Thermospheric Density Derived From Orbit Parameters And Application Research

Thermospheric total mass density is a key parameter for near-Earth space operations,and is also very useful for scientific studies of upper atmospheric behavior.Atmospheric drag exerts a significant influence on near-Earth orbit,and accurate predictions of orbital trajectories require precise knowledge of the ambient density and its spatial and temporal variations.From 1950 s many empirical thermopheric density models were established and improved to adapt to various engineering tasks.Unfortunately,these models often exhibit large density errors of around 15% RMSE or up to 100% during magnetic storm periods.Density errors translated into to orbit errors,adversely affecting applications such as re-entry operations,manoeuvre planning,collision avoidance and precise orbit determination for geodetic missions.Firstly,Onboard GPS observation data of a satellite includes accurate information of velocity and location,which are related closely to atmospheric density.Therefore,these GPS information can be used to derive thermospheric density through integration of differential equation.This paper presents a new method of deriving atmospheric density with a high temporal resolution from precise orbit data of low earth orbiting(LEO)space objects,and also presents the solution procedure of mean motion and the inverse ballistic coefficient which are the two most important parameters for retrieving density.Tiangong-1 and CHAMP are taken as examples to evaluate the effectiveness of the method.The result shows that the GPS-derived density is in good agreement with observed density with the average error 8.6% and 8.4% respectively on 1st January and 24 th February in 2012 for Tiangong-1.This result indicates the method provides an effective and reliable way to obtain extensive and accurate thermospheric density.Secondly,the data of thermospheric total mass density derived from the high-accuracy accelerometer on board the CHAMP and GRACE-A/B satellites during 2001-2008 is analyzed.NRLMSISE-00 is selected as reference model for comparing with derived density.Mean absolute errors and mean relative errors are given between derived density and model values.Reasons rationality for error and some abnormal characters for density are shown.The reliability of the derived density is verified by TLEs density.The main conclusions are as follows: 1)Density obtained by CHAMP is larger than those obtained by GRACE-A/B at 400 km altitude.The error between the density obtained by CHAMP and model value are smaller than that between GRACE-A/B and model.2)There is a global deviation between the density obtained by CHAMP and model.The density variation in the similar space environment and the density obtained by TLEs show that the density in 2001 obtained by CHAMP is lower than those actual values.3)Examples of the density obtained by CHAMP and GRACE-A/B indicate that there are some individual characteristics of the density obtained by the satellites,and they should be analyzed and chose in application.Thirdly,based on thermospheric density from Explore 8 and the CHAMP satellite,long-term decrease in thermospheric density,and relationships between density and five solar flux proxies(F10.7,E10.7,Mg10.7,S10.7,Y10.7),two geomagnetic proxies(ap,Dst)are analyzed.The results show: 1)Thermospheric density derived from Explore 8 and CHAMP indicate a long-term decreased in density.2)Correlation coefficient between density and each solar proxy is greater than 0.97.Without considering the physical meaning of the radiation proxies,each proxy could represent upper atmosphere-relevant flux.3)Compared with ap,Dst has a better relationship with density on beginning time and increased amplitude through analyzing more than 70 geomagnetic storm samples.4)Five storm examples show different detail changes of density which model can‘t describe.5)During dayside,there is a density peak near south latitude 20oS,and in summer and winter,density will show obvious double peak structure in the northern and southern hemisphere,but the peak position is different.During nightside,there is a common feature of density groove region near the equator.Fourthly,density errors of NRLMSISE-00 according to different range of F10.7,different ap index,different latitude and different seasons are given.The result shows: 1)Model exhibit a decreasing error with the increase of F10.7,ap(or Dst index).2)RMSE of model on noon is bigger than that of night.3)Model error structure shows RMSE in northern hemisphere is smaller than that of northern hemisphere,and the minimum error appears near the North Pole.4)The error between model and CHAMP is less than that of model to GRACE-A/B.Finally,thermospheric density calibration is researched preliminarily during geomagnetic storms and calm conditions.The result shows: 1)During geomagnetic quiet period,the error between model and satellite density are relatively stable and the error correction method of sliding error method can be very effective to modify the model.2)Based on the relationship between density and Dst,the change of density during one orbit period could is reflected well by Dst value.3)Every storm event is a special case and the characters are different even though under the same geomagnetic conditions.It indicates only a few indexes cannot fully represent the causes and changes of density.