An accurate estimate of the energy budget of the ionosphere and thermosphere systems is important in satellite drag prediction. Calculating satellite drag during high geomagnetic activity has been a challenge. Radiative cooling by nitric oxide (NO) and carbon dioxide (CO2) regulates the thermospheric density and thermal response to geomagnetic storms and has been shown to be directly correlated with composition and exospheric temperature changes during geomagnetic storms. In this work, we present a sensitivity study of the Global Ionosphere Thermosphere Model (GITM) to solar and geomagnetic activities by comparing the GITM model result with observations from Global-scale Observations of the Limb and Disk (GOLD) and Sounding of the Atmosphere using Broadband Emission Radiometry (SABER) measurements. SABER measurements are used to compare NO and CO2 cooling emission rate, and GOLD measurements to compare exospheric temperature and column integrated O/N2 ratio with GITM model results. We examined two geomagnetic storm events for which we drive GITM with different external inputs such as solar wind, F10.7, and auroral electrojet values to understand and address how well the model reproduces the lower thermospheric NO responses to solar and geomagnetic activity. The comparison suggests that GITM is overestimating the exospheric temperature during the geomagnetic storm period and underestimating the exospheric temperature during geomagnetic quiet time periods. GITM simulation is also run using the recently developed auroral electrojet (AE) driven Feature Tracking of Aurora (FTA) model, which provides the energy flux and average energies in the auroral region. We find that the high latitude electrodynamics, especially auroral participation specified in the FTA model, reduces the storm effect, which GITM overestimates during geomagnetic storm periods.