A study on upper atmospheric joule heating using observations and coupled models and a space weather consequence: Geomagnetically induced currents

Space weather has immediate consequences on today's space borne and ground-based technological infrastructure. In the long term, it, as well, affects the health of the aircrew on aircraft and astronauts on space missions. Thus, it is essential to quantify the extent and limits of space weather effects on Earth and near-Earth space environment. In this work, we study the space weather effects on the upper atmosphere and ground level during geomagnetically disturbed time intervals. Following, this thesis comprises three main parts: 1) investigation and comparison of different methods and data that were separately used and presented for JH in previous studies in the literature, 2) thermospheric neutral density variations that occur in response to the elevated temperatures during geomagnetic storms and substorms, 3) detection and initial quantification of low to mid-latitude geomagnetically induced currents using a magnetotelluric station located at Bozcaada, Çanakkale (in geographic coordinates: 37.5oN, 106oE). JH is the frictional heating that arises from the collisions between the neutrals and ions in the ionospheric levels. Thus, it is crucial to have exact information on both species to be able to calculate JH precisely. Its exact determination depends on the simultaneous knowledge of the ionospheric electric fields, conductances, thermospheric neutral wind and ambient magnetic fields or in a second approach, conductances, ion drifts and neutral wind at the high latitudes. Previously, some studies were also conducted by assuming that the neutral wind contribution is negligible. In our study, we investigate the JH with and without the neutral wind effects. For the calculations not involving the neutral wind, we get use of combinative methods. Electric field data from radars and an MHD model and conductance data from ionospheric electrodynamics (IE) module of an MHD model and an empirical model are combined interchangeably to calculate the JH. The most used indice-based methods are also investigated. Our aim is to present the alternative methods that give the best estimate for the JH rate in the absence of the global observations of the parameters needed for the calculation. Observational electric field data comes from Super Dual Auroral Radar Network (SuperDARN) radars and MHD model output is from the Space Weather Modeling Framework/Block-Adaptive-Tree-Solarwind-Roe-Upwind-Scheme (SWMF/BATSRUS, hereafter will be referred to as BATSRUS) model, whereas the conductances come from BATSRUS-IE and OVATION Prime + International Reference Ionosphere 2012 (IRI2012) models. For the calculations without the neutral wind, we find that with respect to the variations in Akasofu's Epsilon parameter, all combinative methods show a time lag varying between 10 to 20 minutes for the peak JH. However, the proxy method, which is based on the AE index, shows the maximum time lag, which varies between 20 minutes to 1.5 hour. Qualitatively, all methods exhibit similar variations. AE proxies show the highest rate of JH while the combinations with International Reference Ionosphere 2012 (IRI2012) conductances result in the lowest. The differences between the combinations with SuperDARN and BATSRUS electric fields are the smallest when SuperDARN has data in the midnight sector. In the second part of this thesis, we investigate the neutral density estimations from the empirical and physics-based IT models for GEM-CEDAR challenge events. We quantify the amount of errors using the neutral density observations from the Challenging Minisatellite Payload (CHAMP) satellite. We show that coupled ionosphere-thermosphere models and empirical models of the ionosphere give different neutral density climatologies and responses during geomagnetic storms. The storm effects are found as superimposed on the climatology. A performance assessment solely on the neutral density maxima leads to misleading performance assessments. A model overestimating the climatology but underestimating the storm-time response might well end up with a good performance record. Real time estimation of the atmospheric drag heavily relies on estimating the storm-time responses right. Following, we apply several techniques to separate the model climatology and storm-time response of the thermosphere from each other. Baselines for the thermospheric neutral densities are determined using the quiet periods preceding geomagnetic storms for both the CHAMP observations and the models. The data were shifted to baseline levels and effects due to the magnetic storm could be extracted. We obtain increases in skill scores up to 113% for the models, which differ from the quiet time climatology. Physics-based model neutral density variations show similar tendencies with CHAMP data. Physics-based models perform the worst for the low KP events, whereas they perform better for high KP events. Either CME or CIR induced, physics-based model results capture the variations in neutral densities very sufficiently. Empirical models are good in capturing the climatology and the baseline shifts do not affect the storm-time response considerably. However, despite attaining close values to the neutral density peaks, the shapes of the response curves are substantially different. For real time calculations of the satellite drag, we suggest using ensemble approach and background removal methods together to obtain reliable neutral density estimates. Furthermore, using the baseline shift method, we investigate the relationship between Joule heating and storm-time neutral density variations. A simple model using linear regression method is constructed by relating the JH estimations of the physics-based models of the thermosphere with CHAMP satellite neutral density observations. In the last part of our work, we study the magnetic and electric field characteristics of the GICs at mid-latitudes using electric field and magnetic field observations in Turkey during the geomagnetically active intervals. A magnetotelluric station consisting of an electrometer and a magnetometer were set up in Bozcaada, Çanakkale (37.5N, 106E). Several cases that show large electric and magnetic field fluctuations during geomagnetic disturbances are selected. The effects of geomagnetic activity are studied using the time derivatives of horizontal component of the magnetic field and the deviations in the magnetic and electric field components from the quiet background. In magnetic field data, quiet day Sq variations are removed using cubic spline fits. Similarly, the magnitude of the deviations in the electric field are determined by subtracting the background electric field determined by using cubic spline. Corresponding to the strong geomagnetic activity identified using KP and Dst indices, high frequency, strong fluctuations in the magnetic field, its derivatives, and electric field are observed. These fluctuations in horizontal magnetic and electric field are compared with those seen during a magnetically quiet day. The close association between the fluctuations of the time derivatives of the horizontal magnetic field and electric field components are demonstrated. Two types of variations in the electric and magnetic fields corresponding to the different phases of the geomagnetic activity are identified: those corresponding to the initial phase including the sudden commencement and those to the main phase of the geomagnetic storm. The fluctuations in both magnetic field and electric field corresponding to the sudden commencement and the initial phase indicate the effects of magnetopause currents driven by the large solar wind dynamic pressure as associated with the coronal mass ejection (CME) events. Each event shows increased levels of ground level magnetic and electric field fluctuations corresponding to the CME compression at the subsolar magnetopause. High frequency, large fluctuations continue subsequently during the main phase in the presence of the geomagnetic storms. GIC occurrences in our latitudes are shown to be associated with the sudden commencement and main phase of the geomagnetic storm activity. To conclude, in this thesis, we analyse our current capabilities of estimating the thermospheric Joule heating and neutral density variations during geomagnetically disturbed intervals. For the first time, the magnetic and electric field variations during geomagnetically active intervals are determined in the Eurasia region for the detection of GICs. Our study aims to contribute to the worldwide understanding and modelling efforts on the space weather effects on the thermospheric altitudes and the ground level.