Response of the polar cap F region convection direction to changes in the interplanetary magnetic field: Digisonde measurements in northern Greenland

Abstract

Results of ionospheric drift measurements with a Digisonde 256 digital ionospheric sounder located at Qaanaaq, Greenland (87°N, corrected geomagnetic latitude), are presented. Digisonde drift data have been related to the interplanetary magnetic field (IMF) measured by the IMP 8 satellite for 32 days during 1986, 1987, and 1988. Extremely good statistical agreement between these measurements and convection directions derived from satellite instrumentation is demonstrated when the IMF z component is negative. For Bz negative and By negative the daily average convection direction is centered on −12° (anticlockwise) from the antisunward direction. When Bz is negative but By is positive, the convection direction is centered on +36°. These directions differ by 3° and 9°, respectively, from the models of Heppner and Maynard (1987). The variation about these values is ∼ ±20°. The excellent agreement between the Digisonde measurements and models derived from satellite measurements demonstrates the utility of the Digisonde for making ground‐based measurements of the convection direction in the polar cap F region when Bz is south. The convection directions under conditions of positive Bz have also been examined, and we have measured three types of temporal variation in azimuth, namely, an ordered and slowly (OS) varying change in direction, an ordered and quickly (OQ) varying change in direction, and disordered (D) variations in direction. The latter are believed to result from a breakdown of the analysis technique due to velocity shears in the vicinity of polar cap arcs, and we estimate that they account for ∼25% of the measurements when Bz > 0. When Bz is positive and By is negative, our small subset of OS measurements supports the distorted two‐cell model of Heppner and Maynard (1987). The remainder of the measurements show no well‐defined daily average convection direction or diurnal variation. Likewise for Bz positive and By positive, no well‐defined convection direction can be discerned, nor can any diurnal variation. The existence of OQ variations when Bz > 0 suggests that meaningful average statistical convection patterns may be much harder to synthesize than similar patterns when Bz < 0.