Abstract

Recent studies have revealed that the post-storm thermospheric mass density at a fixed altitude could be lower than the pre-storm mass density. This phenomenon has been attributed to an overcooling effect caused by enhanced infrared emissions from nitric oxide (NO). Here we report that thermosphe... Show moreRecent studies have revealed that the post-storm thermospheric mass density at a fixed altitude could be lower than the pre-storm mass density. This phenomenon has been attributed to an overcooling effect caused by enhanced infrared emissions from nitric oxide (NO). Here we report that thermospheric density decrease can take place under both storm and non-storm conditions. Relevant observations are presented during two superstormes, one intense storm, and one non-storm case. Thermospheric neutral density variations observed by GOCE and CHAMP satellites are investigated along with simultaneous measurements of thermospheric nitric oxide density and emissions from TIMED/GUVI and TIMED/SABER, respectively, together with auroral hemispheric power from DMSP, Joule heating rate from the Assimilative Mapping of Ionospheric Electrodynamics (AMIE) algorithm, and solar extreme ultraviolet (EUV) flux from SOHO/SEM and TIMED/SEE. It was found that the apparent storm-time density decrease or "overcooling" is a result of combined effects: an increase in NO cooling and a decrease in both the solar EUV flux and the auroral energy input. It was also found that whether or not a density decrease exists depends on whether there are auroral activities prior to the pre-storm reference day and how the reference day is chosen. The relative contributions of the different drivers (namely, enhanced NO cooling and variations in auroral hemispheric power, Joule heating, and solar EUV at the pre-storm and post-storm days) to thermospheric density decrease remain to be determined. Show less