• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

Development of ground-based search-coil magnetometer systems in the polar regions and studies of ULF Pc 1--2 wave propagation in the ionospheric waveguide

Dissertation
Author: Hyomin Kim
Abstract:
Search-coil magnetometers, which measure time-varying magnetic flux density (dB/dt ) and its direction, have been developed for the observations of geomagnetic pulsations in the ultra low frequency (ULF) range (a few mHz to a few Hz). The design, fabrication, and test/calibration have been performed to detect very weak geomagnetic pulsations with approximately a few pT resolution over the frequency range 0 -- 2.5 Hz and 100 μsec timing accuracy, given a system gain of 4.43 V/(nT·Hz) and 12-bit analog-to-digital converter (ADC) with GPS time stamps. These instruments are deployed in the Polar regions forming high latitude networks and conjugate measurement points between the northern and the southern hemispheres. In addition to the development and installation of the magnetometers, this thesis describes analysis of the data from the magnetometer systems, mainly focusing on ULF wave propagation in the ionospheric waveguide (duct) centered around the electron density maximum near the F2 ionization peak. The Antarctic magnetometer array observes well-defined, band-limited ULF Pc 1-2 waves with poleward spectral power attenuation over a very extensive latitudinal coverage from geomagnetic latitudes of -62° to -87° (over the distance of 2920 km). This is a clear indication of the propagation of the electromagnetic ion cyclotron (EMIC) waves in the ionospheric waveguide. This study focuses on the ducting events by comparing spectral power attenuation factors and polarization patterns. A statistical survey of the events reveals that the attenuation factors are between ∼10 to 14 dB/1000 km and the polarization sense changes as the waves are ducted poleward from the low latitude regions. For a detailed event study, a CHAMP satellite conjunction is presented. During the overflight, a transverse and linearly polarized Pc 1 ULF wave was also found over a limited latitudinal extent (-53° to -61° ILAT), which supports the idea that EMIC waves are injected at low latitudes and ducted in the ionosphere. The results show the observations of ducted waves over such an unprecedented latitudinal extent, which have rarely been measured before, and thus provide very important information about ionospheric wave ducting characteristics.

TABLE OF CONTENTS ACKNOWLEDGMENTS iii LI ST OF FI GURES x LI ST OF TABLES xxii ABSTRACT xxiii 1 INTRODUCTION 1 1.1 Solar-Terrestrial Environment 2 1.2 System Engineering of Magnetometer Project for Space Research 4 1.3 Waves in Space Plasma 5 1.3.1 Generation of Waves in Plasma 5 1.3.2 Types of Plasma Waves 7 1.3.3 Polarization of Waves 11 1.3.4 Geomagnetic Pulsations 12 1.4 Wave Propagation in the Ionospheric Waveguide 15 2 DEVELOPMENT OF SEARCH- COI L MAGNETOMETERS 19 2.1 Introduction 19 2.2 Principle of Operation 20 2.3 Technical Details of the UNH ULF Search-Coil Magnetometer 22 2.3.1 Search-Coil Magnetic Sensor 23 2.3.2 Main Analog Electronics 25 2.3.3 Data Acquisition System 26 2.3.4 System Specification 27 vi

2.4 Tests 30 2.4.1 Polarity Test , 30 2.4.2 Frequency Response Test 30 2.4.3 Resolution Test 35 2.4.4 Deviation Tests 37 2.5 Installation in the Polar Regions 38 3 DUCTI NG CHARACTERISTICS OF P C 1 WAVES AT HI GH LATITUDES ON THE GROUND AND IN SPACE 42 3.1 Introduction 42 3.2 Instrumentation and Data 45 3.3 Observations and Interpretations 48 3.3.1 Spectral Power Attenuation 51 3.3.2 Propagation Speed 54 3.3.3 Polarization 55 3.3.4 Spatial Extent of Wave Injection Region 60 3.4 Summary 67 4 STATISTICAL STUDY OF P C 1-2 WAVE PROPAGATION CHARACTERISTICS IN THE HI GH- LATI TUDE IONOSPHERIC WAVEGUIDE 69 4.1 Introduction 69 4.2 Data Analysis 72 4.2.1 Instrumentation 72 4.2.2 Data Survey 73 4.3 Example Events . . . 75 4.3.1 Example 1: Mar. 5, 2007 76 vii

4.3.2 Example 2: Mar. 24, 2007 82 4.3.3 Example 3: Oct. 7, 2007 . 84 4.4 Statistical Results 91 4.4.1 MLT Occurrences 91 4.4.2 Spectral Power Attenuation and Ionospheric Conductivity 93 4.4.3 Polarization Characteristics 98 4.5 Summary 103 5 SUMMARY AND CONCLUSIONS 106 5.1 Development of Search-Coil Magnetometers 108 5.2 Ducting Characteristics of Pc 1 Waves at High Latitudes on the Ground and in Space 109 5.3 Statistical Study of Pc 1-2 Wave Propagation Characteristics in the High- Latitude Ionospheric Waveguide 110 5.4 Concluding Remarks . . . . 112 APPENDI CES 113 APPENDI X A POLARIZATION ANALYSIS TECHNI QUE 114 A.l Polarization of Plane Waves 114 A.2 Polarization Analysis for Geomagnetic Pulsations 117 A.2.1 Theory 117 A.2.2 Implementation 121 APPENDI X B UNH ULF SEARCH- COI L MAGNETOMETER TECHNI CAL DETAILS 124 B.l Search-Coil Magnetic Sensor Spool and Wiring Diagram 124 B.2 Electrical schematic of Preamp 126 B.3 Electrical schematic of Main Analog Electronics 126 viii

B.4 Magnetometer System Cable Connection Diagram 126 B.5 Frequency Response Characterization of Search-Coil Magnetometer System Using Equivalent Circuit Model 129 B.5.1 Search-Coil Magnetic Sensor Model 129 B.5.2 Passive Low-Pass Filter Model 130 B.5.3 Active Low-Pass Filter Model 131 B.5.4 Modeling of the Overall Magnetometer System 133 APPENDI X C THE CHAMP SATELLITE FLUXGATE MAGNETOMETER DATA COOR DINATE TRANSFORMATION 138 BIBLIOGRAPHY 141 IX

LI ST OF FI GURES 1 The Earth's magnetosphere and its current system (with permission from Anthony T.Y. Lui at JHU/APL) 3 -2 A sketch showing one use of the "system" of ground-based magnetometers, which provides spatially and temporally extended observations of the geo magnetic pulsations 5 -3 Charged particles gyrating around a guiding center in a uniform magnetic field 6 -4 Schematic illustration of frozen-in flux. In MHD fluid flow, the magnetic flux is frozen in to the fluid so that the fluid can move with the field lines. Therefore, the total magnetic flux through surface remains unchanged. ... 9 5 Types of waves in a magnetized plasma 11 6 Profiles of neutral atmospheric temperature and ionospheric electron density with the various layers designated (Courtesy of Bhamer, from Wikipedia public domain) 16 7 Conceptual sketch of the propagation of EMIC waves from the equatorial region in the magnetosphere to the ionospheric waveguide 17 1 Schematic of search-coil magnetometer showing principle of operation based on Faraday's Law of induction. When the magnetic flux density through the coil changes, a voltage (emf) is induced in the coil 21 2 Schematic diagram of the UNH ULF search-coil magnetometer system. . . . 23 x

2-3 UNH ULF search-coil magnetometer system 24 2-4 Preamp board of the UNH ULF search-coil magnetic sensor 25 2-5 UNH ULF magnetic sensor with annotations 26 2-6 UNH ULF search-coil magnetometer main analog electronics 27 2-7 PC 104 module stack in the data acquisition system enclosure. The module consists of analog-to-digital converter (ADC), Microcontroller, GPS receiver, VGA graphic card, power supply board, and main analog electronics 28 2-8 (a) Definition of signal polarity of the UNH ULF search-coil magnetic sensor; (b) test signal for sensor coil polarity test; (c) magnetometer output signal in response to the test signal 31 2-9 A solenoid and an RC circuit for polarity test. This device is used externally near the search-coil magnetic sensor to test the sensor polarity 32 2-10 (a) Square wave test signal generated by the main analog electronics of the UNH ULF search-coil magnetometer system; (b) output signal in response to the rising and falling transition of the test signal 32 2-11 Test setup for frequency response and resolution characterization of the search-coil magnetometer system 33 2-12 Test solenoid used for frequency response and resolution tests 33 2-13 Bode plots of frequency response to the voltage output (top) and to the magnetic field (bottom) of the UNH ULF search-coil magnetic sensor (with preamp), showing that the frequency response is 0 - 10 Hz (specified at - 3 dB corner frequencies). Note that DC magnetic field (when / = 0) is not detected by a search-coil magnetic sensor 34 XI

2-14 Bode plots of frequency response to the voltage output (top) and to the magnetic field (bottom) of the UNH ULF search-coil magnetometer system, showing that the frequency response is 0 - 2.5 Hz (specified at —3 dB corner frequencies). Note that DC magnetic field (when / = 0) is not detected by a search-coil magnetometer 35 2-15 Screenshot of spectrum analyzer showing the search-coil magnetometer re sponse to a 1 Hz (left) and a 2 Hz (right) test signal when the peak is 3.0 dB above the noise floor. This peak intensity corresponds to 8.8 pT and 5.5 pT, respectively 36 2-16 Test setup for search-coil magnetic sensor deviation test 37 2-17 Bode plot of frequency response for deviation tests of the 18 preamps (top) and the 15 main analog electronics (bottom) 38 2-18 (a) Polarity setup of the ULF search-coil magnetic sensor; definition of co ordinate systems applied to the sensor installation and orientation of the sensors for ground installation (b) in the northern hemisphere and (c) in the southern hemisphere 40 2-19 UNH ULF search-coil magnetic sensors (X and Y components) installed un der the ground in Ny Alesund, Svalbard 41 xn

3-1 Map of Antarctica showing the ULF search-coil magnetometer array used for this study. The geomagnetically most poleward station, P5, and the other four stations, PI, P2, SPA and HBA are aligned well with the magnetic meridian. This map also shows the southern hemisphere ground track of the magnetic field lines traversed by the CHAMP satellite from 0835 to 0845 UT on March 23 2007, based on data from the Satellite Situation Center Web utility, available at http://sscweb.gsfc.nasa.gov 47 3-2 Stacked 0 - 1 Hz Fourier spectrograms in the X (north-south) (left panels) and Y (east-west) (right panels) components of the search-coil data showing the ULF Pc 1 waves recorded at the Antarctic stations, Halley (HBA), P2, South Pole (SPA), PI, and P5 from 0800 to 1300 UT on March 23, 2007. The stations are located along the magnetic meridian covering from geomagnetic latitudes of —62° to —87° with the lowest latitude station at the bottom of the plot. Spectral power attenuation as the waves are ducted poleward is clearly registered. X axis data from AGO P5 were unavailable during this interval. Note that the time intervals of Figure 3-3, Figure 3-5, and Figure 3-8 are indicated 49 3-3 Temporal and spectral structures of the ULF Pc 1 waves (called "pearls") measured from HBA in a shorter time scale (from 1020 to 1050 UT). Wave modulation with approximately 3 min period is clearly shown. Note that the event during this period is indicated in Figure 3-2 50 xin

3-4 Power spectra (log10 power versus frequency) of the ULF Pc 1 events observed by the Antarctic search-coil array and wave power attenuations (in dB) over the distance from HBA (in km) at four selected frequencies (0.3, 0.4, 0.5, and 0.6 Hz) during the two time periods, (a) 0915 - 0945 UT and (b) 0945 - 1015 UT. The graphs for each time period display the results from both X and Y components 52 3-5 Propagation time delay in the temporal and spectral structures observed from HBA and SPA. This event period is band-pass filtered over 0.3 - 0.5 Hz for the structures to be seen more clearly. The wave arrival time is delayed by ~ 18 sec between the stations, which are separated by ~ 1600 km. The wave packets and the spectral patterns that are compared for the timing are indicated by vertical lines. Note that the event during this period is indicated in Figure 3-2 55 3-6 Polarization ellipticity (e) of the ULF Pc 1 waves observed by the ground array from 0800 to 1300 UT on March 23, 2007 in a plot of frequency versus time. The ellipticity is shown in a color scale with —1 being LH circular polar ization (negative ellipticity) and +1 being RH circular polarization (positive ellipticity). LP is defined as having |e| < 0.2. Note that the time intervals of Figure 3-3, Figure 3-5, and Figure 3-8 are indicated 57 xiv

3-7 Polarization angle (8) of the ULF Pc 1 waves observed by the ground array from 0800 to 1300 UT on March 23, 2007 in a plot of frequency versus time. The angle change ranges between —90° and +90°. The sign represents the direction of the angle with respect to the magnetic meridian in the north- south direction (X component in the magnetometer data) with positive angle being counterclockwise and negative angle being clockwise. Note that the time intervals of Figure 3-3, Figure 3-5, and Figure 3-8 are indicated 59 3-8 Stacked spectrograms of the Y-component of the HBA search-coil data, the three components (bu, b±g, and ftj.^,) of the magnetic field data and the po larization ellipticities (in three ranges: LP, LHP, and RHP) from the fluxgate magnetometer of the CHAMP satellite during the event in this study. Band- limited ULF waves over the frequency band (~ 0.4 - 0.5 Hz) are observed approximately from 0840 to 0843 UT on Mar. 23, 2007. The satellite cross ings over SPA, P2, and HBA are shown with the arrows. Note that the event during this period is indicated in Figure 3-2 61 3-9 Time-series plot of the two perpendicular components, meridional (b±e) and azimuthal (b±v) perturbations from the CHAMP satellite magnetic field data (first and second panels, respectively) and the three components, Bx, By, and Bz, of the search coil magnetometer at Halley Station, Antarctica (third, fourth, and fifth panels, respectively) during a Pc 1 wave event from 0840 to 0844 UT on Mar. 23, 2007 64 xv

4-1 Stacked 0 - 1 Hz Fourier spectrograms in the Y (east-west) components of the search-coil data showing the ULF Pc 1 waves recorded at the Antarctic stations, Halley (HBA), AGO P2, South Pole (SPA), PI and P5 from 0200 to 0600 UT on Mar. 5, 2007 (Example 1) 78 4-2 Power spectra (log10 power versus frequency) of the ULF Pc 1 events observed by the Antarctic search-coil array and the wave power attenuation (in dB) over the distance from HBA (in km) at four selected frequencies (0.55, 0.62, 0.69, and 0.76 Hz) during the two time periods, (a) 0220 - 0240 UT and (b) 0240 - 0300 UT on Mar. 5, 2007. The graphs for each time period display the results from both X and Y components 79 4-3 Polarization ellipticity (e) of the ULF Pc 1 waves observed by the Antarctic search-coil array on Mar. 5, 2007 in a plot of frequency versus time. Each panel is represented in a color scale with —1 being LH circular polarization (negative ellipticity) and +1 being RH circular polarization (positive ellip ticity). LP is defined as having |e| < 0.2 80 4-4 Polarization angle (6) of the ULF Pc 1 waves observed by the Antarctic search-coil array on Mar. 5, 2007 in a plot of frequency versus time. The angle ranges between —90° and +90°. The sign represents the direction of angle with respect to the magnetic meridian in north-south direction (X component) with positive angle being counterclockwise and negative angle being clockwise. The panels on the left and right present positive (8 > 0) and negative (6 < 0) angles, respectively 81 xvi

4-5 Stacked 0 - 1 Hz Fourier spectrograms in the X (north-south) components of the search-coil data showing the ULF Pc 1 waves recorded at the Antarctic stations, Halley (HBA), AGO P2, South Pole (SPA), and PI from 1900 to 2130 UT on Mar. 24, 2007 (Example 2) 83 4-6 Power spectra (log10 power versus frequency) of the ULF Pc 1 events observed by the Antarctic search-coil array and the wave power attenuation (in dB) over the distance from HBA (in km) at four selected frequencies (0.27, 0.31, 0.35, and 0.39 Hz) during the two time periods, (a) 2010 - 2030 UT and (b) 2100 - 2120 UT on Mar. 24, 2007. The graphs for each time period display the results from both X and Y components 84 4-7 Polarization ellipticity (e) of the ULF Pc 1 waves observed by the Antarc tic search-coil array on Mar. 24, 2007 in a plot of frequency versus time. Each panel is represented in a color scale with —1 being LH circular polar ization (negative ellipticity) and +1 being RH circular polarization (positive ellipticity). LP is defined as having |e| < 0.2 85 4-8 Polarization angle (9) of the ULF Pc 1 waves observed by the Antarctic search-coil array on Mar. 24, 2007 in a plot of frequency versus time. The angle ranges between —90° and +90°. The sign represents the direction of angle with respect to the magnetic meridian in north-south direction (X component) with positive angle being counterclockwise and negative angle being clockwise. The panels on the left and right present positive (9 > 0) and negative (9 < 0) angles, respectively 86 xvn

4-9 Stacked 0 - 1 Hz Fourier spectrograms in the X (north-south) components of the search-coil data showing the ULF Pc 1 waves recorded at the Antarctic stations, Halley (HBA), AGO P2, South Pole (SPA), and PI from 0420 to 0640 UT on Oct. 7, 2007 (Example 3) 87 4-10 Power spectra (log10 power versus frequency) of the ULF Pc 1 events observed by the Antarctic search-coil array and the wave power attenuation (in dB) over the distance from HBA (in km) at four selected frequencies (0.30, 0.38, 0.46, and 0.54 Hz) during the two time periods, (a) 0440 - 0500 UT and (b) 0600 - 0620 UT on Oct. 7, 2007. The graphs for each time period display the results from both X and Y components 88 4-11 Polarization ellipticity (e) of the ULF Pc 1 waves observed by the Antarctic search-coil array on Oct. 7, 2007 in a plot of frequency versus time. Each panel is represented in a color scale with —1 being LH circular polarization (negative ellipticity) and +1 being RH circular polarization (positive ellip ticity). LP is defined as having |e| < 0.2 89 4-12 Polarization angle (9) of the ULF Pc 1 waves observed by the Antarctic search-coil array on Oct. 7, 2007 in a plot of frequency versus time. The angle ranges between —90° and +90°. The sign represents the direction of angle with respect to the magnetic meridian in north-south direction (X component) with positive angle being counterclockwise and negative angle being clockwise. The panels on the left and right present positive (9 > 0) and negative (9 < 0) angles, respectively 90 4-13 MLT distribution of ULF Pc 1-2 ducting events in the year 2007 92 xvm

4-14 MLT distribution of ULF Pc 1-2 ducting events - clear poleward propagation versus irregular propagation 94 4-15 Frequency distribution of two different Pc 1-2 wave propagation types - clear poleward propagation versus irregular propagation. Note that the occur rences of each propagation type are normalized 95 4-16 Spectral power attenuation versus frequency under three ionospheric sunlight conditions 98 4-17 Frequency distribution of Pc 1-2 ducting events under three ionospheric sun light conditions. Note that the occurrences in each sunlight condition are normalized 99 4-18 Ellipticity occurrence percentiles over distance using the data from the Antarc tic array. Note that the events under dark ionospheric condition are not included due to insufficient number of events 100 4-19 Plots showing the relationship between spectral power attenuation and po larization ellipse major axis angle at HBA using the data from the Antactic array 101 4-20 Horizontal spatial distributions of polarization patterns for incident wave (left) and transmitted wave on the ground (right). Circles drawn with a broken line and a solid line are demarcations of the polarization sense and the major axis direction, respectively (after Fujita & Tamao (1988)) 102 A-l An example of rotation of a plane wave at z — 0 as a function of time: right- hand circular polarization as seen in a plane perpendicular to the direction of propagation 115 xix

A-2 An example of representation of a polarized field by the convention in elec tromagnetics. This type of representation is called a hodogram. Clockwise (CW) sense and counterclockwise (CCW) sense of the field trace are desig nated as left-hand polarization (LHP) and right-hand polarization (RHP), respectively. Note that the convention in optics applies in the opposite way - CW is RHP and CCW is LHP 116 A-3 An example of a hodogram from a computer-generated signal of multiple frequencies, amplitudes, and phase angles, which mimics a natural signal. . 117 B-l Spool winding diagram showing how the wire is wound around the spool. . 124 B-2 UNH ULF search-coil magnetic sensor wiring diagram 125 B-3 Electrical schematic of the preamp circuit for the search-coil magnetic sensor. 126 B-4 Electrical schematic of the main analog circuit of the search-coil magnetome ter system 127 B-5 ULF search-coil magnetometer system cable connection diagram 128 B-6 Search-coil magnetic sensor model used in this study 130 B-7 Magnitude and phase frequency response of the search-coil magnetic sensor model 131 B-8 First order passive low-pass filter model used in this study 131 B-9 First order active low-pass filter model used in this study 132 B-10 Magnitude and phase frequency response of the preamp model 134 B- l l Magnitude and phase frequency response of the main analog circuit model (Block 1) 134 B-12 Magnitude and phase frequency response of the main analog circuit model (Block 2) 135 xx

B-13 Magnitude and phase frequency response of the main analog circuit model (Block 3) 135 B-14 Magnitude and phase frequency response of the main analog circuit model (Block 4) 136 B-15 Magnitude and phase frequency response of the main analog circuit model (Blocks 1 to 4) 136 B-16 Magnitude and phase frequency response of the overall magnetometer system model without the search-coil magnetic sensor 137 B-17 Magnitude and phase frequency response of the overall magnetometer system model with the search-coil magnetic sensor 137 C-l Schematic diagram showing coordinate transformation from the CHAMP satellite coordinate system to the new coordinate system used in this study. 140 x x i

LI ST OF TABLES 1.1 The classification of ULF geomagnetic pulsations (Jacobs et al. 1964). ... 13 3.1 Geographic and geomagnetic locations of the Antarctic stations used in this study. Geomagnetic coordinates, dipole l val ues, and MLT MN in UT are ob tained from NASA GSFC Modelweb Website, http://modelweb.gsfc.nasa.gov /models/cgm/cgm.html, for epoch 2007, assuming an altitude of 100 km. . 46 xxn

ABSTRACT DE VE L OP ME NT OF GROUND- BAS ED SEARCH- COI L MAGNE T OME T E R S YS TEMS IN T HE P OL AR REGI ONS AND S TUDI ES OF ULF P c 1-2 WAVE P ROP AGAT I ON I N T HE I ONOS P HERI C WAVEGUI DE by Hyomin Kim University of New Hampshire, May, 2010 Search-coil magnetometers, which measure time-varying magnetic flux density (dB/dt) and its direction, have been developed for the observations of geomagnetic pulsations in the ultra low frequency (ULF) range (a few mHz to a few Hz). The design, fabrication, and test/calibration have been performed to detect very weak geomagnetic pulsations with approximately a few pT resolution over the frequency range 0 - 2.5 Hz and 100 ^xsec tim ing accuracy, given a system gain of 4.43 V/(nT-Hz) and 12-bit analog-to-digital converter (ADC) with GPS time stamps. These instruments are deployed in the Polar regions form ing high latitude networks and conjugate measurement points between the northern and the southern hemispheres. In addition to the development and installation of the magne tometers, this thesis describes analysis of the data from the magnetometer systems, mainly focusing on ULF wave propagation in the ionospheric waveguide (duct) centered around the electron density maximum near the F2 ionization peak. The Antarctic magnetometer array observes well-defined, band-limited ULF Pc 1-2 waves with poleward spectral power attenuation over a very extensive latitudinal coverage from geomagnetic latitudes of —62° to —87° (over the distance of 2920 km). This is a clear indication of the propagation of the electromagnetic ion cyclotron (EMIC) waves in the ionospheric waveguide. This study xxiii

focuses on the ducting events by comparing spectral power attenuation factors and polar ization patterns. A statistical survey of the events reveals that the attenuation factors are between ~ 10 to 14 dB/1000 km and the polarization sense changes as the waves are ducted poleward from the low latitude regions. For a detailed event study, a CHAMP satellite con junction is presented. During the overflight, a transverse and linearly polarized Pc 1 ULF wave was also found over a limited latitudinal extent (—53° to —61° ILAT), which supports the idea that EMIC waves are injected at low latitudes and ducted in the ionosphere. The results show the observations of ducted waves over such an unprecedented latitudinal ex tent, which have rarely been measured before, and thus provide very important information about ionospheric wave ducting characteristics. xxiv

CHAPTER 1 INTRODUCTION The Sun and the Earth are closely linked through a stream of charged particles ejected from the Sun called the solar wind and the Sun's magnetic field carried by the solar wind called the interplanetary magnetic field (IMF). This connection is often called the Sun- Earth connection or the solar-terrestrial environment. The physics involved in the space environment has been of interest not only because its complex nature is intriguing but also because it impacts life on Earth in many ways especially during the space era. For example, a large-scale geomagnetic field perturbation (e.g., geomagnetic storm or substorm) can disrupt or damage technologically complex systems such as electric power and telecommunication systems as well as spacecrafts in orbit. This thesis presents a system design of ground-based instruments which measure the geomagnetic field in the ultra low frequency (ULF) range, called search-coil magnetometers, and the study of ULF geomagnetic pulsations using the magnetometer array. Multiple sets of magnetometers have been constructed to form ground-based magnetometer arrays in the polar regions, which enable us to perform large-scale systematic observations of ULF wave propagation in the ionosphere. Various types of waves are generated in space plasma by electron and ion dynamics associated with the geomagnetic field. They play an important role in accelerating radiation belt particles and transporting magnetospheric energy to the ionosphere. Some of the energy 1

Full document contains 170 pages
Abstract: Search-coil magnetometers, which measure time-varying magnetic flux density (dB/dt ) and its direction, have been developed for the observations of geomagnetic pulsations in the ultra low frequency (ULF) range (a few mHz to a few Hz). The design, fabrication, and test/calibration have been performed to detect very weak geomagnetic pulsations with approximately a few pT resolution over the frequency range 0 -- 2.5 Hz and 100 μsec timing accuracy, given a system gain of 4.43 V/(nT·Hz) and 12-bit analog-to-digital converter (ADC) with GPS time stamps. These instruments are deployed in the Polar regions forming high latitude networks and conjugate measurement points between the northern and the southern hemispheres. In addition to the development and installation of the magnetometers, this thesis describes analysis of the data from the magnetometer systems, mainly focusing on ULF wave propagation in the ionospheric waveguide (duct) centered around the electron density maximum near the F2 ionization peak. The Antarctic magnetometer array observes well-defined, band-limited ULF Pc 1-2 waves with poleward spectral power attenuation over a very extensive latitudinal coverage from geomagnetic latitudes of -62° to -87° (over the distance of 2920 km). This is a clear indication of the propagation of the electromagnetic ion cyclotron (EMIC) waves in the ionospheric waveguide. This study focuses on the ducting events by comparing spectral power attenuation factors and polarization patterns. A statistical survey of the events reveals that the attenuation factors are between ∼10 to 14 dB/1000 km and the polarization sense changes as the waves are ducted poleward from the low latitude regions. For a detailed event study, a CHAMP satellite conjunction is presented. During the overflight, a transverse and linearly polarized Pc 1 ULF wave was also found over a limited latitudinal extent (-53° to -61° ILAT), which supports the idea that EMIC waves are injected at low latitudes and ducted in the ionosphere. The results show the observations of ducted waves over such an unprecedented latitudinal extent, which have rarely been measured before, and thus provide very important information about ionospheric wave ducting characteristics.