Ground-Based Observation,Modeling,and Response To Substorms Of Equatorial Spread F

Equatorial spread F(ESF)is the general term used to describe the manifestation of electron density irregularities in the low-latitude ionosphere.It was first observed as diffuse radio echoes from the ionospheric F layer in the equatorial region.Since then,it has been studied theoretically and experimentally using different types of techniques and has become a hot spot in the research of space weather.ESF may cause the phase or amplitude scintillation of radio signal,which significantly deteriorates the quality of communication as well as navigation system.Hence,it is of great scientific significance to study and model ESF.In this paper,using ground-based coherent scattering radar and digisonde measurements,we compared the statistical characteristics of ESF at America and Asia region,we also analyzed the generation of post-midnight ESF;In addition,a simple and effective ESF prediction model based on machine learning algorithm is proposed;The ESF response characteristics of ESF and vertical drifts to substorms were analyzed.The main work and innovative points of the paper are as follows:1.Using observations of coherent scatter radar and digisonde in Peru and Hainan,China,the statistical characteristics of ESF occurrence were compared,including solar activity variation,seasonal variation and local time variation.The evolution characteristics of post-midnight ESF during geomagnetic quiet days are analyzed.The main conclusions are as follows:(1)There was a significant seasonal difference in the occurrence of ESF between the two sectors.In America sector,the occurrence was higher in the spring equinox than in the winter solstice,while the highest occurrence occurred in the winter in the Asian sector.(2)The local temporal variation of ESF was different.The peak of ESF occurred at 2100 LT in the n sector and earlier in Asia,with the highest incidence at 2000 LT.(3)Two post-midnight ESF cases were reported.The unusual spread-F pattern presented distinct feature from those observed at post-sunset,with spread F appearing first at the higher-frequency edge of the F-layer trace to further develop into a mixed spread F.2.During geomagnetic disturbance,the ionospheric plasma density,electric field and current,and atmospheric wind change in complex ways.Hence,the growth of ionospheric ESF during perturbation period is very complicated.In this paper,the response characteristics of ESF to substorms in the American sector were analyzed using the Incoherent scatter radar(ISR)and coherent scattering radar(JULIA)data from the Jicamarca Radio Observatory over the past 18 years(2000-2017).The main findings are shown below:(1)There is a strong correlation between the occurrence of ESF and substorms,and the impacts may last for more than 12 h.The perturbation vertical drift follows the same pattern as the ESF occurrence rate;The correlation coefficient between the AE index and the occurrence rate of ESF is greatest when the lag time is 3–3.5 h,indicating that the effect of disturbance dynamo electric fields(DDEF)may be more substantial than that of prompt penetration electric fields(PPEF)during a substorm.(2)The response of equatorial spread F to substorm shows local time dependence: For the post-sunset(1800–2400 LT)sector,the occurrence rate of equatorial spread F increased to a maximum at about 0.5 h and decreased to a minimum value around 3.0 h after the substorm;while for post-midnight(0000–0600 LT)sector the occurrence rate showed an opposite response.The local time dependence is due to the opposite perturbation drift caused by PPEF and DDEF.3.The influence of ESF on radio wave signals limits the application of communication and satellite navigation systems.Modeling and forecasting ESF is of great engineering value.In this paper,using the JULIA data of Jicamarca,an ESF model is constructed.Taking into account the variations of ESF with season,local time,altitude,solar activity level and geomagnetic level,the results are as follow:(1)Proposed model based on XGboost can capture the nonlinear characteristics of ESF.The accuracy of the verification set is 90%,and the accuracy of the test set is about 81%.The performance of the test set is obviously superior to the other three models with good generalization ability.(2)The model can predict the morphology of ESF,the accuracy of ESF prediction is about 81%,and the accuracy of daily occurrence rate is more than 90%.(3)The accuracy of the model is affected by local time,height and F107.By using Shapley technology,the physical characteristics of ESF can be deeply understood.We also found that local time,together with altitude,H’ F and solar activity levels,are several factors that have a great impact on ESF.