CHAPTER II MST RADAR TECHNIQUE AND INDIAN …shodhganga.inflibnet.ac.in/bitstream/10603/101573/8/08...28 CHAPTER II MST RADAR TECHNIQUE AND INDIAN MST RADAR SYSTEM 2.1 Introduction

28

CHAPTER II

MST RADAR TECHNIQUE AND INDIAN MST RADAR SYSTEM

2.1 Introduction

2.2 MST Radar technique

2.2.1 The Refractive Index of the Target and its Fluctuations

2.2.2 The Radar Equation

2.3 MST Radars in Atmospheric Studies

2.4 The Indian MST Radar

2.5

2.4.1 Antenna Array and Feeding Network

2.4.2 Transmitter System

2.4.3 Receiver and Signal processor

2.4.4 Exciter and Radar Controller

2.4.5 Data Processing

Indian MST Radar in atmospheric studies

2.6 Gravity wave Experiment

2.7 Conclusions

29

2.1 Introduction

The Mesosphere-Stratosphere-Troposphere (MST) radar is a high power

coherent pulse Doppler radar capable of mapping the structure, vector wind fields and

turbulence in the atmosphere with very high temporal and spatial resolution. The

MST radar consist of a two-dimensional phased antenna array, a set of high power

transmitters with appropriate feed network, T/R switches, a phase coherent receiver

with quadrature channels, a signal processor consisting of two identical channels of

AID converter, decoder and integrator, a computer interface and a computer with

essential peripherals and software support.

MST radar provides estimates of atmospheric winds on a continuous basis

with high temporal and spatial resolution, which is important in the study of the

various dynamical processes of the atmosphere. MST radar uses the echoes

obtained over the altitude ranges of 1-100 km to study winds, waves, turbulence and

atmospheric stability. Echoes below 50 km arise primarily due to neutral turbulence

whereas above 50 km, the echoes are due to irregularities in the electron density. In

the height ranges 30-60 km, density of the atmosphere as well as electron density, are

very low and the echoes are very weak, resulting in a gap region in most of the MST

radars. For probing this region, MST radar along with Rawinsonde, Dropsonde,

Rocketsonde, Lidar and Meteor radar could be used.

Woodman and Guillen (1974) studied the lower atmosphere using the incoherent

scatter radar, which is used to probe the ionosphere. They could obtain echoes from

the variation in the refractive index of the clear air. The contribution of MST radars

in the study of the structure and dynamics of middle atmosphere was reviewed by

Rottger ( 1980). MST Radars can be utilized for observing wind, waves and

turbulence (Gage and Balsley, 1978). Balsley and Garello (1986) analysed the short

period wind fluctuations over poker Flat, Alaska using the Poker Flat MST Radar.

The vertical velocity power spectra was studied by Ecklund et al. (1986) using poker

flat MST Radar.

30

Using Indian MST Radar wide variety of observations were carried out

during past few years by a number of scientists. Some of the important topics are

study of gravity waves and tidal waves, tropopause detection, study of unstable layers,

convection events and ionospheric irregularities. In this work Indian MST Radar was

operated in ST mode to study the velocity profiles and wave activity.

2.2 MST Radar Technique

MST (Mesosphere -Stratosphere - Troposphere) technique is usable in all

weather conditions being unaffected by precipitation or cloud cover. MST radars

make use of scattering from small scale structure in the atmospheric refractive index

with scales of the order of one half the radar wavelength (Rao, 1990).

t..

(',.)

a,

l"J L.

r = ct/2

ltime t

�. fo

Figure 2.1 Principle of a pulsed Doppler radar (Rottger, 1989 )

31

Usually the MST radars as well as the incoherent scatter radars apply the

conventional pulse modulation technique. Figure 2.1 presents the principle of a

pulsed Doppler radar (Rottger, 1989). A short radar pulse is transmitted and the back

scattered radar echo from a range r is received after a time t. Sampling the received

echoes from different ranges given r = c t I 2, where c is the velocity of the radar

signal. Usually the power of the Doppler spectrum is computed for signals received

in ce1iain range gates and the basic parameters like total power P, Doppler shift fd and

the spectrum width cr are deduced. In addition, further useful parameters can be

determined from the particular shapes of Doppler spectra.

2.2.1 The Refractive Index of the Target and its Fluctuations

In the case of atmospheric radars, the target is the earth's atmosphere. The

characteristics of the atmosphere seen by radio waves in the absence of liquid water is

expressed in terms of refractive index n defined as n = � (Sato, 1989 a) where c is V

the speed of light in free space and v is the velocity of the radio wave in air.

Microscopic changes of n in space cause refraction or reflection. Major contribution

to n at frequencies of HF through UHF bands are expressed approximately as (Basley

and Gage, 1980)

__ 3.75(10- 1 e) + 7.76(10-5 P)n-1

T2

T (2.1)

Where e is the partial pressure of water vapour and P is the total atmospheric pressure

in units of mb, T is the absolute temperature, Ne is the number density of electrons

and Ne is the critical plasma density. The first term represents the contribution from

water vapour and is of importance in the lower atmosphere. The partial pressure of

water vapour becomes negligibility small above tropopause. The second term due to

dry air becomes dominant in this region. The third term gives contribution from free

electrons. This term is negligible below about 50 km, but is dominant at ionospheric

heights above 80 km.

32

In the absence of total reflections, scattering from fluctuations in the refractive

index n dominates the received echo of the atmospheric radar. Statistical fluctuations

of the electron density due to random thermal motion of electrons and ions can be

strong enough in the ionosphere to cause detectable scattering. This component is

called incoherent scattering because, the scattered wave from individual electrons are

random in phase, so that they add up incoherently. Received echo power is then

proportional to the number of electrons illuminated by radar.

The major source of scattering in the lower and middle atmosphere is by the

fluctuations due to atmospheric turbulence. Here the main component is due to

coherent scattering in contrast to the incoherent scattering in the ionosphere. The

main difference of the coherent scattering from incoherent scattering is that the

fluctuation of n is caused by motion of air parcels, each of which contains a large

number of molecules and electrons which contribute to the scattered electric field

coherently in phase. As a result, scattered echo power is roughly proportional to the

square of the number density of the scatterers instead of the liner proportionality of

the incoherent scattering. This substantial enhancement in echo power is the basis for

the MST radar technique and the observation of the neutral atmosphere with a

relatively small system compared to power full incoherent scatter radars.

2.2.2 The Radar Equation

A relation between transmitted and received echo power is called the radar

equation. If we transmit a radio wave of power P1 out of an omni-directional antenna

which radiates the power into all directions with uniform strength, the density of the

power passing through a unit area located at a point sufficiently far from the antenna

and perpendicular to the direction of propagation is given by (Sato, 1989 a)

(2.2)

33

where r is distance of the point from the transmitting antenna. The antenna used for a

radar usually has a strong directivity with which a narrow region can be illuminated.

Hence the equation can be modified as

p,G, pi

=

4nr 2(2.3)

where Gt is the directional gain of antenna. Now consider a target, which is located at

this point which intercepts the power and scatter it in various directions. The density

of the scattered power P 5 per unit area at distance r from the target is expressed as

P = P; (J's -,-2

4nr (2.4)

where a is the effective area of the scatterer. If we receive the scattered power with

an antenna, which has a capability of collecting all power passing through an effective

area Ac, the received power Pr is given as

(2.5)

where L is the loss factor, which represents various attenuations of received

signal due to antenna, transmission line, etc. Thus

(2.6)

This equation gives the received echo power from a given target by a radar,

and is called radar equation. There is a universal relation between Gt and Ae (Silver,

1951) which is given as G1 =

4�1e

where 11. = !:_ is the radar wave length for a monostatic radar, the radar equation can f

be reduced to

34

(2.7)

This equation allows m choosing appropriate transmitter power Pt and

effective antenna area Ae for a given target with a scattering cross section er at a range r.

The above equation applies to a single target. If there are more than one target

in the same volume V of the air observed by a radar, the electric field receiver is

expressed as the sum of the electric field components caused by individual scatterers.

For a uniformly distributed target, V is determined by the spatial resolution of the

radar. For a radar with a circular antenna, it is expressed in terms of the half power

beam width of the antenna Sh in radians and the size of the range cell r. Thus

(2.8)

Probert (1962) expressed a relation connecting beam width of the antenna and the

gain of antenna G1• The relation is

(2.9)

Where a is non-dimensional factor which concerns the non uniformity of illumination

of the antenna.

is the effective diameter of the antenna. Thus the radar equation for distributed target

may be written as

P, Ae

:ra 2 !1r Ln

Pr =

64r2

(2.10)

35

where n is the volume reflectivity defined as the scattering cross section per unit

volume.

2.3 MST Radar in Atmospheric Studies

The MST radar technique can be considered as having evolved from the

pioneering work of Woodman and Guillen (1974). Since then, the technique has been

used by a number of observers to deduce a variety of important properties like wind,

waves turbulence and stability of the atmosphere over increasingly greater height

ranges. Results obtained from such observations have been helpful in a number of

disciplines including meteorology, atmospheric dynamics global circulation, gravity

wave and turbulent studies. Gage and Balsley (1978) have discussed the historical

perspective of technique, while Balsley and Gage (1980), Harper and Gorden (1980),

and Balsley ( 1981) have considered the potential of the technique for middle

atmospheric studies. Related wind measurement techniques have been utilized by

Gregory et al ( 1979), Walker ( 1979), Harper and Gorden (1980) and Gage and

Vanzandt ( 1981 ).

Following the first MST radar studies reported by Woodman and Guillen

(1974), several MST/ST radars have been constructed which are devoted to

atmospheric studies. A significant advancement in data continuity was achieved

following the construction of the Poker Flat MST radar in Alaska (Balsley et al.,

1980). Measurements by Poker Flat MST radar has revealed highly variable short

tenn fluctuations attributed to internal gravity waves (Ecklund et al., 1981; Gage et

al., 1981 ). Rottger (1987) has investigated various gravity wave sources using VHF

radars. Balsley et al. ( 1984) studied the seasonal variation in the VHF echoes

obtained from the mesosphere and lower thermosphere using the Poker Flat MST

radar.

36

Table 2.1 (Woodman and Guillen, 1974) gives a list of MST, ST and

incoherent scatter radars. In the table SA stands for spaced antenna capability. M, S

and T for mesosphere stratosphere and troposphere respectively (M) for D region in

in-coherent scatter mode and I for ionosphere thermosphere incoherent scatter mode.

Table 2.1 Existing MST, ST and IS radars

Radar Location Frequency Antenna Average Altitude Beam

(MHz) Gain Power Coverage Directions

(dB) Aperture Product

(Wm2)

Arecibo Puerto Rico 2380 75 1 * 10 10 ST Bistatic

Arecibo Puerto Rico 430 61 6 * 109 I (M) ST Multi

Arecibo Puerto Rico 46.8 12 5 * 107 MST Multi

Chung Li Taiwan 52 29 1 * 107 ST 5, SA

Fairbanks Alaska. USA 220 40 1 * 106 ST 5

Flatland Illinois. USA 40.5 27 4 * 108 ST 5

Indian MST Tirupati, India 53 36 7 * 109 MST 6

Jicamarca Peru 49.9 44 1 * 10 10 IMST Several, SA

MU Radar Japan 46.5 34 4 * 108 IMST 1657, SA Ponope

Christmas Pacific 49.8 32 5 * 106 ST 1, 3

PROUST France 935 51 7 * 106 ST 1, bistatic

SOU SY W. Germany 53.5 31 7 * 107 MST Multi, SA

SOU SY, Norway 53.5 35 7 * 107 MST 4 mobile

Sunset Colorado, USA 40.5 24 6 * 106 ST 5

Urbana Illinois, USA 40.9 29 2 * 107 MST Several

2.4 The Indian MST Radar

37

A major MST radar has been established at Gadanki near Tirupati (Lat

13° 27' 34" N. Long 79° 10' 34" E, MSL 190 m), in the state of Andhra Pradesh in

India. The radar has been developed in two phases. In the initial phase, it was

operated in low power ST mode using partial power aperture of the system and later

the final phase of the development of the full MST radar has been completed (Rao et

al., 1994 b ). The system can work in ionospheric coherent back scatter mode also.

The specifications of the MST radar system are given in Table 2.2. The Indian MST

radar is a highly sensitive VHF phased array radar operating at 53 MHz with an

average power aperture product of 7 x 108 Wm2• Figure 2.2 shows the simplified

Table 2.2 MST Radar System Specifications

SYSTEM

Operating Frequency 53 MHz

Peak Power Aperture Product 3 * 1010 W.m2

Height range 5 to 100 Kms

Spatial Resolution

Range 150 m (pulse width)

Angle 3° (Beam width)

Velocity resolution 0.1 m/sec

Time resolution 0.5 minute

Wave form Selectable pulse widths and PRF's including pulse compression

Pulse compression Psuedo random coding ( complimentary BPSK code sequence of Baud length = Im sec)

Signal Processing Real Time Digital (FFT based)

SUBSYSTEMS

Antenna Phased array with 1024 crossed Yagi elements

Gain 36 dB (nominal)

Beam width 30

38

Beam positions Zenith, ± 20° off Zenith in EW and NS Directions

Side lobe -20 dB

Size 130m * 130m

Transmitter Coherent; modular with variable pulse width and PRF

Peak power 2.5MW

Duty ratio 2.5%

Pulse width Selectable 1 to 32 m sec.

Receiver 2 Channel ( 1 & Q) coherent

Overall gain llOdB

Dynamic range 70 dB

Coho stability 1 * 1010 (short term)

Data acquisition & Signal Processing Real time, computer controlled

Data resolution 12 bits

Sampling rate 1 MHz per channel

No of range gates Up to 256 512 (Design goal)

No. of points for spectral estimation 64 to 512

Velocity resolution 0.1 m/sec

Signal enhancement by coherent integration 20 dB (nominal)

Spectrum integration period Selectable from 5 sec to 10 min. in steps

System computer 32 bits super Micro Computer

Operating system Real time Unix (RTU)

Online memory 2 M Bytes (expandable)

Storage Hard disk, Floppy and Mag. Tape

Display CRT with colour graphics

Hard copy Printer, Plotter

block diagram of the radar system. The system comprises of high resolution antenna

array high power transmitters, transmit receive switch a signal processor consisting of

two identical channels of ND converter, decoder and integrator, a computer interface

and a computer with essential peripherals and software support. The detailed

specification of the Indian MST radar is given by Viswanathan (1986) and Rao et al.,

( 1995). Figure 2.3 is the functional block diagram of Indian MST radar. The

39

following sub-sections present a brief description of the functioning of the various

sub-systems of the radar.

2.4.1 Antenna Array and Feeder Network

The phased antenna array consists of two orthogonal sets, one for each

polarization, of 1024 three-element Yagi-Uda antennas arranged in a 32 x 32 matrix

over an area of 130 m x 130 m ( Rao, 2001). The array is illuminated in either of the

polarizations using 32 transmitters of varying power, each feeding a linear sub-array

of 32 antennas. The feeder network consists of two orthogonal sets, one for each

polarization, of 32 parallel runs of center-fed series structures. The RF power from a

transmitter is fed to a 3-dB in phase power divider (combiner for reception) and

distributed along the sub-array through appropriate couplers of the feeder line. For

the modified Taylor distribution adopted for the aperture, a directive gain of about

37 dB, a half-power beam width of 2.62° and a first side lobe level of -20 dB could

be realized.

ANTENNA ARRAY PHASING NElWORK

11Ui\VIT CH& PREAMPLIFlERS

I I I I I I I I I I I I I I I I I I

--+1 TRANSMITTER MODULES

WAVEFORM GENERATOR

RECEIVER

COHERENT OSCIL LATOR

Q

SIGNAL PROCESSOR

&FIT

Simplified Block Diaaram of Indian l.\l:IST Radar at Tirupati

GRAPHIC >------< TERMINAL

1------1 TAPE DRIVE

GRAPHIC 1------1 PRINTER

FLOPPY1------1 DRIVE

1------1 HARD DISK

Figure 2.2 Simplified block diagram oflndian MST radar at Tirupati.

1:32 DIVIDER

MODULATOR CODF.R

IF S\'NTHESIZER

YAGI

ANTENNA ARRAY

FIEDER NETWORK [fAVLOR ILLUMINATION)

POLARJZATION S'WITCHFS

REFERE:r\CE OSCILLATOR

DISTRIBUTED

l'IIASE SIIIIT CONTROL

IIUST COMPUTER

PRINTER MACiNEllC TAPE

32:1 COMB,\INER

LOCAi. l'lmn�<;SOR

GRAPHICS DISPLAY

BROADBAND

IF-AMP.

STC

PAT (0·60 dB)

IF-AMP. CHAIN

QUAD MIXER

vmrn

AMP.

ADC(12 111T) 2 Nos.

l>ECUUl:I{

INTF.GRATION

Fig 2.3 Functional block diagram oflndian MST radar (Rao et al., 1995)

40

41

The antenna pattern has been characterized in the receive mode by

recording the radio source Virgo-A (3C 274) using the phase switching

interferometer technique of Ryle (1952). Based on the measurements taken on

several passes of the radio source during 23 May - 10 June 1993, it is found that the

. beam pointing accuracy is better than 0.2°

and the 3dB beam width is in the range

of 2.8°

to 3°

· The radar beam can, in principle, be positioned at any look angle, but

it is currently programmed to sequence automatically any combination of 18 look

angles over a range of 20°

from zenith in the NS and EW planes with a resolution

of 1 . The phase angels for transmit and receive beam for the 18 beam positions are

stored in four EPROMS, each serving eight transmit and receive channels. A local

processor 8085A), located in each of the four transmitter huts, adds the phase read

from the EPROM to the calibration phases and provides the control signals to 8-bit

phase shifters. The calibration phases for the 32 transmit and receive channels are

measured periodically by running a test signal through the channels and comparing

its phase to that of a reference signal.

2.4.2 Transmitter System

A total transmitter power of 2.5 MW (peak) is provided by 32 transmitters

ranging in power from 15 kW to 120 kW, each feeding a sub-array of 32 Yagis. It

has four amplifier stages and associated power monitoring and controlling, and safety

interlock circuits. The amplifier chain consists of a solid state amplifier (SSA), pre­

driver (PDR), driver (DR), and high-power amplifier (HPA). The SSA module has

four stages and for an input signal of 0.25 to lmW, it provides an output power

ranging from 25 to 100 W with a bandwidth of about 8 MHz. The PDR, DR and

HPA operate in class C, employing varian triodes 3CX1500 A7, 3CPX1500 A7 and

3CPX5000 A 7 in a grounded grid configuration. The output powers for the three

stages range from 300-1200 W, 3-15 kW and 36-120 kW and the corresponding

bandwidths are 3.5, 3.2 and 2 MHz, respectively.

42

The input to the transmitter is a low-level (lmW) pulse-modulated (coded/un

coded) signal at 53 MHz generated by a mixer which receives as inputs a 5MHz

pulse-modulated signal and an appropriately phase-shifted 48 MHz local oscillator

(LO) signal. The transmitters can operate up to a duty cycle of 2.5%, limiting the

total average power to about 60 kW. It is possible to transmit both coded and un

coded pulses with pulse repetition frequency (PRF) in the range 62.5 Hz to 8 kHz,

keeping the duty cycle from not exceeding the limit. The un coded pulses can be

varied in pulse width from 1 to 32 s in multiples of two. The coded pulses are either

16 or 32 baud bi-phase complementary pairs (AB A B) with a baud length of 1 µs,

providing a range resolution of 150 m. The use of AB along with the complementary

pair AB facilitates the removal of system de bias, if any.

The output of the transmitter is connected to an antenna sub-array through a

transmit-receive (T/R) duplexer and a polarization selection switch. The duplexers,

which serve to switch the antenna array between the transmitters and the receiver

channels, are realized by means of distributed and lumped hybrid couplers, and PIN

diodes. The duplexers have an insertion loss of 0.5 to 1 dB and provide an isolation

of 50 dB. For polarization selection, vacuum relays of Jennings make with an

insertion loss of 0.1 dB are used.

2.4.3 Receiver and Signal Processor

The signal processing details are given in figure 2.4 (Viswanathan, 1995). The

front end units of the receiver, consisting of a blanking switch, a low-noise

amplifier (LNA), and a mixer-preamplifier for each of the 32 channels, are located

in the four transmitter huts, eight in each of them. The LNA is a 53-MHz tuned

amplifier with a gain of 24 dB and a bandwidth of 4 MHz. The output of an LNA

is mixed with an appropriately phase shifted 48 MHz LO signal and amplified in a

mixer-preamplifier having an effective gain of 7 dB. The IF outputs from the 32

channels are combined and amplified, in a broadband modular amplifier, with a

gain of about 15 dB. The signal then goes through a sensitivity time control (STC)

circuit providing a fixed attenuation of 20 dB up to a selectable range over which

43

the signal tends to be saturated. The output of the STC passes through a

programmable attenuator of 0-60 dB with 10 dB steps and IF amplifier chain with a

gain of 60 dB and a bandwidth of 1. 7 MHz. The IF signal is now split into two and

applied to a pair of quadrature mixers which mix them with 5 MHz LO signals

having quadrature phases of 0°

and 90°. The quadrate signals from the mixers are

fed to two identical channels of low pass filter (LPF) and video amplifier to obtain

the two bipolar video signals of A cos and A sin at the output. The receiver has

an overall gain of about 120 dB and a dynamic range of70 dB.

MST Radar

Antenna Array

Receiver Control

Processor

T.S.G

Host Computer

Radar

controller

Power Spectrum Spectral

Averaging Moments

Tape Hard disk Graphic

Cons

Fig. 2.4 Signal processing details of MST radar (Viswanathan, 1995)

44

The quadrature (I and Q) outputs of the receiver are limited to ± 5 volts and

given to a preprocessor unit consisting of two identical channels of analog to digital

converter (ADC), decoder, coherent integrator, and a common interface. The ADC

is of 12 bit resolution to match the dynamic range of the receiver and of 500 ns

conversion time to meet adequately the requirement of 1 MHz sampling rate. The

decoding operation essentially involves cross correlating the incoming data from

the ADC with the replica of the transmit code and it is implemented by means of a

16 bit, 32 tap correlator I transversal filter chip. The decoded signal is coherently

integrated for a specified number of pulses to effect a significant reduction in the

data volume without in any way compromising the information to be derived from

the signal. A 16 bit parallel interface multiplexes the integrated output from I and

Q channels and transfer the data to the host computer (Mascomp-MC5600) for

further processing. The computer performs FFT of the complex time series on-line

and the power spectra, integrated for a specified number, are recorded on a

magnetic tape. There is option, however, to record the raw data of complex time

series instead of power spectra, if so desired. For on-line monitoring, the power

spectra can be displayed on the graphics console of the host computer in a selected

format.

2.4.4 Exciter and Radar Controller

The exciter unit generates all the RF and timing and control signals for

various sub-systems of the radar. It comprises a master reference oscillator, a two­

channel frequency synthesizer, a phase-locked oscillator, a P-controlled bi-phase

coder and a timing signal generator. The master reference is a 5 MHz oven­

controlled crystal oscillator (HP l 508) with a short-term stability better than 1 part

in 10 10• The two channel synthesizer provides two 5 MHz channels with

programmable relative amplitude and phase. It provides LO signal to the receiver,

IF to the transmitter coder and reference and simulator signals for phase calibration.

The phase locked oscillator (PLO), operating at 48 MHz, serves as LO for up­

conversion while transmitting and down-conversion while receiving. The bi-phase

coder generates a 5 MHz complementary coded pulse taking the 5 MHz signal from

45

two channel synthesizer and the complementary code sequence from an 8085A

processor. The timing signal generator (TSG) is a programmable multi-channel

pulse generator, supplying control signals for synchronizing the operation of the

various subsystems of the radar. The output from the TSG includes Tx and Rx gate

signals, duplexer signal, coder and ADC sample clocks, and control signals to

preprocessor and signal simulator.

ANALOO

CHANNEL

-------------------------------------�---------------I

PREPROC ESSOR I

DECODER INTEGRATOR

INTERFACE

I I I

ANALOO r::::--i Q -C-H-AN-N -EL--

�---: DECODER INTEGRATOR

TlMlNO

CONTROLS

TIMINC IICNAL CENlRATOR

I I I I

I

L-··----··, I I

CONTROL COMMUl'«CATION

TESTCARD

I

r ................ L I I I

(CCO

L-------------- -- ---------�-�-�-�-�------------

Block dia11ram of preprocessor

RADAR CONIROU.IR

ROSI" PARALLEL INl'ERFACE

Figure 2.5 Block diagram of preprocessor of Indian MST radar (Patro et al.,

1989)

The radar operates under instructions from an IBM PC-AT based radar

controller (RC) which executes an experiment according to the data given in the

form of an experiment specification file (ESF). The main function of the RC is to

set up, control, and synchronize the operations of various sub-systems during the

normal operation of the radar. The main sub-systems, functioning under RC

46

control are timing signal generator (TSG), bi phase coder, preprocessor, host

computer and the four satellite processors located in the transmitter huts. The RC

communication with TSG, coder and preprocessor is established through IEEE 488

interface and with host computer and satellite processor through RS 232. Figure 2.5

depicts the preprocessor of the radar (Patro et al., 1989). On any experiment, once

the radar parameters are specified in the form of ESF, the RC takes over the

operation of automatic data acquisition and completes the run without any need for

intervention

2.4.5 Data Processing

The Doppler spectra, recorded on magnetic tape, are analyzed off-line for

parameterization of the spectrum following a method given by Riddle ( 1983), as

extended suitably by Anandan ( 1997) for adaptive signal processing. Figure 2.6

shows the on line and off line data processing details of the radar. The method

involves (1) the removal of de, (2) estimation of the average noise level, (3) the

removal of interference, if any, (4) incoherent integration (further to whatever done

on-line), (5) signal identification through an adaptive technique, and (6)

computation of the three low-order (0th, 1st and 2nd) moments. The de

contributions from non-fading clutter and uncancelled system biases, if any, are

eliminated by notching out the zero frequency and averaging the two adjacent

Doppler bins to interpolate for a new zero frequency value. For estimating the

average noise level, an objective method developed by Hildebrand and Sekhon

(1974), which is widely used, has been adopted here. This technique is based on the

statistics of a Gaussian random variable and the expected relationship between mean

and variance for the spectrum of a white noise source. The noise level thus

determined is subtracted from the received power for each Doppler bin. Any

interference band that might run through the entire range window, as experienced

often, is subtracted out by estimating it in a range bin where it dominates the real

signal. At this stage any incoherent integration of the spectra, further to that already

carried out on-line, is implemented if so required to improve the signal detectability,

,:1:

47

although at the expense of time resolution. Now, for each range bin, the signal is

identified through an adaptive technique and its window is determined by noting all

the contiguous points that are above zero level. The three low-order moments are

computer then through numerical integration using the expressions given by

Woodman ( 1985). The three moments represent the signal strength, the weighted

mean Doppler shift and half-width parameters of the spectrum. A typical Doppler

spectrum of the east, west, zenith, north and south beams are given in figure 2.7 .

............................................................................................................................................................................. , ,·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·-·

On-line I Off-line Processing

I-Channel �---�

Signal Processor 1.,. ----1.... Decoder � Coherent _..___..,

I �

Normalization Windowing (I & Q) � Integrator 1

IQ-Channel �---�

Time Series ................................................. , .......................................................................................................................... ·

Noise level Estimation ._

Spectrum Cleaning

Incoherent Averaging

�

Fourier Analysi &

Power Spectrum

·-·-·-·-·-·-·-·-·-·-··

Power Spectrum

Moments UVW 1� Zonal, Meriodonal, Vertical. wind velocity

Total Power, Mean Doppler, Doppler Width

Off-line Processing

Fig.2.6 Processing steps for extraction of parameters (Anandan, 1997)

2.5 Indian MST Radar in Atmospheric Studies

48

Using Indian MST radar at Gadanki in the ST mode of operation, wind

velocity measurements were carried out by many workers. Study of momentum flux

and turbulence parameters were investigated by Jivrajani et al., (1994). Narayana

Rao et al., ( 1994 a; b) studied the refractivity turbulence structure constant and

turbulent energy dissipation rate with Indian MST radar. Characteristics of unstable

modes around a shear layer was studied by Mini et al., (1994). The Indian MST

radar is also used for probing the ionospheric irregularities. Viswanathan et al.,

( 1994) and Rao et al. ( 1994 b) studied plasma irregularities in the E region for the

first time using simultaneous day time observation made during February March 1994

by the MST radar and VHF back scatter radar at Trivandrum. Dynamics of the

equatorial spread-F was observed by Patra et al., (1994) using MST radar in

ionospheric coherent back scatter mode.

A weakening of tropopause and associated enhancement in troposphere

stratosphere exchange was observed on some nights and explained as due to enhanced

turbulence caused by strong wind shears (Jaya Rao et al., 1995). This phenomenon is

reported to be most conspicuous under enhanced convection (Jain et al., 1997). Using

the radar wind data, preliminary studies have been made of the various aspects of the

lower and middle atmospheric dynamics, including gravity waves, tides and

equatorial waves (Narayana Rao et al., 1997; Jivarajani et al., 1997; Sasi et al., 1997).

Narayana Rao et al (1997) have derived eddy dissipation rates using the radar data

collected on 17 June 1994. The values are found to vary with height in the range of

10·<> to 10 ·3 m2 s·3 with maximum occurring in the height range of 13 to 16 km. The

possibility that temperature profile can be derived from MST radar data of vertical

winds has been pointed out by Rottger (1986) and demonstrated by Revathy et al

( 1996), using the data taken at Gadanki. The derived temperature profile was found

to be in good agreement with the radiosonde observations.

49

The mesospheric echo characteristics have been studied by Dutta et al ( 1997)

and winds and turbulence by Sasi and Vijayan (1997). The most intense echoes were

in general , confined to a band of 70 - 75 km. The echo characteristics suggested that

they were of turbulent scattering rather than of Fresnel reflection. The radar back

scatter from meteor trails has been studied by Raghava reddy and Muraleedharan Nair

( 1998) using the Indian MST radar. A fairly large number of meteor echoes have been

detected over the observational windows.

The MST radar at Gadanki has been operated in ionospheric coherent back

scatter mode for mapping the structure and dynamics of the E region and Field

aligned irregularities (F AI). It was shown by Krishna Murthy et al ( 1998) that the

observed drift velocities below 95 km are driven by neutral wind and the meridional

wind component derived from the drift velocity is found to be consistent with the

theoretical neutral wind models. The field aligned F region irregularities were studied

by Patra et al ( 1997) and Rao et al ( 1997).

2.6 Gravity Wave Experiment

The measurement capabilities of the MST radar technique include winds,

waves and turbulence. Most of these measurements require a knowledge of the

power spectrum of the echoes returned to the radar from different layers of the

atmosphere. The spectrum which represents relative echo power density versus

Doppler shift, can give information about various parameters including the total

received power, the mean Doppler shift introduced by the atmospheric scatterers on

various beams and the distribution of velocities about the mean value (spectral width).

The wind velocity is estimated from the mean Doppler shift, on each of the five

beams of the MST radar, as explained in chapter 3.

The experiment was conducted on 4 days, viz., 25- 28 August 1999, as apart of

gravity wave campaign project by S.R. Prabhakaran Nayar and K.Revathy as

investigators. These experiments were conducted to study the short period gravity

waves and the associated momentum transport. The details of the system operation

and data acquisition during 25 - 28 August 1999 are given below.

50

The system was operated in the ST mode at 53 MHz with an average power

aperture product of 4.8 x 106 Wm2 and a beam width of 3°. The inter pulse period for

the observation was 1000 µs (Pulse repletion frequency = 1000 Hz). The data was

obtained using coded pulse with a pulse width of 16µs and hence a range resolution of

0.15 km was obtained. The MST radar was operated continuously for about seven

hours on each day during 25 -28 August 1999 in the 5 beam mode (East, West,

Zenith X, North and South). In the process of scanning the atmosphere, the beam

positions are switched in the following sequence: east, west, zenith-y, zenith-x, north

and south. The system takes about 33 seconds to obtain the Doppler spectra for 183

range bins, for each beam position. Thus, in a scan cycle of nearly 3 minutes, the

Doppler spectra for 183 range bins are obtained in the east, west, zenith-x, north and

south beam directions. The two oblique beam pairs ( east - west and north - south )

were at 10°

off zenith direction.. On each day the observation was carried out

approximately during 0900 - 1630 hrs LT, for approximately 7 hours. The small data

gaps in the time series of data were filled using linear interpolation in height and time

based on linear averaging of data points on either side of the data gap. Though the

observation was carried out with in the altitude range of 3.6 km to 30.9 km, the

analysis was limited to 3.75 km to 20 km range to have good SNR

The observation scheme of these days is as described in table 2.3. The

interpulse period (IPP) was selected to be 1000 µs so that, a pulse repetition frequency

(PRF) of 1000 Hz was obtained. The number of coherent integrations (Ne) done is

128. The maximum detectable frequency was obtained by the equation,

PRF fmax =

2N,.

Thus, a maximum Doppler frequency of 3.90625 Hz was obtained. The maximum

velocity that can be observed becomes 10.94m/s. The beam dwell time (T), that is,

the time required for obtaining one frame of spectra can be obtained by the expression

T = IPP X Ne X NFFT

51

...:." .. ::

~ -::

, I. J.J 12 .,~ \. 95 0.9"l.00 0.98 \.9S 2.9J J.ea

(l a f' r l ~ R (II:):.-\ ":"'~" F\(ht ....

II 10I ~.JS3.r~

11 8.eS ;:.\ a.1O "1/ 7.35C 6.60f 5.85

(y'",~.10U53.60

-J.912.Sl- us 0.9ro.OO 0.9~ 1.95 1.93 3.S8o {J P P ll: R (H,\

(c) (d)

'J

I I I.J.912 9J.I.SS-0.''lJ.CQ 0 9' I ~S =)):?fa

DO"PlE~("'Z)

r'#\TE:2~nno Nl0x Tlrt.• ~ :1 ..;:

A

I :·3.912.9),\.250.900.000.98 \.95 2.SJ3.aa

DIJPPlEA(H"L'f, T::: 71 n ,'g'} S lOx 1'lm. ! 5'Z-1:Z:

I,,I I I I I

-1.Sl~ ).i.: :;: ;'J: :": C .:: 1 ~5 1.SJ! :.:C' : ? ? l E F. (f-'l:

Z.~ Ti",., 5·33, II

~ ~ '\-,. '"

~ ~s~.~~

::: :':

~! ....- :s'/

G, !:

E ~ :~

H"I

-J.SI2.S~ 1.95 O.S'll.~Q 0.98 1.952.93 3.P3DO f' P lEn (Ill)

,)/lTE, 11 n ISP Zy Tin", 5·32,36

Figure 2,7 Typical Doppler spectrum of the six beams (Kusuma et aI.,200l)

52

Table 2.3 - Observation Scheme

Pulse Width 16 µs

Transmission mode coded

Range resolution 0.15 km

Inter Pulse Period 1000 µs

No. of beam positions 5

Receiver attenuation level OdB

No. of coherent integrations 128

No. of FFT points 256

No. of incoherent integrations 1

Output mode of data spectral moments

Maximum Doppler Frequency 3.90625 Hz

Maximum Doppler velocity 10.94 mis

Frequency resolution 0.03051 Hz

Velocity resolution 0.08542 mis

Time required for one spectrum 32.77s

53

Here, NFFT gives the number of FFT points used to obtain the Doppler

spectra which is equal to 256. A beam dwell time of 32. 77 was obtained. Since the

frequency resolution is equal to 1/T, the Doppler spectrum showed a frequency

resolution of 0.03051 Hz and velocity resolution of 0.08542 mis. Thus, Doppler

spectra were obtained between± 3.90625 Hz, with a resolution of 0.03051 Hz. Since

we are interested in the study f short period fluctuations in the lower atmosphere, the

number of incoherent integrations carried out during the experiment is limited to one.

This allowed a time resolution of less than 3 minutes.

The experiment duration and the number of scan cycles are detailed in table

2.4. On 25 august 1999 observations have been made during 0928 to 1608 hr 1ST.

On 26 August 1999 we have data from 0912hrs to 1626 hr 1ST. On August 27, 1999,

the experimental duration is from 0859 hrs to 1634 hr 1ST. On 28 August 1999 the

data obtained is from 0900 hrs to 1601 hr 1ST. On all the days of observation there is

a data gap of approximately 10 minutes at the end of each 38 cycle due to the

magnetic tape change.

The Doppler spectra were stored in magnetic tapes of 1600 BPI and 2400 ft

usmg MASSCOMP 5600 computer system. Later, the three moments of each

Doppler spectrum were obtained on an off-line basis, using the software ADP

developed by Anandan (1997 ),NMRF , Gadanki. These data files give information

about the date, time, beam position, scan cycle etc and presented as an array giving

the height, zeroth moment, first moment, second moment, noise level and SNR (dB).

2. 7 Conclusion

The MST radar technique utilizes very high frequency radar in the Doppler

mode to determine the drift velocities of back-scattering elements whose nature

depends on the region begin scanned. The echoes received by such radars from the

troposphere and lower stratosphere are caused by refractive index variations due to

54

Table 2.4 Experimental duration

25 August 1999 26 august 1999 27 August 1999 28 August 1999

9 : 28: 49 to !6: 8 9: 12: 49 to 16: 8: 59: 6 to 16: 34: 9:0 :7 to 16: 1: 3 : 13 26:25 20

( 128scan cycles) (132 scan cycles) ( 146scan cycles) ( 134scan cycles)

density fluctuations associated with the neutral atmosphere. In the mesosphere, the

echoes are produced by scattering due to fluctuations in the free electron density

associated with turbulence. With the MST radar techniques it is possible to obtain

three dimensional velocity fields. MST radars have limited horizontal coverage but

can produce data with high temporal and spatial resolution in a given locality. These

radars are thus appropriate for studying high frequency components of the motion

field, such as gravity waves and tides. They can be used to trace meteors, study

cyclones, monsoons, thunder storms etc. Ionospheric studies can also be made with

this radar.