A diurnal variation in thermospheric temperature in the geomagnetic polar cap at solar minimum (January 1987) has been observed using high spectral resolution measurements of the O(1D) emission line made with the Fabry-Perot interferometer located at Thule, Greenland (76¿32'N, 68¿45'W, &Lgr;=86). The temperature maximum in the diurnal cycle occurs at 1600 UT, which is very near the magnetic local noon period in the norther hemisphere. Data from the wind and temperature spectrometer on the Dynamics Explorer 2 satellite show a similar diurnal variation in the northern polar cap thermospheric temperature for near solar maximum conditions. In an attempt to find the mechanisms responsible for the variation, we have used the diagnostics package developed for use with the NCAR thermospheric general circulation model (TGCM), which allows the tracking of individual parcels of gas in both time and space by interpolation through the model grid. At each of the predicted loci of the parcels we have examined the physical factors which influence the thermal balance by decomposing the thermodynamic equation into its constituent terms.

By tracing the trajectory of a parcel backward in time and space from the location of Thule (i.e., well inside the geomagnetic polar cap), we have been able to show, for a model run with input parameters pertaining to day 314, 1976 (i.e., several weeks from solstice near solar minimum), that the observed diurnal temperature variation is attributable to the following factors: (1) the degree of solar heat input that a parcel temperature variation is attributable to the following factors: (1) the degree of solar heat input that a parcel experiences en route to the polar cap, which is dependent on whether or not it crosses the solar terminator, (2) the route a parcel takes through the polar cusp, i.e., whether the parcel skirts the edge or passes directly through the center determines the total quantity of heat added by soft particle impact within the cusp, (3) the time duration between maximum heat input from the combination of solar, cusp, auroral and Joule sources and the time of arrival overhead at Thule. In addition, we have shown (1) that the hot geomagnetic polar cap reported by Hays et al. (1984) is primarily due to transport of heat into the polar cap by parcels that were heated by auroral, Joule and cusp sources, and (2) that the hydrodynamical variations in the winter high-latitude regions from solar maximum to solar minimum are insufficient to mask the thermodynamical effects related to the offset of the geographic and geomagnetic poles. ¿ American Geophysical Union 1988