Solar magnetic fields and the dynamo theory M. Dikpati * High Altitude Observatory, National Center for Atmospheric Research, 3450 Mitchell Lane, Boulder, CO 80301, USA Received 1 November 2004; received in revised form 19 April 2005; accepted 19 April 2005 Abstract Unlike EarthÕs dipolar magnetic fields, solar magnetic fields consist of wide ranges of length-scales and strengths, and interest- ingly, they evolve in a cyclic fashion with a 22-year periodicity. A magnetohydrodynamic dynamo operating in the Sun is most likely responsible for producing the solar magnetic activity cycle. While the first solar dynamo models were built half a century ago, recent views differ significantly from those models. According to widely accepted present concepts, the large-scale solar dynamo is of flux- transport type, which involves three basic processes: (i) generation of toroidal fields by shearing the pre-existing poloidal fields by differential rotation (the X-effect); (ii) re-generation of poloidal fields by lifting and twisting the toroidal fluxtubes (the a-effect); (iii) flux transport by meridional circulation. This class of dynamos has been successful in explaining many large-scale solar cycle fea- tures, including a particularly difficult one – the correct phase relationship between the equatorward-migrating sunspot belt and the poleward drifting large-scale, diffuse fields. The dynamo cycle period in such models is primarily governed by the meridional flow speed near the bottom of the convection zone. After briefly reviewing the historical background, we will present the successes of flux- transport dynamos, including their predictive capability. For example, we will demonstrate how the meridional circulation plays a key role in governing the SunÕs memory about its own magnetic field, and how a flux-transport dynamo-based predictive tool can explain the cause of the very slow polar reversal in the so-called ‘‘peculiar’’ cycle 23 compared to those in cycles 20, 21 and 22. We will close by presenting explanations for certain long-term variability using these models, such as, what may have maintained the observed cyclic variation in slow solar wind flow during Maunder minima, in the presence of near zero solar activity. Ó 2005 COSPAR. Published by Elsevier Ltd. All rights reserved. Keywords: Sun: magnetic fields; Activity cycle; Dynamo theory; Cycle prediction 1. Introduction The sunspots have been known to humanity for about 2000 years. Their rising and waning every 11 years is the best known manifestation of the solar cycle. The cyclic evolution of sunspots is generally represented in the Ôbutterfly diagramÕ (see Fig. 1), which is a time–lati- tude diagram illustrating that, at the beginning of a cycle the sunspots appear in a latitude belt at 35°, and with the progress of the cycle, the latitude belt in which the sunspots appear migrates equatorward (see the equator- ward-migrating butterfly wing in Fig. 1). It is most likely that an oscillatory dynamo in the Sun is driving the solar activity cycle. First solar dynamo models were developed about half a century ago (Parker, 1955). The solar dynamo models have evolved significantly since then in order to accom- modate observational constraints. There are great many reviews on this topic with varying foci. An extensive re- cent review is by Ossendrijver (2003). Accepting the fact that there is no unified model yet to explain both the cyclic and non-cyclic evolution of solar magnetic fields, we focus in this article on the large-scale solar dynamo models which are seeking physical explanations for the longitude-averaged solar cycle features occurring during decadal as well as centennial time-scales. 0273-1177/$30 Ó 2005 COSPAR. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.asr.2005.04.061 * Tel.: +1 303 497 1512/1594; fax: +1 303 497 1589. E-mail address: dikpati@hao.ucar.edu. www.elsevier.com/locate/asr Advances in Space Research 35 (2005) 322–328