934: \bibitem[\protect\citeauthoryear{Miesch}{2005}]{miesch2005}
935: Miesch, M. S. 2005, Living Rev. Sol. Phys., 2, 1,
936: \url{http://solarphysics.livingreviews.org/Articles/lrsp-2005-1/}
937:
938: \bibitem[\protect\citeauthoryear{Mininni \& G\'omez}{2002}]{mininni2002}
939: Mininni, P. D., \& G\'omez, D. O. 2002, ApJ, 573, 454
940:
941: \bibitem[\protect\citeauthoryear{Moffatt}{1978}]{moffatt1978}
942: Moffatt, H. K. 1978, Magnetic Field Generation in Electrically Conducting Fluids
943: (Cambridge: Cambridge Univ. Press)
944:
945: \bibitem[\protect\citeauthoryear{Nandy}{2004}]{nandy2004}
946: Nandy, D. 2004, in Proceedings of SOHO 14/GONG 2004, Helio- and Asteroseismology:
947: Towards a Golden Future, ed. D. Danesy (ESA SP-599), 241
948:
949: \bibitem[\protect\citeauthoryear{Nandy \& Choudhuri}{2001}]{nandy2001}
950: Nandy, D., \& Choudhuri, A. R. 2001, ApJ, 551, 576
951:
952: \bibitem[\protect\citeauthoryear{Nandy \& Choudhuri}{2002}]{nandy2002}
953: Nandy, D., \& Choudhuri, A. R. 2002, Science, 296, 1671
954:
955: \bibitem[\protect\citeauthoryear{Nandy \& Martens}{2007}]{nandy2007}
956: Nandy, D., \&, Martens, P.C.H. 2007, Adv. Space Res., in press
957:
958: \bibitem[\protect\citeauthoryear{Ossendrijver et al.}{1996}]{ossendrijver1996}
959: Ossendrijver, A. J. H., Hoyng, P., \& Schmitt, D. 1996, A\&A, 313, 938
960:
961: \bibitem[\protect\citeauthoryear{Ossendrijver}{2003}]{ossendrijver2003}
962: Ossendrijver, M. 2003, A\&A Rev., 11, 287
963:
964: \bibitem[\protect\citeauthoryear{Parker}{1955}]{parker1955}
965: Parker, E. N. 1955, ApJ, 121, 491
966:
967: \bibitem[\protect\citeauthoryear{Schatten}{2005}]{schatten2005}
968: Schatten, K. 2005, Geophys. Res. Lett., 32, 21106
969:
970: \bibitem[\protect\citeauthoryear{Sch\"ussler \& Schmitt}{2004}]{schuessler2004}
971: Sch\"ussler, M., \& Schmitt, D. 2004, A\&A, 421, 349
972:
973: \bibitem[\protect\citeauthoryear{Sheeley}{1991}]{sheeley1991}
974: Sheeley, N. R. 1991, ApJ, 374, 386
975:
976: \bibitem[\protect\citeauthoryear{Solanki et al.}{2002}]{solanki2002}
977: Solanki, S. K., Krivova, N. A., Sch\"ussler, M., \& Fligge, M. 2002, A\&A 396, 1029
978:
979: \bibitem[\protect\citeauthoryear{Svalgaard et al.}{2005}]{svalgaard2005}
980: Svalgaard, L., Cliver, E. W., \& Kamide, Y. 2005, Geophys. Res. Lett., 32, 1104
981:
982: \bibitem[\protect\citeauthoryear{Tobias}{1997}]{tobias1997}
983: Tobias, S. M. 1997, A\&A, 322, 1007
984:
985: \bibitem[\protect\citeauthoryear{Tobias et al.}{2001}]{tobias2001}
986: Tobias, S. M., Brummell, N. H., Clune, T. L., Toomre, J. 2001, ApJ, 549, 1183
987:
988: \bibitem[\protect\citeauthoryear{Wang et al.}{2002}]{wang2002b}
989: Wang, Y.-M., Lean, J., \& Sheeley, N. R. 2002, ApJ, 577, L53
990:
991: \bibitem[\protect\citeauthoryear{Wilmot-Smith et al.}{2005}]{wilmotsmith2005}
992: Wilmot-Smith, A. L., Martens, P. C. H., Nandy, D., Priest, E. R., \& Tobias, S. M.
993: 2005, MNRAS, 363, 1167
994:
995: \bibitem[\protect\citeauthoryear{Wilmot-Smith et al.}{2006}]{wilmotsmith2006}
996: Wilmot-Smith, A. L., Nandy, D., Hornig, G., \& Martens, P. C. H. 2006, ApJ, 652, 696
997:
998: \bibitem[\protect\citeauthoryear{Yoshimura}{1978}]{yoshimura1978}
999: Yoshimura, H. 1978, ApJ, 226, 706
1000:
1001: \end{thebibliography}
1002:
1003: \begin{figure}
1004: \plottwo{f1a.eps}{f1b.eps}
1005: \caption{Streamlines of the meridional circulation profile in our model (\emph{left}), and latitudinal velocity profile across a radial cut at $\theta=45^\circ$ (\emph{right}). Dots on the streamlines show yearly positions for particles with $v_0=25\mpsec$, starting from the squares at $\theta=45^\circ$ and moving anti-clockwise.} \label{fig:merid}
1006: \end{figure}
1007:
1008: \begin{figure}
1009: \epsscale{0.7}
1010: \plotone{f2.eps}
1011: \caption{Example
1012: diffusion profiles as a function of $r$ for the toroidal ($\eta_t$) and poloidal
1013: ($\eta_p$) fields. The dotted line shows the location of the tachocline. Here
1014: $\eta_0=2.0\times 10^{12}\cmsqpsec$, $\eta_1=0.04\times 10^{12}\cmsqpsec$, and
1015: $\eta_2=1.0\times 10^{12}\cmsqpsec$.} \label{fig:eta}
1016: \end{figure}
1017:
1018: \begin{figure}
1019: \epsscale{0.7}
1020: \plotone{f3a.eps}
1021: \plotone{f3b.eps}
1022: \caption{Two example
1023: solutions: (a) with $v_0=20\mpsec$ and $\eta_2=2.0\times10^{12}\cmsqpsec$
1024: (characterizing a diffusive flux-transport dominated SCZ); and (b) with the same $v_0$
1025: but $\eta_2=0.5\times10^{12}\cmsqpsec$ (characterizing an advective flux-transport
1026: dominated SCZ). In each case black lines are contours of toroidal field $B$ at the base
1027: of the convection zone (solid lines for positive values, dashed for negative). The
1028: grayscale in the background shows surface radial field strength $B_r(r=R_\odot)$, with
1029: white for positive and black for negative. The same contour levels are used in both
1030: plots.} \label{fig:fields}
1031: \end{figure}
1032:
1033: \begin{figure}
1034: \plotone{f4.eps}
1035: \caption{Dependence of
1036: cycle duration on the meridional circulation speed $v_0$. Each line style corresponds
1037: to a different value of $\eta_2$ (the poloidal diffusivity in the convection zone) as
1038: given in the legend.} \label{fig:dur_v0}
1039: \end{figure}
1040: \begin{figure}
1041: \plotone{f5.eps}
1042: \caption{Dependence of
1043: cycle amplitude on the meridional circulation speed $v_0$. Each line style corresponds
1044: to a different value of $\eta_2$ (the poloidal diffusivity in the convection zone) as
1045: given in the legend.} \label{fig:amp_v0}
1046: \end{figure}
1047: \begin{figure}
1048: \plotone{f6.eps}
1049: \caption{Dependence of cycle amplitude on the poloidal diffusivity $\eta_2$ in the
1050: convection zone. Each line style corresponds to a different value of the meridional
1051: circulation speed $v_0$, as given in the legend.} \label{fig:amp_diff}
1052: \end{figure}
1053:
1054: \begin{figure}
1055: \hbox{
1056: \includegraphics[width=0.46\textwidth]{f7a.eps}
1057: \includegraphics[width=0.54\textwidth]{f7b.eps}
1058: }
1059: \caption{Transition
1060: between the advection-dominated and diffusion-dominated regimes. In (a) asterisks
1061: indicate the flow speeds $v_0$ corresponding to turnover of cycle amplitude for fixed
1062: values of $\eta_2$ (inferred from the simulations shown in Figures~5 and 6). The bold
1063: line shows the transition point that may be inferred from simple theoretical comparison
1064: of circulation and diffusion timescales. Panel (b) shows the diffusion timescale
1065: $\tau_{\textrm{D}}$ as a function of $\eta_2$ (bold line), and circulation timescales
1066: $\tau_{\textrm{C}}$ for selected speeds $v_0$ (horizontal lines), as defined in the
1067: text.} \label{fig:regimes}
1068: \end{figure}
1069:
1070: \begin{figure}
1071: \begin{center}
1072: \epsscale{0.47}
1073: \plottwo{f8a.eps}{f8f.eps}\\
1074: \plottwo{f8b.eps}{f8g.eps}\\
1075: \plottwo{f8c.eps}{f8h.eps}\\
1076: \plottwo{f8d.eps}{f8i.eps}\\
1077: \plottwo{f8e.eps}{f8j.eps}
1078: \caption{Comparison
1079: of poloidal fields in advection-dominated (left column/panels a to e) and
1080: diffusion-dominated (right column/panels f to j) regimes. Each row corresponds to a
1081: time through the solar cycle, running from one cycle minimum to the next. Solid lines
1082: show clockwise field lines and dashed lines show anti-clockwise field lines. Also
1083: indicated is the base of the SCZ at $0.71R_\odot$.} \label{fig:adpoloidal}
1084: \end{center}
1085: \end{figure}
1086:
1087: \begin{figure}
1088: \begin{center}
1089: \epsscale{0.45}
1090: \plottwo{f9a.eps}{f9f.eps}\\
1091: \plottwo{f9b.eps}{f9g.eps}\\
1092: \plottwo{f9c.eps}{f9h.eps}\\
1093: \plottwo{f9d.eps}{f9i.eps}\\
1094: \plottwo{f9e.eps}{f9j.eps}\\
1095: \caption{Comparison
1096: of toroidal field in advection-dominated (left column/panels a to e) and
1097: diffusion-dominated (right column/panels f to j) regimes. Each row corresponds to a
1098: different time through the solar cycle, running from one cycle minimum to the next.
1099: Grayscale contours show toroidal field strength, with black corresponding to the
1100: strongest negative field and white to the strongest positive toroidal field. Also
1101: indicated is the base of the SCZ at $0.71R_\odot$.} \label{fig:adtoroidal}
1102: \end{center}
1103: \end{figure}
1104:
1105: \begin{figure}
1106: \epsscale{0.8}
1107: \plotone{f10.eps}
1108: \caption{Ratio of poloidal field $\left|\bpol\right|$ at base of convection zone
1109: ($r=0.715R_\odot$) to that at the surface ($r=R_\odot$), measured as a function of
1110: diffusivity at latitudes $30^\circ$ (solid lines) and $60^\circ$ (dashed lines). Thick
1111: lines correspond to runs with $v_0=20\mpsec$ and thin lines to runs with
1112: $v_0=30\mpsec$. } \label{fig:polratio}
1113: \end{figure}
1114:
1115: \begin{figure}
1116: \plotone{f11.eps}
1117: \caption{Cycle-to-cycle correlations in the diffusion-dominated regime (run 1), between radial flux
1118: $\Phi_r(n)$ and (a) toroidal flux $\Phi_{\textrm{tor}}(n)$, (b) $\Phi_{\textrm{tor}}(n+1)$,
1119: (c) $\Phi_{\textrm{tor}}(n+2)$, and (d) $\Phi_{\textrm{tor}}(n+3)$. The Spearman's rank
1120: correlation coefficient is given along with its significance level for 275 cycles. All
1121: magnetic fluxes are in units of $10^{25}\mx$.} \label{fig:cordiff}
1122: \end{figure}
1123:
1124: \begin{figure}
1125: \plotone{f12.eps}
1126: \caption{Cycle-to-cycle correlations in the advection-dominated regime (run 2), between radial flux
1127: $\Phi_r(n)$ and (a) toroidal flux $\Phi_{\textrm{tor}}(n)$, (b) $\Phi_{\textrm{tor}}(n+1)$,
1128: (c) $\Phi_{\textrm{tor}}(n+2)$, and (d) $\Phi_{\textrm{tor}}(n+3)$. The Spearman's rank
1129: correlation coefficient is given along with its significance level for 275 cycles. All
1130: magnetic fluxes are in units of $10^{25}\mx$.} \label{fig:coradv}
1131: \end{figure}
1132:
1133: \clearpage
1134: \begin{deluxetable}{lrrrrrrr}
1135: \tablewidth{0pt}
1136: \tablecaption{Cycle-to-cycle correlations \label{tab:correlations}}
1137: \tablehead{ &
1138: \multicolumn{3}{c}{$\Phi_r(n)$ for run 1} && \multicolumn{3}{c}{$\Phi_r(n)$ for run 2} \\
1139: & \multicolumn{3}{c}{(diffusion-dominated)} & \phantom{---} & \multicolumn{3}{c}{(advection-dominated)} \\
1140: & \multicolumn{2}{c}{$r_\textrm{s}$} & \multicolumn{1}{c}{$r_\textrm{p}$} && \multicolumn{2}{c}{$r_\textrm{s}$} & \multicolumn{1}{c}{$r_\textrm{p}$} }
1141: \tablecolumns{8}
1142: \startdata
1143: $\Phi_{\textrm{tor}}(n)$ & 0.185 & $99.8\%$ \phantom{--}& \textit{0.287} && 0.653 & $100.0\%$ \phantom{--}& \textit{0.778} \\
1144: $\Phi_{\textrm{tor}}(n+1)$ & 0.737 & $100.0\%$ \phantom{--}& \textit{0.706} && 0.805 & $100.0\%$ \phantom{--}& \textit{0.851}\\
1145: $\Phi_{\textrm{tor}}(n+2)$ & -0.040 & $49.1\%$ \phantom{--}& \textit{0.028} && 0.356 & $100.0\%$ \phantom{--}& \textit{0.546}\\
1146: $\Phi_{\textrm{tor}}(n+3)$ & 0.195 & $99.9\%$ \phantom{--}& \textit{0.202} && 0.237 & $100.0\%$ \phantom{--}& \textit{0.417}\\
1147: $\Phi_{\textrm{tor}}(n+4)$ & 0.036 & $44.5\%$ \phantom{--}& \textit{0.056} && 0.183 & $99.8\%$ \phantom{--}& \textit{0.357}\\
1148: $\Phi_{\textrm{tor}}(n+5)$ & 0.107 & $92.3\%$ \phantom{--}& \textit{0.073} && 0.214 & $100.0\%$ \phantom{--}& \textit{0.316}\\
1149: \enddata
1150: \tablecomments{Correlation coefficients and significance levels for peak surface radial flux $\Phi_{r}$ versus peak toroidal flux $\Phi_{\textrm{tor}}$ for 275 cycles from stochastically forced dynamo simulations.}
1151: \end{deluxetable}
1152:
1153:
1154: \end{document}
1155: