close

Вход

Забыли?

вход по аккаунту

?

The environment of transitory microwave angels in the lower troposphere

код для вставкиСкачать
NOTE TO USERS
This reproduction is the best copy available.
UMI
®
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
THE ENVIRONMWT OF TRANSITORY MICROWAVE
ANOEIS IN THE LOWER TROPOSPHERE
*y
Morley Bruoe Bell
Submitted in partial fulfillment
of the requirements for the degree of
Master of Soienoe
Faoulty of Graduate Studies
The University of Western Ontario
London, Ontario
1962
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: EC45158
INFORMATION TO USERS
The quality of this reproduction is dependent upon the quality of the copy
submitted. Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleed-through, substandard margins, and improper
alignm ent can adversely affect reproduction.
In the unlikely event that the author did not send a com plete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
UMI
®
UMI Microform EC45158
Copyright 2006 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Approved for the
Approved for the
Department of Physios
Department of Physios
by the Advisory Committee
c'
by the Examining Committee
y').
r. //
”' / W ffcdjJL
f. / s ■
17
& ft
Datet November 9, 19^2
ii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ABSTRACT
A study of the fine-soale structure of the refraotive index of the
air in the lover troposphere has been investigated through the use of a
low-powered radar of speoial design.
The radar is directed vertically to
examine that p u t of the troposphere which lies between 150 and 1500
meters above the ground.
A number of modifications were made in the radar
to improve its sensitivity and reliability inoluding the installation of a
system for optical viewing of the region probed by the radar beam and the
incorporation of an anti-correlation deteotor at the radar receiver.
In
addition, a study has been made on methods of improving the sensitivity
of the radar.
Interpretation of the observations is related to local surface
measurements of air temperature, relative humidity, air pressure and wind
speed taken at the radar site, supplemented by information on the air
mass, frontal position and oloud oovsr as supplied by the Meteorological
Branoh of the Department of Transport, Canada.
The analysis of the limited number of observations showed no apparent
association between transitory reflections and surface air temperature,
relative humidity, air pressure and wind speed.
An association between
the inoidenoe of transitory reflections and frontal sons and the appearanoe of cumulus oloud was noted.
These observations are interpreted as
indicating that the reflections arise at changes in the refraotive index
of the air that occur over height intervals of no more than a few centi­
meters Cof the order of the radar wavelength), such as may be produoed by
strong wind sheer within a frontal zone or by erosion at the edges of a
rising air bubble in a oonveotive thermal below oumulus clouds.
iii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACOOWLEDOaCENTS
The author is greatly indebted to Dr. R.J. Uffen, Acting Stead of the
Department of Physios, to Professor R.L. Allen, Chairman, and to the
present Departmental Head, Dr. P.A. Forsyth, F.R.S.C., for making available
the facilities of the department.
He also wishes to thank Dr. D.R. Hay, supervisor of this candidate's
researohy for his expert advice on the many experimental and theoretical
problems enoountered throughout.
Special thanks are extended to the Rational Research Counoil of Canada
for the Bursary awarded this oandidate during the first twelve months of
his research.
Additional financial assistance and experimental equipment
proffered by the Defenoe Research Board are also gratefully acknowledged.
Far their advice and their time pertaining to electronic and photographic matters y appreciative thanks are extended to Mssrs. D. Rumbold,
T. Davidson and the late S. Thomson.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTESTS
ABSTRACT....................
....... iii
ACKNOWLEDGEMENTS..........................
LIST OF ILLUSTRATIONS
iv
................... ........
viii
x
LIST OF PLATES..............................................
Chapter 1.
Introduction
1
....
Section 1.1
Historical Background.........
1
Section 1.2
Theories of the Echoing Meohanism
Section 1.3
Methods of Angel Investigation.
Section 1.4
Subject Division and Statement ofthe Problem.••
...
...
2
4
6
...
8
Section 2.1
Introduction..... ............
8
Section 2.2
The F.M. Radar...........
8
Section 2.3
Modification to the Radar.
Chapter 2.
Chapter 3*
Modification of the P.M. Radar
........
The Tracking Looal-Osoillator
13
.....
15
Seotion 3*1
Introduction. ...........
15
Seotion 3*2
Relations ana Features for F.M............
15
Seotion 3*3
Estimation of Required Bandwidth.
16
Seotion 3*4 Theoretical Sensitivity
Section 3*5
The Looal-Osoillator System
Seotion 3*6
The Balanced Mixer and A.M..
Seotion 3*7
Sources of A.M..........
Chapter 4.
18
•....
.....
18
..........
20
22
The Optical Viewing System......
31
Seotion 4*1
Introduction....................................
31
Seotion 4*2
Specifications of the Optioal Viewer.....
31
Section 4*3
Design C o n s i d e r a t i o n s . . . . . . 33
Seotion 4.4
The Viewing System
Seotion 4*5
Viewer Performance
Chapter 5*
The Anti-Correlation Device..
....
36
......
38
........
Seotion 5*1
Introduction............
Seotion 5*2
Requirements of the Anti-Correlator.
Seotion 5*3
Design Considerations.
44
...........
.....
..............
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
44
44
46
Chapter 6 . Experimental Observations with the 6770 MC/s Radar...
52
Seotion 6.1
Organisation of Bata...........
52
Seotion 6.2
Teohnique of Analysis..........................
53
Seotion 6.3
(a)
(b)
(c)
(d)
(e)
Association of Angel Incidence with Surface
Variables...... ....... .
Variation of Angel Incidence with Surface Air
.....
Temperature.
Variation of Angel Inoidenoe with Relative and
Specific Humidity.......
Association of Angel Inoidenoe with Baronetrio
...............
Pressure.
Association of Angel Inoidenoe with Air Mass......
Association of Angel Inoidenoe with Wind Speed....
61
61
63
63
66
66
Seotion 6 .4
Association of Angel Inoidenoe with Air Structure
Above the Badar................................
68
Seotion 6 .5
Singular Transitory Phenomena..................
71
Seotion 6 .6
(a)
lb)
(o)
(d)
Chapter 7*
Association of Angel Inoidenoe with Other
Variables.................
Variation of Angel Inoidenoe with Height••.•••....
Variation of Angel Inoidenoe with Month..........
Variation of Angel Inoidenoe with Angel Duration..
Inoidenoe Fluctuations within the Recording Interval
Interpretation of the Experimental Observations.
Seotion 7*1
Introduction.
....
72
72
72
72
75
77
77
Seotion 7.2 Interpretation of the Effect of Surface Variables 77
Seotion 7*3 Interpretation of the Effect of Air-Mass
Association....................................
(a) Soundings in a Single Air-Mass.
.....
lb) Transitory Angels Associated with Fronts..........
(o) Transitory Angels Associated with Low, Cumulus
Clouds.•••••....
(d) Singular Transitory Phenomena
.......
78
79
Seotion 7.4
(a)
(b)
(oj
Chapter 8.
Interpretation of the Association of Angel In­
......
oidenoe with Height and Duration.
Association of Angel Inoidenoe with Height*.••....
Association of Angel Inoidenoe and Angel Duration
Interpretation of Inoidenoe Fluctuations within
the Recording Interval...............
77
77
78
Remarks and Recommendations.
Seotion 8.1
.....
81
82
84
85
Introduction...........••••••••••••••.......
Seotion 8.2 Remarks....
Recommendations.,
81
....
....
vi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
85
85
87
Appendix I........................ .... .......... ........ .
90
Appendix II.............. ..................... ............
92
Appendix III......................................... .....
94
BIBLIOGRAPHY..............................................
96
VITA
vii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OP ILLUSTRATIONS
Figure 2.1
Blook Diagram of Entire PM Radar................
9
Figure 2.2
Salient Features of the PM Radar................
11
Figure 2.3
Development of the beat-note
11
Figure 3.1
FM Frequency Spectrum............
17
Figure 3.2
The Loo&l Oaoillator ...............................
19
Figure 3.3
The Looal Oscillator with AX
23
Figure 3*4
AM Introduoed by the Microwave Filter ..............
25
Figure 3*5
Frequenoy Response of Microwave Filter .............
26
Figure 3.6
Klystron Modes
28
Figure 3.7
Increase in Simultaneous Fluctuations
Figure 4.1
The Periscope Principle.........................
35
Figure 4.2
Determination of Optimum Focal Length
..........
37
Figure 4*3
The Viewing System
Figure 4*4
Timing Cirouit For Automatic Camera Control
.....
43
Figure 5*1
Examples of Coherent and Incoherent Radar Echoes ...
45
Figure 5*2
10 Channel Correlation Deteotor
49
Figure 5.3
Analysing and Recording Unite Showing Anti-Correlator
50
Figure 6.1
Superposition of Recording Charts...............
55
Figure 6.2
Air-Mass Type and Associated Fronts..........
57
Figure 6.3
Simultaneous Fluctuations
........
59
Figure 6.4
High-Frequency Signals
.............
60
Figure 6.5
Random Fluctuations
........... .
....
.....
.........
......
30
40
...........
..••••
6o
Figure 6 .6
Reflections from Birds above Radar .................
62
Figure 6.7
Average Angel Inoidenoe vs. Temperature ............
64
Figure 6 .8
Angel Inoidenoe vs. Relative Humidity ..............
64
viii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 6 ,9
Angel Inoidenoe vs. Speoifio Humidity
Figure 6.10
Angel Inoidenoe vs. Pressure
Figure 6.11
Angel Inoidenoe vs. Air Hass
......
64
....
65
....
65
Figure 6.12 Association of Average Angel Inoidenoe with Wind Speed 67
Figure 6 ,1 3
Incidence vs. Height Profiles
Figure 6 .1 4
Association Between Angel Inoidenoe and Frontal Zone
....
Figure 6 .1 5
Angel Inoidenoe vs. Height
Figure 6,1 6
Angel Inoidenoe vs. Month
Figure 6.17
Angel Inoidenoe vs. Angel Duration
Figure 6 .1 8
Inoidenoe Fluctuations for Individual Recording
Intervals
........
....
69
70
73
..........
........
73
74
76
Figure 7.1
The Formation of Cumulus Clouds by Connective Lifting 80
Figure 7.2
Interpretation of Angel Inoidenoe vs. Height Curve.
82
Figure 8.1
AM Compensator for Removal of AM Introduced by
Microwave Filter
.......
88
ix
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OP PLATES
Plat* I
Field of View of Sky Viewer .........................
39
Plate II
Relative Field of View of Radar and Sky Viewer .......
42
Plate III
Bird Photographed through Sky Viewer .................
%
x
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER I
INTRODUCTION
Station 1.1 Historical
Background
The aim of the research described in this thesis is basically to
extend the present knowledge of the small-scale structure of the tropos­
phere.
Reflections from inhomogeneities in this region result in
anomalous eohoes that appear on radar receivers when there is no visible
scattering source present.
The atmosphere surrounding the earth is divided into several distinct
layers.
The layer lying olosest to the earth's surface is called the
troposphere and extends upwards to a height of approximately six miles.
(Berry, Bollay and Beers, 1943).
It is in this layer that weather with
its associated turbulence, temperature fluctuations, pressure changes and
precipitation occurs.
Here, air heated by contact with the earth, rises
and is replaced by cooler air.
In so doing it carries with it evaporated
water which ultimately condenses as the surrounding temperature and
pressure deorease with height.
Adding to the turbulence caused by these
updrafts there are also associated horizontal winds which are the more
oommon air currents encountered on the earth's surface.
Through this region too, fly today's modern aircraft whose fore­
runners, in World War II, were responsible for the sudden need to probe
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2
this region with the then recently-discovered, techniques of radar.
It was from sensibly clear regions in this layer that inexplicable
radar echoes were first reported in the mid-thirties.
It is common
knowledge today that war-time radar operators often reported mystifying
instances of anomalous radar propagation.
Others told of the detection
of strange and confusing echoes from regions of the troposphere in whioh
no aircraft were flying.
Since World War II radar sets have become in­
creasingly more powerful and more sensitive with the result that such
eohoes from the troposphere have become common.
Watson-Watt et al.» (1936) and Mitra (19 3b) oonducted researoh on
the structure of the ionospherio £ region employing the technique of
pulsed radio sounding at vertical inoidenoe.
In addition to the expected
ionospherio eohoes (from approximately 100 kms altitude) many eohoes were
received from regions far below the ionosphere.
Observations made by
Collwell (19 37) and Friend (1948) told of reflections whioh oame from
altitudes of from 6500 to 39»000 feet.
Such echoes were later termed
microwave "angels'* since they were refleoted electromagnetic energy of
a few centimeters wavelength from invisible regions*
■SaaUPtt U 2
Theories of tha Echoing Maohanlam
Watson-Watt et al. (1936) suggested that the intense "patches of
ionisation" resulting from thunderstorm activity could be a possible
source of reflections.
Qish and Booker (19 39) later demonstrated that
ionisation in the lower atmosphere was not sufficient ( 5x10
were needed) to explain the observed reflections.
ions/oo
Piddington (1939)
suggested that gradients in the permittivity of the air could cause
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3
sufficient energy return at the wavelengths used, to explain the echo
amplitudes obtained.
Early work emphasised only the observed features of angels.
Little
effort vas directed towards understanding their physical origin (Friend,
1949) Friis, 1947).
Measurements of fluctuations in refractive index
employing the radiosonde method served only to enhance the controversy
over possible theories whioh attempted to explain the source of such
reflections, for the measured gradients remained at least orders of
magnitude too small to account for the observed angel reflections.
It
soon beoame apparent that more preoise measurements on the physical
structure of the air in the troposphere were needed and some of the
methods used to obtain these will be discussed in detail in seotion 1.3.
The most extensive survey of angel studies was made by Plank (195^
and 1959).
He acknowledged that some are associated with insects and
small birds in the air) however, many oould not be attributed to
particulate matter.
This was also confirmed by Harper (1957) * A-t the
same time Plank presented the theory that angel eohoes are reflections
from refractive index gradients associated with convective bubbles or
thermals.
In early investigations the different types of eohoes were grouped
according to their appearance on a radar soreen (i.e. ring eohoes, dot
echoes, etc.). Alternatively, they have been classified according to
duration) that is, those having periods of less than one minute were
termed transitory while echoes lasting longer were oalled persistent
angels.
Only the transitory reflections will be investigated here.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A recent investigation at the University of Western Ontario Cover a
full year) (Hay and Reid, 19*>1) has shown that the inoidenoe of angels
increases with the type of air mass in the order continental Arctic,
maritime Arctic, maritime Polar, maritime Tropical and that the predominant
duration is about one second*
temperatures down to 20* F.
Here, transitory angels were observed for
However few angels of short duration were
observed in the range from 30° to 50* F.
Additional evidence indioates
that angel incidence is heaviest under conditions when temperature is
high, winds are light and humidity is high (Flank, 195^)»
It has also
been determined (Hay and Reid, 19^1) that the reflection ooeffioients
-14
of transitory angels are less than 10
.
Most workers in this field now agree that angel eohoes are due to
variations in the refractive index of the air in the troposphere.
Hay and Reid (19^1) have further considered a means by whioh the
vertical depth of a reflecting irregularity may be estimated*
The results
of this consideration show that a transitory angel may be attributed to
a radar reflection at a refraotivity gradient, found by experiment to
8xiat in the troposphere, if the depth of the reflecting stratum does
not exoeed more than one or two radar wavelengths.
This seotion has indicated some of the theories that exist whioh
have attempted to explain the source of these miorewave phenomena*
Al­
though there have been many different interpretations of the origin of
angels, all are, at present, only speculative.
SflQUonJjJ
Mat3»fla of Aaaai InYeaUfiaUaa
As indioated in a previous section the early reports of angels came
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5
from investigators who vers using pulsed radars of comparatively long
wavelength (Watson-Watt, 1936% Mitra, 1936).
Studies of tropospherio
soundings hy pulsed radio waves were later extended from the original
medium frequency phase to include the use of high-powered, sharply-beamed
microwave radar systems (Friend, 1949).
The experiment described in this
thesis employed a low-powered frequency modulated radar of special design.
This represents a departure from earlier methods.
Equal performances are
possible from the two types of radar (Qnanalingaa, 1954)•
To aohieve this
speoial techniques must be employed whioh provide for a reduction in both
leakage signal (signal transfer occurring directly between transmitting and
receiving antennas) and the interference of parasitio amplitude modulation.
When such techniques are included, extreme receiver sensitivity can be
achieved (Onanallngam, 1954} Ridenour, 1947).
The physical details and operating procedure of the FM radar are
described in chapter II.
However there are several points in the ex­
perimental procedure used throughout this experiment whioh differ from
that of the previous investigation.
These will be indioated here.
In the previous study (Reid, 19^0) the associated meteorological
parameters such as temperature, barometric pressure, relative humidity,
etc. were supplied for a site six miles east of the radar installation.
For the present experiment all information of this type was measured
immediately prior to or during eaoh recording interval and all measurements
were taken on the roof of the laboratory housing the radar.
A motor-
driven psyohrometer was used to measure relative humidity* the Alnor
Velometer wind speed indicator was used to determine the wind speed
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
6
during each sounding period while a calibrated thermometer mounted on the
laboratory roof was used to record the temperature Immediately before
each sounding interval.
Cloud oarer was noted here, also, following the
method outlined in Berry, Bollay and Beera (1945) •
In addition to this a 14~power, war surplus, rangefinder was used to
determine the height of clouds that appeared above the radar during a
sounding period.
This instrument has accurate distance measurement from
500 to 20,000 yards.
Further, examples of bird eohoes with photographic
correlation were obtained from the vise of the viewing system.
Finally,
the overall sensitivity of the radar was inoreased.
seotion i,a
SttMtgt ElTiflifltt wi.SMmrnXjit
As mentioned above, a low-powered FH radar was employed in this study
of transitory microwave angels,
A previous investigation (Held, 19&0)
using this same radar has indicated that several aspeots of the experimental
tsohniq.ua must be improved if the desired increase in radar sensitivity
is to be realised,
Several modifications are necessary too, if an improve­
ment in the reliability of this radar is to be made.
These are disouased
in detail in Chapter II, but will
be outlined here in order that a state- _
ment of this writer's problem may
be included in thischapter.
If the reflections arising from
from the records, it is necessary
insects and birds are to be eliminated
to inoluds a system whioh will enable the
radar operator to view and to photograph that region of the sky scanned by
the radar.
Such a system would also be valuable in making a study of the
effect of oloud activity on angel inoidenoe.
In addition to this some
means must be devised to eliminate large fluctuations in the background
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
noise level.
The means hy whioh these problems were solved are described
in Chapters III and IV.
This writer's research problem is therefore, as
follows*
(1) Additional information on the mieroetruoture of the troposphere
was to be sought using the M radar desoribed in Chapter II.
Particular
interest was to be paid to the observation of transitory angels and their
environment*
(2) In addition to this study the following modifications were to be
made.
a.
The construction of a viewing system whioh would enable the
re..Ur operator to observe and to photograph that region of the sky scanned
by the radar.
This would include a timing device which could be operated
manually or automatically to control the exposure time and repetition
frequency of the 35am frame.
b.
The design and construction of an anti-correlator to eliminate
instabilities such as klystron frequency drifts (as outlined in section 2*3)
inherent in the system.
Included in this thesis is an analysis of the limitations in radar
sensitivity as they arise in the tracking local-osoillator system.
A
theoretical treatment of the FM spectrum is considered in this analysis in
Chapter III.
Chapter VII contains an interpretation of the experimental
and theoretical work.
Chapter VIII.
Conclusions and reooamendations are presented in
An alphabetical bibliography follows the Appendices.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER II
MODIFICATION OF THE IM RADAR
section 2.1
;fairadaq.1tl9n
This chapter contains a brief description of the low-powered
M
radar
and the associated experimental technique as it was when taken over by the
author.
The improvements neoessary to inorease the overall sensitivity
and reliability* as outlined in Chapter I* are considered in detail here.
Saotion 2.2 Thfl FM R ftflfiJ
A blook diagram of the complete radar system is shown in figure (2.1).
Details are given elsewhere (Reid* 1960).
The operating characteristics
of this radar are reproduced in table 1 below.
la b le -l*
CharftQ-toriatiflB. -s i ..tha..$7.7Q..attZa,JFK itedar
Operating transmitter center frequency
6770 mc/s
Operating transmitter frequency deviation
1 mo/s
"Sawtooth" modulation frequenoy
100 ops
Power output
310 aw
Theoretical minimum detectable signal
2 5x 10" 19w (-I86dbw )
.
Antenna Characteristics*
Diameter of parabolic refleotors
10 feet
Antenna gain
44.1 db
Antenna beamwidth
1.
0*
8
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10ft PARABOUC REFLECTOR
3 0 mc/s
SIG- GEN-
lOdb
DIRECTIONAL
COUPLER
XTAL
UNILINE
ISOLATOR
2 0 db
DIRECTIONS
COUPLER
FREQUENCY
METER
F X R C 4 0 IA
DETECTOR
XTAL
IN23A
GYRALINE
ROTATOR
C R C- R670
AUDIO
KLYSTRON
X 26 0
VAR- ASSOC.
AMPLIFIER
KLYSTRON
POWER SUPPRO- 801A
100 cp«SAW-TOOTH
OENERAT
XTAL
SHORT SLOT
HYBRID
TEE
VARIABLE
ATTENUATOR
PR-D- II7 B
UNILINE
ISOLATOR
POWER
SUPPLY
MICROWAVE
POUND
SYSTEM
SHORT
SLOT HYBRID
TEE
FILTER
EDIN DCAMPLIFIERS
10 CHANNELS
MULTI­
METER
RECORDER
DRIVER
CIRCUITS
NARROW
BAND
AMPLIFIERS
EDIN
RECORDERS
10 CHANNELS
AUDIO
HIGH-PASS
FILTER
ANTICORRELATOP
DETECTOR
8 AUDIO
AMPLIFIER
XTAL
MIXER
IN 23 EM
3 0 M C /S IF
AMPLIFIER
XTAL
MIXER
IN 23 EMR
3 0 MC/S I F
PRE-AMPLIFIER
FIG- 2-1 BLOCK DIAGRAM OF ENTIRE F-M- RADAR
<
10
In this system a reflex klystron is employed to produoe the 6770 mo/s
transmitted wave.
Frequency modulation of the transmitted signal» with
a modulation index of 10,000, is used to establish the range to a re­
flecting target according to conventional practice (Qnanalingam, 1954)*
Parasitic amplitude modulation present with the transmitted M
signal is
reduced to a tolerable level by means of a feedback network (Eugen, 1958)*
This will be discussed in detail in Chapter III.
The antennas are set-up
for sounding at vertical incidence as indicated in fig. (2.2).
It should
be noted here that the region of interest in the experiment extends from
150 to 1500 meters directly above the radar.
This region is represented
by the shaded area on the diagram and is common to both antenna beams.
With an 9K radar auoh as is used here, the range information is
oontained in the frequency domain.
The process by whioh the extraction
of this information occurs is depicted in fig. (2.3).
Basioally it
involves the direct comparison of the transmitted and received signals.
When such a comparison is made in a suitable detector Figure (2.3) shows
that a beat-note will result equal to the difference frequency that exists
at that instant between the transmitted and received signals.
This beat-
note frequency will be a function of the heigit of the reflecting region
or more simply, will be proportional to the time the received signal has
taken to travel to the target and be returned to the reoeivex-delay
time (-'T) (Ridenour, 1947).
Qnanalingam has shown that for delay times much greater than the
modulation period, we can consider the echo beat frequency to be constant
over the entire modulation cycle.
Two oases arise from consideration of
the number of beat note cycles (nQ) occurring in the modulation period.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11
LOCAL
OSCILLR
EM.
RECEIVER
EM .
TRANSM'R
FIG. 2.2 SALIENT FEATURES OF TH E
F .M . RADAR
transm itted
—-re c e iv e d
ft
A
F' ^ // /
] 711 r
/ /
|
/
i
/ /
j
/ //
r
/
j
x
/ /
iA F
/ /
j
y /
/
*/
J jL
X
X
/
X
X
X
*
t
0
0
0
■—
FIG. 2 .3
0
0
*
•
X
/
0
0
DEVELOPM ENT
*f —{*>■ i ^
O F THE B E A T -N O T E
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12
First, when an integral number of cycles are present coherence results.
The heat note frequency is then a simple multiple of the modulation
frequency. In the seccnd oase, when nc is non-integral, phase discon­
tinuities exist. The heat note is then periodic in T.
In this case
several heat frequencies will result each a multiple of the modulation
frequency.
In the case when an integral number of oyoles are present,
an integral multiple of the modulation frequency ia produced every 150
meter increase in target height. Thus a 100 ops note is produced whan
the reflecting target lies at 150 meters, a 200 cps note is produced
for a target at 300 meters, etc., and the range of the target is thus
indicated hy the frequency of the beat-note produced.
In the non-integral
oase the range is determined hy the height producing the heat note with
the largest amplitude.
A
table showing range versus beat-note frequency
is given below*
Tatea 2»
Bange (meters)
fianm Vargaa Jaafritoto
Belay time (secs)
Beat Freq. (ops)
nv
150
1
100
1
300
2
200
2
450
3
300
3
600
4
400
4
750
500
900
5
6
600
5
6
1050
7
700
7
1200
3
800
8
1350
9
900
9
1500
10
1000
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13
Narrowband filters peaked to the harmonios of the 100 ops modulation
frequency enable the beat-notes to be separated into channels.
These in
turn display the range information. As is shown in the radar block diagram,
the output from eaoh filter is oonneoted through a separate d.o. amplifier
to its own chart reoorder. A more complete desoription of the radar is
given elsewhere (Reid, 19&0).
SaeUoa„,2«3 StodlflgfttlQna to ttw Radar
A previous investigation (Hay and Reid, 1960) has shown that several
aspeots of the existing experimental technique should be Improved before
thi3 radar and its assooiated equipment are used again.
The value of a viewing system whioh would permit the radar operator
to observe continuously or to photograph the sounding region is evident.
Previously the operator has had to rely on intuition when eliminating
reflections from physical objeots (i.e. birds, insects, aircraft etc.)
from the records.
Suoh a system would also enable one to examine any
correlation whioh may or may not exist between angel reflections and
meteorological prooesses suoh as cloud formation (water vapour condensation).
Since this study of the angel environment will not consider the reflections
from physical objects as angels, this system is necessary to provide the
radar operator with a reliable means for their elimination.
In the investigation conducted by Reid, (1958-1960), interpretation
of the recordings proved to be a most difficult and tedious task. Due to
short term fluctuations in the receiver noise signal, the desired echoes
have, at times, been almost entirely masked. Although time-consuming,
recognition of the true echoes is still possible since the undesired
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
14
fluctuations ooour in all ohannels simultaneously, and can be ruled out
using this fact alone.
There are many sources from whioh these unwanted signals oould arise.
Some of those suggested as a result of previous work are} frequency drifts
in the transmitter, fluctuations in power supplies and changes in leakage
signal occurring as a result of antenna vibrations. Whatever their oause,
they must be eliminated if a significant increase in radar sensitivity is
to be realised.
Although other technical modifications were carried out, only the
two mentioned above are discussed here. Specifications on the design of
the optical viewer stipulated that it conduct a beast of light from the
laboratory roof to the radar approximately twenty feet below} that the
angular field of view be suffioient to give a olear picture of the area
scanned by the radar} that the resolution be suffioient to reoord the
physical objects previously mentioned when they appear in the sounding
region.
Specifications on the anti-oorrelator (device to eliminate undesireable fluctuations) were merely that it detect simultaneous fluctuations
in all channels and provide some means to remove them.
The methods by
whioh these problems were approached are given in Chapters IV and V.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER III
THE TRACKING LOCAIrOSCILLATOE
fa c tio n V1
Intattduqtlqn
This chapter presents an analysis of the local oscillator system
in whioh the source of undesired signal fluctuations is investigated.
These fluctuations are shown to occur as a result of incidental AM
present on the PM carrier.
It will he shown that AM* arising in the
local osoillator system, can seriously limit the sensitivity of an PM
radar of the type used here. Relations and features of PM are reviewed
at the outset and the aotual frequency speotrua of the transmitted signal
is established as a result.
section
i.g ,Bftlfttagna.-aB4 featurta XouM
This section provides a review of some of the basic features of PM.
The general equation for the instantaneous value ^ of an unmodulated
sinusoidal carrier is
It « Ia sin (n_t + $)
where
■ 2*F
P • carrier frequency
In FM there are many side-current pairs as opposed to one pair in AH.
Here the number of sidebands is determined by the ratio AP/f (where f is
the modulation frequency). However, an increase in the number of sidebands
15
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
16
does not necessarily mean a corresponding inorease in the bandwidth W*
occupied by the sideband spectrum (Bund* 1942)*
Bow the instantaneous value of the frequency modulated signal can
be represented by
1^ » IB sin (xxt + p sin wt)
where
ft is the modulation index dp/f
The expansion of this equation in terms of the sideband pairs is possible
by means of Bessel functions of the first kind.
+ Jg (p) [ sin(-Q.+ 2w)t - sin(o.- 2w)t]
+
+
(p) [ sin(n. + 3w)t - sin(o.- 3w)t]
.....
+ Jn (p) £sin(CL+ nw)t + (“l)n sin(n.- nw)tjj
If dp and f are small oompared to the oenter frequency P.
Section i. i BgttBatton of Baqttkrtfl Banartdth
Figure 3*1 shows the frequency spectrum of fM signals of the same
peak frequency deviation but of different modulating frequencies.
All
amplitude values are shown as positive quantities sinoe the purpose of
this table is solely to determine the required bandwidth.
The polarity
of the various Bessel factors is therefore of no concern.
For f - 100 ops
and 0 • 10*000 there will be approximately 10*000 upper side-frequencies.
This would be impossible to calculate.
For the case considered here*
however* the frequency spectrum whioh gives a representation very close
to the true one is shown in diagram 3.1 (e).
Thus the bandwidth is two
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
17
A F = 1 ,0 0 0 ,0 0 0 c/s
jJjLlU
i
- 1.-J
f = 1 0 0 , 0 0 0 c /s s in e
W = 2.4 m cl s
F ♦ AF
F -A F
(q)
I
I
A F = 1,00 0 ,0 0 0 c /s
f = 2 0 0 0 0 0 c / s sine
W = 2.8 m c / s
I
(b)
f = 8 4 0 0 0 c/s
W =232mc/s
si ne
sm ew ave
A F = 1 ,0 0 0 ,0 0 0 c Is
f —> 0
W = 2 m c/s
(d)
s a w to o th
s
i
(e)
FIG-3-I
FM
FREQUENCY
SPECTRUM
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
18
megacycles.
SflfiUga ,3t.4 Theoretical Senaltlirltv
The available noise power at the receiving antenna is given by
Pn - (p)(kT)CB)
where
Pn is in watts
-21
kT » 4x10
watts per cycle bandwidth
B » receiver bandwidth in cycles per second
The measured overall noise figure of the receiver is 8 db.*
Per overall
coherent deteotion it has been shown that the effective reoeiver band­
width is the minimum reoeiver bandwidth (Primioh, 1956) • This is deter­
mined by the bandwidth of the narrow-band filters tuned to the harmonios
of the modulation frequenoy, eaoh of which has been set at 10 ops.
The
minimum deteotable signal is therefore determined to be
Pn - 2.3x10 ^ watts
- -1 8 6 dbw
section 1.5 S m Local Oscillator Sy»t«B
The coherent tracking local oscillator system is shown in Figure (3.2).
It is composed basically of a signal generator* a modulator and a micro­
wave filter.
Double sideband signals are generated in the balanced
modulator for use in the local osoillator system.
The carrier is suppressed.
The main points in the modulator axe indicated below.
It consists of a
short-slot hybrid tee with a timed detector mount on eaoh of the oollinear
arms.
The use of one reversedrpolarity diode paired with one of forward
* Private communication. J.W.B. Day Defenoe Res, Tel. Est. Ottawa, 1958.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
OSCILLATOR
LOCAL
O l— T5
FIG- 3-2
THE
o o
o
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
20
polarity allows the modulation signal to be supplied in the same phase to
eaoh diode resulting in out-of-phase conduction.
The auppressed-carrier
conditions suggested theoretically can be realised very olosely in practice
if matohed crystal diode pairs are used and the diode mounts are carefully
balanced as described in the appendix.
It has been observed that carrier
rejection is somewhat dependent on modulation signal amplitude due to the
dependence of crystal impedance on modulation level (Mackey, 19^1).
The microwave filter is located between the balanced modulator and
the balanced mixer as shown in Fig. ( 3*2).
This unit is designed to
permit passage of the upper sideband in the local oscillator signal
spectrum and eliminate the lower one.
Its passband centered around 6800
Mcps is shown in figure (3*4).
Station 3*6 The Balaaati Miitr and, Mvlituii EfeMatton
The EM radar information is derived by mixing the signal from the
receiving antenna output u with the local signal u obtained directly
r
8
from the transmitter.
These signals invariably contain parasitic amplitude
modulation whioh oan arise from several souroes (Ridenour). Perhaps the
most common source lies in the klystron itself since its output power is
not constant with frequency.
As mentioned earlier, however, this oan be
reduced to a tolerable level through the use of an AM compensator and a
klystron frequency-atabilisation unit. (Reid, 1960j Johnston, 1962).
In section 3*4 the theoretical sensitivity was determined.
It is
known however, that the limit of sensitivity in an EM radar system of this
type is net due to thermal noise, but instead, is due to miorophonios and
other sources- of AM (Ridenour, 1947) • Thus suoh interfering AM must be
reduced to a very low level if high sensitivity is to be achieved.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
As
21
mentioned at the beginning of this seotion, the radar information is
obtained by mixing the signal from the receiving antenna output
with
the local oscillator u . It was also stated that undesirable AM is
3
generally associated with the M
signals uf and ug. Furthermore, all
suoh AM is periodic at the modulating frequency and will have components
at the modulating frequency
and its haxmonios 2fn,
their amplitude decreasing as the frequency increases.
•. • etc., with
Thus the AM will
produce a low-frequency spectrum whioh falls within the amplitude
response of the useful beat-note signal.
of AM is undesirable (Ismail, 1935)*
ance of u ^
s
UAM
For this reason the presence
It is now apparent that the appear­
(the deteoted AM of the looal oscillator signal), u ^ , and
r
mixer output produoes overwhelmingly strong low-frequency
*
sr
interference and makes it impossible to use simple mixers to detect the
radar information.
It is important to note at this point that the greatest
interference is due to u ^
since the signal from the looal osoillator is
s
always much greater than the signal from the receiving antenna. (Ismail,
1955).
A reduotion in the effect of this AM may be accomplished by the use
of a balanced mixer whioh is capable of balancing out the AM in the
detector output.
A balanced hybrid mixer (Tyrrell, Edwards, Miller, 1947)
is employed in the radar in question to minimise reoeiver noise resulting
from looal osoillator signal injection.
The looal osoillator signal is
applied to one input of the hybrid junction.
Any received signals are
applied to the other.
A pair of matched (forward and reversed) crystal
converters mounted in
tunable holders are connected to the output ports
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
22
of the hybrid junction.
The two crystal outputs are applied in-phase and
in parallel to the single-sided I.F. pre-amplifier unit.
Looal osoillator
noise is essentially suppressed (Edwards, 1947)*
Additional precautions must be taken however, to ensure a minimum
of AM on the signals entering the mixer since the AM oannot be oanoelled
out entirely with a balanoed mixer.
Experimentally it has been determined
by this author that even small changes in the amount of AM present intro­
duce very large changes in the loir-frequency signals in the audio amplifier
output.
SflQtton 'St7 ggttrMa of AaplUudg Modulation
Occurrence of parasitio AM in the klystron output has previously been
eliminated.
Thus if AM is associated with the signals entering the balanoed
mixer, it must ooour during signal transfer through the coherent tracking
local osoillator system or through the antenna leakage path.
It has been
determined experimentally by this author that the latter is not important
when the klystron mode is free from amplitude modulation.
Thus the looal
osoillator system must be investigated more fully for possible AM sources.
Figure ( 3.3) shows a block diagram of the radar with its associated
looal osoillator system.
The AM present at several significant points was
determined after an extensive examination in the lab (with the AM com­
pensator removed) and has been indicated here.
Large AM signals are seen
to arise in the looal osoillator system.
As a result of this study an examination of the balanoed modulator
was carried out.
balanoed.
For linear operation the modulator must be properly
Since no general balanoing procedure was available a satisfactory
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
FIG- 3-3
THE
L OCAL
OSCILLATOR
23
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
24
one was worked out and is included in the appendix*
When this procedure
is followed the hybrid tee modulator has been found by this writer to be
sufficiently linear so as to aooept the frequency modulated carrier with­
out adding additional AM to it.
The spectral distribution and bandwidth of the transmitted wave was
determined in section 3*3*
A portion of this frequency speotrum is applied
to one input of the balanoed modulator and a 30 Mops carrier signal is
applied to the orystal inputs.
The fourth hybrid arm transmits sideband
spectra whioh are replicas of the M
from it —
1947).
filter,
carrier speotrum but displaced 30 Mops
the carrier being suppressed in the modulation process (Ridenour,
Thus only the upper and lower sidebands arrive at the microwave
the response of whioh is shown in Figure ( 3*5)♦
Its passband is
designed to remove the lower sideband while permitting the upper one to
reaoh the balanoed mixer input.
An examination of the response curve in Figure ( 3.5) reveals that if
the upper sideband has a center frequency of 6800 Mo/s, the introduction
of AM in the microwave filter is unavoidable.
presented in Figure (3*4)«
An indication of this is
Furthermore, slight frequency shifts will
produce fluctuations in its amplitude.
It is suggested here, as a result
of many experimental observations, that the Pound Stabiliser does not
completely compensate for transmitter frequenoy shifts and that this there­
fore, is a possible source of simultaneous fluctuations.
In experiment
this writer has found the miorowave filter to be the source of greatest
AM introduction.
In praotioe it is possible to produce additional AM by
one or more of the means suggested earlier, whioh would in effeot eliminate
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2?
crystal detector
output at filter Input
i ------
0709-5
0770-5
frequency
(mc/s)
crystal detector
output at filter output
0709-5
0770-5
frequency
(mc/s)
FIG-3*4 A M P L I T U D E
MICROWAVE
MODULATION INTRODUCED BY THE
FILTER
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
26
relative
power
6798
6800
6802
6804
frequency (mc/s)
FIG- 3-5
FREQUENCY RESPONSE OF
MICROWAVE
FILTER
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
27
or oancel that introduced by ths filter.
This vould be a means of solving
the problem if it were not for the frequenoy oh&ngeB.
Hence if the un~
desired fluctuations are to be removed some method of eliminating the
AM or the frequenoy shifts must be found.
Previously in this ohapter it has been assumed that the presence of
AM was the chief limit to sensitivity.
In practice, this is not the oase.
Actually, if complete removal of AM at the klystron output is assumed, the
existence of a constant amount of AM becomes important only when the lowfrequency components it possesses are so large that saturation occurs in
the latter stages of the IF and audio amplifiers.
A more troublesome
problem is found to exist when fluctuations in this AM ooour— especially
fluctuations with periods of ths order of one saoond since they are
difficult to distinguish from angel echoes.
Suoh fluctuations oan be
brought about as a result of shifts in the transmitter frequency as was
demonstrated above.
It will now be shown that changes in the klystron
cavity resulting from local temperature changes can result in output peak
frequenoy changes oapable of produoing angel-like fluctuations in the out­
put channels.
As was indicated earlier, an AM compensator has been incorporated in
the transmitter to remove AM present in the klystron power output.
The
proper functioning of this unit is, however, subject to certain specifica­
tions.
For example, it has been found experimentally by this author that
if all AM is to be removed at this point, the output mode must be symmetric
about its oenter frequenoy.
Figure ( 3.6).
Sample klystron output modes are shown in
Here, diagram (a) depicts suoh a symmetric output mode.
The AM compensated form of this mode is shown in (b).
Application of the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
28
UNCOMPENSATED
COMPENSATED
POWER
6 7 6 9 .5
6 7 7 0 .5
6 7 6 9 .5
(b)
(a)
67695
6 7 7 0 .5
677Q 5
6 7 6 9 .5
6 7 7 0 .5
FREQUENCY
(mc/s)
(d)
(C)
FIG-3-6
KLYSTRON
MODES
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
29
compensating unit to a mode-shape which is other than symmetrio, suoh as
that of diagram (o), results in an output mode as shown in (d).
The
amplitude of the AM components present at the modulation and harmonio
frequencies are reduced hut not eliminated, sinoe the resulting waveform
is still periodic in the modulation frequency.
Thus if the AM compensator is to remove AM as previously assumed, the
operating frequenoy must remain at ths peak of the mode.
If this is the
case, even slight changes in the klystron cavity dimensions oamwt he
tolerated for these will alter the output peak frequenoy and prevent com­
plete removal of AM.
Suoh changes have heen observed experimentally.
They
oan he eliminated in part by readjustment of the klystron hut cannot he
prevented in this manner.
In addition to this they are not controlled
by the Found Stabilisation system (Johnston, 1962) whioh oan only compensate
for changes in the operating frequenoy.
Thus, although the Pound system
may ensure that the operating frequenoy remains at the oenter of the
modulation sweep, it cannot prevent the operating frequency from deviating
from the peak of the mode.
Assuming that AM oan he introduced to the transmitted mode as
described above, ths following processes are suggested as a possible
source of the simultaneous fluctuations.
First, the transfer of signal directly between the transmitting and
receiving antennas could conceivably be modulated by changes in the
diffraotion pattern existing between the two antennas.
initiated by vibrations caused by the wind.
These could be
This process is unlikely
however, since observations of the recordings reveal many instances in whioh
suoh fluctuations apparently ooamenoe at will, without a corresponding
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30
Channel
Channel
J u ly 1 0 , 0 :2 2 pra
Channel
Channel
^:Miu
J u ly 1 0 , 0 : 2 3 pm
Fig. 3.7
Increase In Simultaneous Fluctuations
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
increase in wind gustiness.
An example of this is presented in Figure 3.7.
A second method by which the fluctuations could ooour would be as a
result of reflections from looal air turbulence in the vicinity of the
antennas. This is a suggested possibility only.
At present it is impossible to work without a certain amount of
parasitic AH.
Thus, the radar sensitivity, as determined by the thermal
noise level, oannot be attained. A possible method for removing the AM,
or at least the source of the simultaneous fluctuations, is inoluded in
the recommendations.
This chapter has presented an investigation of the possible origins
of the incidental AM whioh appears on ths FM carrier wave, with special
consideration to AM as it arises in the local osoillator system. Fluctu­
ations in the interfering AM, due to transmitter frequency shifts and
klystron cavity temperature changes, have been shown to plaoe a limit on
receiver sensitivity by produoing large low-frequenoy fluctuations in the
reoeiver output.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER IV
THE OPTICAL VIEWIRQ SYSTEM
£flaU<ax,.4.l In.teaflftfilAan
Previous investigators have reported eohoea from birds and inseots
that closely resemble angel echoes (Plank, 195^, 1959j Crawford, 1949 J
Reid, 1961).
Suoh reflections (i.e. from physioal objects) will not be
considered angels in this study and will be removed from the records.
This chapter considers an optical viewing system which will provide a
reliable means for the elimination of all suoh signals.
The possibility of using this system to exclude echoes from clouds
is indicated.
On the other hand, the detection of olouds is now possible
and their eventual use as indicators of air currents is desired.
2ftft£ifia,.,.4e2 .SMgififlfttiaafl of thn flirtInfti ay.tea
An angular field of view for the system was ohosen arbitrarily at 7*.
Although the region to be viewed was not oritioally specified it must, of
course, be large enough to allow for errors in alignment when the system
is erected.
It is desirable, on the other hand, to have as small a field
of view as possible if physioal objects are to be detected throughout the
entire sounding region.
It is sufficient to specify here that the final
field of view must allow for observation of birds throughout the entire
sounding region and inseots at least in the lower regions.
32
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33
The Bain apeoifioation on the system is that it prosent the region of
the sky viewed by the radar at a point sufficiently close to the radar so
that it may be visible to the operator at all times.
Further, there must be some way of indicating the time and date on the
frame to permit ex&ot correlation between photographs and the chart
recorders associated with the radar reoeiver.
,Station 4» 3 Baalai.flflaalfocatifflaa
There are three possible means by which the above could be aooomplished.
The most straight-forward of these would be to construct a tube without
lenses.
If it was desired to produce an image on a ground glass plate or
photographic emulsion, a telescopic lens system might be considered.
The
third method would be to employ the lenses in the periscope prinoiple to
conduct the required field of view from the laboratory roof to the
vicinity of the radar.
The advantages and disadvantages of eaoh of these
systems will now be investigated.
To obtain a 10° field of view using the first method an objective
aperature of approximately three feet would be required.
In addition to
this, an elbow in such a tube would necessitate the use of a large frontsurfaced mirror.
Consideration of a telesoopio system reveals that there are two types
available - namely the Galilean and Astronomical.
Further consideration
reveals that, for both types, a greater field of view will be obtained
with a straight tube without lenses (Sears, 1949)*
Hence neither of these systems will provide an adequate field of view
with an arrangement that oan be constructed in the laboratory.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The perl-
34
scope principle will therefor* he considered, in more detail* (Fig. 4.1).
The periscope has several important features whioh should he mentioned.
The main feature* as far as our oase is concerned, is that it is oapahle
of fitting a telescope with a comparatively large field of viev into a
long narrow tube (Jacobs, 1943)*
Of importance too* hut to a lesser
degree* is the faot that the field is uniformly illuminated.
In most
optical systems* in order to determine the field of viev* it is necessary
to consider the location and sise of the entrance pupil and window.
periscopes this is not the oase.
Here*it is possible to determine
With
in
advanoe what may he accomplished with a given number of lenses in a certain
tube without previous computation of the system.
In comparing two systems
the following formula may he usedt
L • kd*n/aV*
(Jacobs, 1943)
where L is the distanoe between the extreme lenses of the system* k is a
constant of proportionality* d is the diameter of the tube* n is the number
of lenses employed* a is the external aperature of the system and 7 is the
diameter of the field of view.
When a large number of lenses is used*
numerous solutions are possible) other factors being equal the best solution
is that of employing the weakest lenses.
aberrations and Petaval curvature.
uniformly spaced.
This is desirable to reduoe
In most practical oases the lenses are
For this condition the weakest lenses are obtained for
a solution in whioh a lens is looated in the plane of eaoh aperature stop
(Jacobs* 1943).
A system utilising three lenses was designed using the
above formula.
The resulting system is shown in figure 4* 3 and has been
designed for a 7° field of view.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
35
LENS
PERISCOPE
TU B E
FIG* 4-1
THE
PERISCOPE
P RI NC I P L E
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
36
seotion d.A !Eha Ylming .SgatflBB
In addition to the perisoope tube, a separate system must be included
to provide for photographing of the resulting image.
In that the above
system lends itself easily to use with a telephoto lens, this will now be
considered.
Suoh a lens must have a similar field of view for most
effioient usage of the light available at the perisoope eyepiece.
(Note:
the focal length of this lens was not oritioal sinoe the camera available
contained a variable position focal plane). Figure 4.2 was used to de­
termine the optimum focal length.
For a lens with a 7® field of view the
effect of increasing the fooal length has been indioated.
Thus if the
entire field of view is to be photographed the fooal length must be less
than 145mm*
However, if all the emulsion is to be sensitized, a lens of
at least 245mm must be employed.
Resolution limits as determined by the
grain size of the film, suggest that a lens of at least 196mm fooal length,
be employed.
tfhen the resolution of a photographic emulsion is investigated the
limit is found to be due, not to individual grains, but to dusters of
silver halide grains (Mees, 1952).
If the actual grain size determined
resolution then the system previously desoribed would produce an image of
a fly at 1500 meters whioh would be equal in size to a silver halide grain
(a few miorons in diameter). Instead, the grainy effect visible to the
naked eye is due to clusters of such crystals and depends on the emulsion
and development technique used (Kodak).
For a typically fine-grained
emulsion (Panatomio-X) the resolution is given as 150 lines per mm.
Thus
the image produced must be 1/150 of a mm if it is to be observed in the
grain.
Thus the limit of resolution as determined by the emulsion grain
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37
145 mm
1
196 mm
•245 mm
35mm
frame
FIG-4-2
DETERMINATION O F O P T I M U M F O C A L L E N G T H
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
38
would permit observation of an object the sise of a sparrow at the upper
regions of sounding area.
An image of a clock faoe was introduced into the periscope tube by
means of a collimating lens system and directed to one corner of the 35nm
frame with a small front-surfaced mirror.
The intensity of the image
produced is easily controlled by varying the wattage of the light source
illuminating the clock faoe.
The date was recorded by placing a date card
in front of the clock for one exposure at the start of each recording
interval.
In addition to this a Kodak A-25 red filter was obtained for oloud
photography.
Before the system will beoome operational* a unit to control the
exposure time and repetition frequenoy of the photographs must be con­
sidered.
The design of this device followed closely that reported in
'Eleotronics'* (1946).
A few simple modifications were necessary and the
revised circuit is shown in Figure 4*4*
A switch was introduced here as
shown* to permit manual control of the repetition frequenoy.
It may be
necessary to make even further oomponent changes if it is desired in the
future to reduce the rate at whioh exposures are taken.
With this completed the required lenses and tubing was obtained and
the entire viewing system constructed.
as shown in figure (4. 3).
The completed system was assembled
Several tests were carried out at this time to
determine its perfoxmanoe.
Section A. 5 YXWtVS girfOrMBM
In order to determine the actual field of view of the system* a
photograph was taken before the periscope was mounted vertically.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
This
39
photograph was also used to examine the uniform illumination over the
entire field of view.
The approximate angular field is readily determined
when the distance to the figures is known.
Knowing also the height of the
vertical post apparent at the right of the picture, the angular field was
calculated and found to he 7.5°.
Plate I
Field of View of Sky Viewer
The uniform illumination over the entire field of view is evident
from the photograph above.
After vertioal installation of the periscope was completed, a tethered
balloon was sent aloft above the radar to align the two systems.
Additional
observations through the viewing system at the time of the balloon flight
gave evidence that both systems were directed at the same region in the
sky.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40
VIEWER F IE L D (7°)
RADAR BEAM (1°)
■LENS
MIRROR
COLLIMATOR
CLOCK FACE
•FILTER
TELEPHOTO LENS
35MM CAMERA
FIG. 4 .3
THE
V IE W IN G
SYSTEM
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
41
Plate II indicates the relative angular fields of view as displayed
by the radar and viewing system.
The region circled in blaok represents
the radar antenna beam* assuming both antennas to be coincident.
Sinoe this
is not the exaot situation the region aoanned by the radar will be a circle
at one height only and at all other heights must be represented by two
circles.
However sinoe the antennas are in faot beside eaoh other, this
one oirole will approximate the aotual oase very closely.
This chapter has outlined the disadvantages and advantages of three
different viewing systems.
The design whioh best fulfills the specifi­
cations for the optical system in this experiment has been ohosen - namely
the periscope, and shown to perform satisfactorily under test.
Additional
examples of physioal objects, photographed during the aotual experiment,
are shown in Chapter VI.
Approximate exposure time in seoonds
oloud (light)
blue sky
filter in
1/20
l/4
filter out
1/100
1/20
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
42
Plate II
Relative Field of View of Radar and Sky Viewer
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
43
o
c.
o
CONTROL
-M o
OJ
I 10
CIRCUIT
FOR
Oq
AUTOMATIC
CAMERA
XT
3
^0
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
FIG. 4.4
TIMING
o>
CMPTER V
THE AHTI-GOBBELATION DEVICE
Sagiion S t
Sample records obtained by Reid, 19&0 are reproduced In figure 5*1*
It Is evident from this diagram that there are simultaneous fluctuations
occurring in all ohannels whioh resemble the angel eohoes very olcsely.
Indeed, on some occasions the angel reflections were completely mashed
by these undesired changes In signal level.
The possible causes of such unwanted signals were outlined in section
2.3.
Power supply fluctuations, changes in the antenna leakage signal
(i.e. transfer of signal directly from transmitting to receiving antenna),
klystron frequenoy shifts and incidental amplitude modulation as outlined
in Chapter III, are all potential souroes.
Much time would be saved in
the analysis of recordings and the validity of the results would be
greatly increased, if a device were available to eliminate such inter-*
ferenoe.
This Chapter deaoribes the design and construction of such a
unit.
aaatton
B agtttaffltata o f Ifra A a tl-C g rrtla to r
The fundamental requirement on a device to remove the simultaneous
fluctuations is that it accomplish this without eliminating a particular
44
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
450 M
300 M
g M p s s
A.
b.
INCOHERENT PRECIPITATION ECHOES RECORDED
JULY l3 /* 6 0 , 1 0 :30 A.M. E.S.T.
HIGHLY
FLUCTUATING
ECHO AMPLITUDES CLEARLY EVIDENT.
THIS SAMPLE
IS CHARACTERISTIC OF ECHOES FROM
SNOW , R A IN ,
DRIZZLE, FOG , H A Z E , E T C .
F Ig 5 « 1
EXAMPLES OF
COHERENT
COHERENT ANGEL ECHOES AT VERTICAL I N CIDENCE
JUNE 3 /> 6 0 , 2 : 2 0 P.M . E.S.T. SMOOTH
AMPLITUDE VARIATION IS USUAL FOR
ANGELS
OBSERVEDWITH NAR RO W -BEAM MICROWAVE
RAD ARS
AT V E R TIC A L IN C ID E N C E .
AN D
INCOHERENT
RAD AR
ECHOES.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
46
type of angel or destroying valuable information about individual angels
auoh as their amplitude, duration, height, etc.
a
further analysis of
previous reoords indloates the period of the unwanted signals to be of
the order of one second and as mentioned in Chapter II, they always begin,
reach a maximum and end at precisely the same time in all ohannels.
This
last faot is of major lmportanoe and one upon which the eventual design
of the anti-oorrelator will depend.
Sinoe the ezaot source of these signals (if indeed, only one source
is responsible) cannot be determined, the unit considered here must be
designed to eliminate fluctuations from one or all of the above-mentioned
sources.
Although it has been shown in Chapter III that a large number
are occurring due to the presence of parasitic amplitude modulation and
klystron frequenoy shifts, others may be occurring due to fluctuations
in power supplies.
Hence if this unit is to remove fluctuations from all
sources it must be incorporated at a point in the system after which suoh
fluctuations can no longer originate - namely, after the ten narrow-band
filters.
After the received signal has been directed into different
ohannels the production of fluctuations that will occur simultaneously in
all channels is impossible.
Safi-felon ‘it, 3 Pflaiga SQPaifltrattQaa
It was mentioned earlier that the only trait peculiar to the undesired
signals was the simultaneity of their ooourrenoe in all ohannels.
faot alone forms the chief design oriterion for this unit.
This
Further, it
perhaps should be mentioned that the amplitude is found to decrease with
ohaimel number.
However, this has been off-set by a corresponding gain
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
47
increase in each of the channel amplifiers, (see section 7 .4 , on radar
sensitivity curve).
far as the unwanted
Thus the amplitude is constant
in eachchannel
as
signals are concerned. On the other hand angel echoes
have "been shewn to ooour in from one to three or four adjacent ohannels
simultaneously, with the amplitude a maximum for the channel representing
the height olosest to the height of the reflecting region, and decreasing
for adjacent ohannels above and below.
It is necessary, therefore, to design a device which will be composed
of
(a) a sensing oirouit to detect an increase in amplitude occurring
simultaneously in all ohannels.
(b) a feedbacknetwork to be controlled by the sensing
oirouit and
which will eliminate the increase deteoted in (a)•
Ssnalng Qixoulli
Several alternative methods of sensing were considered first and will
be mentioned here.
An attempt to use the signal from channel eleven
(representing an height above the sounding region) was discarded when an
examination of this signal indicated a large noise background.
In addition
to this, all angels occurring at this height would also be Bensed and
returned in the feedbaok oirouit inoorreotly.
A second method attempted
to employ a simple feedbaok oirouit from the output to input of eaoh
channel d.o. amplifier.
This was proven difficult for several reasons.
First, the output signal from eaoh amplifier is 150 volts above ground
level.
Seoond, an attempt to use the B voltage in eaoh amplifier as a
common reference point, initiated large drifts in their output voltage.
The final design of a sensing oirouit is shown in figure 5.2.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Its
48
operation is as outlined 'below.
The output signal from eaoh of the ten
ohannel amplifiers is direoted through coupling oapaoitors to separate
thyratrons as shown.
The value of the oapaoitors was chosen so as to
permit the passage of the unwanted signals but to reject simultaneous
signal increases of the order of one minute.
Persistent angels have been
observed in all ohannels simultaneously.
Only when a signal is received on the oontrol grid of all the
thyratrons will they conduct.
fluctuations are sensed.
It is in this way that the unwanted
Aoourate oontrol of the bias on eaoh oontrol grid
permits detection of extremely low-amplitude fluctuations.
with a specially designed bias supply.
This is aohieved
When sensing ooours, the resulting
current flow is used to operate the feedbaok network.
Before this system
can be made operational however, some means of stopping the conduction
must be determined.
The mere application of a decreasing signal on their
oontrol grids does not extinguish gas tubes after conduction has been
initiated.
voltage.
this.
This is most easily accomplished by removal of the plate
Thus an additional thyratron is inoluded in the oirouit to do
Being biased so that it will oonduot as soon as plate voltage is
supplied, the above is accomplished by providing plate voltage for this
tube from the sane souroe which operates the feedbaok network.
Thus
when an inverted signal is applied to the oontrol grid of this tube, it
is held below 'out-off• while the feedbaok oirouit is eliminating the
undesirable signal and then turns the feedback oirouit off by conducting
as soon as the fluctuation has passed.
T h t F B lfllftflfc C lg fflllt
The feedbaok oirouit is shown in figure 5*3*
It applies a 'short'
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
49
2D21
From output o f
Edin amipl
p lifier
n u m b e r 10
,
^
0 -15 v
884
3
(inverted)
O
1\
2
F IG . 5 . 2
10
CHANNEL
CORRELATION DETECTOR
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
50
WAVE ANALYSER
CHANNEL 2
WAVE ANALYSER
CHANNEL 1
FIG. 5 .3
EDIN D.C.
AMPLIFIER
NO. 2
NVERTER
ANALYZING AND
SHOWING
ED IN RECORDP
NO. 2
ANTI-CORRELATOR
RECORDING
UNITS
ANT I-CORRELATOR
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
51
to the inputs of the Edin amplifiers until the unwanted fluctuation has
ceased.
Notei no short is applied to the input of the extinguishing tube.
The entire anti-*oorrelator system was incorporated in the radar
system in such a manner that it could be turned off and on with a simple
switch.
Under test it was found to operate exaotly as described and the
removal of simultaneous fluctuations of very low amplitude was observed.
Due to an oscillation in oh&nnel three (see Recommendations, Chapter VIII)
this unit has not yet been used during An experiment.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER VI
EXPERIMENTAL OBSERVATIONS WITH THE 6770 MC/3 FK BADAB
Sm U qb £»1
This chapter presents experimental observations of the miorowave
angel environment obtained from 70 recording intervals during the months
of May, June and July of 1962.
The radar installation permits only
vertical sighting of radio reflections.
Operating intervals were eaoh of
thirty minutes duration and included soundings taken in the morning* after­
noon and evening for 25 reoording days.
Observations on days of heavy
overoast or incipient preoipitation were not included in the study.
Several examples of soundings taken during a period when frontolysis
(the breaking-up of an air mass boundary) was occurring sure lnoluded also.
Two examples of reflections from birds are given shoving the characteristic
signal return for suoh objects.
Finally* soundings were taken with all
air masses within the radar range exoept continental Arctic (there were
no oases when this air mass came sufficiently far south).
Installation of an optical viewing system permitted photographs of
the sounding region to be taken every 30 seconds as desoribed in Chapter
IV.
A rangefinder desoribed in Chapter I* was employed to determine oloud
heights on the oocasion that they were present.
The technique of identifying radio reflections is desoribed in
52
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
53
section 6.2.
Only those signals which oould be observed, olearly above
random noise were included*
The duration of eaoh was determined as out­
lined in the following seotion.
As demonstrated in seotion 2.2, the range
is determined from that spectrum oomponent for which a peak amplitude is
recorded.
Information on surface air temperature» air pressure and relative
and speoifio humidity was taken on the laboratory roof immediately prior
to or preoeding eaoh recording interval.
The prooedure involved here was
outlined in Cb*”ote~ I.
Frontal positions were estimated as outlined in
the following seotion.
All additional surface air information was supplied
for a site six miles east of the radar installation* by Mr. D. Soott,
Meteorologist at the London Weather Office.
Seotion 6.2 Teohnloue of Analysis
Analysis of the recorder charts was accomplished with the aid of a
light-table.
In this process the traces from two channels are superimposed
over a fluorescent soreen.
This procedure is pictured in Fig. 6 .1 . Here
the recorded signal from ohaxmel seven or eight has been superimposed on
the signals from ohannels one to five inclusive.
Mo two ohannels were
oompared by superposition unless they were separated by at least two
ohannels.
This prevented individual angels from being over-looked if
signal components appeared in more than one channel.
The trace in channel one is typioal of the simultaneous fluctuations
desoribed previously and it is evident that the same background signal is
present in channel seven.
Departures from this trace are easily seen when
a comparison of this type is made.
Some typioal angel reflections are
also indicated in Figure 6 .1 .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
54
Figure 6 .3 is an example of simultaneous fluctuations when no angels
are present.
Here, ohannels one and seven are oompared and found to have
almost exact traces.
The duration of a particular refleotion was measured from the point
at whioh the two traces separated to the instant at which they were again
coincident.
Eaoh snail soale division represents 0.2 seconds.
EhQtQBBHMQ E lla Analysis
The film used throughout the recording periods was developed at DESS,
Ottawa and was analysed on a speoially designed 35 mm film viewer.
Here,
eaoh frame is viewed individually after projection onto a frosted-glass
screen approximately 40 on square.
Objects such as insects and birds are
readily detectable by their characteristic appearance.
Whereas dust
specks on the negative appear dark, inages of physical objects appear
light.
In the event of soratohes on the emulsion however, this is not
the oase.
Whenever any doubt existed, a microscope was used to examine
then more closely.
Scratches are easily deteoted with the aid of a mioro-
soope.
It is pointed out here that dust specks that might appear from tine
to time on the lenses in the viewing system, do not produce an image in
the focal plane of the camera and hence do not appear on the photographs.
The near point (for an object to be in focus) is approximately 100 feet
above the objective end of the viewer (determined experimentally).
An example of a bird photographed at an height of approximately 500
feet above the radar is shown in Plate III enlarged 18 tines.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
55
Channel
F IG . 6.1
S U P E R P O S IT IO N O F R E C O R D IN G
CHARTS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
56
,j j
*
Plate III
Bird Photographed Through Sky Viewer
Determination of Frontal Positions
The Canadian olinate is the result of complex seasonal movements
of four main air masses.
These are as followss continental Artio,
maritime Aratio, maritime Polar, maritime Tropical.
The simplest
atmospheric model for the various air masses is depioted in figure 6.2.
Although the physical properties of eaoh are quite different, all four
are considered horisontally uniform and suffer little modification hy
contact with one another.
As a consequence, the Boundary layer Between
air masses is preserved for long periods at a time (Berry, Bollay and
Beers, 1945)*
The transition region Between different air masses is known as a
frontal sons or simply a front.
A typioal frontal seme may extend
thousands of miles in the horisontal.
In the vertical however, the
thickness is usually Between 1000 and 3000 feet, (godson, 1951 and Sawyer,
1955)*
The names given to the fronts are those associated with the oolder
of the air masses whioh lie on opposite sides of the frontal sons.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
57
A IR -M A SS TYPE AND ASSOCIATED
FRONTS
polar
front
1■
mT
mP
arctic
sc- f r o n t
mA
FIG. 6.2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
58
FCA (frontal contour aloft) charts are compiled from country-wide
upper air soundings by the Dept, of Transport, Meteorological Branch,
Maiton, Ontario.
Frontal positions used in this researoh were inter­
polated from such maps.
Frontal heights below 1000 meters were only
estimated since the charts are inaccurate in this region.
As a result
of this inaoouraoy, an exact correlation between certain areas within a
front and angel activity could not be made.
Figure 6 .3 is a comparison of the simultaneous fluctuations in
ohannels 7 and 1.
Here no angels are present.
This exact correspondence
between both ohannels is necessary if angels are to be determined reliably
by viewing overa light-table as desoribed earlier.
Figure 6 .4
signals.
depictsan example of many high-frequency,low-amplitude
These were too many to tabulate individually and have been
previously considered as originating from sources other than that of
angels (miorophonios in the klystron etc.).
As mentioned in Chapter II,
the thermal noise level cannot be attained with amplitude modulation present
on the carrier wave.
To show that the high-frequency signals of Figure
6.4 are not due to receiver noise it la only necessary to remove the
klystron modulation.
It has been determined experimentally that they are
produced only when the modulation is applied and hence are not due to
receiver noise.
Furthermore, suoh high frequenoy signals are not always
present as is seen fron Figure 6 .5 .
Figure 6 .5
is also typioal of records obtained forperiods
high angel inoidenoe.
of very
Even when two adjacent ohannels are considered,
as here, little or no correlation is observed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
59
Tl ME
SIGNAL AMPLITUDE
F I G . 6-3
SIMULTANEOUS
FLUC TU A TIO N S
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
60
T
FIG. 6 - 4
F IG .6.5
T
H IG H -F R E G U E N C Y
RANDOM
SIGNALS
FLU C TU A TIO N S
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
61
Examples of reflections fro* birds are shown in Figure 6.6.
Figure
6.6 (a) represents the signal produced when a bird was observed at a n
height of approximately 300 asters above the radar.
Here the maximum
component appears in channel 2 (representing an height of 300 asters).
The presence of a large signal in ohannels one to five is possible due
to the large signal return from suoh physical objects.
Figure 6.6 (b)
is an example of a reflection from a bird at an height of approximately
30 meters.
aiatlP B
3 AaaoelaUpB o f A n a l .Ira lflta g t *lUQ. 3u cfact Y arlaU ea
This section presents information obtained from the 70 recording
intervals included in this study. An attempt is aade to present the data
in a manner whioh will facilitate comparisons to be aade with the previously
obtained results of Beid.
All aeteorologioal variables have been recorded
on the laboratory roof immediately prior to or during eaoh recording in­
terval.
Results whioh are not in agreement with results obtained by
previous investigators are presented first.
Results whioh show a notable
trend or whioh are in agreement with earlier work are then presented in
aore detail,
(a)
Variation of angel Incidence with Surface Air Temperature
An indication of the variation of average angel inoidenoe per two
minute intervals with temperature is shown in Figure 6.7.
Eaoh bar
represents the average angel inoidenoe per two minutes of sounding time
for all recording periods at one speoifio temperature.
The solid bars
represent reflections of durations from 3 to 60 seoonds while the out­
lined bars are representative of all transitory reflections observed.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 6.6
Reflections from birds above the radar
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
63
An irregular association between angel incidence and temperatures
from 60®P to 90*F is indicated.
(b) Association of Average Angel Inoidenoe with Relative and Specific
Humidity
Angel inoidenoe as effected by Relative humidity is shown in Figure
6.8.
A alight inorease in inoidenoe with humidity is suggested.
Again,
angels with durations between 3 and 6o seconds are indicated by the
darkened portion of the graph.
A similar inorease of inoidenoe with
relative humidity is suggested here.
Angel inoidenoe as affected by specific humidity is presented in
Figure 6.9.
Here only an irregular association between inoidenoe and
specific humidity is observed.
(o) Association of Angel Inoidenoe With Barometric Pressure
Angel inoidenoe and barometric pressure are plotted in Figure 6.10.
An irregular variation is indioated.
Some additional observations associated with this graph will be men­
tioned here in order that they may be used in the interpretation later.
An inorease is observed at the low pressure end of the graph*
This has
been found to be entirely due to the passage of a front on May 23* A
large number of radio reflections were reoeived from the region of this
front, as well as from regions below it.
The peaks at 29*20 and 29*30 inches of mercury are due in a large part
to two phenomena.
A front undergoing frontolysis was stationary above
the radar, and within the sounding region, for the period beginning July 7
and ending July 9*
Many angels were recorded from the region of this
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
^allan qe is
■angefs o f d u r a t i o n
> 3 seconds
AVERAGE
ANGEL
12'
I N C ID E N C E
( p e r 2 m in .)
rf
8-
i
!-;
II
60
70
90
80
E M P E R A T U R E (°F )
FIG. 6-7
AV ER AG E
ANGEL
i N C ID ENCE VS. TEMPERATURE
AVERAGE
ANGEL
20i
IN C ID EN C E
f
(pe" 2 m i n . )
-
n
10*1
! 1
0-
NO
RELATIV E H U M I D I T Y
FIG. 6 - 8
T2
S P E C IF IC HUMIDITY
A N G E L INCIDENCE VS. F I G . 6 - 9 ANG. INCID. VS.
R E L A T I V E HUMIDITY
SPECIF. HUMID .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
65
AVERAGE
ANGEL
NCIDENCEi
(per 2
min.) 8” ]
29.00
2 9.2 0
2940
2960
BAROMETRIC PRESSURE
(inches
of
m ercury)
FiG-6-10 ANGEL INCIDENCE VS- PRESSURE
A
A
AVERAGE [
ANGEL 12i
INCIDENCE!
(per 2 min.).
4-!
cA
mA
m*P
rnT
AIR MASS
FIG-6-11 ANGEL INCIDENCE
VS- AIR MASS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
66
■boundary.
The seoond phenomenon ooourred after the rainfall on July 9.
This was the only example in whioh soundings were taken at a time when
the ground was wet.
(d) Association of Angel Inoidenoe with Air Hass
No data was obtained for oA air.
The association between angel in­
oidenoe and the remaining air masses suggests a slight inorease for mp
air.
It is noted here* for use in later interpretations, that mT air
was within the radar sounding region in the month of May only.
The number of recording periods in eaoh air mass is indicated below*
Air Hass
oA
mA
mP
mT
Sounding Intervals
0
20
27
14
(e) Association of Angel Inoidenoe with Wind Speed
The association between angel inoidenoe and wind speed is given in
Figure 6.12.
No soundings were taken for wind speeds of 20 mph.
irregular variation with inoidenoe is apparent.
Again an
A slight decrease is
observed for high wind speeds and again at low (5 mph) speeds.
Angels of
durations between 3 and 6o seoonds show increased activity at 0 wind speeds
and again at 15 mph winds.
This completes the comparisons between angel inoidenoe and surface
variables.
The parameters investigated in this comparison to determine a
possible relation between eaoh parameter and the inoidenoe of transitory
microwave reflections were as follows* surface air temperature, relative
humidity, specific humidity, barometric pressure, air mass and wind speed.
An interpretation of these comparisons will be presented later.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
El
AVERAGE
ANGEL
INCIDENCE
13 D U R A T I O N
(1 t o 6 0 s e c )
I
(3 t o 6 0 s e c )
D U R A T IO N
I T U r l D U R A T I O N (3 t o 6 0 s e c ,
u n d e r w e t-g ro u n d co n d ­
itions)
( pe r 2 m in ,)
10
WIND SPEED
(m ph)
FIG-012
ASSOCIATION
OF AVERAGE ANGEL INCIDENCE WITH WiNDSFEED
CD
•^1
68
SftotlQn
AaaoQlaUon of basal iMlflaagi .irlth Air Straatttcrifrm
^bftJShdu
The analysis reported in this section is inoluded so that the effects
of olouds and fronts cm the angel environment oan be investigated.
Store
the comparisons are made over individual recording intervals as opposed
to the comparisons of section 6.3 which inoluded the results of the en­
tire experiment in eaoh graphical representation.
The six graphs shown belov indicate observed variations in inoidenoe
profiles.
It is apparent from these examples that while seme profiles
show reflection concentrations at one height only (July 11 and July 16 )
others indicate several such localisations of angel activity (May 22
and July 10).
In the results reported belov these will be referred to
as regular and irregular profiles respectively.
A more detailed interpretation of these profiles will be presented
in Chapter VII.
As was mentioned previously a continuous graph was plotted shoving
the position of fronts and clouds relative to the recording intervals.
A
portion of this graph is reproduced in figure 6.14. All information
pertaining to the air structure above the radar and which was taken from
the continuous ohart, is inoluded in the following tablet
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
VERSUS
HEIGHT
O
P R O F ILE S
-»
ANGEL
INCIDENCE
(no./2 m in )
69
FIG. 6-13
INCIDENCE
O
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
HEIGHT
(meters)
1500-
5 00 -
0000
0600
1200
1800
0000
0600
1200
1800
T IM E (h o u rs )
1500-
500July 16
FIG. 6-14 ASSOCIATION
J u ly 17
AVERAGE ANGEL I NCIDENCE (per 2 min.)
BETWEEN ANGEL INCIDENCE AND FRONTAL Z O N E
3
71
2ktmia1ilQn ./ar,,gatrrA lr atamtitta JPata
Example* (1)
(2)
Angels In One Air mass.
Transitory angels associated with cumulus oloud in one
air mass*
( 3)
Transitory angels associated with fronts.
Example number
1
2
3
Sounding Intervals
9
9
12
Intervals with Angels
9
9
12
Intervals with Regular profiles
8
0
1
Intervals with Irregular profiles
1
9
11
Intervals with oumulus clouds present
0
9
0
Intervals with fronts present
0
0
12
Position of angel occurence
all
heights
below
olouds
in frontal
sone
Slttiloa o**? Singular Transitory Phenomena
On July 9 many transitory reflections of long duration (3 to 6o
seconds) were observed during a period when wind speed was approximately
15 mph.
This example represents the only case in which the ground was
vet during the sounding interval.
On July 17 a series of soundings was taken in vhioh profiles with
perturbations were obtained for single air mass conditions.
These intervals
were also unique in that they supplied the only example in whioh details
in a transitory angel inoidenoe profile were found to persist for more
than one recording period.
On the other handy the formation of this
persistent irregularity took less than one hour.
A section of the continuous chart included in Figure 6 .1 4 illustrates
the above case.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
72
.arafeLflnJu6
A a a fts la U g a q£
Lassl ja 9 ifta a ffa .iftth , Q th ftr VasA&laUB
This section presents Information on the aaaooiation of angel in­
oidenoe with the following non-related variablea: height of reflecting
region, duration of reflections, and month.
In addition to this,
information on inoidenoe fluctuations within the individual recording
intervals is included.
(a) Variation of Angel Inoidenoe with Height
The ohange in angel inoidenoe with increasing distance above the
radar is shown in Figure 6 .15 . Here the average inoidenoe per two minute
interval is plotted for eaoh channel height for all sounding intervals
inoluded in this study.
An inoidenoe peak is found to occur at an height
of 300 meters with a decrease in inoidenoe above and below this height.
Additional inoidenoe profiles for individual reoording intervals
are given in Figure 6.13.
An interpretation of these curves is presented
in Chapter VII.
(b) Variation of Angel Inoidenoe with Month
This relation is depicted in Figure 6.16.
Angel activity is found to
inorease during the month of July.
(o) Variation of Angel Inoidenoe with Angel Duration
The association of inoidenoe with reflection duration is presented
in Figure 6 .17 . Here both variables are plotted on logarithmic soales
with all reflections observed during the entire experiment included.
In addition to this the results of a simultaneous experiment in­
vestigating persistent angels (Johnston, 1962) have been inoluded.
The
curves are straight lines in both oases and the line joining them is
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
73
FIG-6-15 ANGEL INOIDENOE VS- HEIGHT
AVERAGE 3.
ANGEL
INCIDENCE
(per
2
min.)
fi
o
300
600
900
1200
HEIGHT
(m eters)
AVERAGE 24'D
ANGEL
INCIDENCE
(per
2 min.)
I’IN
MONTH
FI0-6-1G ANGEL INCIDENCE VS- MONTH
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
74
I
o ^
0 O c/>
O -r4
O
/ qt
CD
© O© /
CM
0
©
0)
© /
/
//
c
o
u
o
o
0
c/)
o
o I—
C
o
/
-&->
(/)
©
/
/°
c
JC
/' 2o
o
“ 3
‘2 <
q:
3
Q
3
O
»
<
5
Q
0
LlJ
P
0
/
o <
o
_J
L'J
O
c£-
/
CO
/
>
L eJ
o
/
/
2
L'J
Q
O
.-o
O'"
I'J
<
6
0
j _rr"rn r
o
o
o
(J
— I^
~grrrp
r
|ll I l“T
O
<r lu
Q
D f)
^1-
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
75
found to be continuous.
It is noted here that the large inorease apparent for reflections of
one seoond duration is not real since this point is representative of all
reflections with durations less than one seoond as well.
(d) Inoidenoe Fluctuations Within the Eeoording Interval
Graph 6.18 is a comparison of angel incidence variations observed
during recording intervals for three separate occasions. Here the
total
number of reflections from all heights has been summed for eaoh two
minutes of reoording time.
The average or mean 2-minute-inoidenoe for
the entire reoording is indicated by the dashed line through eaoh graph*
In addition to this the mean deviation (MD) about this average has been
calculated and is inoluded for eaoh of the three oases.
An osoillatory fluctuation with periods from five to ten minutes is
observed within each 30 minute reoording interval.
This chapter has presented an analysis of the data obtained through­
out the experiment.
Two different approaches to the analysis have been
described) one considering all the data to determine the assooiation
between angel inoidenoe and surface variables, another in whioh individual
reoording intervals were reviewed.
The first method permitted the comparison of air temperature,
humidity, air pressure, wind speed, and air-mass with average angel in­
oidenoe.
The seoond method allowed for investigations of the effect of
air-mass boundaries within the radar range, wet ground conditions,
frontolysis, and the convective lifting associated with cumulus cloud
formation.
In addition to this, information on the maximum deviation
from mean Inoidenoe in a typical reoording period was presented ae was
angel duration and height.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
so-
INCIDENCE
( p e r tw o
MD = 0 . 3 8
min.)
MEAN
40-i
MD = 0 .3 0
20i
-M E A N
40
MD = 0 .2 S
to
24
- O
RECORDING TIME
(minutes)
FIG*6-18 INCIDENCE
RECORDING
FLUCTUATIONS
FOR INDIVIDUAL
INTERVALS
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER VII
INTERPRETATION OF THE EXPERIMENTAL OBSERVATIONS
Seotion 7.1
Introduction
This chapter presents an Interpretation of the results reported
in Chapter VI. The interpretation oust he considered tentative in
view of the United period of study.
aaatton T.2 Inltngitation of the SftfaQt of Sugfaot YaclafrUa
A slight inorease was indicated for the association between trans­
itory angel inoidenoe and relative humidity.
The associations of angel inoidenoe with surfaoe air temperature,
air pressure, air mass, specific humidity and wind speed produced
variations which were irregular. The need is suggested here for the
consideration of nonrsurfaoe variables.
SttsUon 7. \ IntwpntoUan
Aig-Maaa AeaoQiation
(a) Soundings in a Single Air-Mass
Angel inoidenoe profiles obtained under olear-air oonditions
usually show a random or uniform scattering of angel sources within
the sounding region. However a few oases with irregular inoidenoe
profiles were observed, indicating that such looalisation of reflections
can ooour without the presence of olouds or frontal sones.
77
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The ooourrenoe of angola In all olaar-air soundings, vary oftan
with an high mean inoidenoe, suggests that although localisation of
reflections may he aooounted for hy environmental situations most often
associated with clouds or fronts, the actual angel sources may he merely
altered hy these factors and not necessarily horn of them.
(h) Transitory Angels Associated with Fronts
Godson C1951) and Sawyer (1955) give a picture of a frontal sone
which has the following characteristics* the sone is usually some 2000
to 5000 feet in depth* generally the air within the zone is very dry
exoept at levels below 2000 or 3000 feet.
In addition to this, the sone
has been found to possess strong wind sheer which is discontinuous at the
boundaries.
While refractive index profiles indicate a uniform refraotivity
lapse rate within the frontal zone, abrupt changes in lapse rate are
usually observed at the upper and lower boundaries.
Irregular inoidenoe profiles were almost always observed when fronts
were within the radar range.
Although the inoidenoe localisations oould
not be associated with definite parts of the front, suoh irregularities
could always be attributed to the frontal sone.
(o) Transitory Angels Associated with Low, Cumulus Clouds
The atmosphere loses ; heat into outer spaoe by infra-red radiation
at a rate sufficient to cool it about ten Centigrade degrees every week
(Ludlam and Scorer, 19&0) • This loss is balanced by the stirring upwards
of air from the sun-warmed ground.
Although the upward movements may
occur in many different ways, perhaps the most oommon fair-weather prooess
is that of oonveotive lifting (Weather Ways, 1957).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
79
Convection ooours in the atmosphere when it is heated at the earth's
surface, as when the ground is warmed in sunshine*
In this process,
large volumes of air rise from the surfaoe layers and penetrate into and
mix with the oooler air above (Ludlam and Scorer, 19^0).
Above the
condensation level (the level at which the air becomes saturated) these
volumes become visible as cumulus olouds.
The process described above is depicted in Figure 7*1*
At the con­
densation level the volume becomes saturated and a cumulus cloud forms)
the bulging upper surfaoe of the cloud shows the position of the rising
air but the base of the oloud is defined by the condensation level and
is strikingly flat.
Concentrations of angel eohoes have been observed on all oooasions
when vertical (cumulus) olouds were present within or above the sounding
region.
An association between such echo concentrations and the in­
visible volumes of arising air is therefore suggested.
This is as
suggested by Plank (1956)•
(d) Singular Transitory Phenomena
As a result of the sounding interval on July 9, wet-ground conditions
appear to favour the production of longer-duration angels.
agreement with the results of Reid (19&0).
This is in
While approximately one-third
of the angels with durations greater than three seconds were obtained
for aero wind speeds, almost as many were obtained during this unique
reoording interval when the ground was wet.
No explanation of the persistent detail observed in the July 16-17
inoidenoe profiles is apparent at this time.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
80
Condensation level
A
\____
If
\
FIG. 7.1
The formation
lifti ng
of cumulus clouds by convective
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
81
Saotion i-A.
latarpgttetloB ot thi AasaslaUpa gf Anal iagidiagf ,.¥itb
galgfat and PuxaUon
(a)
In order to determine the effect of height on angel inoidenoe, it
is necessary to consider all factors that might influence the detection
of angels in eaoh channel*
First, consider a flat reflecting area filling at least the first
Fresnel sone of the antenna beara (Atlas, 1960) • Suppose now that this
target is sowed upwards from the radar through the sounding region*
The received signal power is found to decrease in a l/r* relationship,
where »r* is the height of the reflecting region (Kerr, 1951) • The
oean signal powers associated with eaoh channel are indicated in Figure
7*2 (a), (assuming the same gain in eaoh ohannel).
It has been determined
experimentally that the ohief noise componenta in eaoh ohannel arise as
a result of the amplitude modulation present on the FM carrier*
The AM
is observed to approximate a 'sawtooth* waveform as indioated in Figure
7*2 (b). Henoe the amplitude of the noise components in eaoh channel
are obtained from the Fourier harmonic components present in this saw­
tooth waveform.
The amplitude of the Fourier harmonics is found to
decrease as l/n, where n is the ohannel number.
Their power will thus
decrease as l/n?»
Figure 7*2 (o) shows the mean angel signal power and its associated
noise components for eaoh channel*
Now in order that the charts may be analysed on a light-table as
described in seotion 6.2, the individual ohannel &iins were adjusted so
that the simultaneous fluctuations had the same amplitude in eaoh.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The
/
82
SIGNAL ‘
POWER
(d b )
POWER
-
20
-
10
CHANNEL NO.
40-
40signal
noise
20
FREQUENCY
( m c/s)
-
36-
■32
- O
c
CHANNEL NO.
FIG. 7 2
INTERPRETATION OF ANGEL IN C ID E N C E
VERSUS HEIGHT CURVE
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
83
result Is shown In Figure 7.2 (d)*
Observation of the recorder charts reveals that the mean signal
amplitude is approximately constant with each channel as predicted from
Figure 7.2 (d). This suggests that no convergence or divergence of
the radio waves takes place upon reflection.
A small decrease in the signal-to-noiae ratio is observed at 150
meters (ohannel 1) in Figure 7.2 (d). This is due to non-uniform
illumination when the target is within the Fresnel region of the antenna
(Silver, 1949).
Thus, at least for heights above 150 meters, the inoidenoe profiles
presented earlier give a oorreot picture of the angel concentrations at
different heights.
The large decrease apparent in the experimental
results for channel 1 cannot be explained at this time.
(b) Association of Angel Inoidenoe and Angel Duration
Reid (19<>0) has shown the average transitory angel duration to occur
around 1 seoond.
In that investigation the duration was determined to
the nearest 0.2 seconds.
noarest seoond.
This author has reoorded duration only to the
The reason for this will now be explair.od.
It becomes apparent in the chart analysis that many of the reflections
of durations less than one second are barely visible above the background
noise level.
Indeed, there is no means of determining at present that
this background noise is not made up of a vast number of shorter duration
reflections which would indicate an even finer microstructure in the
troposphere (see Figure 6 .4 ).
Further investigation of this must remain
until the ultimate radar sensitivity is achieved.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
84
A logarithmic decrease in Inoidenoe with increasing duration is
indioated in Figure 6,17*
The results obtained from a simultaneous
experiment (Johnston, 1962) in whloh persistent angels were investigated
allows us to extend this curve from 6o seconds to 17 minutes.
Favourable
agreement between the two experiments is observed which indicates that
there is no sharp dividing line between transitory and persistent angels
and henoe no obvious reason for division as far as aotivity is concerned.
The faot that a continuous spectrum of angel durations is obtained
suggests that a similar spectrum may be found in the angel source itself.
This might indicate an association with the scale size of turbulent
eddies in the troposphere.
(o) Interpretation of Inoidenoe Fluctuations within the Reoording Interval
Gossard (1953) reports oscillations in air pressure and wind speed
with periods from 5 to 10 minutes.
A barograph and damped anemometer
were used in that experiment and the evidence obtained was found to favour
waves as opposed to drifting oonvection oells as a possible souroe of the
perturbations.
It is suggested here that the inoidenoe fluctuations observed within
eaoh reoording interval might be assooiated with these 'gravity waves',
as th e y a re c a lle d .
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER VIII
REMARKS AND RECOMMENDATIONS
Section 8.1
intoMtoatlan
This ohapter presents a summary of the conclusions derived from the
results of the experiment deeorihed in the previous chapters.
In addition)
several ohanges in experimental equipment and technique are suggested.
SftgtlQn S. 2 Remrftfl
The following conclusions were drawn from the experimental results.
Although lack of sufficient data prevented the investigation of a
diurnal variation in angel inoidenoe) this writer reports the deteotion of
night-time eohoes in agreement with Plank hut in disagreement with Reid's
results.
Further disagreement with Reid is seen in the faot that 20^ of
the eohoes reported here were of a duration in excess of three seconds.
Hie reports less than % in this range.
of analysis used hy Reid.
This is possibly due to the method
It was determined hy this author that angels
with durations in excess of three seconds oannot he deteoted unless the
method of ohannel superposition is employed.
No regular aseooiation between surfaoe air temperature from 6o to 90
degrees F (an he reported.
Angel inoidenoe was found to deorease at high wind speeds as previously
reported by Flank (1956) and Reid (1960) •
85
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Average angel inoidenoe was found to exhibit a decreasing distribution
with height.
Individual inoidenoe profiles often gave evidence of con­
centrations of reflections with height.
Such localisations were found, in
general, to he attributable to activity within frontal zones or activity
associated with oumulue cloud formation.
The air volumes or 'thermals'
involved in convective lifting of the air are considered as possible
sources.
This supports the suggestion by Scorer and Indian (1953) and
Plank (1956).
A similar experiment investigating persistent angel phenomena
(Johnston, 1962) has shown that when persistent angels ooour, the abovementioned localisation of reflections is usually observed in the transitory
inoidenoe profile.
Random or uniform angel distributions are associated with clear-air
single air-mass conditions, however a few profiles with perturbations
have been observed when only one air mass was within the sounding region
and skies were clear.
Vet ground conditions have been found favourable for the production
of long duration transitory reflections in agreement with the results of
Beid (1960).
A continuous spectrum of angel durations has been observed.
As a
result, it has been suggested that this might be indicative of a similar
spectrum in the eddy scale aise of atmospheric turbulence associated with
the angel source.
Average inoidenoe fluctuations were found to be 31^ of the mean
inoidenoe for an individual reoording interval.
In addition to this, the
inoidenoe fluctuations were found to exhibit a sinusoidal nature with
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
87
periods of from 5 to 15 minutes.
A possible connection to atmospherio
pressure inicropulsations is suggested.
RECOMMENDATIONS
As a result of the experiment described in the previous chapters,
several recommendations are suggested here, which might be of value in
future investigations of this kind.
Equipment modifications designed to
inorease the radar sensitivity are strongly urged.
Additional changes in
experimental techniques and analysing procedures are mentioned.
Three plans are suggested here as possible methods for Increasing the
radar sensitivity.
(a)
The first of these is designed to eliminate the incidental AM intro**
duoed hy the microwave filter of the looal oscillator system (see Chapter
III). This could be accomplished by inoluding an AM compensator of the
type described by Reid (1960), immediately after the microwave bandpass
filter and before the balanced mixer (see Pig. 8.1).
As indicated in
section 3*7, proper operation of the AM compensator necessitates that the
AM be symmetrical about the center frequency.
Figure 3.
suggests that it
is only necessary to inorease the radar frequenoy from 6770 to 6772 mo/s
to ensure that the upper sideband, now at 6800 me/s, is located at the
center of the filter passband.
(b)
Secondly, the possibility of fluctuations arising from thermal changes
in the klystron oavity should be reduced by employing a more stable
klystron.
The Varian VA-244B Reflex Klystron which is conduction oooled,
long-lived and extremely stable, is suggested as a possible replacement.
(o)
Further, it was suggested in Chapter III that the Pound stabilisation
unit was not completely eliminating klystron frequenoy shifts.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Suggested
88
G y po l i n e
AM
Ferrite
Rotato r
M ic ro w a v e
Co mpensator
C rysta I
D etector
Frequency
M e te r
B alanced
Mixer
FIG. 8.1
AM Compensator for removal of
introduced by Microwave F ilte r
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
AM
89
improvmentB in this regard art given by Johnston (19&2) •
With respect to the time of year in which to conduct further exper­
imentsy the spring and summer seasons were again found very favourable
for angel studies as far as this geographio location is oonoemed.
It is
suggested here also that further transitory investigations should include
evening soundings as well as the usual daytime recording intervals.
The lengthening of recording intervals to at least two hours is
suggested if information on short-term environmental changes is desired.
Furthermore» it is urged that all meteorological synoptic features such
as temperature) wind speed) pressure etc.) be reoorded on a continuous
basis so that the angel incidence fluctuations reported here) might be
investigated more fully.
The investigation of micropulsations in the air pressure (gravity
waves) is suggested as a possible study to be included in future work.
Finally) an attempt must be made to eliminate low-frequency
oscillations (one to ten cycles per seo.) that ooour from time to time in
several channels of the reoording system.
It is believed that these arise
as a result of the received signals beating with the harmonics of the linefrequenoy - especially the fifth harmonic which is approximately the same
frequency as the channel three signal.
A 60 ops notch filter has been
incorporated at the output of the audio amplifier) and a sisable reduction
in the oscillations was observed.
Further reductions must be made) however)
if the anti-correlation device is to function properly.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX I
This appendix contains the circuit diagram of the 100 ops sweep
generator which was modified by this author in the course of the experiment*
Cirouit diagrams of all other units associated with the 6770 mops radar are
included in (Seid, 19611 Appendix I).
90
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
91
o
<r
uj
2
UJ
a
UJ
UJ
£
v>
al
u
o
o
2< o
CM
o
<0
-£>-s2-S
v - / a,</>
m*
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX I I
This appendix outlines the method devised for tuning the balanced
modulator.
Steps three and four may he omitted In a routine balancing
prooedure hut must he included after crystals have been replaoed.
Procedure for 'failing the Modulator
(1)
Adjust the klystron frequency so that 6770 mops appears at the peak
of the mode.
(2)
Adjust the frequency meter in the Found system to ensure that the
region of the mode being swept is located symmetrically about the
peak frequency.
Notes
If major adjustments are required in the above steps, then 1 and 2
should be repeated several times.
( 3)
Adjust the 100 ops sweep generator for a one megacycle frequency
deviation in the transmitted signal.
(4)
With a frequency meter and crystal detector connected immediately
adjacent to the microwave filter input, adjust the tuning stubs on
the balanced modulator until the sideband frequencies are deteoted
when the frequency meter is adjusted.
Depression of the carrier may
also be observed in this manner.
(5)
Beoonneot the local oscillator to the balanced mixer.
(6)
Remove the receiving antenna wave-guide and replace with a 'dummy* load.
92
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
93
(7)
Preoiae tuning of the modulator is now accomplished, using the radar
receiving system as a deteotor, hy adjusting for minimum signal
level on the multi-meter.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
APPENDIX III
Calculation of the affect of nonunifora target illumination in the
Fresnel region of the antenna.
The diffraction integral for the optioal-Fresnel region may be written
as
TT
SI
p
where ((•
1 a-^8
21
E ~ J P (f »>|)
a (Cos* + i0 . s) df dn
) are the co-ordinates of a point in the aperature
(x,y,z) are co-ordinates of field point P, and
Up is the diffraotion field, the field over the aperature is
designated by P(^ »>i). (Silver, 1949)
This equation differs from the expressions of the Fresnel field
generally found in the literature in the presence of the tern i .s whioh
arises from a nonuniform phase distribution over the aperature.
The method of Fresnel cones used extensively in optics affords a
simple physical basis for understanding the effects that are observed in
the Fresnel region.
Using this method the amplitude of the field along
the axis is found to pass through maxima and minima, the maxima coming at
points that subtend an odd number of Fresnel zones.
To make a more quantitative evaluation we must oonsider the aotual
values of the field intensity and the gain.
For this purpose we will
start from the Fresnel approximation equation given above, which in the
94
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
95
prosent oase using polar oo-ordinates, takes the fora
a
£ 2
/ *
/*
O
where ^ - <o oos $«, -v\ K
sin pf '
» $ 0 " aperature field distributionfthis is integrated for
unifora illumination giving
U - 2j sin (
P
4K
) e
» amplitude at P
This corresponds to radiated power per unit solid angle
r m XaJ.
4R
A « wa*
The total power radiated V
the aperature is simply ^ ( € /p)
a» whence
the gain is
0 „ Ai|
X
a
x
The faotor (sin s/x)* expresses the ratio of the gain measured at a distance
E to the gain 0 of the true Fraunhofer field at infinity.
For a reflecting
region at 150 meters above the radar.
■ f c - °-95
= 0.5 db.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
BIBLIOGRAPHY
Atlas, D., 19^0, Possible key to the dilemma of angel eohoes. J. Meteor.
11, 95-103.
Berry, F.A., E. Bollay and N.R. Beers, 1945, BanfltiQQlC OfMgtgQXOlQgY,
MoGraw-Hill Book Co., Ino., Hew York.
Colwell, B.C., and A.W. Friend, 1937, Tropospherio ware reflections,
laiaaat, & , 473-474.
Crawford, A.B., 1949, Radar reflections in the lower atmosphere, Proo.
la&aJL., 11, 404-405.
Edwards, C.F., 1947, Microwave Converters, Proo. I.R.E.. VS. 1181—1191•
Friend, A.V., 1949, Theory and praotioe of tropospherio sounding by
radar, Proo. I.R.E.. Jl, 116-138.
Friis, H.T., 1947, Radar reflections from the lower atmosphere, Proo.
1*2*1., 15, 494-495.
Gish, D.I., and H.Q. Booker, 1939, Honexistenoe of continuous intense
ionisation in the troposphere and lower stratosphere, Proo. I.R.K.,
21, 117-125.
Gnanaling&m, S., 1954, Apparatus for the detection of weak ionospherio
eohoes, Proo. I.R.E.. 10.1, Ft. 111, 243-248.
Godson, W.L., 1951, flMMrt, J, S83T,. KfttttttCPlt -Sag., 11, 633 .
Harper, V.G., 1957, Angels on centimetrio radars oaused by birds, Batura
1&2, 847-849.
Hay, B.R., and W.M. Reid, 19^2, Radar angels in the lower troposphere,
Q&a*.
, M , 128-138.
Bund, A., 1942, Frequency modulation, McGraw-Hill Book Co., Ino., Hew York.
Ismail, M.A.W., 1955, A study of double modulated F.M. radar, Verlag Leeman
Zurioh. 21.
Ismail, M.A.W., 1956, A precise new system of F.M. radar, Proc. I.R.E.,
M , 1140-1145.
96
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
97
Jaoobs, D.H., 1943* Fundamentals of Optical Engineering, McGraw-Hill Book
Co., Ino., Hew York.
Johnston, B.W., 1962, M.Sc. Thesis, University of Western Ontario, London,
Ontario*
Kerr, D.E., 1951» Propagation of Short Badlo Waves. McGraw-Hill Book Co.,
Ino., Rev York.
Maokey, B.C., 1962, Some characteristics of microwave balanced modulators,
Trans. X.R.B. » MTTV10. Ho. 2. 114—117*
Mees, C.E.K., 1952, The theory of the photographio process, The Macmillan
Co., Hew York.
Miller, 3.E., 1947, Considerations in the design of centimeter-vave radar
reoeivers, Proo. I.B.E..
340-351.
Mitra, S.K., 1936, Letter to the editor. Mature. 137. 867.
Fiddington, J.H., 1939, The scattering of radio waves in the lower atmos­
phere. Proc. I.R.S.. 21, 75>757.
Plank, V.G., 1956, A meteorologLoal study of radar angels, Qeophva. Bea.
Paners 62, A.F.C.R.C.
Plank, V.G., 1959» Spurious eohoes on radar, a survey, Gaophvs. Bea. Parers.
Jl2t A.F.C.R.C.
Primioh, B.I., 1956, Relative merits of various types of microwave reoeivers,
BEL ParajflQ't*
Reid, W.M., 1961, Thesis, University of Western Ontario, London, Ontario.
Ridenour, L.W., 1947, Radar system engineering, MoGraw-Hill Book Co., Inc.,
Hew York and London.
Sawyer, J.S., 1955, Iondon Meteorological Office. Qeophva. Mem. 12(96^.
Sears, F.W., 1949, Optics, Addison-Wesley Press, Ino.
Scorer, R.S., and F.H. Ludlam, 1953, Bubble theory of penetrative connection,
ftimpt. .T. Bov. Meteor. Soc..
94-104*
Soorer, B.S., and F.H. Ludlam, 1960, Cloud study, John Murray, London.
Silver, S., 1949, Microwave antenna theory and design, McGraw-Hill Book Co. Ino.
Tyrrell, W.A., 1947, Hybrid circuits for microwaves, Proo. I.R.E..
1294-1306.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
98
Watson-Watt, R.A., L.H. Baihbridge-Bell, H.P. Wllkina and B.C. Bowen,
1936, Return of radio waves from the middle atmosphere, Nature. 137.
866
.
Weather tfava. 1957, Meteorological Branoh, Dept, of Transport, Canada.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
VITA
NAME*
Morley Bruoe Bell
BOHN*
Orillia* Ontario* Canada* 1937
EDUCATED*
Primary*
Hobart Public School* S.S. 3* Medonte* Ont.
Canada, 1943-1951
Seoondary*
Midland Dietriot High School* Midland* Ont.
Canada* 1951-1955
MidlandrPenetanguishene Dist. High School*
Midland* Ont.* Canada 1955-1956
University*
University of Western Ontario, London* Ont.
Canada 1956-19*2
Degress*
B.Sc.* 1960
3CH0LAHSHIPS*
national Research Counoil Bursary* 1960*61
national Besearoh Council Studentship* 1962-63
APFOBfTMENTS*
Research Assistant* Defenoe Besearoh Board,
Ottawa* Ontario* Suaaer 1960
Demonstrator* U.W.O.* 196061
Demonstrator* U.W.O.* 19*1-62
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Документ
Категория
Без категории
Просмотров
0
Размер файла
4 212 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа