A longstanding mystery about Jupiter has been the straightness and steadiness of its weather-layer jets, quite unlike terrestrial strong jets with their characteristic unsteadiness and long-wavelength meandering. The problem is addressed in two steps. The first is to take seriously the classic Dowling-Ingersoll (DI) 1-1/2-layer scenario and its supporting observational evidence, pointing toward deep, massive, zonally-symmetric zonal jets in the underlying dry-convective layer. The second is to improve the realism of the model stochastic forcing used to represent the effects of Jupiter's moist convection as far as possible within the 1-1/2-layer dynamics. The real moist convection should be strongest in the belts where the interface to the deep flow is highest and coldest, and should generate cyclones as well as anticyclones with the anticyclones systematically stronger. The new model forcing reflects these insights. Also, it acts quasifrictionally on large scales to produce statistically steady turbulent weather-layer regimes without any need for explicit large-scale dissipation, and with weather-layer jets that are approximately straight thanks to the influence of the deep jets, allowing shear stability despite nonmonotonic potential-vorticity gradients when the Rossby deformation lengthscale is not too large. Moderately strong forcing produces chaotic vortex dynamics and realistic belt-zone contrasts in the model's convective activity, through an eddy-induced sharpening and strengthening of the weather-layer jets relative to the deep jets, tilting the interface between them. Weak forcing, for which the only jet-sharpening mechanism is the passive (Kelvin) shearing of vortices (as in the "CE2" or "SSST" theories), produces unrealistic belt-zone contrasts.

The simplest interior magnetic field B_i that can explain the observed uniform rotation of the Sun's radiative envelope is an axial dipole stabilized by a deep toroidal field. It can explain the uniform rotation only if confined in the polar caps. The field must be prevented from diffusing up into the high-latitude convection zone, whose slower rotation must remain decoupled from the radiative interior. This paper describes new analytical and numerical solutions of the relevant magnetohydrodynamic equations showing that such confinement and decoupling is dynamically possible by means of a laminar "magnetic confinement layer" at the bottom of the tachocline. With realistic values of the microscopic diffusivities, a weak laminar downwelling flow U~10^{-5}cm/s over the poles is enough to enforce exponential decay of B_i with altitude, in a confinement layer only a fraction of a megameter thick. Downwelling in the polar tachocline is implied both by helioseismic observations, combined with elementary dynamics, and by theoretical arguments about the "gyroscopic pumping" that would spread differential rotation downward in the absence of B_i. Our confinement-layer solutions are the first to take account of all the relevant physical effects in a self-consistent mathematical model. The effects include magnetic diffusion, baroclinicity and stable stratification (thermal and compositional), Coriolis effects, and thermal relaxation. We discuss how the confinement layers at each pole might fit into a global dynamical picture of the solar tachocline. That picture, in turn, suggests a new insight into the early Sun and solar lithium depletion.

The global-scale interior magnetic field needed to account for the Sun's observed differential rotation can be effective only if confined below the convection zone in all latitudes, including the polar caps. Axisymmetric nonlinear MHD solutions are obtained showing that such confinement can be brought about by a very weak downwelling flow U~10^{-5}cm/s over each pole. Such downwelling is consistent with the helioseismic evidence. All three components of the magnetic field decay exponentially with altitude across a thin "magnetic confinement layer" located at the bottom of the tachocline. With realistic parameter values, the thickness of the confinement layer ~10^{-3} of the Sun's radius. Alongside baroclinic effects and stable thermal stratification, the solutions take into account the stable compositional stratification of the helium settling layer, if present as in today's Sun, and the small diffusivity of helium through hydrogen, chi. The small value of chi relative to magnetic diffusivity produces a double boundary-layer structure in which a "helium sublayer" of smaller vertical scale is sandwiched between the top of the helium settling layer and the rest of the confinement layer. Solutions are obtained using both semi-analytical and purely numerical, finite-difference techniques. The confinement-layer flows are magnetostrophic to excellent approximation. More precisely, the principal force balances are between Lorentz, Coriolis, pressure-gradient and buoyancy forces, with relative accelerations and viscous forces negligible. This is despite the kinematic viscosity being somewhat greater than chi. We discuss how the confinement layers at each pole might fit into a global dynamical picture of the solar tachocline. That picture, in turn, suggests a new insight into the early Sun and into the longstanding enigma of solar lithium depletion.

High-latitude laminar confinement of the Sun's interior magnetic field is shown to be possible, as originally proposed by Gough and McIntyre (1998) but contrary to a recent claim by Brun and Zahn (A&A 2006). Mean downwelling as weak as 2x10^-6cm/s -- gyroscopically pumped by turbulent stresses in the overlying convection zone and/or tachocline -- can hold the field in advective-diffusive balance within a confinement layer of thickness scale ~ 1.5Mm ~ 0.002 x (solar radius) while transmitting a retrograde torque to the Ferraro-constrained interior. The confinement layer sits at the base of the high-latitude tachocline, near the top of the radiative envelope and just above the `tachopause' marking the top of the helium settling layer. A family of exact, laminar, frictionless, axisymmetric confinement-layer solutions is obtained for uniform downwelling in the limit of strong rotation and stratification. A scale analysis shows that the flow is dynamically stable and the assumption of laminar flow realistic. The solution remains valid for downwelling values of the order of 10^-5cm/s but not much larger. This suggests that the confinement layer may be unable to accept a much larger mass throughput. Such a restriction would imply an upper limit on possible internal field strengths, perhaps of the order of hundreds of gauss, and would have implications also for ventilation and lithium burning. The solutions have interesting chirality properties not mentioned in the paper owing to space restrictions, but described at http://www.atmos-dynamics.damtp.cam.ac.uk/people/mem/papers/SQBO/solarfigure.html