Global induction by Sq and Dst sources in the presence of oceans: bimodal solutions for non-uniform spherical surface shells above radially symmetric earth models in comparison to observations

We pursue in this paper the following goals. (1) Detailed study of the ocean effect on geomagnetic fields excited by Sq and Dst sources in spherical earth models. The investigated period range is from 6 hr to 10 days. Oceans and continents form a thin shell of variable conductance above a layered deep structure, allowing for leakage currents from and into the shell. (3) Development of an interpretation scheme for the estimation of the resistivity pi of the lithosphere, (3) Application of this scheme to estimate rho(1) from a comparison of calculated and observed Sq harmonics at coastal sites. We conduct a number of model calculations with simplified hemispherical shells and with realistic shells for the world's oceans and continents, deriving their variable conductance from the ocean depth at sea and assuming a uniform value of 400 S on land, with inclusion of continental shelfs. From these model studies we draw the following conclusions. (1) Variations of rho(1) in the uppermost 100 km of the deep structure significantly change the vertical magnetic surface field Z near oceans. (2) Ocean effects in the horizontal magnetic surface fields X and Y are quite negligible on land for Sq. Therefore, radially symmetric earth models can be used to determine the external Sq source field from a spherical harmonic analysis of Y and Y on land. (3) With realistic shell models it is possible to reproduce the anomalous coast effect in Z of Sq, A two-step interpretation scheme is presented to estimate the resistivity rho(1) of the lithosphere. In step 1 we determine the spherical harmonic coefficients of the external Sq sourer field using the observed Sq variations in X and Y and a presumed earth model of uniform shell conductance to separate their external and internal parts. In step 2 we derive model fields with various trial values of rho(1) for a realistic surface shell excited by the external source field from step 1. We choose the most likely value of rho(1), that for which the rms misfit between observed and calculated Z variations is smallest. Our observational data are the first four Sq harmonics of the four equinoctial months of 1964 as derived from the mean daily variations of quiet days at 76 observatories. Z observations at 12 coastal observatories lead us to the conclusion that a global estimate of lithospheric resistivity rho(1) is of the order of 10 000 Ohm m.