Motivations and Contributions - 各類高空短暫發光現象的超低頻與極低頻至甚低頻電波特性

Chapter 1 Introduction

1.3 Motivations and Contributions

The motivation of this dissertation is to systematically examine the electromagnetic signatures of various types of TLEs. The events were observed either in the Taiwan ground TLE campaigns or recorded globally by the ISUAL experiment onboard the FORMOSAT-2 satellite. Sferics that are associated with TLEs in the ULF to ELF/VLF bands form the foundation of this dissertation in the study of the characteristics of the causative discharges. The contributions of this thesis work are the followings:

 An earlier, lightning-damaged Lulin ULF recording system are repaired,

re-calibrated and improved to achieve a broader recording frequency range (up to ~1 kHz). A new Cingcao ELF/VLF recording system is installed with the aims to closely examine the TLE-associated discharges and with a higher time resolution.

Data from both recording systems are used for the cross-checking and the deduction of the electromagnetic signatures of TLE-associated discharges.

 A simple method for signal reconstruction is introduced to properly re-analyze the dataset recorded by the previous notch-filtered Lulin system. From examining the response of the notch-filtered system to the synthetic waveforms from a signal generator and 20 sferics associate with ISUAL sprites, the reconstructed method were found to consistently recovers the phase and the part of amplitude that are most critical for the data analyses. Using the reconstructed sferics, a tighter linear

relationship between the CMCs of the sprite-inducing CGs and the brightness of sprites is found.

 With the renovated Lulin ULF and the newly-constructed Cingcao ELF/VLF recording stations, many associate sferics of the TLEs observed in the ground TLE

campaigns are recorded. Among them, ULF and ELF/VLF sferics of fourteen

negative gigantic jets occurring near the southern Taiwan are recorded and analyzed.

The features in the associate ULF and ELF/VLF sferics could be used to clearly identify the discharging sources are from cloud discharges or gigantic jets

themselves. Furthermore, the variations of features in the sferics from the negative GJs can be divide into three types as “tree-like”, “carrot-like” and “tree-carrot-like"

GJs and correspond well with the stages discerned in the optical images, which include the initiating lightning, the leading jet, the fully-developed jet and the trailing jet.

 The Lulin ULF and Cingcao ELF/VLF systems are important supporting facilities for the space-borne instrument: ISUAL. From June 2009 to June 2012, both systems have operated continuously, and have recorded more than ten thousand sferics associate with ISUAL TLEs among a much larger set of sferics data. With the radio data, the TLE detectability in the ULF/ELF/VLF bands, the polarity distribution, the averaged waveforms, the rise and decay time, and the frequency spectra of the TLE-associated discharges are analyzed.

Chapter 2 Instruments

2.1 Optical TLE Ground Campaigns

2.1.1 Overview

Since 2001, the NCKU (the red square in Figure 2-1) group has carried out yearly summer optical TLE campaigns at various locations in the southern Taiwan [Su et al., 2002; Hsu et al., 2003]. The main observation sites are located in three regions. The first one is in the mountain areas including Ali Mountain of Taiwan near the Central Ridge with an altitude of 2,413 meters and the Lulin observatory with an altitude of 2,862 meters (the yellow cross in Figure 2-1), for they often have clear sky and un-obstructed views to the surrounding areas. When the East Asian rainy season approaches in the spring, these two locations are favorable for the observation of TLEs occurring over Fujian province of China and the nearby coastal areas [Su et al., 2002; Chou et al., 2011].

When the large convective systems and typhoons approach from the south or east of Taiwan during the summers, the Kenting site (the yellow circle in Figure 2-1) at the southern tip of Taiwan has the best view to thunderstorm systems in these two directions [Su et al., 2003]. Finally and thirdly, the roof of the Physics Departmental building at the NCKU campus will be used if there is not enough time to safely get to the

abovementioned remote sites [Su et al., 2002; Huang et al., 2012]. Over the past decade, more than fifteen hundred TLEs have been captured with various types of instruments include as a medium-speed camera (Photron SUPER 10K FASTCAM, 250 FPS at resolution of 512 × 480 pixels), spectrometer, and video-rate low-light-level Watec CCD cameras equipped with different band filters. The observational instruments and strategies have been improved over the time and have helped to increase our knowledge of TLEs.

Two kinds of platforms are currently used by the NCKU group to carry out ground TLE observations.

Figure 2-1. Map for the vicinity of Taiwan. The overlapping markers are: the yellow cross - the ULF station locates at the Lulin observatory, the green triangle - the ELF/VLF station locates at the Cingcao elementary school, the red squares - remote-controllable optical platforms at Kinmen, Ponghu, NCKU, and Taitung (from west to east) and the yellow circle is Kenting at the very southern tip of Taiwan for TLE ground campaign.

2.1.2 Portable Platforms

From over a decade of experience in carrying out the TLE ground campaigns, several manual operated platforms have been developed to assist TLE ground

observations. They usually contain two to six well-aligned cameras on a single tripod and have digital video recorders or laptop computers to record the standard NTSC frame rate (30 frames or 60 image fields per second) video data. Each camera consists of a WATEC NEPTUNE 100 CCD or WATEC 902H ULTIMATE CCD and equips with several lens in different focal lengths ranging from 8 mm (FOV: 46.2° (Horizontal) × 34.7° (Vertical)) to 50 mm (FOV: 7.5° (Horizontal) × 5.6° (Vertical)). The standard configuration for the portable platform contains three cameras all equipped with 12 mm/f1.2 lens (FOV: 30.8°

(Horizontal) × 23.1° (Vertical)) and red, blue, and open filters to feature a relatively wide FOV and observe TLEs through three optical bands. The image fields from the cameras are further passed through GPS synchronized time code inserters to imprint the

millisecond accuracy time codes on the video frames. These portable observational platforms are relatively easy to be relocated and appropriate for TLE observations at mountainous areas like the Lulin observatory (the yellow cross in Figure 2-1) or at seashore sites like Kenting (the yellow circle in Figure 2-1).

2.1.3 Remote-Controllable Platform

Before 2010, several prototypes of the remote-controllable platform had been constructed, tried and abandoned. The first field-deployed platform was installed in 2010 and quickly brought many benefits. For example, the system is operatable through internet, and features remote-controllable power, rotation or tilt of cameras, and adjustable trigger setting and internet to transfer of video data. The outdoor unit is waterproof, so the observation can be carry out until it is start to rain. If the weather condition temporally improves and windows of opportunity still exist, the platform can be put back to operation in less than one minute. The ground TLE campaign becomes much more manageable with the remote-controllable system for it exerts a lesser demand on the human resource. From 2010 to 2012, ~600 TLEs had been captured in our ground campaigns carrying out through the remote-controllable platforms [Huang et al., 2012].

They contributed 73% toward the captured events in 2010 and rose to ~100% in 2012. Up to the present, remote-controllable platforms have been installed at NCKU (2010),

Taitung (2011), Ponghu (2012), and Kinmen (2013); as shown by the red-square markers in Figure 2-1. Each remote platform consists of two cameras, a digital spirit level and compass; hence the elevation and the azimuth angles of the recorded events can be accurately determined. The camera comprises of a WATEC NEPTUNE 100 CCD or WATEC 902H2 ULTIMATE CCD and 12 mm/f1.2 lens (FOV: 30.8° (Horizontal) × 23.1°

(Vertical)) optical lens. One of the cameras is fitted with a red-pass filter (passing band:

570-2700 nm), while the other contains no filter. The video footages are imprinted with GPS time codes, and are stored in the computer hard disc and to be transferred back to NCKU later.

2.2 ISUAL Payload onboard the FORMOSAT-2 Satellite

2.2.1 Overview

The ISUAL onboard the FORMOSAT-2 satellite is the first space-borne experiment dedicating to a long-term survey of TLEs [Chern et al., 2003; Chen et al., 2008] that has a design mission life of 5 years. The FORMOSAT-2 satellite was launched in May 2004 and featured 14 daily re-visiting, low-earth sun-synchronized orbits at 891 kilometer altitude. The ISUAL TLE observation adopts an eastward limb-viewing strategy and the typical distance for an ISUAL event ranges from 2,300 to 4,000 kilometers. In the northern summers and winters, the ISUAL surveyed zone roughly spans the region between 45°S to 25°N and 25°S to 45°N latitudes, respectively [Chen et al., 2008]. After the first five years of the ISUAL experiment (June 2004 to May 2009), ~144 thousand events were recorded with ~15 thousand TLEs were identified in 1,503 operation days.

The average ISUAL event rates are ~96 triggers and ~10 TLEs per day. The health of ISUAL remains excellent and an extended mission has been planned to the end of 2014.

The main sensor packages in ISUAL include an intensified CCD imager (imager for short), a six-channel spectrophotometer (SP), and a dual-band array photometer (AP) which are funded by NSPO in Taiwan and are built through an international collaboration between NCKU group (Taiwan), the University of California at Berkeley (UCB, USA), and Tohoku University. The specifications and features of each sensor will be presented in the following sub-sections.

2.2.2 The ISUAL Imager

Imager is one of the most important instruments in ISUAL that provides the

morphological structure and the temporal evolution of the triggered events. The Imager is fitted with a rotating filter wheel that contains six switchable band-passing filters (N₂1P, 762nm, 630nm, 557.7nm, 427.8nm and a blank). The full width at half maximum (FWHM) and the center wavelength of the band-passing filter are 633.4-750.9 (692.15;

N21P) nm, 759.4-767.1 (763.25) nm, 627.8-634.8 (631.3) nm, 555.7-561.7 (558.7) nm, and 426.4-431 (428.7) nm, respectively. The N₂1P filter is the most common one for the routine TLE surveys [Chen et al., 2008; Hsu et al., 2009], while sometimes other filters will be selected on demand [Kuo et al., 2011; Kuo et al., 2012]. Each image frame has the resolution of 512 pixels × 128 pixels and a FOV of 20° (Horizontal) × 5° (Vertical), as shown in Figure 2-2. On triggering, the Imager will take six (up to eight) consecutive frames of image data and each frame typically has a 29 milliseconds integrating time and one millisecond dead time for data processing.

Figure 2-2. FOVs of the ISUAL Imager (red rectangle), SP (red rectangle) and AP (black rectangle).

At the beginning stage of the ISUAL observations in 2004, the imager was

programmed to keep a frame (img0) aligning with the event trigger and then take another five consecutive image frames (img1-5). Beginning from May 2006, the Imager was re-programmed to keep a pre-trigger frame (img0) with the purpose of recording the

lightning activity prior to the event trigger (img1). To ensure the data volume remain the same, the number of consequent images after the trigger has been fixed at four (img2-5) and the total number of image frames for an event trigger remains at six.

2.2.3 The ISUAL Spectrophotometer (SP)

The function of the ISUAL spectrophotometer (SP) is to provide high-temporal spectra of the ISUAL events, while it has limited spatial resolution. The ISUAL SP contains six identical and independent photometers: SP1 (broad-band filter; 150-280 nm;

far ultraviolet, N₂ LBH band), SP2 (narrow-band filter; 335-341.2 nm, centered at 337 nm; N22P(0,0)), SP3 (narrow-band filter; 387.1-393.6 nm, centered at 391 nm;

N₂⁺1N(0,0)), SP4 (broad-band filter; 658.9-753.4 nm; N₂1P), SP5 (narrow-band filter;

773.6-783.4nm; centered at 777.4 nm; OI emission in lightning), and SP6 (broad-band filter; 250-390 nm; middle ultraviolet, N₂2P and N₂⁺1N). They are all well co-aligned and share the same boresight FOV (20° (Horizontal) × 5° (Vertical)) of the Imager. As shown in Figure 2-2, the radiative emissions in the FOV of SP (red rectangle) are filtered, integrated and sampled at 10 kHz. The rapidly sampling SP provides a much higher temporal resolution data (0.1 ms vs. 29 ms) than that is available from the Imager.

Moreover, the SP also serves as the trigger source to activate the Imager from moving the data in the circular buffer to a temporary memory area. When the increments (the

following value minus the previous value) in SP1, SP2, and SP6 all exceed the threshold settings of 8, 160, and 160 counts, the trigger pulse is issued by the spectrophotometer.

With the rise of the trigger flag, the Imager data (6-8 image fields), the spectrophotometer data (24 ms before and 181 ms after the trigger moment), and the array photometer data (8 ms before and 232 ms after the trigger) are saved and packed along with the other

event information.

2.2.4 The ISUAL Array Photometers (AP)

Array photometers (AP) detect the high-temporal spectral variation along the vertical direction at a slightly higher spatial resolution than those from the ISUAL SP.

The array photometers contain two identical and independent modules which observe in the blue (370-450 nm) and the red (530-650 nm) bands, respectively. Each module comprises of 16 vertically-stacked-anode photometers that equally divide a FOV of 22°

(Horizontal) × 3.6° (Vertical). The individual FOV of a AP channel is 22° (Horizontal) × 0.23° (Vertical); see Figure 2-2. On triggering, the AP samples the incoming photons at a rate of 20 kHz for the period between 8 milliseconds before and 10 milliseconds after the trigger, and then the sampling rate drops down to 2 kHz for 10 to 232 milliseconds after the trigger.

2.3 ULF Magnetic Field Recording System

2.3.1 Overview

Early in the Taiwan ground TLE campaign seasons after 2001, many sprites [Su et

al., 2002; Hsu et al., 2003; Chou et al., 2011], halos, elves [Hsu et al., 2003; Chang et al.,

2011], blue jets [Chou et al., 2011] and gigantic jets [Su et al., 2003; Chou et al., 2011;

Huang et al., 2012] were captured. Soon the NCKU group realized that, besides optical

band images, the radio sferics from the associate lighting and/or the TLEs themselves would be very important for these TLEs to be properly studied. The Lulin ULF magnetic field recording system [Wang et al., 2005] was installed by Dr.Yun-Ching Wang in

August 2003. The system located at Lulin observatory (the yellow cross in Figure 2-1;

longitude: 120.87361°E and latitude: 23.46861°N) in the Yushan National Park of Taiwan to be away from human activities and the noises radiated by the power grid. With the ISUAL payload on FORMOSAT-2 satellite starting to perform the global survey of TLEs in July 2004, the number of recorded TLEs began to accumulate and we started to

wonder about the detectability of the TLE-associated sferics in the Lulin ULF system, the polarity ratios for each type of TLEs, and the signatures of the associate sferics for each type of TLEs. The details on the analyses of the associate sferics of TLEs and what can we learn from the tasks will be presented in Chapters 3, 4 and 5.

2.3.2 Re-build the ULF Magnetic Field Recording System

The core of the Lulin ULF system is a pair of EMI-BF4 magnetic coils; see Figure 2-3. One coil was deployed in the direction parallel to the Earth magnetic field (H;

north-south) and another antenna is in the perpendicular direction (D; east-west) to the Earth’s magnetic field [Wang et al., 2005; Huang et al., 2011; Huang et al., 2012]. The frequency range and the 3-dB frequency corners of the antennas are 0.0001 Hz to 700 Hz, and 0.3 Hz and 500 Hz, respectively; see Figure 2-9. The indoor units of the ULF system used to include a signal modulator, an amplifier, a power supply, and a GPS time clock.

Transmission lines of 100-meter length are used to transfer signals and power to/from the indoor units. Block diagram and photographs of the Lulin ULF system are shown in Figure 2-3 and Figure 2-4. In order to filter out the power grid emissions at 60 Hz and its harmonics above 100 Hz, a signal modulator, with a 1-Hz 2-pole Butterworth-type high pass filter, two 100-Hz 4-pole Butterworth-type low pass filters, and two

Butterworth-type notch-filters (60 Hz and 120 Hz with the quality factor, Q=6), was

inserted into the signal chain. The detailed block diagram and frequency response of the Lulin ULF system are shown in Figure 2-5 and Figure 2-6, respectively. The notch-filter design was assisted by the Tohoku University group (Prof. Yukihiro TAKAHASHI and Dr. Mitsuteru SATO, both now with the Hokkaido University) and manufactured by Dr.

Wang. This signal modulator results in a very slow and tardy output signal, as denoted by the red dashed-line in Figure 2-7. The illustrative sferics that originated from ~3,000 kilometers away was recorded by the Lulin ULF station on 21 May 2011. The original waveform is denoted by the blue line (Figure 2-7), but when it passes through the signal modulator with output represented by the red dashed line in Figure 2-7), its amplitude decreases, the rising edge of the signals is pushed back, and the waveform becomes highly distorted. The un-foreseen signal distortion caused by the signal modulator greatly hinters our effort in finding the associate sferics of the ISUAL recorded TLEs and

lightning. In September 2009, we finally disabled the signal modulator followed the suggestions of Prof. Steve Cummer of Duke University, USA. With the new signal chain, we start to record the sferics directly from the antennas and sampled them at 5 kHz. The electric grid noises are digitally removed during the post-processes of the sferics. This higher sampling rate enables us to use the full frequency band of the antennas (see Figure 2-9) and can be used to investigate lightning discharges with a millisecond time

resolution [Cummer, 2006]. Other side effects of the signal modulator will be discussed further in Chapter 3.

Figure 2-3. Photographs of the Lulin ULF magnetic field recording system: (clockwise from the upper left) the two EMI-BF4 magnetic coils, the wooden housing of the EMI-BF4 magnetic coils, the installation location (the blue circle) of the coils and the control units of the coils, and the indoor units and recording computer.

Figure 2-4. Block diagram of the Lulin ULF recording system. The signal modulator was removed in September 2009. The usage of the amplifier is selectable and the degree of amplification selected will depend on the level of the background noise.

Figure 2-5. Block diagram of the Lulin notch-filtering signal modulator. The signal modulator is located behind the ULF magnetic antennas to filter out the power grid emissions and improve the SNR of the signals.

Figure 2-6. The measured frequency responses of the signal modulator in the old Lulin ULF recording system for north-south (blue line) and east-west (red dashed line) directions.

fc=100Hz

Figure 2-7. Sferic associates with an ISUAL sprite observed on 21 May 2011; the positive CG lightning that emitted the sferic was ~3 Mm away from the Lulin ULF station. This sferic was selected to highlight the effects due to the signal modulator. The un-filtered sferic (the blue line) is fed through the signal modulator (red dashed line). To see effects due to the modulator more clearly, the filtered sferic was scaled up by a factor of four.

2.3.3 The Magnetic Coil Calibration

When a lightning discharge occurred near the antennas, the induction current may sometimes be too strong to endanger the coils. Over a decade after the installation of the Lulin ULF system, the pre-amplifiers inside antennas had been damaged four times due to nearby lightning events. The most severe case occurred on 19 September 2008 and the antennas were out of service for five months. We carried out the repair works in our laboratory when the cost of the replacement units turned out to be too high. However, there is the nagging doubt about how the performance of the repaired units stacks up to those from the factory? Hence, we further carried out an in-lab calibration for the repaired BF4 coils. The calibration works were performed by Mr. M.S. Hsu with the assistance of the author and other fellow students; the calibration works formed the core of M.S. Hsu’s master thesis [Hsu 2009]. In order to generate a uniform magnetic field

needed for the BF4 coils calibration, a 2.77 meters solenoid with 865.6 turns of copper wires was constructed; see Figure 2-8. The diameter of the copper wire and the solenoid are 0.32 and 16.68 centimeters, respectively. The non-uniformity in the generating magnetic field is less than 1% near the axial region. After a series of measurements, the frequency responses of our repaired units (the blue square and the red circle lines) and the factory-shipped units (the black asterisk line) are shown in Figure 2-9; the response curves are closely matched. Thus, it is a good indication that the antennas were well-repaired and ready for service.

Figure 2-8. Photographs taken during the EMI-BF4 coil calibration processes; (left): winding of the solenoid for the generation of the uniform magnetic field to be used in the calibration; (right): the calibration of the repaired antennas.

Figure 2-9. Calibration curves (the blue line with squares and the red line with circles) for the two repaired

在文檔中各類高空短暫發光現象的超低頻與極低頻至甚低頻電波特性 (頁 44-0)