RADAR is the acronym for Radio Detection And Ranging. The. radar invention has its roots in the pioneering research during

Transcription

1 1 1.1 Radar General Introduction RADAR is the acronym for Radio Detection And Ranging. The radar invention has its roots in the pioneering research during nineteen twenties by Sir Edward Victor Appleton in UK and Breit and Tuve in USA (1925) on the detection of ionization layers in the upper atmosphere. The radar works on the principle that when a pulse of electromagnetic waves is transmitted towards a remotely located object, a fraction of the pulse energy is returned through either reflection or scattering, providing information on the object. The time delay with reference to the transmitted pulse and the received signal power provide respectively the range and the radar scattering crosssection of the target detected. These classes of radars are known as pulse radars. In case the target is in motion when detected, the returned signal is Doppler shifted from the transmitted frequency and the measurement of the Doppler shift provides the line-of-sight velocity of the target. The radars possessing this capability are referred to as pulse Doppler radars. In addition to the above, if the location of the target is to be uniquely determined, it is necessary to know its angular position as well. The radars possessing this capability employ large antennas of either phased array or dish type to generate narrow beams for transmission and reception. Two major radars of this kind used for scientific research are the phased array radar of Jicamarca and dish antenna radar of Arecibo. Two important parameters that characterize the capability of radar are its sensitivity

2 2 and resolution for target detection. The sensitivity is determined by the peak power-aperture product and the resolution by the pulse volume which depends on the pulse length and the radar beam width. There are several variants to the above type of pulse radars that have been developed with varying degrees of complexity to meet the demands of application in various fields, eg: Over The Horizon (OTH) radars, Imaging radar, Synthetic Aperture Radar (SAR), Doppler Weather Radar (DWR), Precipitation radar, Atmospheric radars are given by (1), (2), (3). 1.2 Atmospheric Radars Radar can be employed, in addition to the detection and characterization of hard targets, to probe the soft or distributed targets such as the earth s atmosphere. The atmospheric radars of interest to the current study are known as clear air radars and they operate typically in the VHF ( MHz) and UHF (300 MHz 3GHz) bands are given by (4). The turbulent fluctuations in the refractive index of the atmosphere serve as a target for these radars. There is another class of radars known as weather radars which serve to observe the weather systems and they operate in the SHF band (3-30 GHz) by (5). A major advance has been made in the radar probing of the atmosphere with the realization in early seventies, through the pioneering work by (6), that it is possible to explore the

3 3 entire Mesosphere-Stratosphere-Troposphere (MST) domain by means of a high power VHF backscatter operating ideally around 50 MHz. It led to the concept of MST radar and this class of radars has come to dominate the atmospheric radar scene over the past few decades by (7), (8), (9). MST radar is a high power phase coherent radar operating typically around 50 MHz with an average power-aperture product exceeding about 5x10 7 Wm 2. Radars operating at higher frequencies or having smaller power-aperture products are termed as ST (Stratosphere-Troposphere) radars. In arriving at an optimum radar frequency for MST application, the main considerations are the frequency dependence of radar reflectivity for turbulent scatter and possible interference from other sources of sporadic nature. The weak radar reflectivity of the turbulent scatter coupled with a requirement of a few tens of meters of range resolution has called for the application of pulse compression and advanced signal and data processing techniques. Nowadays Radars were extensively used for obtaining the wind information (wind profiler) and the techniques which are discussed in the thesis also, leading to the application to wind profiler, hence more details on this subject is discussed further.

4 4 1.3 Wind Profilers Measurements of atmospheric parameters are a first step towards understanding the Earth s atmosphere. Our knowledge of the state of the Earth s atmosphere is highly dependent on the instruments we use for measuring it. The parameters of primary interest are temperature, pressure, humidity, precipitation and wind. Apart from routine surface measurements of these parameters, significant development has been made in the last five decades in measuring these parameters as a function of space, time and height. Introduction of the radiosonde and subsequent improvements in our weather forecasts testify to the validity of this observation. With a network of nearly 800 radiosonde stations worldwide, the twice-perday manually launched radiosonde has become the standard source for upper air data. Advances in our understanding of the weather are dependent on obtaining measurements with a high resolution in both time and space. For this network of radiosonde stations, temporal resolution is sufficient for large scales; however our increased awareness of the importance of smaller scales, inadequacy of radiosondes for obtaining the dense set of observations required for short range forecasting and the desire for greater resolutions has encouraged the development and acceptance of ground-based system as an alternative measuring device. Development of remote sensing technology offers a solution to this problem. It is now possible to

5 5 measure vertical profiles of the wind on a nearly continuous basis with accuracy better than that normally obtained with radiosondes. 1.4 History of development of Wind Profilers The development of wind profiler technology over the past two and half decades is an out growth of research work done with radars designed to probe the ionosphere. Many of the researchers who participated in the early development of wind profiler technology were ionospheric physicists redirecting their efforts for the studying of the lower neutral atmosphere. In 1960s techniques were developed to measure radial velocities of the ions and electrons in the ionosphere using the Doppler principle. These techniques were later used to examine the atmosphere below the ionosphere (the troposphere, stratosphere and mesosphere). The first published Doppler radar wind measurements in clear air were made independently by (10),(11) using large, steerable radars at Wallops Is., Virginia, and at Deford, England. Early results which helped to stimulate the development of wind profiling in the lower atmosphere were reported by ionospheric physicists (6) using an ionospheric research radar at Jicamarca, Peru. The success of Woodman and Gúillen in obtaining strong continuous echoes between 10 and 35 km altitude inspired the use of other ionospheric radars to probe the troposphere and stratosphere. These include the Chatanika, Alaska radar by (12) and the Arecibo, Puerto Rico radar by (13). The success of the Jicamarca observations

6 6 motivated the development of radars designed specifically to measure winds. The first generation of wind-profiling radars included the small 40 MHz radar at Sunset, Colorado by (73), the 53.5 MHz SOUSY radar in Germany by (14), and the 41 MHz radar at Urbana, Illinois by (15). Finally summarized by (16) Doppler radar capability for probing the atmosphere, wind profiling radars have been used successfully for meteorological research and they have been considered for routine operations by (17). Among the several research radars developed in the late 1970s and early 1980s, the Poker Flat radar in Alaska by (18) and the MU radar in Japan (19), (20). deserve special mention. Other important research radars have been developed in many countries, including Australia by (21), (22), Canada by (23), (24), France by (25), (3), (26), India by (27), Taiwan by (28), and Wales by (29). Also, the Trans-Pacific Profiler Network by (30), (31) consisting of several UHF or VHF radars near the equator from Peru to Indonesia, has been established to study atmospheric dynamics in the equatorial region. 1.5 Technique and Principle of Wind Profilers There are two wind profiling techniques that are commonly used to determine the three components of velocity vector (Vertical, Zonal and Meriodonal) namely, the Doppler Beam Swinging (DBS) method and Spaced Antenna (SA) method. The DBS method uses a minimum of three radar beam orientations (Vertical, East-West, and North- South) to derive the three components of the wind vector by (32). In

7 7 the SA method, the backscattered signal is received by three noncoplanar antennas, located usually at the corners of a right angle triangle. The horizontal velocity and the characteristics of the ground diffraction pattern and thereby that of the scattering irregularities can be obtained through the full correlation analysis by (21). The SA method was applied to measure the wind velocities in the middle atmosphere by (33) and lower atmosphere by (34). The most commonly used technique for wind profiling is the Doppler technique. There are related methods that use multiple spaced receivers: Radar Interferometry (RI) or Spatial Interferometry (SI) and Imaging Doppler Interferometry (IDI). Wind profiling radars use electromagnetic waves to look into atmospheric properties. These radars depend upon the scattering of electro magnetic (EM) energy by minor irregularities in the index of refraction of the air. The index of refraction is a measure of the speed at which EM waves propagates through medium (atmosphere) for wind profiling. A spatial variation in the index of refraction encountered by a propagating EM wave causes a minute amount of the energy to be scattered in all directions and most of the energy incident on the refractive irregularity propagates through it. A small portion of this scattered energy is returned to the radar site, where it can be received and analyzed. The primary scattering mechanism that gives rise to the echoes observed by 'clear-air' radars is scattering from turbulent irregularities in the radio refractive index by (77). Maximum

8 8 backscattering power occurs when irregularities are about half the size of the radar wavelength (Bragg scattering). Tracking refractive index irregularities, which are carried by the wind, reveals wind information. The profiler computes height by using the time interval between transmission of the pulse and reception of the return signal. However, wind speed and direction are determined by using the Doppler principle. A wave will shift in frequency because of the motion of the target relative to the observer. The received frequency higher than the transmitted frequency indicates that the wind is moving towards the profiler. The received frequency lower than the transmitted frequency indicates the wind is moving away from the profiler. The amplitude of waves received at radar is used to estimate the reflectivity and the phase of the waves is used to estimate the radial wind velocity. The radial velocity of scattering particles is determined from their observed phase difference between successive radar pulses. The radar receiver determined the in-phase (I) and quadrature phase (Q) signals with respect to the transmitted. By calculation of the Fourier Transform of I and Q signals as a function of transmitted pulse number, the Doppler velocity spectrum can be obtained. The Doppler spectrum provides the meteorological information content from which almost all measurements are made by Doppler wind profilers. The Doppler spectrum is determined from the output of a fast Fourier transform (FFT). The FFT, which is calculated

9 9 on a special digital signal processing (DSP) card that is part of the data processing subsystem of the profiler, is characterized by the number of spectral points in the FFT, the maximum unambiguous velocity of the spectrum and the resolution of the spectrum. The Doppler spectrum is characterized by its three spectral moments. The 0 th moment is called reflectivity, 1 st moment is radial velocity, and is found by measuring the time variation of reflectivity i.e. how the targets move and the 2 nd moment of power distribution is called spectrum width, and is found by measuring the time variation of radial velocity. Basically, this is an indicative parameter for turbulence in the atmosphere. Doppler wind profilers are being used principally for atmospheric sounding in several frequency bands. The frequency of operation of a wind profiler is based on the size of the irregularities present in the atmosphere. Because back scattering occurs preferentially from irregularities of a size about one half the wave length of the probing radio wave. Therefore profilers must operate with a frequency whose wavelength is proportional to the size of the irregularities present in the atmosphere. Since irregularities in the atmosphere exist in the size ranging from a few centimeters to many meters, the wave length must also be in the centimeter to many meter range. Generally wind profilers operate in VHF ( MHz) and UHF (0.3 3 GHz) bands. Much of the early work with wind profilers was accomplished at VHF primarily in the MHz frequency

10 10 band. In the lower VHF close to 50 MHz, profilers are used primarily for observation of the free troposphere and the lower stratosphere. The smaller irregularities (cm size) are abundant only at the lower heights. The higher frequency waves therefore are not backscattered as effectively at the greater heights as are those at the lower frequencies. Profilers that operate at the lower frequencies can probe higher atmosphere than those operating at the higher frequencies. In general, atmospheric radars are not able to probe the very bottom part of the atmosphere due to its larger size and the pulse width being used for transmission. 1.6 Applications of Wind Profilers It is important to recognize that wind profilers do much more than simply measure the wind. Measurement of the vertical wind is a unique capability of wind profiling radars. Since they provide almost continuous measurement of wind over a range of altitudes, they provide detailed information on vertical structure and wind variability. Vertically directed beams provide for direct measurement of vertical motions. In addition to wind and wind variability, the Doppler wind profiler provides measurement of signal strength and Doppler spectral width. Signal strength is influenced by the intensity of inhomogeneities in the radio refractive index, which depends upon the gradients of atmospheric temperature and humidity as well as the intensity of turbulence. In addition, profilers are being adapted to a

11 11 variety of platforms so that soundings can be made where they are needed most. Wind profilers are expected to have a growing impact upon weather forecasting, atmospheric research environmental pollution monitoring, climate and mesoscale modeling, air traffic control and many more. It is therefore important that the wind measurements of these radars are both accurate and reliable. While wind-profiling systems continue to evolve and profilers are being used increasingly for diverse observations, there is no doubt that profilers will be an important part of any observational program of the future. 1.7 Importance of signal processing in wind profiler technology Although profilers have proved their worth in many areas, the pace of future expanded use is, somewhat dependent on the speed with which remaining limitations can be overcome. Improvements in the timeliness and quality of wind profiler products will depend on improving the signal processing. There has been an explosive growth in digital Signal Processing theory and applications over the years. Flexibility and versatility of digital techniques grew in the front-end signal processing and with the advent of integrated digital circuitry; high speed signal processors were developed and realized. Radar signal processing continued to grow in the recent years by keeping the future developments in mind and with better digital capability. Significant contributions in DSP in Radar have been in MTI processing, Automatic Detection and extraction of signal, Image

12 12 reconstruction, etc. The objective in a signal processing problem is to process a finite number of data samples and extract important information which may be hidden in the data. This objective is usually achieved by combining the development of mathematical formulations with their algorithmic implementations (either software or hardware) and their applications to real data. Profiler data can be contaminated by returns from things other than the atmosphere. Progress has been made in signal processing that can reduce the negative impact of this contamination. The current signal processing methods generally follow the assumption in the existing algorithm that the atmospheric return is the only signal present. Signal processing begins with time series acquisition of data. Analysis of the raw timeseries is carried out in either the frequency or the time domain. Each domain has advantages over the other, with the choice often being a consideration of which atmospheric parameters are being observed. The frequency domain allows the determination of echoes originating from different types of scatterers, providing the frequencies involved are sufficiently separated. For example, clear-air and precipitation echoes, which are sufficiently separated in the spectra at VHF or the frequency domain, can be used to identify and remove undesirable signals. The computation of wind velocity is important in the atmospheric radar data processing. This leads to wind profiling radars used in meteorological research and they have been considered for

13 13 routine operation by (16), (17). Profilers were deployed in the National Oceanic and Atmospheric Administration (NOAA) network by the National weather service USA by (35). Profilers are expected to have a growing impact on weather forecasting, environmental pollution monitoring climate and meso - scale modeling, air traffic control etc.. It is therefore, important, that the wind measurement of these radars be both accurate and reliable (36). The measurement error is one of the main reasons for generating erroneous information on wind. This is generated from the estimation of moments from the Doppler Spectrum. There are various approaches adopted to remove the erroneous points. Estimation of moments by peak detection algorithm and taking consensus averaging is one of the approaches adopted in wind profilers by (37),(38)and(39). Consensus averaging may reduce these errors, but limit the temporal resolution and allow some erroneous data to pass unchecked by (40). The first-guess technique algorithm approach by (40) is a powerful method for reducing the probability of false detection, but it requires a prior knowledge of the wind profile coupled with interactive quality control. To provide the prior knowledge information additional collocated facility like radiosonde is required. The algorithm based on first-guess followed with neural network is a good method for eliminating the ground clutter and detection of signals in low SNR region by (41). Most of these algorithms are based on the statistical methods applied to the velocity

14 14 profiles derived from the moments estimated by the usual single peak detection method. There is an approach adapted at Gadanki radar to derive the wind vector that works basically on estimation of moments through an adaptive method. This method which is different from the other methods of moments estimation used for the data processing considers a number of candidate peaks for the same range gate and tries to extract the best peak, which satisfies the criteria chosen for the adaptive method of estimation. The adaptive moments estimation algorithm developed by (42) is used for the extraction of three spectral moments from the spectrum data with the expressions given by (43). This technique has significant advantage in terms of better height coverage compared to the conventional single peak detection method. As elucidated in the earlier sections the extraction of reflected power and mean Doppler shift (frequency) is the most important and critical part of the signal detection and processing of the wind profilers/mst radar. In general the signals received from the atmosphere are very weak having very low SNR. It is also observed that sometimes the received signals are contaminated with interferences with complex characteristics. So sophisticated signal detection and extraction techniques are required for this kind of radars. There are number of theories and algorithms which are developed for different situations available in the literature. Here in the present study, few new approaches for signal analysis and filtering, its algorithm development and testing is reported.

15 Organization of Thesis This section summarizes the contents of this thesis and the details of each chapter are given below. Chapter-2 presents a detailed description of observational instruments used for this study. Details and system description of MST radar is presented with block diagrams. Method of data acquisition and signal processing steps involved are also explained. The signals received from MST radar are sometimes contaminated with interference signals received from other objects or generated with in the system through arcing of high power devices. There are multiple interference bands with different characteristics observed in the power spectra, which contaminate the wind information and other atmospheric signals. An autonomous interference detection and filtering algorithm has been developed for removing interference bands generated in the Doppler spectra of Mesosphere-Stratosphere-Troposphere (MST) radar signals. The technique, implemented with the MST radar at Gadanki is based on identifying the interference like band signals using a statistical signal variance approach for fixing the amplitude threshold through which detecting the interference frequency and design an adaptable notch filter to filter the undesired frequency bands. The autonomous interference detection and filtering algorithm is applied to various

16 16 cases and it is found that interference signals could be removed effectively, leaving behind original signals. Chapter-3 discusses this algorithm in detail and demonstrated its performance with different type of observations. Data analysis is a necessary part in pure research and practical applications. Imperfect as some data might be, they represent the reality sensed by us; consequently, data analysis serves two purposes: to determine the parameters needed to construct the necessary model, and to confirm the model constructed to represent the phenomenon. Unfortunately, the data, whether from physical measurements or numerical modeling, most likely will have one or more of the following problems. (a) The total data span is too short; (b) The data are non-stationary; and (c) The data represent nonlinear processes. Although each of the above problems can be real by itself, the first two are related, for a data section shorter than the longest time scale of a stationary process can appear to be non-stationary. Facing such data, we have limited options to use in the analysis. The type of analysis to be applied when system (Generation process) is unknown is highly subjective. The method, which adopts is the trial and error, which supports best output in a given situation. The echoes received from MST radar which represents atmospheric background information is considered to be generated through a

17 17 random process. There are different approaches adopted by different people for analyzing the data and interpret the result. Most of the approaches aim to enhance Signal to Noise Ratio (SNR) for improving the detectability. The most common approach is the FFT, which is simplest and straightforward among all the methods. Historically, Fourier spectral analysis has provided a general method for examining the global energy-frequency distributions. As a result, the term spectrum has become almost synonymous with the Fourier transform of the data. Partially because of its prowess and partially because of its simplicity, Fourier analysis has dominated the data analysis efforts since soon after its introduction has been applied to all kinds of data. Although the Fourier transform is valid under extremely general conditions, there are some crucial restrictions of the Fourier spectral analysis: the system must be linear; and the data must be strictly periodic or stationary; otherwise, the resulting spectrum will make little physical sense. The available data are usually of finite duration, non-stationary and from systems that are frequently nonlinear, either intrinsically or through interactions with the imperfect probes or numerical schemes. Under these conditions, Fourier spectral analysis is of limited use. For lack of alternatives, however, Fourier spectral analysis is still used to process such data. The uncritical use of Fourier spectral analysis and the insufficient adoption of the stationary and linear assumptions may give misleading results.

18 18 The Fourier spectrum defines uniform harmonic components globally; therefore, it needs many additional harmonic components to simulate non-stationary data that are non-uniform globally. As a result, it spreads the energy over a wide frequency range. Fourier spectral analysis uses linear superposition of trigonometric functions; therefore, it needs additional harmonic components to simulate the deformed wave-profiles. Such deformations are the direct consequence of nonlinear effects. Whenever the form of the data deviates from a pure sine or cosine function, the Fourier spectrum will contain harmonics. As explained above, both non-stationarity and nonlinearity can induce spurious harmonic components that cause energy spreading. The consequence is the misleading energy-frequency distribution for nonlinear and non-stationary data. A new data analysis method based on the empirical mode decomposition (EMD) method, which will generate a collection of intrinsic mode functions (IMF). The decomposition is based on the direct extraction of the energy associated with various intrinsic time scales, the most important parameters of the system. Expressed in IMF s, that they have well-behaved Hilbert transforms, from which the instantaneous frequencies can be calculated. Thus, we can localize any event on the time as well as the frequency axis. The decomposition can also be viewed as an expansion of the data in terms of the IMFs. Then, these IMFs, based on and derived from the data, can serve as the basis of that expansion which can be linear or

19 19 nonlinear as dictated by the data, and it is complete and almost orthogonal. The new method of data analysis for non linear non stationary process is called Hilber-Huang Transform (HHT). An algorithm for HHT based data analysis of MST radar back scattered echoes has been developed and detailed analysis is carried out. Chapter-4 will give in detail the algorithm implementation and case studies with atmospheric signals. Radar Interferometer technique in spatial domain basically provides information on the angular position of the discrete scatters and their aspect sensitivity to the radar backscatter in the vertical plane containing the interferometer baseline. It is also possible to derive the drift of the scatterers along the direction of the baseline by tracking their positions as function of time. The technique has been used extensively for studies on plasma irregularities in the ionosphere. A study on lower atmospheric radar signals during a severe convective event using spatial domain interferometry technique is carried out. Necessary algorithms were developed for the analysis of signals received from radar. During convective event atmosphere is highly turbulent and wind field will be changing very fast in time and space. There is a strong updraft and down draft of mass observed along with heavy precipitation during the event. This will lead into generation of multiple echoes within the radar observational volume. Due to its complexity the extraction of information is very difficult and

20 20 could not reveal desired information by using conventional technique such as Doppler Beam Swinging and the signal extraction through power spectral analysis. In the chapter-5, we are reporting an approach using Space Domain Interferometry (SDI) technique to extract the information. Using SDI, coherence and phase of the bimodal signal has shown unique characteristics which helps in identifying the velocity of clear air and precipitation echoes. Large number data sets were tested with this technique and found to be successful in identifying the multiple peaks echoes in a highly disturbed atmospheric environment. This is the first time this kind of analysis is carried out for extracting the wind information during the disturbed atmospheric condition. In conclusion, the chapter-6 summarises the study carried out for the thesis. The advantage of using the new adaptable filtering technique and its usefulness for wind profiler is discussed. The new analysis approach for non-linear non- stationary signals with emphasis on atmospheric radar back scattered echoes are also discussed. SDI is an approach for lower atmospheric observation when the medium is highly turbulent and non stationary. A discussion about this is given in the chapter. The future scope on these studies is also discussed in this chapter