Determining the Most Appropriate Solar Inputs for Upper Atmospheric Density Models

The solar spectral irradiance in the UV/EUV range is essential to characterize the amount of solar energy the upper atmosphere receives, and this is important in particular for the determination of satellite drag through the use of thermosphere models such as DTM or MSIS. However, because of the lack of long-term and continuous observations, various proxies for solar activity are used instead in upper atmosphere models. Finding more appropriate solar inputs has become crucial for specifying the effect of air drag on satellites and on debris. Using almost 17 years of mean densities derived from orbit perturbation analysis from the French geodetic satellite STELLA at about 820 km altitude and a selection of solar activity proxies, we provide a quantitative assessment of the performance of each proxy. We do this on different time scales as the performance of the proxies is found to be quite different for short (days) and long (months) time scales, as well as for low and high solar activity. This study shows which solar inputs are most appropriate at each time scale and provides guidance on the choice of a better set of inputs for operational density models.