The Feature Analysis and Modeling of Upper Atmospheric Midnight Density Maximum

Abstract

The features of upper atmospheric midnight density maximum (MDM) around low geographic latitudes are studied based on neutral mass densities data at altitudes 360–480 km, derived from the accelerometer measurements aboard on the three polar orbiting satellites CHAMP (CHAllenging Minisatellite Payload), GRACE-A (Gravity Recovery and Climate Experiment-A), and SWARM-C (The Earth's Magnetic Field and Environment Explorers-C). The MDM appears during the local times from 23:00 to 02:00 LT (Local Time), whose peak locates at the low latitudes within ${15}^{\circ }$ and two valleys locate at the middle latitudes between ${35}^{\circ }$ and ${45}^{\circ }$ on both hemispheres separately. The structure of MDM drifts toward the southern hemisphere overall. The MDM's amplitude decreases with increases in altitude and solar radiation level. The seasonal effect weakens the MDM's amplitudes around the summer and winter solstices, while the amplitudes around the spring and autumn equinoxes are extremely significant due to the slight seasonal difference between both hemispheres. Three atmospheric density models DTM2000 (Drag Temperature Model 2000), NRLMSISE00 (US Naval Research Laboratory Mass Spectrometer and Incoherent Scatter Radar Extended atmosphere model), and JB2008 (Jacchia-Bowman 2008 model) are used to simulate the MDM along these three satellites' orbits, and compared with the observations. It is found that the JB2008 model is failed to describe the MDM, and the other two models underestimate the MDM's amplitudes at altitudes 360 km and 480 km: the simulated amplitudes by the DTM2000 model are 46% and 53% of the observed amplitudes, respectively, and only 33% and 26% for the NRLMSISE00 model. These three models are also failed to depict the MDM's variation with altitude, solar radiation level, and seasonal effects. In order to correct the model prediction, a 6th-order Legendre polynomial of geographic latitude, coupled with arguments of local time and altitude, is designed to fit the MDM signals from the three satellites' observations. In terms of amplitude and phase of the MDM, the fitting results agree with the observations very well, and the correlation coefficient is 0.923. It indicates that this empirical polynomial could be helpful to the density model correction and high accuracy prediction of spacecrafts in low Earth orbits.