Possible reasons for underestimating Joule heating in global models: E field variability spatial resolution and vertical velocity

t is important to understand Joule heating because it can significantly change the temperature structure, atmosphere composition, and electron density and hence influences satellite drag. It is thought that many coupled ionosphere-thermosphere models underestimate Joule heating because the spatial and temporal variability of the ionospheric electric field is not totally captured within global models. Using the Global Ionosphere Thermosphere Model (GITM), we explore the effect of the electric field temporal variability, model resolution, and vertical velocity differences between ion and neutral flows on Joule heating in a self-consistent thermosphere/ionosphere system. First, the response of Joule heating to a step change in the externally driven electric field has been studied. While Joule heating is strongly affected by the convection electric field, both neutral winds and electron densities can significantly alter the spatial distribution of the Joule heating. Owing to the ramping up of neutral winds, there is a temporal variation of the Joule heating energy deposition rate when the electric field is constant. Second, we compare the calculated neutral gas heating rates when GITM is run with three different temporal variations of the electric fields, having the same temporally averaged electric field (E) but different standard deviations (σ E ). The neutral gas heating rate increases with the electric field temporal variability, and due to the feedback of the neutral winds and electron densities, the percentage increase is different from σ E 2/(E)2, which is normally used to describe the effect of electric field temporal variability on the Joule heating. Third, comparison of the neutral gas heating rate with different model resolutions shows that at 200 km altitude, the polar average neutral gas heating rate increases by 20% when the latitudinal resolution increases from 5° to 1.25°. This is due to the model's ability to better capture small-scale features in the electric field and particle precipitation. Last, inclusion of the vertical velocity difference (which is neglected in many models) is less significant than the other two factors and appears to be negligible at high latitudes. While the magnitude of the neutral gas heating rate at middle and low latitudes is smaller than that at high latitudes, the relative importance of the vertical velocity difference is larger, and the contribution can reach 15% of the averaged Joule heating at middle and low latitudes.