Ideally, in the convection problem we should start at the bow shock and solve self-consistently for energy, momentum and plasma transfer in the complete downstream system. The state of the art is to study individual regions (magnetosheath, tail lobes, high-beta plasma sheet, and Bipolar region) and make assumptions about the boundary conditions connecting them to neighbouring regions. Plasma sheath convection can be approximated by the model of a blunt body immersed in a supersonic or super-Alfvenic flow with some degree of success. Improvements need to include the effects of the daysidemerging and the nightside expansion fan sections of the magnetopause on flow in the plasma sheath. Convection in the Bipolar region is fairly well understood in terms of the calculation scheme proposed by VASYLIUNAS (1972), but a self-consistent determination of the latitude of the high-latitude boundary is needed as well as a better understanding of boundary conditions. Understanding of convection in the high-beta plasma sheet and lobe is still very poor. The energy source is electromagnetic and can be represented as Poynting vector flow from the highlatitude magnetopause. We are still very uncertain as to the energy sink and the plasma sources and sinks in the plasma sheet. In the tail lobes, selfconsistency is needed between the field-line topology (including twists and open-closed configuration) and the location of and merging rate at X lines. Non-MHD processes are important at the boundaries between all the above regions.
The System and the Ideal SolutionA measure of our understanding of a physical system, such as the magnetosphere, is the extent to which we can predict its behaviour starting from basic physics (in this case, plasma physics) and a set of initial and boundary conditions. In this paper, I shall attempt to evaluate how well we understand the magnetosphere. Figure 1 shows the system of interest. For the purposes of this paper, it is divided into the five regions indicated in the bottom half of the figure. These are the solar wind, the magnetosheath, and then within the magnetosphere: the tail lobes, the high-beta plasma sheet and the Bipolar region. The high-beta plasma sheet consists of closed magnetic flux tubes which are substantially distorted from a dipole-like form. Boundaries between these regions include: (i) the bow shock which, because it is a fast shock, is thin and clearly defined. (ii) the magnetopause which, in its simplest form, consists of a dayside merging configuration (X-line, slow shocks, magnetic islands?) connected to a downstream slow expansion fan which is not a sharp discontinuity, (iii) the plasma sheet boundary layer, also not a sharp discontinuity, which may be the result of merging at the nightside X line shown in the figure and (iv) the boundary between significantly distorted field lines and almost-Bipolar field lines. The two X-lines shown in the figure, one on the 183