On periodic water-waves and their convergence to solitary waves in the long-wave limit

Abstract

It is possible to infer general properties of fluid flow in the core, but there is a large class of flows that fit even perfect data. This class can be reduced in size by making simplifying hypotheses about the flow and electrical conductivity of fluid in the core. For a hypothesis to be reasonable it must be possible to test it against observation; if the test proves favourable the hypothesis is adopted. Three such hypotheses are suggested here: the familiar one of perfect electrical conductivity, and two new ones, those of purely toroidal motion, and steady motion. The first two make it possible to find one component of core flow everywhere. In principle the third hypothesis allows both components to be found. A map of one component for the period 1959-74 shows significant north-south flow over Indonesia and rapid motion in a swath running from beneath North America, across Africa and down to the southern Indian Ocean. With longer-term data it may be possible to find both components of the flow.