Super Dual Auroral Radar Network (SuperDARN) data is a powerful tool for space science research. Traditionally this data has been processed using a routine with known limitations. A large issue preventing the development and implementation of new processing algorithms was the lack of a realistic test dataset. We have implemented a robust data simulator based on physical principles which is presented in Chapter 2. The simulator is able to generate SuperDARN data with realistic statistical fluctuations and known input Doppler velocity and spectral width. Using the simulator to generate a test data set, we was able to test new algorithms for processing SuperDARN data. The algorithms which were tested included the traditional method (FITACF), a new approach using the bisection method (FITEX2), and the Levenberg-Marquardt algorithm for nonlinear curve fitting (LMFIT). FITACF is found to have problems when processing data with high (> 1~km/s) Doppler velocity, and is outperformed by both FITEX2 and LMFIT. LMFIT is found to produce slightly better fitting results than FITEX2, and is thus my recommendation to be the standard SuperDARN data fitting algorithm. The construction of the new midlatitude SuperDARN chain has revealed that nighttime, quiet-time plasma irregularities with low Doppler velocity and spectral width are a very common (> 50% of nights) occurrence. Following on previous work, we have conducted a study of nighttime midlatitude convection using SuperDARN data. First, the data are processed into convection patterns, and the results are presented. The drifts are mainly zonal and westward throughout the night. The plasma drifts also display significant seasonal variability. Additionally, a large latitudinal gradient is observed in the zonal velocity during the winter months. This is attributed to processes in the conjugate hemisphere, and possible causes are discussed. During my graduate studies, we have been part of the development of a software package for enabling and accelerating space science research known as DaViTpy. This software package is completely free and open source. It allows access to several different space science datasets through a single simple interface, without having to write any code for reading data files. It also incorporates several space science models in a single install. The software package represents a paradigm shift in the space science community, and is presented in Appendix A. / Ph. D.

Medium Scale Travelling Ionospheric Disturbances (MSTIDs) are quasi-wavelike structures in ionospheric density that can be sensed using Global Navigational Satellite Service (GNSS) Total Electron Content (TEC) techniques and coherent scatter radars such as the Super Dual Auroral Radar Network (SuperDARN). MSTIDs, especially those observed during quiet times and on the night side, have been known to be driven by electrodynamic instability processes, such as the Perkins instability. In this work, SuperDARN is used in conjunction with GNSS TEC data to investigate MSTIDs during a major geomagnetic storm on September 7-8th, 2017. The interval of this study is in the North American region between 23UT and 3UT, during the peak of the storm, when Kp reached 9. MSTIDs during the interval were investigated by previous studies. However, the roles of electrodynamic instability processes and atmospheric gravity waves (AGWs) in driving the MSTIDs were not determined. GNSS TEC fluctuations associated with the MSTIDs were strong, reaching up to half of background TEC. In SuperDARN, MSTID signatures were observed in power measurements. Meanwhile, SuperDARN line-of-sight (LOS) plasma velocity corresponding to MSTID structures exceeded $pm$500 m/s. This systemic change in the polarity of SuperDARN LOS velocities is indicative of strong polarization electric fields and therefore driving electrodynamic instability processes. This work therefore presents signatures of storm time electrified MSTIDs in mid-latitude North America. / Master of Science / The upper atmosphere contains a region called the ionosphere, where ionized gas called plasma exists. This plasma can be sensed using satellites and ground-based receivers. Specifically, Global Navigational Satellite Service constellations, such as GPS, are good candidates for this technique. This method yields a column density measurement of electrons and is known as GNSS TEC. Most of the time, GNSS TEC is used in a low resolution format, but a high-resolution format is available. This high-resolution GNSS TEC allows for smaller structures in the ionosphere to be investigated. Ionospheric plasma can also be sensed using ground based radar systems, such as the Super Dual Auroral Radar Network (SuperDARN). Combining GNSS TEC and SuperDARN allows for investigation of disturbed structures in the Ionosphere. These structures include wave-like behavior, with time scales under 30 minutes, called Medium Scale Travelling Ionospheric Disturbances (MSTIDs). When these MSTIDs are investigated during times where the Sun is especially active, some unexpected results are found. Most importantly, SuperDARN radars see plasma velocity behave as if it is affected by MSTID structures. This suggests that the buoyancy force which drives the MSTIDs is an electric force instead of a pressure gradient. This behavior has been shown before, but only at night times, specifically when the Sun is not as active. Therefore, this work presents a new kind of MSTIDs.

Total electron content (TEC) data measured from ground-based GPS receivers is compared to HF backscatter from ionospheric irregularities obtained by Super Dual Auroral Radar Network (SuperDARN) radars.  We present the first observations of a recurrent region of anomalous enhanced TEC at mid-latitudes over North America and attempt to characterize its frequency of occurrence.  Next, we examine the relationship of convection electric fields to the formation of a polar cap tongue of ionization (TOI) from mid-latitude plumes of storm enhanced density (SED) during a geomagnetic storm on 26 September 2011.  A channel of high density F region plasma was transported from the dayside ionosphere and into the polar cap by enhanced convection electric fields extending to mid-latitudes.  After the solar wind IMF conditions quieted and the dayside convection electric fields retreated to higher latitudes, an SED was observed extending to, but not entering, the dayside cusp region.  The source mechanism (enhanced electric fields) previously drawing the plasma from mid-latitudes and into the polar cap was no longer active, resulting in a fossil feature which persisted for several hours as it elongated in magnetic local time. Finally, we discuss ground surface effects on the HF backscatter observed by four SuperDARN radars. Monthly ground scatter occurrence rates are calculated for comparison with Arctic sea ice boundaries derived from satellite observations, showing reduced backscatter from regions covered by ice. / Master of Science

Ionospheric convection dynamics is an important window for understanding the coupling of the solar wind and interplanetary magnetic field to the Earth's ionosphere and upper atmosphere. In this study, we use measurements of ionospheric convection made by the SuperDARN radars to investigate the role of interhemispheric magnetic conjugacy in magnetosphere-ionosphere coupling and study the large-scale interactions between the magnetosphere and ionosphere. SuperDARN radars cover large geographic regions in both hemispheres and have a dataset spanning more than a decade, making them ideal for such studies. We begin in chapter 2 with an analysis of the degree of interhemispheric conjugacy exhibited in a Sub-Auroral Polarization Stream (SAPS). We present simultaneous observations of a SAPS event in both hemispheres made by mid-latitude SuperDARN radars with magnetically conjugate fields-of-view. An interhemispheric comparison of the characteristics of the SAPS channel reveals that the channel was conjugate in terms of potential variations across the channel even though substantial differences in latitudinal width and electric fields were observed in the channel. In chapter 3, we use interhemispheric SuperDARN observations of high latitude ionospheric convection in the noon-dusk sector to investigate the effects of IMF By penetrating into the closed magnetic field line region. The observations support the existence of an IMF By associated interhemispheric potential difference and field-aligned current system resulting in the generation of the interhemispheric asymmetries in ionospheric convection. Four events are analyzed in this study and the strength of interhemispheric currents associated with IMF By are estimated. Moreover, the strength of the interhemispheric currents is found to depend on the magnitude of IMF By, proximity of the currents to open-closed field line boundary, ionospheric conductivity and magnetic local time. In chapter 4, we use data from the mid-latitude SuperDARN radars between Jan-2011 and Aug-2012 to compile a database of SAPS events spanning about six hours in magnetic local time. The event database is used to analyze the average spatial variations in the occurrence rate and velocities of the SAPS channel under different geomagnetic conditions. An empirical model based on Dst-index is then developed to estimate the occurrence rate of SAPS at a given latitude and magnetic local time. / Ph. D.

First time ionospheric refractive index values have been determined at mid latitudes using frequency switched SuperDARN plasma convection velocity estimates. Previous works have found a disparity between high latitude SuperDARN plasma convection velocities and those made by other devices. It was noted that the scattering volume’s refractive index was being neglected when estimating plasma convection velocities, meaning a correction factor was needed in order to more accurately reflect other measurements. Later work proposed a solution which implemented frequency switching in SuperDARN radars and determined a single correction factor based off of many years of data. We present case study driven research which applies the principles of these previous works to mid latitudes in an attempt to determine the refractive effect in mid latitude SuperDARN plasma convection velocity data by examining frequency switched quiet time ionospheric scatter. It was found that the 1/2 hop ionospheric scatter exhibited little to no measurable refractive effect (n ∼ 1), while the 11/2 hop ionospheric scatter tended to exhibit measurable refractive effects (n ∼ 0.7). This is then expanded to a storm-time 1/2 hop ionospheric scatter case study. It was again found that the refractive effects were nearly negligible (n ∼ 1), indicating that the 1/2 hop plasma convection velocities reported by mid latitude SuperDARN radars only require a very small correction factor, if any at all. / Master of Science

The Super Dual Auroral Radar Network (SuperDARN) is a global network of coherent high frequency (HF) radars located in the polar, high- and mid-latitudes of both the Northern and Southern hemispheres. This thesis deals with close-range SuperDARN echoes (oblique HF backscatter from the lower part of the ionosphere). The aim of this thesis is to shed light on the origin of these echoes. Previous studies have been content to propose explanations for the origin of these echoes without thorough checking of the proposed mechanisms against constraints available from various radars and other important information. For the purpose of clarifying the situation, a chain of SuperDARN radars in the Northern and Southern hemispheres and several years of daily statistics have been used. This has allowed for several findings. Notably, the close-range SuperDARN echoes show diurnal and seasonal variations and their properties with respect to signal-to-noise-ratio, Doppler velocity and Doppler width vary. Three distinct populations of close-range HF backscatter have been established: (1) a morning population (0400-0700 LT), (2) a midday summer population (0800-1300 LT) and (3) a pre-midnight (2100-2300 LT) population. The morning population is associated with meteor trails which are observed to be peaking near local dawn as expected, and already suggested by previous research. High latitude SuperDARN radars also had echoes (pre-midnight population) with higher Doppler velocities than the others yet the Doppler velocities are smaller than that expected from auroral E region echoes. Given the time and location of this population of echoes, it has been concluded that they are a special class of high latitude E region echoes at high aspect angle which have been termed ``high aspect irregularity region" echoes in the past. Lastly, the midday summer population was found to be too high for polar mesosphere summer echoes and too early for plasma instabilities. It is proposed that these SuperDARN echoes are produced either from contribution from meteors trails or by neutral turbulence which is suspected (from other work) to be present near 100 km. The properties of the midday summer population resembles those of meteor trails as they have the same power, and the same altitude and have high summer occurrence as expected for meteors. Their late morning occurrence could be due to particular look direction of individual radars which may change the occurrence statistics in the presence of meteor showers. With respect to neutral turbulence, the drift of the midday summer population is similar to that of neutral wind.

The Sub-Auroral Polarization Stream (SAPS) is a narrow, intense and persistent westward (sunward) ionospheric convection flow channel observed equatorward of the auroral electron precipitation boundary, predominantly on the nightside. Previous studies have identified disturbed-time SAPS to be a geomagnetic activity dependent phenomenon, which exhibits average pre-midnight and post-midnight velocities of 1000 m/s and 400 m/s respectively. Numerous studies have reported even narrower and more intense westward plasma flows called SAIDs to be embedded within SAPS channels, especially during substorm recovery phases. Quiet-time SAPS studies, although relatively few, have shown these SAPS to be associated with much weaker velocities and to be influenced by substorm intensifications. However, these studies have been limited in their ability to make simultaneous measurements of SAPS flow velocities over many hours of MLT. The recent expansion of SuperDARN radars to middle latitudes facilitates unprecedented large-scale observations of SAPS over 10 hours of MLT with high temporal and spatial resolution. In this thesis, we first examine the spatial and temporal dynamics of one quiet-time SAPS event, using the mid-latitude SuperDARN radars. The SAPS was identified as elevated flows lying equatorward of the auroral electron precipitation boundary specified by the NOAA POES satellites. We demonstrate the L-shell fitting technique to analyze the dynamics in the strength and direction of the two-dimensional SAPS flow velocities at three separate magnetic longitudes. The quiet-time SAPS event thus examined lasted for over 4 hours in UT and extended over 10 hours of MLT, as is commonly observed for disturbed-time SAPS. However, the decrease in SAPS peak latitudes and peak velocities with MLT and MLon respectively, observed for disturbed-time SAPS, was not observed for this event. We also find the dynamics of the enhancements in the quiet-time SAPS peak velocities, to correlate well with that of substorm intensifications identified using the CARISMA magnetometers. We then identify numerous such conjunctions between quiet-time SAPS and substorms to infer that quiettime SAPS were almost always associated with substorms and their durations were well bounded by that of the substorms for most cases. Next, we extend this analysis over to a statistical study of quiet-time and disturbed-time SAPS events identified over two years. From this study, we find quiet-time SAPS to occur between the relatively narrow nightside MLT range of [18, 4] whereas disturbed-time SAPS was found to occur between the broader nightside MLT range of [15, 5]. We also find the occurrence percentage of quiet-time SAPS to be at its highest between the narrow latitude range of 60-66⁰, while disturbed-time SAPS was observed to occur within a much broader latitude range of 55-66⁰. Finally, the calibration and validation of a control card used in the SuperDARN radar transmitters, is discussed. / Master of Science

Ultra-low frequency (ULF; 1 mHz - 1 Hz) waves are known to play an important role in the transfer of energy from the solar wind to Earth's magnetosphere and ionosphere. The Super Dual Auroral Radar Network (SuperDARN) is an international network consisting of 35 low-power high frequency (HF: 3-30 MHz) coherent scatter radars at middle to polar latitudes that look into Earth's upper atmosphere and ionosphere. In this study, we use Doppler velocity measurements obtained by the SuperDARN radars and coordinated spacecraft observations to investigate the occurrence statistics and driving mechanisms of ionospheric ULF waves. We begin in Chapter 2 with a case study of Pi2 pulsations which are short-duration (5-15 min) damped geomagnetic field oscillations with periods of 40-150 s. Simultaneous observations of Pi2 pulsations from THEMIS spacecraft, midlatitude SuperDARN radars, and ground magnetometers, together with analysis of their longitudinal polarization pattern and azimuthal phase propagation, confirmed that they are consistent with a plasmaspheric virtual resonance excited by a longitudinally localized source near midnight. In Chapter 3, to further investigate the overall occurrence of ionospheric ULF signatures, a comprehensive statistical study was conducted using an automated detection algorithm to identify ionospheric signatures of Pc3-4 and Pc5 waves over 7 years of high time resolution SuperDARN radar data. Specifically, we have investigated their spatial occurrence, frequency characteristics, seasonal factors, and dependence on solar wind and geomagnetic conditions. We note two particular findings: (i) an internal wave-particle interaction source is most likely responsible for Pc4 waves at high latitudes in the duskside ionosphere; and, (ii) a source associated with magnetotail dynamics during active geomagnetic times is suggested for Pc3-4/Pi2 waves at midlatitudes in the nightside ionosphere. These findings are further expanded in Chapter 4 which investigates the hypothesis that internal wave-particle interactions are an important source for generation of these waves. A case study of long-lasting poloidal waves was conducted using coordinated observations with the GOES and THEMIS satellites to examine the generation and propagation of waves observed in the dayside ionosphere by multiple SuperDARN radars. The source of wave excitation is suggested to be bump-on-tail ion distributions at 1-3 keV. Collectively, these research findings provide better constraints on where and when ionospheric ULF waves occur, their source mechanisms, and how they might affect magnetospheric and ionospheric dynamics. / Doctor of Philosophy / Earth’s magnetic field, approximates that of a bar magnet. It is an effective barrier to charged particles originating directly from the Sun and protects us against harmful space weather influences. The geomagnetic field lines can oscillate in ultra-low frequencies (ULF: 1 mHz - 1 Hz). These natural oscillations of closed magnetic field lines, analogous to vibrations on a stretched string, are also called geomagnetic pulsations or ULF waves. The interaction between matter and electromagnetic fields emitted from the Sun and the Earth’s outer atmosphere and magnetic field form a magnetic shield named the Earth’s magnetosphere. ULF waves play a key role in the transfer of energy from outside this shield to regions inside it, including Earth’s upper atmosphere and ionosphere (a region extending from about 60 km to 1000 km above the Earth’s surface). In this study, we use Doppler velocity measurements obtained by the Super Dual Auroral Radar Network (SuperDARN) radars and coordinated spacecraft observations to investigate the occurrence statistics and driving mechanisms of ionospheric ULF waves. We begin in Chapter 2 with an event study of a type of irregular pulsations (Pi2) which are short-duration (5-15 min) damped geomagnetic field oscillations with periods of 40-150 s. Simultaneous observations of Pi2 pulsations from NASA THEMIS spacecraft, midlatitude SuperDARN radars, and ground magnetometers, together with further analysis of wave spectra and propagation, confirmed their driving mechanism as a type of magnetic resonance, analogous to striking a bell. In Chapter 3, to further investigate the overall occurrence of ionospheric ULF signatures, a statistical study was conducted using an automated detection algorithm to identify ionospheric signatures of ULF waves over 7 years of high time resolution SuperDARN radar data. Specifically, we have investigated their spatial occurrence, frequency characteristics, seasonal factors, and dependence on solar and geomagnetic activity. We obtained findings regarding the different driving sources of waves observed in different regions. The findings are further expanded in Chapter 4 which investigates the generation of waves through energy exchange with charged particles. A case study of long-lasting (2-3 days) waves was conducted using coordinated observations with the GOES and THEMIS satellites to examine the generation and propagation of waves observed in the dayside ionosphere by multiple SuperDARN radars. The source of wave excitation is suggested to be unstable particle distributions in the magnetosphere. Collectively, these research findings provide better constraints on where and when ULF waves occur, their source mechanisms, and how they affect dynamics in the geospace environment.

The total solar eclipse over the continental United States on August 21, 2017 offered a unique opportunity to study the dependence of the ionospheric density and morphology on incident solar radiation. Unique responses may be witnessed during eclipses, including changes in radio frequency (RF) propagation at high frequency (HF). Such changes in RF propagation were observed by the Super Dual Auroral Radar Network (SuperDARN) radars in Christmas Valley, Oregon and in Fort Hays, Kansas during the 2017 eclipse. At each site, the westward looking radar observed an increase in slant range of the backscattered signal during the eclipse onset followed by a decrease after totality. In order to investigate the underlying processes governing the ionospheric response to the eclipse, we employ the HF propagation toolbox (PHaRLAP), created by Dr. Manuel Cervera, to simulate SuperDARN data for different models of the eclipsed ionosphere. Thus, by invoking different hypotheses and comparing simulated results to SuperDARN measurements, we can study the underlying processes governing the ionosphere and improve our model of the ionospheric responses to an eclipse. This thesis presents three studies using this method: identification of the cause of the increase in slant range observed by SuperDARN during the eclipse; evaluation of different eclipse obscuration models; and quantification of the effect of the neutral wind velocity on the simulated eclipse data. / Master of Science / The ionosphere is the charged layer of the upper atmosphere, which is generated and sustained by sunlight ionizing neutral particles to form a plasma. In the absence of sunlight, ions and electrons can recombine into neutral particles. The total solar eclipse over the continental United States on August 21, 2017 offered a unique opportunity to study the dependence of the ionospheric density and plasma motion on sunlight as the period of the eclipse is much shorter than night. Observations of the ionosphere during past eclipses indicate that unique ionospheric behavior may be witnessed during eclipses, including changes in radio wave propagation for radio waves in the high frequency (HF) regime. Such changes in radio propagation were observed by the Super Dual Auroral Radar Network (SuperDARN) ionospheric HF radars in Christmas Valley, Oregon and in Fort Hays, Kansas during the 2017 eclipse. At each site, the westward looking radar observed an increase in distance that the radio waves traveled before they were reflected back to the radar during the eclipse onset followed by a decrease in this distance after totality. In order to investigate the mechanisms that produce these observed effects, we employed the HF propagation toolbox (PHaRLAP), created by Dr. Manuel Cervera, to simulate radio propagation and generate simulated SuperDARN data for different models of the eclipsed ionosphere. Thus, different models can be tested by comparing simulated data to measured data. Hence, we can study the underlying processes governing the ionosphere and improve our model of the ionospheric responses to an eclipse. This thesis presents three studies using this method to: identify the cause of the increase in the distance radio waves traveled during the eclipse; evaluate different models of change in eclipse magnitude over time; and investigate the effect of the neutral wind velocity on the simulated eclipse data.

Software defined radio (SDR) is a rapidly developing field enabled by continuing improvements in digital electronics. Software defined radio has been used extensively in communication systems due to its flexibility and cost effectiveness. Recently, SDR has been incorporated into radar systems, particularly for ionospheric research. This study investigated the benefits and design of a high frequency (HF) SDR receiver for the next generation of Super Dual Auroral Network (SuperDARN) radars. This work analyzed digital beamforming and waveform design approaches that would be enabled by the adoption of a SDR based radar design and found that these techniques could improve the performance of SuperDARN radars. This work also developed a prototype receiver to demonstrate the feasibility of a SDR based SuperDARN radar. The hardware selection for this receiver leveraged low-cost commercial off-the-shelf software defined radios and amplifier designs supplemented by custom filters. The software implementation utilized GNU Radio, an open source SDR and signal processing platform, to process and record receiver data. A prototype was successfully designed and constructed using the Red Pitaya software defined radio. This prototype included a 4 channel receiver which was evaluated in the laboratory setting and tested at the Blackstone, Virginia radar site. A comparison of results from the prototype receiver and the existing hardware showed promise for the use of this platform in future ionospheric research. / M.S. / Software defined radio (SDR) is a rapidly developing field which uses software to perform radio signal processing traditionally accomplished by hardware components. Software defined radio has been used extensively in communication systems due to its flexibility and cost effectiveness. Recently, SDR has been incorporated into radar systems, particularly for space science research. This study investigated the benefits and design of a SDR receiver for the next generation of Super Dual Auroral Network (SuperDARN) radars. This work analyzed radar design approaches that would be enabled by the adoption of a SDR framework and found techniques that could improve the performance of SuperDARN radars. This work also developed a prototype receiver using low-cost commercial off-the-shelf software defined radios to demonstrate the feasibility of a SDR based SuperDARN radar. A prototype was successfully designed and constructed using the Red Pitaya software defined radio. This prototype was evaluated in the laboratory setting and tested at the Blackstone, Virginia radar site. A comparison of results from the prototype receiver and the existing hardware showed promise for the use of this platform in future space science research.