Estimation of Lightning Location from Single Station Observations of Sferics

Transcription

1 Electronics and Communications in Japan, Part 1, Vol. 90, No. 1, 2007 Translated from Denshi Joho Tsushin Gakkai Ronbunshi, Vol. J89-B, No. 1, January 2006, pp Estimation of Lightning Location from Single Station Observations of Sferics Isamu Nagano, 1 Satoshi Yagitani, 1 Mitsunori Ozaki, 1 Yoichi Nakamura, 1 and Kazutoshi Miyamura 2 1 Graduate School of Natural Science and Technology, Kanazawa University, Kanazawa, Japan 2 PFU Limited, Kahoku, Japan SUMMARY A system for lightning location from single station observations has been developed for localizing lightning (distance and direction) at rather short distances (several hundred kilometers). In this system, a vertical component of the electric field and two horizontal components of the magnetic field of the VLF wave of the sferics are observed. The sferics are usually observed as a pulse train. The first pulse is used for estimation of the direction. The second and subsequent pulses are multiple reflections between the ionosphere and the ground. The differences in arrival time of these reflected waves are determined by the propagation distance between the ground and the ionosphere. The reflection altitude of the ionosphere and the horizontal distance to the lightning location are estimated from the arrival time difference of more than two pulses. In this research, the electromagnetic pulses radiated from the lightning are first analyzed by the full wave method, considering the ionosphere, free space, and the ground, in order to evaluate the location estimation method. The lightning location is estimated by calculating the space discharge sferics waveform at the observation point. Observations were also carried out with the lightning location system for estimation of lightning locations Wiley Periodicals, Inc. Electron Comm Jpn Pt 1, 90(1): 25 34, 2007; Published online in Wiley InterScience ( DOI /ecja Key words: sferics; full wave method; ionosphere; lightning location detection. 1. Introduction Various kinds of natural electromagnetic noise can be observed around the earth. Of these, sferics due to lightning discharge are electromagnetic noise generated in the earth s atmosphere. Sferics are electromagnetic pulses with large amplitude radiated by lightning discharges. Most of these pulses propagate in a waveguide formed by the ground and the ionosphere consisting of an anisotropic medium. Some of the low-frequency components pass through the ionosphere and become whistler modes due to interaction with plasma. The energy propagates along paths close to the geomagnetic field lines of force of the earth s magnetic field and reaches a point on the other half of the sphere geomagnetically conjugate to the point of incidence with the ionosphere [1 4]. In fact, low electromagnetic waves (whistlers) with low-frequency components (several hundred hertz to 10 khz) radiated from lightning have been observed by rockets and scientific satellites at low altitude. Sferics propagating in the waveguide formed by the ionosphere and the ground can propagate over a long distance as tweek discharges. In tweeks, the frequency components near the low-frequency cutoff (1 to 2 khz) of the broadband impulse wave over several kilohertz are observed late in time. Also, several whisker-shaped responses appear at harmonic frequencies corresponding to the cutoff frequency if the waveguide is determined by the reflection altitude of the ionosphere [5 7]. From the theoretical analysis of Miyamura and colleagues [8], it is known that tweek discharges appear at relatively short distances (several hundred kilometers). In the present research, sferics at rather Wiley Periodicals, Inc.

2 short distances are identified by analyzing means of the electromagnetic pulse trains (tweek discharges) arriving as a result of reflection between the ionosphere and the ground. There are two types of evaluation methods for lightning location, those using the multiple locations and those using one location. In those using multiple locations, sferics are measured simultaneously at more than two locations. The lightning location is identified by using the intersection method from the directions of arrival (MDF: magnetic diction finder). The TOA (time-of-arrival) identifies the lightning location from the time difference of arrivals in simultaneous measurement at more than three locations [9 11]. Presently, power companies and related organizations all over the world identify lightning locations in real time by means of these multiple point observation techniques and publicize the results on the Internet. However, in these methods, observation equipment must be placed at more than two locations, so that the facilities become very large and costly. On the other hand, it is possible to identify lightning locations by using information on the sferic pulses contained in the waveform obtained from observation of the sferics at one location [5]. Although localization from one point is not as accurate as localization from multiple points, the system is simpler and less expensive. To date, identification from one location is practically implemented because the complete propagation mechanics of the sferics is undetermined and the signal analysis of observed sferic waveforms is difficult. Against this background, the present authors are engaged in developing a lightning location system based on one-point observation of sferics that is small in size and light in weight. In this paper, the propagation of sferics is analyzed by the full wave method. The sferic waveforms obtained at rather short distances (within several hundred kilometers) are used to investigate whether estimation of the lightning location is possible. A lightning location system using one-point observation of sferics was constructed and lightning locations were identified by using observed sferic waveform data. 2. Method of Estimation of Lightning Location First, we explain the method of estimating the lightning location (distance and direction from the sferic observation point) with one-point observation of sferics Estimation of distance The sferics radiated from a lightning strike propagate as a spherical wave in space and can be decomposed into a Fig. 1. Propagation model of sferics. direct wave (ground wave) which directly propagates along the ground surface and a reflected wave (space wave) which propagates by multiple reflections between the ground and the ionosphere. It is assumed that the reflected wave propagates over multiple paths due to the short distance. If the propagation model of the sferics is as shown in Fig. 1, the times required for propagation of the sferic pulses from the lightning striking point to the observation point for the direct wave and the reflected wave are Here d is the horizontal distance from the strike point to the observation point, c is the speed of light, and n is the number of reflections from the ionosphere. Although the reflections of VLF waves by the ionosphere are different in practice, depending on whether O-mode waves or X-mode waves are involved, so that the reflection altitudes are different, it is assumed that the difference is small, so that the altitude of all reflections by the ionosphere can be taken as h. If sferics propagate according to this propagation model, the sferic pulses observed at the observation point are as shown in Fig. 2. The time interval of pulse arrival expressed in terms of Eq. (1) is as follows for the pulse after n reflections: Here T n is the time interval of the pulse between the (n + 1)-th and n-th reflections. If the reflection altitude h of the ionosphere is known, the propagation distance d is obtained from If there are more than two values of the pulse interval T, or if more than three pulses are confirmed, simultaneous equations can be obtained by substituting the direct wave and (1) (2) (3) 26

3 amplitude values of B x and B y are A x and A y, then the azimuth angle θ in the horizontal plane, that is, the angle of the magnetic field vector seen in the counterclockwise direction, is (8) Fig. 2. Sferic pulses. the pulse intervals of two arbitrary pulses into Eq. (3). Then the propagation distance d and the ionospheric reflection altitude h can be obtained as follows: The direct wave becomes a ground surface wave that does not propagate at the speed of light [12]. In order to eliminate this error, the propagation distance d and the ionospheric reflection altitude h are derived without using T 0, yielding (4) (5) (6) (7) If the direction of the magnetic field vector is found, the direction of propagation of the sferics differs by either +90 or 90. Although there is an ambiguity of 180, the direction of propagation of the sferics can be found from the relationship of the Poynting flux if we know the polarity of the electric field of the direct wave of the sferics. Thus, the direction of arrival of the sferics can be found. Note that the direct wave propagates with linear polarization, while the ionospheric reflections of the second and subsequent pulses change to elliptical polarization from their linear polarization after reflection, so that the magnetic field vector changes in time. Therefore, the direction of arrival of the sferics cannot be found accurately if the second and subsequent pulses are used. 3. Sferic Propagation Analysis by Full Wave Method Next, the effectiveness of the estimation method explained in Section 2 is confirmed by theoretical calculations of the propagation characteristics of the sferics. For the analysis, the spherical full wave method is used [13]. Treating a lightning discharge between a thundercloud and the ground as a propagating transmission line current, the latter is expressed in terms of such parameters as the current magnitude, direction, current length, and current velocity. For analysis, the ground free space ionosphere model is considered, as shown in Fig. 3. The ionosphere is assumed to be a cold plasma consisting of 2.2. Estimation of direction The direction of arrival of sferics is estimated by using the horizontal magnetic field components (east west component B x and north south component B y ) and the vertical electric field component (E z ) in the east west and north south components of the received waveform of the sferics. First, from the vector sum of the two horizontal magnetic field components of the first pulse (direct wave) of the sferics, the direction of the horizontal magnetic field components at the receiving point can be found. If the wave Fig. 3. Calculation model. 27

4 electrons and neutral particles. The plasma is a planar layered anisotropic inhomogeneous medium in which the electron density and the collision frequency of the electrons with neutral particles vary in height. In the full wave method, the effects of the permittivity of the ground and the earth s geomagnetic field vector are taken into account. The lightning discharge takes place within free space near the ground. The coordinate system is chosen in such a way that the south north direction is the y axis (with the north as positive) and the east west direction (with east as positive) is the x axis and the height direction is the z axis for the earth s geomagnetic coordinates. The origin is at the point on the ground plane immediately below the initiation point of the lightning discharge. The lightning discharge current source has an arbitrary time waveform that can be decomposed into frequency components by the Fourier transform. The electromagnetic field at an arbitrary altitude at each frequency can in principle be derived if the electromagnetic field radiated from the source and the electromagnetic field in the ionosphere and on the ground are connected by means of the boundary conditions. The waveform of the lightning sferics is spherical and consequently the electromagnetic field inside the ionosphere cannot be derived analytically. Hence, the spherical wave is represented by a superposition of elementary plane waves and the electromagnetic field intensity in the ionosphere for each elementary plane wave is derived by planar full wave analysis. In planar full wave analysis, the anisotropic inhomogeneous medium (the ionosphere) is divided into homogeneous multilayers and the solutions to Maxwell s equations for plane electromagnetic waves in each layer are connected by the boundary conditions, followed by numerical calculations [14]. The elementary plane waves calculated in this manner are once again combined as a spherical wave by the spherical full wave method. If the inverse Fourier transform with respect to time is applied to the spherical electromagnetic wave derived for each frequency component, the sferic waveform at an arbitrary location can be derived [8]. For the electromagnetic field analysis, there are methods of direct solution by differencing Maxwell s equations, and methods of direct solution of the wave equations with boundary conditions, such as the full wave method in which the inhomogeneous medium is modeled by homogeneous multilayers. An example of the former is the FDTD method. In this method, the analysis domain is flexibly modeled by three-dimensional cells. However, when the propagation analysis for the lightning sferics is performed, the analysis domain becomes so huge (several hundred kilometers) that the number of time steps and the number of spatial cells become enormous due to the stability condition [15]. Further, when the behavior of the electromagnetic field in an inhomogeneous anisotropic dispersive medium such as the ionosphere is considered, differencing becomes complex and many absorbing boundary layers are needed outside the analysis domain. Hence, an extremely large amount of computation memory is needed for the analysis. In contrast, the full wave method has fewer degrees of freedom than FDTD for modeling of the analysis domain, because the latter is divided into many horizontal layered media. However, in the present analysis domain, the ionosphere can be treated as a horizontal layered medium and the effect of the curvature of the earth does not need to be taken into account. Therefore, there is no problem in using the full wave method. In the full wave method, the electromagnetic fields solved in each layer are connected by the boundary conditions at the interface between the layers so that Maxwell s equations are satisfied completely in the analysis domain. Therefore, the complex absorbing boundary layers employed in the FDTD method are not needed. Rigorous electromagnetic field analysis is then possible in an inhomogeneous medium. Further, since the electromagnetic fields can be derived independently at each analysis point in the x y plane [8], numerical calculations limited to arbitrary points are possible without computing the entire domain as in the FDTD method, so that the computational efficiency is superior. 4. Estimation of Lightning Location from Sferic Propagation Waveform By using the spherical full wave method described in Section 3, estimation of the lightning location from an analysis waveform is attempted. The lightning discharge current model is assumed to flow vertically between the thundercloud and the ground with positive polarity, with a discharge initiation altitude of 5 km and a length of 5 km. The parameters of the lightning discharge are listed in Table 1. Here m is the direction vector of the transmission line current flowing in the direction (m x, m y, m z ) t, where t denotes the transpose. A model of the ionosphere is shown in Fig. 4. The altitude profile of the electron density is based on the model Table 1. Parameters for pulse current of a lightning discharge 28

5 Fig. 4. Profiles of electron density and collision frequency versus altitude used in the calculation. Fig. 5. Calculated sferic waveforms. Table 2. Parameters for the full wave calculation at night in the middle latitudes of the northern hemisphere [16]. The collision frequency of electrons and neutral particles is assumed to be proportional to their density in the atmosphere for the calculation [17]. The other parameters used for the spherical full wave method are listed in Table 2. In the spherical full wave method, the electromagnetic field cannot be obtained because the integral of the elementary plane wave diverges near the wave source height. Therefore, the sferic waveform is derived for an altitude of 30 km and then the location of the lightning discharge is estimated. The estimation method is evaluated by comparing the result with the propagation times of the pulse propagating in the model of the ionosphere and ground, as shown in Fig. 1. Figure 5 shows the numerical results for the sferic waveform at an altitude of 30 km at a location of 200 km south of the wave source. The diagrams from top to bottom show the vertical electric field component (E z ), the east west magnetic field component (B x ) and the north south magnetic field component (B y ). It is assumed that the lightning discharge takes place at time t = 0. These waveforms show that the propagation times of the direct wave ( direct ), the first reflection ( first ), and the second reflection ( second ) are read. The first pulse of the magnetic field component (B x ) perpendicular to the direction of propagation (north south) is used to estimate the time of arrival of the direct wave. When the direct wave is reflected by the ionosphere, it becomes elliptically polarized, so that a magnetic field component can be observed in the direction of propagation. Hence, the time of arrival of the reflected wave is derived from the B y pulse. 29

6 5. Lightning Location Detection System Fig. 6. Propagation model of sferic pulses. For comparison, the propagation path corresponding to the analytical model is shown in Fig. 6. Here the altitude of the ionosphere is 85 km, which is the reflection altitude of the plane wave, because the main component of the main pulse of the lightning sferics has a frequency of about 10 khz. Also, it is assumed that the waves propagate at the speed of light. The propagation time read from the analytical waveform and those for each pulse in the propagation model in Fig. 6 are compared in Table 3. When the propagation times for the direct wave are compared, that of the direct wave obtained by full wave analysis is longer than that obtained in the propagation model. This is attributed to the fact that the direct wave becomes a ground surface wave, whose propagation is slower than the speed of light [12]. Also, when estimation of the horizontal distance is attempted from the first and second reflections, the estimated distance from the analytical waveforms is km compared with a true value of 200 km (the error is 13.5%). Also, since the direct wave has only components perpendicular to the direction of propagation in this model, the estimation error with regard to the direction is 0%. Hence, we see that the source location can be estimated within a range of error of about several tens of kilometers. Hence, location estimation is attempted from the observed sferic waveform. Although lightning spherics have wide frequency ranges ranging from several hertz to several megahertz, the authors receiver observes mainly the ELF and VLF bands, in which the waves propagate by multiple reflections between the ionosphere and the ground. When a stepped leader with a weak intensity starts among the clouds and approaches the ground, an upward mobile streamer departs from the ground. At the moment when the two join, large amounts of charge are injected into the stepped leader path, so that a strong main lightning strike (called the feedback strike) is generated. Subsequently, there may occur multiple strikes, in which a second lightning strike is generated, if charges are supplied to the peak of the discharge path [5]. In the present system, the feedback strike with the largest lightning sferic is used as the object of observation. Figure 7 presents a block diagram of the observation system for sferics. In identification of the lightning location, it is necessary to observe the horizontal magnetic field components parallel to the ground and the vertical electric field component normal to it. Hence, in order to observe the magnetic field components in the south north and east west directions, a horizontal cross loop antenna with a diameter of 0.8 m is used. For observation of the vertical electric field, a vertical dipole with a length of 2 m is used. The circuit consists of a preamplifier and a bandpass filter. The observation frequency range is 1 to 14 khz. The A-D converter uses 16-bit quantization and the sampling frequency per channel is khz. Further, the output voltage waveforms obtained from the receiver, including the antenna systems, must be calibrated in order to obtain the electromagnetic waveform of the sferics by taking account of the transfer function of the receiving system. Figure 8 presents the process. The received voltage waveform v o (t) from each antenna is subjected to the fast Fourier transform (FFT) and the frequency Table 3. Propagation times of sferics estimated from the propagation model and the full wave calculation result Fig. 7. Observation system of sferics. 30

7 Fig. 8. Calibration of E and H. spectrum is found. The frequency spectrum obtained by the FFT contains both positive and negative frequency components. The original signal is a real number so that the negative frequency components are redundant and not needed. Hence, the negative frequency components are set to 0. Since the power is reduced to 1/2, the positive frequency components are doubled. Multiplication by the inverse of the transfer function and application of the inverse FFT (IFFT) yield the electromagnetic waveform. 6. Results of Evaluation of Lightning Location A lightning location system (LLS) was used for evaluation of the accuracy of the present system. The LLS is a lightning location identification system operated by power companies and related organizations. In the region targeted for the present study, the sensors were placed at four locations in Hokuriku and Gifu districts. The lightning location was identified by the arrival directions and arrival time differences for the sensors of each sensor. These sensors were synchronized at 60 ns and the identification error was about 2 km [18]. Figure 9 shows the sferic waveform observed at 1 hours, 18 minutes, 30 seconds on November 21, 2002 (JST). The figure shows the vertical components of the electric field (E z ), the east wave magnetic component (B x ), and the south north magnetic component (B y ) from top to bottom. The lightning location estimated from the observed waveform by the present system ( Estimation in Fig. 10) and the LLS identified location ( LLS ) provided by the Japan Meteorological Association are shown in Fig. 10. For estimation of the distance, the peak points ( ) of B y in Fig. 9 are used. Estimation is performed with Eqs. (6) and (7) for the reason that the sferics propagating along the ground surface becomes a surface ground wave with a velocity less than the speed of light [12]. When the lightning location is estimated by the present system, the distance is km and the direction is 49.1 clockwise from magnetic north. Fig. 9. Observed waveforms (November 21, 2002, 01:18:30 JST). Fig. 10. Estimated lightning location compared with LLS (November 21, 2002, 01:18:30 JST). 31

8 The vector estimation error of lightning location by the present system relative to the distance to the location estimated by the LLS is 12.5%. Lightning location estimation was similarly performed for other sferics (about 100 examples) for which the pulse train was clearly confirmed, yielding 17 examples that could be compared with the LLS. For these cases, the locations were estimated with an accuracy in the vicinity of 10%. In the present estimation method, the selection of the peak values of the sferic pulses and the number of pulses significantly affect the estimation results. However, most of the sferic pulses have extremely complex waveforms, not like that shown in Fig. 9 in which the pulses after the direct wave appear at equal intervals. The causes of such complexity include overlap of the sferic pulses from multiple lightning strikes, discharges between the clouds, the effect of the discharge path orientation, differences in the ionospheric reflection altitude, and the distance between the lightning location and the receiving point. Therefore, it is not easy to select the peak values of the sferic pulses. In the future, we plan to develop a method of selecting the optimum peak values and the number of peaks in the sferic pulses for highly accurate lightning location. In addition, the acceptable distance for estimation by the present estimation method must be confirmed by the full wave method. 7. Conclusions In this research, a system for lightning location detection from one-point observations was constructed and its accuracy was evaluated by theoretical calculations and observation. We attempted to estimate the lightning location by the sferic observation at one location by using the spherical full wave method to analyze the sferic pulses. We found that it was possible to estimate the location with an error of about 10%. Sferics generated by lightning at rather short range were observed at one location and identification of the lightning location from the waveform was attempted. Depending on the type of lightning, the waveform was sometimes too broken up to allow identification. However, the location can be estimated with an error of about 10% for types of lightning in which the pulse train is sharp and clear. In the future, based on sferic propagation analysis by the full wave method and the proposed location estimation procedure, we plan to study in detail how the lower ionospheric layer, the ground conductivity, and the lightning discharge parameters affect lightning location estimation. In real observation, many sferic waveforms exhibit complex pulse trains. In such cases it will be necessary to consider a quantitative evaluation method for selection of the peak values of unclear sferic pulses. Further, it will be necessary to theoretically calculate the sferic waveforms at the actual observation altitude (ground) and to compare the theoretical waveforms to the observed ones. We also plan to develop a more accurate, transportable lightning location system. Acknowledgments. The authors thank Mr. T. Komonmae and Mr. T. Takezono, former students at Kanazawa University, who participated in the development of the system. REFERENCES 1. National Polar Research Institute. Science of the South Pole, aurora and upper atmosphere. Corona Press; Hayakawa M. Symphony from space. Corona Press; Maeda K, Kimura I. Electromagnetic wave theory. Ohm Press; Hayashi I. Plasma engineering. Asakura; Sao K. Sferics: The electromagnetic radiation of lightning. Seizando; Hayakawa M, Ohta K, Shimakura S, Baba K. Recent findings on VLF/ELF sferics. J Atmos Terr Phys 1995;57: Baba K, Ota K, Tomomatsu T, Hayakawa M. Frequency dependence of wave characteristics of tweek discharge. IEICE Trans 1991;J74-B-II: Miyamura K, Nagano I, Yagitani S. Full wave calculation of the VLF electromagnetic waveforms radiated from lightning discharge. IEICE Trans 1997;J80-B-II: Cummins KL, Murphy MJ, Bardo EA, Hiscox WL, Pyle RB, Pifer AE. A combined TOA/MDF technology upgrade of the U.S. national lightning detection network. J Geophys Res 1998;103: Cummins KL, Krider EP, Malone MD. The U.S. national lightning detection network and applications of cloud-to-ground lightning data by electric power utilities. IEEE Trans Electromagn Compat 1998;40: Kise W, Ito M, Nomoto K, Kosuge Y, Asano F, Watanabe S. Prediction support method for lightning flash using case-based retrieval by multiple sensor data. IEICE Trans 2000;J83-B: Komonmae H, Nagano I, Yagitani S, Takezono N. Determination of lightning locations by VLF spherics observations with a single station. Tech Rep IEICE 2001;101: Nagano I, Kitagishi Y, Yagitani S, Mambo M, Kimura I. Mapping of VLF intensities in the ionosphere by a diople antenna on the ground. IEICE Trans 1991;J74- B-II:

9 14. Nagano I. Electromagnetic propagation in an inhomogeneous medium. Radio propagation in space by rocket satellite observation. Hoyu Publications; Uno T. Electromagnetic fields and antenna analysis by FDTD method. Corona Press; Rowland HL, Fernsler RF, Bernhardt PA. Breakdown of the neutral atmosphere in the D region due to lightning driven electromagnetic pulses. J Geophys Res 1996;101: Nagano I, Mambo M, Kimura I. Estimation of collision frequency in the upper D and E regions from LF wave by means of a rocket experiment. J Geomag Geoelectr 1982;34: /lls.html AUTHORS (from left to right) Isamu Nagano (member) graduated from the Department of Electrical Engineering, Kanazawa University, in 1968, completed the M.S. program in 1970, and became a research associate in that department. He was appointed a professor in the Department of Electrical and Information Engineering in 1987, and is now a professor in the Graduate School of Natural Science and Technology. In , he was on leave at the Jet Propulsion Laboratory, USA, as an NRC Researcher. He was engaged in the development of electromagnetic field calculation methods in anisotropic inhomogeneous media, methods of measuring the electron density in the D layer by VLF waves, and plasma wave measurement setups for satellites. He holds a D.Eng. degree. He received a Tanakadate Award from the Earth Electromagnetic and Earth Planetary Society in 1987 and a NASA Group Achievement Award in In 2000, he received a Hokkoku bunka Award. He is a member of the Institute of Electrical Engineers of Japan, the Institute of Television Engineers, the Society of Geomagnetism and Earth, Planetary and Space Sciences, and the American Geophysical Union. Satoshi Yagitani (member) graduated from the Department of Electrical and Information Engineering, Kanazawa University, in 1988, completed the M.S. and doctoral programs in 1988 and 1993, and became a research associate in the Department of Electrical and Information Engineering. He is now an associate professor in the Graduate School of Natural Science and Technology. In , he was on leave as a visiting scholar at the University of Minnesota, with support from the Ministry of Education. He has been engaged in research on magnetospheric plasma wave analysis by scientific satellite and computer simulation, in the development of low-frequency wave motion observation equipment for the NOZOMI Mars Expedition Satellite, and in research on position estimation for low-frequency sources. He received a Sangaku Renkei Suishin Ishikawa Award in He holds a D.Eng. degree, and is a member of Society of Geomagnetism and Earth, Planetary and Space Sciences, and the American Geophysical Union. Mitsunori Ozaki graduated from the Department of Information Systems, Kanazawa University, in 2003 and entered the M.S. program. He has been engaged in research on electromagnetic field analysis of radiation from underground current sources. Yoichi Nakamura (member) graduated from the Department of Electrical and Information Engineering, Kanazawa University, in 2002, completed the M.S. program in 2004, and joined NTT DoCoMo. His student research focused on electromagnetic field analysis of sferics in the lower ionosphere. 33

10 AUTHORS (continued) Kazutoshi Miyamura (member) graduated from the Department of Electrical and Information Engineering, Kanazawa University, in 1991, completed the M.S. program in 1993, and joined PFU, Ltd. He enrolled in the doctoral program in 1994, and completed the program in His student research dealt with electromagnetic field calculations of VLF waves in the ionosphere. 34