The response of the thermosphere and ionosphere to the famous January 10, 1997, geomagnetic storm is simulated using the thermosphere-ionosphere-electrodynamics general circulation model with realistic, time-dependent distributions of ionospheric convection and auroral precipitation as inputs. The simulation results show a dominant positive storm phase of increased F layer electron density over much of the northern winter hemisphere, but a negative storm phase with reduced electron density at middle and low latitudes is also evident in the simulation. The coexistence of both positive and negative storm phases is a result of the complex dynamical and chemical interactions between charged particles and neutral gases. The impulsive magnetospheric energy inputs via auroral precipitation and Joule heating generate traveling atmospheric and ionospheric disturbances (TADs and TIDs) which propagate from the northern auroral zone to lower latitudes and penetrate well into the Southern Hemisphere. The simulation results demonstrate that positive storm phases are caused primarily by enhanced auroral precipitation over high latitudes and by TIDs at middle and low latitudes. Globally speaking, composition changes in terms of enhancements in the N2/O ratio are mainly responsible for negative storm effects. However, although there is some correlation between increases in N2/O and decreases in the F layer critical frequency ƒoF2 in the winter hemisphere during the storm main phase and early recovery phase, the overall changes in ƒoF2 are also determined by other processes, such as the ionization production associated with enhanced auroral precipitation and the variations associated with TIDs. In the low to middle-latitude region changes in ƒoF2 approximately anticorrelate with changes at the height of the F layer electron density peak (e.g., hmF2) at 70°W during the storm main phase as well as its early recovery phase. This is attributed in part to the relation that exists between meridional wind velocity and vertical shear of that velocity for aurorally produced TADs.