We present an analysis of a series of observations of the auroral/polar regions of Jupiter, carried out between September 8 and 11, 1998, making use of the high-resolution spectrometer, CSHELL, on the NASA InfraRed Telescope Facility (IRTF), Mauna Kea, Hawaii; these observations spanned an "auroral heating event." This analysis combines the measured line intensities and ion velocities with a one-dimensional model vertical profile of the jovian thermosphere/ionosphere. We compute the model line intensities both assuming local thermodynamic equilibrium (LTE) and, relaxing this condition (non-LTE), through detailed balance calculations, in order to compare with the observations. Taking the model parameters derived, we calculate the changes in heating rate required to account for the modelled temperature profiles that are consistent with the measured line intensities. We compute the electron precipitation rates required to give the modelled ion densities that are consistent with the measured line intensities, and derive the corresponding Pedersen conductivities. We compute the changes in heating due to Joule heating and ion drag derived from the measured ion velocities, and modelled conductivities, making use of ion-neutral coupling coefficients derived from a 3-D global circulation model. Finally, we compute the cooling due to the downward conduction of heat and the radiation-to-space from the H3+ molecular ion and hydrocarbons. Comparison of the various heating and cooling terms enables us to investigate the balance of energy inputs into the auroral/polar atmosphere. Increases in Joule heating and ion drag are sufficient to explain the observed heating of the atmosphere; increased particle precipitation makes only a minor heating contribution. But local cooling effects - predominantly radiation-to-space - are shown to be too inefficient to allow the atmosphere to relax back to pre-event thermal conditions. Thus we conclude that this event provides observational, i.e. empirical, evidence that heat must be transported away from the auroral/polar regions by thermally or mechanically driven winds. © 2005 Elsevier Inc. All rights reserved.