Electrodynamics And Radiative Cooling In The Ionospheric And Thermospheric System During The Geomagnetic Storm

The ionosphere and thermosphere are important components of space weather,covering the atmospheric region from about 60 to 1000 km above the earth surface.During the geomagnetic storm,a large amount of energy is injected from the solar wind and magnetosphere.These energies inject into the polar region of the upper atmosphere in the form of particle precipitation and Joule heating,causing severe disturbances in the composition,density,temperature and circulation of the global thermosphere,along with significant changes in ionospheric parameters.At the same time,the ionosphere and the thermosphere exhibit complex variations during the geomagnetic storm due to the coupling effects of the ionosphere and the thermosphere.Geomagnetic storms and their ionospheric weather effects have always been hot research topics in the field of space physics and ionosphere physics.Many scientific problems still need further study.In this thesis,we focus on the response characteristics of the ionosphere and the thermosphere to geomagnetic storms,and use the ionosphere-thermosphere coupling model to study the response process of the ionosphere-thermosphere system to the main and recovery phases of the geomagnetic storm.Moreover,a machine learning method is used to develop the thermosphere nitric oxide?NO?emission model.The main results of this thesis are as follows:1.Exploration of the mechanism of the nightside ionospheric disturbances during the magnetic stormDuring the main phase of the geomagnetic storm,the prompt penetration electric field?PPEF?and neutral wind variations caused by traveling atmospheric disturbances?TADs?can cause dramatic disturbances in the F₂ peak height of the ionosphere.Previous studies focused on the response of the dayside ionosphere to the geomagnetic storm,while the variations in the nightside ionosphere are not well understood.The Thermosphere-Ionosphere Electrodynamics General Circulation Model?TIEGCM?reproduced the main features of the nightside ionosphere observed by the ionosonde on October 28-29,2003.We isolated the effect of the electric field and the neutral winds on the ionosphere through controlled model experiments.The controlled model experiments demonstrated that the continuous disturbances and long-term uplift of the nightside ionosphere are mainly caused by TADs generated in the polar region,while the PPEF could modulate the nighttime ionospheric variations near the geomagnetic equator.Based on the simulated global neutral winds,it was found that the storm-time TADs had a significant hemispheric asymmetry caused by the hemispheric asymmetry of energy injection in the polar regions.The stronger TADs in the summer hemisphere can cross the equator to the winter hemisphere,reversing the equatorward wind of the winter hemisphere and resulting in a decrease of the ionospheric peak height.When the geomagnetic activity was weak on October 28,the TADs generated in the southern hemisphere?summer hemisphere?can still cross the equator and case wave-like disturbances of nighttime ionospheric peak height in the northern hemisphere.Analysis of the electric potential map during the geomagnetic storm revealed that the nighttime PPEF was westward with the southward BZ.When the BZ turned from southward to northward,the PPEF remained westward before midnight,while the PPEF reversed to eastward after midnight.This caused the ionosphere after midnight to get further uplifted by the combination of the eastward PPEF and the equatorward wind under the northward BZ condition.2.The contribution of NO emission to the recovery of thermosphere density and temperature during geomagnetic stormsEnergy injection in the polar region can dramatically increase the temperature of the thermosphere.The enhanced NO emission takes away a lot of heat during the recovery of the thermosphere.NO emission is likely to be the main contributor causing the overcooling of the thermosphere density.A series of works have been conducted on the basis of the theoretical model TIEGCM to study the recovery process of the thermosphere during the geomagnetic storm recovery phase,but the theoretical model failed to reproduce the observed overcooling of the thermosphere.This may be caused by the underestimation of NO emission in the TIEGCM.We applied different chemical reaction coefficients of N?²D?+O₂ into the TIEGCM.Through model experiments,we found that the simulated NO emission with the temperature-dependent reaction coefficient was more consistent with the observations.In addition,this simulation experiment also reproduced the thermosphere overcooling during the storm recovery phase.Comparing the simulation with different reaction coefficients,it was found that there was no significant difference in the total NO emission between different simulation experiments under the given geomagnetic activity conditions.The simulated total NO emission tends to balance with the energy of Joule heating and particle precipitation.The difference in the evolution of thermosphere temperature was likely caused by the difference in the NO emission per mass.According to the temperature evolution at different altitudes,it was found that the density variation at Challenging Minisatellite Payload?CHAMP?orbit was consistent with the temperature variation of the NO generation source.The temperature overcooling at the NO source had an important contribution to the density overcooling at the CHAMP satellite orbit.The thermosphere density overcooling is not only controlled by the temperature,but also modulated by the neutral wind.3.Improvement of the storm-time NO emission prediction based on the machine learningNO emission is the main cooling source for the thermosphere during the geomagnetic storm.Accurate prediction of NO radiation is important for understanding the thermosphere response to geomagnetic storms.Deep learning has already been successfully applied in space physics,and this technology can perform well in extracting effective features from a large amount of data.Based on the NO emission data during 2002-2015 observed by the Sounding of the Atmosphere using Broadband Emission Radiometry?SABER?satellite,we used a three-dimensional convolutional neural network?CNN?to construct a three-dimensional model of NO emission.To solve the missing data issue,we used a regularized loss function,which can effectively eliminate the influence of missing values on model training.Subsequently,we used a three-dimensional CNN to train a three-dimensional model of NO emission?NOE3D?.NOE3D well predicted the three-dimensional distribution of NO emission.Compared with TIEGCM,the error in NOE3D was smaller,and NOE3D had better correlation with observations.During geomagnetic storms,the prediction of NOE3D was also better than the TIEGCM.The NOE3D generally reproduced the variation characteristics of the SABER NO emission.Thus,machine learning technique has greatly improved the accuracy of storm-time NO emission prediction.