磁暴期间热层大气密度变化

A statistical study on the response of thermospheric total mass density to geomagnetic storms

基于CHAMP卫星资料,分析了2002—2008年267个磁暴期间400km高度大气密度变化对季节、地方时与区域的依赖以及时延的统计学特征,得到暴时大气密度变化的一些新特点,主要结论如下：1）两半球大气密度绝对变化（δρa）结果在不同强度磁暴、不同地方时不同.受较强的焦耳加热和背景中性风共同作用,在北半球夏季,中等磁暴过程中夜侧和大磁暴中,夏半球的δρa强于冬半球;由于夏季半球盛行风环流造成的扰动传播速度快,北半球夏季日侧30°附近大气,北（夏）半球到达峰值的时间早于南（冬）半球.而可能受半球不对称背景磁场强度所导致的热层能量输送率影响,北半球夏季强磁暴和中磁暴个例的日侧,南半球δρa强于北半球;春秋季个例中日侧30°附近大气,北半球先于南半球1～2h达到峰值.2）受叠加在背景环流上的暴时经向环流影响,春秋季暴时赤道大气密度达到峰值的时间最短,日/夜侧大气分别在Dstmin后1h和2h达到峰值.至点附近夜侧赤道大气达到峰值时间一致,为Dstmin后3h;不同季节日侧结果不同,在北半球冬季时赤道地区经过更长的时间达到峰值.3）日侧赤道峰值时间距离高纬度峰值时间不受季节影响,为3h左右.在春秋季和北半球冬季夜侧,赤道大气密度先于高纬度达到峰值,且不同纬度大气密度的峰值几乎无差别,表明此时低纬度存在其他加热源起着重要作用。

During geomagnetic storms,the coupling magnetosphere-ionosphere-thermosphere system is a rather complex phenomenon,and the thermospheric mass density exhibits large deviations from the climatological behavior upon the conjunct effect of Joule/particle heating,Lorentz force,thermal expansion,upwelling,and horizontal wind circulation.Due to different weight effects,thermospheric responses might vary with different storms,and even for the same storm case resulting from unlike methods of data process.In order to know more about theseasonal,magnetic local time（MLT）and latitude dependencies and the time delay characteristic of the thermospheric response to geomagnetic storms,we investigate the thermospheric response to 267 geomagnetic storms in which the Dst minimum,Dstmin,is below-50 nT during 2002—2008.The data of thermospheric mass density normalized to 400 km is derived from the highaccuracy accelerometer on board the CHAMP satellite.Each orbit is first divided into an ascending and a descending half,which are subdivided into five latitudinal segments,namely±60°,±30°and 0°.In order to investigate the dependence of MLT,density data are sorted into 4different MLT sectors：05∶00MLT to 09∶00MLT as the dawn sector,10∶00MLT to 16∶00MLT as the noon sector,17∶00MLT to 21∶00MLT as the dusk sector,and 22∶00MLT to 04∶00MLT as the night sector.To investigate seasonal variations,the available data are subdivided into three local seasons：the northern hemisphere winter（December-February,DJF）,combined equinoxes（March-May,MAM,and September-November,SON）,and the northern hemisphere summer（June-August,JJA）.Dstminis used to identify four categories of geomagnetic storms：weak storms（-30〉Dstmin≥-50nT）,moderate storms（-50〉Dstmin≥-100nT）,intense storms（-100〉Dstmin≥-200nT）and great storms（Dstmin〈-200nT）.By this means the effects of magnetic local time,latitude,season and intensity of storm are separated.Since the quiet-time density（ρq）shows much dependence on the solar activity,season,and local time,the density deviation from quiet-time values,rather than the total storm-time density（ρ）itself,seems better suited for describing the storm effect.There are two ways to define the deviation,one is the absolute difference（δρa =ρ-ρq）,and the other is the percentage difference（δρr =δρa/ρq）.As there is no general agreement on which expression is more appropriate,both the absolute and the percentage variations for each event are presented to produce a complete picture.Considering that the MSIS model underestimates the total mass density in the crest region resulting from its missing double peaks at low latitudes completely,the CHAMP measurements from the day prior to the storm is taken as a quiet-time reference density.The thermospheric mass density reacts after geomagnetic activity with a delay time,which is expected to depend on latitudes,MLT and seasons.Besides the superposed epoch comparisons for different conditions during storms,in which epoch time zero is chosen as the time of Dstmin,time delays between Dstmin and maxima of densities which are divided into different season,latitude,and MLT,have been computed for each storm event and the statistical result accounted for the biggest proportion describes quantitatively the time intervals.Besides some characteristics that have been mentioned in previous research,our statistical results reveal some new or more detailed variations about the responses of thermospheric mass density to geomagnetic storms,and the main conclusions are as follows：1）The absolute enhancements of thermospheric density during storms show a north-south asymmetry dependence on both the intensity of storms and the magnetic local time.In the northern hemisphere summer,for great storms and the nightside of moderate storms,controlled by higher Joule heating rates and prevailing summer-to-winter winds,stronger density enhancements occur in the summer hemisphere.On the dayside of northern hemisphere summer,due to the faster propagation of the disturbance from high to low latitudes in the summer hemisphere,the thermospheric density enhancements happen near 30 degree in the northern（summer）hemisphere peak ahead of those in the southern（winter）hemisphere.While probably affected by the higher rate of the energy transferred to the thermosphere partly dependent on the strength of the background magnetic field which is weaker in the Southern hemisphere due to shifted position of the dipole in positive Z-direction,on the dayside of northern hemisphere summer during intense and moderate storms,δρaof the southern（winter）hemisphere was stronger than that of the northern（summer）hemisphere,and on the dayside near equinoxes for most storms,the thermospheric density enhancements near 30 degree of the northern hemisphere peaked 1～2hahead of that of the southern hemisphere.2）Thermospheric densities of low latitudes enhancing after that of high latitudes during storms,the delay time during great storms is shorter than that of other weaker storms,and the time-lag during nightsie is shorter than that of dayside,indicating that propagation of energy deposited in polar regions to lower latitudes seems faster in the night-side sector during stronger storms.Only for great storms,the percentage differenceδρrof dayside sector in low latitudes is higher than that of high latitudes,and the density of low latitudes peaks earlier than that of high latitudes,implying some other heating source in low latitudes play an important role during great storms.3）Affected by the storm-time disturbance-driven thermospheric meridional circulation,the thermospheric density enhancements of the equator approach their maxima fastest at the equinoxes,and the time delay relative to Dstminis 1h,2hfor the density of dayside,night-side,respectively.At the nightside either in summer or in winter,the thermospheric density of the equator tends to peak 3hafter Dstmin. While for the dayside,the time interval that thermospheric density at the equator approached its maximum is dependent on seasons,and it is shortest for the northern hemisphere winter.4）At dayside,the thermospheric density enhancement at the equator tends to peak after 3h the density of 60 oapproached its maximum,which is independent of seasons.While at nightside of equinoxes and northern hemisphere winter,the thermospheric density at the equator tends to peak before that of high latitudes done,meanwhile the density enhancement maxima of those latitudes were comparable,implying some other heating source working. Although the thermospheric density at the equator tends to respond with 0～2hdelay relative to the response of Dst index during most storms,while in some cases,the density at the equator enhances before the Dst index responded.