Indian Journal of Radio & Space Physics Vol. 24, August 1995, pp. 166-177
r
Characterization of LOS signal fading. at 6/7 GHz over Assam Valley and its association with tropospheric parameters: An observational study
Sanjay Sharma, K I Timothy. M Devi & A K Barbara Department of Physics, Gauhati University, Guwahati 781 014 Received 25 July 1994; revised 20 December 1994; accepted 1 May 1995
Fading features of microwave signals (6/7 GHz) over Milmilia-Durgasarovar, Maopet-Durgasaro- ~ var and Laopani-Habaipur links of Assam valley are characterized in ternlS of fade depth and fade
duration distributions. The multipath occurtence factors for all the three links are defined after sett- ing up fade distribution equation for each link. The fades over Milmilia-Durgasarovar link only are associated with meteorological and atmospheric variabilities received through two sodar units and a number of sensors placed over this link. The worst month fade depth characteristics over this link are associated with tropospheric ducting.
1 Introduction different terrain conditions of Assam valley. The
Microwave fading on line-of-sight (LOS) path is selected links are:
cause? by lar~e clhanges .in refractivity of the p~o- (1) Milmilia-Durgasarovar (P & T) 6 GHz pagatIng medIum. A major source of propagatIon (2) Maopet-Durgasarovar (P & T) 6 GHz
?utage ?n ~crowave LOS link is ~ultipa~ fad- (3) Laopani-Habaipur (Railway) 7 GHz mg, which IS caused by steep negatIve radIo re-
fractive index (RRI) gradient of the atmosphere1.2. For assocIating troposphenc condItIons ~th -\
For a large positive RRI gradient, the fades in fades, the temperature, pressure ~d humIdity microwave signal are on the other hand caused by s.ensors and two sodars a!e place? m one of the obstruction of the LOS path by the earth's bulg- ~ paths, where one .radioson.de IS also operat.ect ing3. Under normal atmospheric situations when a [~Ig. 1(~)]. The P~T ~s ar~ ~ the Kamrup d~s- signal follows a two-ray path geometry, a stable t~ct while the Railway.link IS ~ the Nagaon .dis- signal is received at the receiver site. But when tnct of the valley. T.he info~atIo~ of the vanous layer structures or turbulences in the atmosphere parameters of these links are given m Table 1.
are formed in situations of large refractive index ...,.
gradient of the l1ledium, the reflected wave from Table I-link infonnatIon of the three different links
layers and turbulences also reach the antenna Link parameter MiI.- Mao.- La~.- ~ with varying phases and amplitudes. This leads to D. Sarovar D. Sarovar Habalpur ' a constructive and destructjve interference causing Distance, km 40.2 64.4 55.8
fades of different depths and frequencies4.5. The Frequen~y, GHz 6.4 6.0 7.13 diurnal and seasonal variations of fades are there- Antenna height, m
fore very closely related to varying meteorological Tx 50 70 80
conditions6-S. Parameters like the path geometry, R. 50 70 80
antenna height, transmitting frequency, etc. also .
influence fadings9. Antenna gain, dBm
The basic aim of this paper is to characterize Tx 43.3 44.8 39.5
the fade depth and fade duration of microwave R. 43.3 44.8 35.5
signals over different terrains of Assam valley and HASL J
to associate the fading phenomenon with the at- T 105 1660 80
mospheric variabilities 'afid then to receive the dy- ~ 228 228 80 namics of the system.
fuwer*,dBm 40 40 30
2 Links infonnation
h
.
f.
links mad HASL Height from average sea levelT e selection ~ mIcrowave were so e .With respect to 1 mW signailevel that the propagatIon aspect could be addressed to
SHARMA et al: 6/7 GHz FADINGS OVER ASSAM & TROPOSPHERIC PARAMETERS 167
,
,"-f. S ;
J 0 '
p. ...,
~ ,...'
n,.' i
" 1'. .' \
'" '..,"..,. ,-
po r' .~ ~
~ ...1' -(
~ C,. ""","'"
A t' J
27" 8 HUT AN) ./""
..-, ,.,.'" ~ '"
0 ..."'... ; ' --'-'-'-'- ,.,.",-,-- I ,--. """- ""--'--- .",I'-'~
..J , -,- ./
-< I .""
"-0, ,-
"'z' !. 1'.1
ww.' A !. Iv Q
~ m,l -1.). ~
,+",' \~OI.'2.3 :I..\.~~-I~",\.I.~ 1/ '" po
\ ~~ l,AOPANI(IX) t' (,t-
:I: " '" .-, MIL ILI~\~ ,fI' ,.. '",-""t-~Y] ) ...'\
~
l IJ ~ t-~ r.r "'-"""""""""1 ~~ V;-' 1\)~I.~-1.1 )~_.,;
5 I, '7 ~ ~p.\V '
0 '.., ;) ", ~p / , !.ODAR
/ ! 'j'" M E G HAL A Y ~~~ .~., (" 2 TEMP. SEN!.OR
" "",,) 3 HUMIDITY !.EN!.OR
, \ .
",-,-.-,-.- -'-' 1,., '- ~ J,., .'" l, PRE!.!.URE !.EN!.OR
25" .""<. (I 5 RADIO!.ONOE
0' ..
..,,~.\,..I! \) ~
J n
, I \"
I I h.~
J .r
i f': (~
~ Q 25 so' 75 1 .I \"",.J
f -..~ km '('
:1' L.., r'
'"
9'.0' 93. 0' ~ 0-
Fig. 1 (a)- Locations of microwave link terminals, sodars and meteorological instruments.
3 Terrain features and Fresnel zone clearance are sat, shagun, pine and abundant foliage like The terrain features over link paths have been bamboo,
collected from sources like National Remote' Th Le aopanI- ' H ba mpur ' link lies
.
m N agaon di s-Sensmg Agency (NRSA), Department of Flood '
d ath f 5 5 kIn Thi link S
( ) & tnct an covers a po, s run
i Control, Assam, and P&T Ref, 10. The leatures I ' t ' al th K iIi " hi h
"' " over a pam erram ong e op nver w c
of each terram are descnbed separately, '" , .'
h Th Milmili" D link f _11 ' G IS crossmg this link path at three pomts, T e ter-
e a- urgasarovar i1US m uwa-. thi link , d 'th ulti' t d I d
h 'Cal LOS ' ' t' ram over s IS covere WI c va e an
atl- cutta lll1crowave commumca Ion ~ t
path and passes over the major swamps of Dee- and ores,
por Beel as well as over Kukurmara Beel systems The path profiles of the three links at two at- which lie further south to the link Also it runs mospheric condjtions (K= 4/3 and 2/3) are along the Brahamputra river to the north as well shown in Fig, 1 (b,c,d). These two K values were as along the Kulsi river to the south. The yeserve selected to receive the effects of the earth's curva- forest in the link path is populated with sal and ture on first Fresnel zone at normal and worst sagoon trees, In summer most of the belt is propagation conditions, The Fresnel zone clear-~
covered with paddy fields and monsoon water, ance for all the three links were then examined at
~his link thus passes mainly over swamps and these two atmospheric situations, It is clear from low-lying areas. the figures that, except for Laopani-Habaipur link, The Maopet-Durgasarovar link falls in Guwa- the first Fresnel zone is free from terrain obstruc- hati-Shillong microwave path and runs to the tion even for the worst propagation condition south from Durgasarovar, This terrain is basically (K = 2/3), However in Laopani-Habaipur link the hilly and is covered with evergreen as well as with first Fresnel zone is obstructed at midpath of the deciduous forests, Over this link, the major trees link due to the earth's bulging when K = 2/3,
168 INDIAN J RADIO & SPACE PHYS, AUGUST 1995
~ 234
~ :~;~~~~~ -2~ ~ :~~~~~~~
~ 175 ~ 175
~ 117 ..01 117:C 11
71
117X 5& 58.- ~
5'
5&0 2. ,. 41 0 2.15
PURGAsAROVAS\. MII..~ILA DUROASARovA~ MILM'tA
(RX) DI5TANCI;(km) (TX) tRw) DiSTANCE lkm) tT~)
MILMILIA -OURGA SAROVAR LINK (K- 4/3) MILMILIA -OURGASAROVAR L IN K (K = 2/3 )
(b)
1660
1510 1660
1510
e 1.2.08
--1206
:c 01 906 ---!E
-c; :E 900
X 01
604 :I: 604-
..
302 302
0 f6 31 its
MAOPET 6 0 16 32. 48 64-
(Tx) OISTANC!" (i<m) bURGA5AROVAR MAOPET PI5rANCE(km) OU~GASAROVAR
(Rx) (Tx) (Rx)
MAOPET- DURGASAROVAR LINK (K- 4/3) MAOPET-OURGASAROYAR LINK (K- 2/3)
I ;~~~~~~~~ 1 i 0
(C)e120-120 -E~ 80 ";40 ~ 80 -1 ~60 - 20
I ~~=~sl
60X .;40 40
0 10 20 50 40 50 55 X
lAOP~NI HABAIPUR 0 15 20 ~O 40 50 55
(
) E(L. ) (R ) LAOP~NI HABAIPUR
...TIC Dl5TANC "m x (Tx) DISTANCE (km) <,R~)
LAOPANI- HABAIPUR LINK (K_4/3) LAOPANI- HABAIPUR LINK (K= 2/3 )
(d)
Fig. 1 (b,c,d)- Path profiles and Fresnel zone clearance over the three links for K = 4/3 and 2/3 [(a), Milmilia-Durgasarovar;
(b) Maopet-Durgasarovar, and (c) Laopani-Habaipur].
1
4 Fade measurement and data analysis fluctuations up to O.l°C could be detected both To derive the fade characteristics of the signal for dry and wet temperature sensing devic~s. Two from the links, the automatic gain control (AGC) sodar \!nits, one at the Gauhati University and the output of the receiver was fed to recorders with a other at Mirza (at the midpath of this link), were recorder time constant of 500 ms. But for fades also operated to receive the atmospheric boun- of lower depths the frequency response was bet- dary layer conditions. These systems are placed ter. Therefore we kept the integration time of the over Milmilia-Durgasarovar link. Therefore the signal conditioner at 100 ms. The supporting fade character information over this link only will meteorological parameters like temperature, pres- be associated with different atmospheric and sure and humidity were received through systems meteorological parameters to drive the physics of developed by us. The radiosonde data (at the the system.
midpath of the Milmilia-Durgasarovar link) col- It is also highly desirable to examine the fade lected by !MD, Guwahati, were also utilized in characteristics in different seasons over a terrain the analysis. The temperature and humidity infor- and it is therefore necessary to define the seasons
mation were also received by a 20 m high tower over a locality through measurements of variables installed at the Gauhati University. Temperature like temperature, humidity and rainfall. An ela-
SHARMA et al.:6/7 GHz FADINGS OVER ASSAM &TROPOSPHERIC PARAMETERS 169 borate analysis on these parameters over the links
was made by one of the authors II and the seasons were then defined as follows:
Summer-June, July, August
Post-monsoon-September, October, November Winter - December, January, February
Pre-monsoon-March, April, May
The amplitude (signal envelop voltage level L) data collected over all the links for the period 1991-93 provided the basic information for the evaluation of fade characteristics. The relevant parameters for such a study are the probability P(L) that the amplitude v(t) will be below a speci- fied signal level L, and the average duration of the fades below L, i.e. t(L). These parameters are all functions of L. The fade data were then grouped at hourly duration. The fade depth with- in one hour period was coded at intervals of 10 dB and the time duration a particular signal le- vel remains below a specified value was found out. The average fade duration at any fade depth was obtained from the ratio of total time at or be- low the fade depth level to the number of fades of that depth". The data were then analysed in terms of probability distribution of fade depths.
Statistical analyses of the signals received under multipath propagation have also been carried out by a large number of workers+" ". A general for- mula for the probability distribution of fades in this condition is given by the following relation- ship'?
... (1) where L, Amplitude in linear meausre; K, Factor for climatic conditions; Q, Factor for terrain con- ditions; R, Envelop voltage normalized to the non-faded signal level;
f,
Frequency in GHz; d, Path length in km; Fn, Path clearance factor;B, C,X, Constants.
or
P(R
<
L)= EU ... (2)where E is known as the multipath occurrence factor".
The multipath occurrence factor was then cal- culated from the fade distribution. The diurnal and seasonal variations in the occurrence of fades and the distribution of fade durations at specified dB levels were also found out.
5 Results
5.1 Types offading
The fade character .analyses over these links in- dicate presence of different types of microwave
dB -10
o
-30 -20 -10
o
-35 -20 -10
o o
-10
(d) 07 TIME (hrs)
Fig. 2-Sample records of different types of fadings over mic- rowave links under study [(a) Slow and shallow fades.
(b) Deep and rapid fades, (c) Slow and deep fades, and (d) Fast fades modulated over slow fades].
fades. Representative patterns over Milrnilia, Maopet and Laopani links are presented in Fig. 2(a,b,c,d).
Figure 2(a) shows slow and shallow types of fadings observed in all the three links where depth varies from - 2 to - 10 dB. Figure 2(b) shows rapid and deep fades. In this type of fading the depth varies from - 10 to - 30 dB and signal crosses the median level 4 to 15 fades/h. The preferential time of occurrence is early morning hours. Figure 2(c) shows slow and deep fades. In this type of fading the depth varies from 10 to 30 dB and signal crosses the median level 1 to 3 times per hour. ThG preferential time of occurr- ence is pre- or post-midnight hours. Figure 2(d) shows a typical fade pattern where fast fades are modulated over slow fadings. The observed fade
170 INDIAN J RADIO & SPACE PHYS, AUGUST 1995
patterns ovef these three links are given below: fades, if any, are then examined over the links' Link 1 Link 2 Link 3 and for this purpose fade depths within 1 to
1. Slow~w* 1. Slow~w* 1. Slo~ow 45 dB have been taken into account. Fades over T 2. Rapid, deep 2. Rapid, deep 2. Rapid, deep* these links show a clear diurnal character. It is
3. Slow,deep* 3. Slow,deep* 3. Slow,deep obse.rved that occurrence of noc~rnal f~dings is 4. Fast fading 4. Fast fading 4. Fast fading relatively more than that of daytime fadmgs and modulated modulated modulated fade patterns also vary at different times of the over slow over slow over.slow day. Deep fades are detected more in the early fading fading fading morning hours [Fig. 2(b,c)] during all the seasons
" except for summer where fades are more often
Indicates the dommatlng feature of the fadmg detecte m pre-ffil ru tours. d
..
d. gh
h It IS so to.
al be 5.2 Probability offade distribution noted that pre-noon fades are fast with low depth For this analysis the fade depths are classified (scintillation type) whereas the early morningas explained in section 4. As the fade depth en- fades are rapid and deep. Figure 4(a) shows the -1 countered for multipath phenomenon can be quite diurnal occurrence pattern of fades over Milmilia-
severe at times, it is essential to know multipath Durgasarovar link, where the fading is observed occurrence factor [as defined in Eq. (2)] over a to be basically a nocturnal phenomenon with terrain. So the fade distribution analyses would preferential period of occurrence in post-midnight basically be made with multipath types of fades, hours except during summer months when fades and the distribution of fade depth at 10 dB inter- are seen more often in pre-midnight hours. This vals is plotted by taking all the data collected over diurnal feature is also maintained over Maopet- the study period. The probability distribution plot Durgasarovar path [Fig. 4(b )]. However we note a so obtained over Milmilia-Durgasarovar link is change in the variation of fade pattern over Lao-
shown in Fig. 3(a), where the distribution follows pani-Habaipur link [Fig. 4(c)], where the fades are Rayleigh pattern for deep fade region. The multi- largely detected during pre-midnight hour~ in all
path occurrence factor (e) for this link is then de- the seasons.
terDlined and is fixed at 0.609. The fade distribu- tion over this link can therefore be defined by
p= 0.609 L2. The slope of this line is found to be 5.4 Seasonal variation of fade occurrence
11 dB/decade of occurrence of fades, where the Besides the diurnal changes in fadings, the sig- link cutoff (45 dB below normal in this case) is nal also suffered fadings and attenuation depend- observed for 0.004% of the total time. Probability ing on season. For this analysis too, fades of occurrence over Maopet-Durgasarovar link also depths within -1 to -45 dB have been considered.
follows the Rayleigh distribution for deep fade re- To bring out the seasonal character, the average gion as shown in Fig. 3(b). Multipath occurrence fade time for each day and then for each month factor is found out to be 0.714, leading .to a dis- are evaluated. The seasonal percentage of fade tribution equation given as P=0.714 L2. The occurrence is then calculated for each season, slope of the curve representing the distribution from the total data time and the total seasonal equation is 11 dB/decade of probability and the fade time. The variations in fade occurrence over link cutoff (at 45 dB below normal level) is ob- all the three links at different seasons are shown served for 0.005% of the total time. The proba- in Fig. 5. It is interesting to note a difference in bility of occurrence of fades over Laopani-Habai- seasonal characteristics of P&T link~ of Milmilia- pur link is shown in Fig. 3(c). Here also, the fade Durgasarovar and Maopet-Durgasarovar and distribution curve follows the Rayleigh pattern for Railway link of Laopani- Habaipur .We observe deep fade region. Similarly, the multipath occurr'" that over Milmilia-Durgasarovar link, the proba- ence factor (e). is calculated for this link and the bility of fade occurrence is maximum during win-
equation of distribution curve is received (by us- ter seasQn, where fadings are detected over 30% ) ing this value of e) as P=0.774 L2. The slope of of the winter period. This is followed by summer
this curve is 16 dB/decade of probability of oc- months where over 20% of the total summer data currence and the link cutoff (at 43 dB below nor- show fadings. The seasonal variation of fade .(}G- mal level) is observed for 0.008% of the total currence over Maopet-Durgasarovar link shows
time. similar fade occurrence characteristics as that
over Milmilia-Durgasarovar path, with maximum 5.3 Diurnal variation of fade occurrence fades during winter (34% of occurrence) followed
The temporal variations in the occurrence of by pre-monsoon months. The seasonal fade char-
SHARMA et al: 6/7 GHz FADINGS OVER ASSAM & TROPOSPHERIC PARAMETERS 171
'1 c( c(
tJ) \J')
~ ~
U u
m m
c( <t.
V/ 10 V/
IlJ IlJ 10
~ 1 P=0-609L2 ~ P=0-714L2
a. a. 1
w w
0 a
y ~ 0.1 w
:c( 0 0.1
u.; ~
}::
w
0.01 w.:: "~ 0 -0 1
"0"- '0
u.. u.
0 0-001 0 0.001
...0.00
..
20.0() 40.00 60.00;! 0-00 20-00 40-00 60'00FADE DEPTH (dB) FADE DEPTH (dB)
(a) (b)
~
«
tJ) tJ)~
~ ' ---0 0-- Experimental Curve« ,
~ 1 0 '\ Best Fit Curve
-,
,
:I: \
~ ..-1 a.
w P = 0 -774L.:2a
~ 0'1
~
w 0 -01
~ .--0
i=
u.0 0'001
eo! 0'00 20-00 40-00 60-00
FADE DEPTH (dB)
(c)
Fig. 3-Fadedepthdistriburionover(a) Milmilia-Durgasarovar link- (b) Maopet-Durgasarovar link- and (c) Laopani-Habaipur link.
172 INDIAN J RADIO & SPACE PHYS, AUGUST 1995
w 6
u
z
~ 60 50 r
0:=> 5Q
~ 4
0 40
w
0
30Q 3
u. c{ 2 2
0 u. 1
;-.1
6 12 18 24 6 12 18 24
TIME (hrs) TIME (hrs) ~
(0) (b)
UJu
Z a SUMMER
UJ 60
0:0: X POST-MONSOON
=>
uU 4 0 WINTER
0
UJ A PRE-MONSOON
Q
~ 20
0u.
.
..:6 12 18 24
TIME (hrs) (c)
Fig. 4-Diumal variation of fade occurrence over (a) Milmilia-Durgasarovar link, (b) Maopet-Durgasarovar link, and (c) Lao- pani-Habaipur link.
~ Ie 0 acteristics of Laopani-Habaipur link on the other ~
~ 3 5 -hand show that the probability of fade occurrence
~ 30 -is maximum during summer months followed by
u 2 5 ~ ~ ~ pre-monsoon months.
u '"
0w 20 5.5 Average fade duration
~ 1 5 ~ The average fade duration at any fade deptlt le-
LI; 1 0 ~ vel is obtained from the ratio of total time at or
~ I; below the fade depth to the number of fades at
;; 5 ~ % ~ ~ ~ that fade level. The average fade duration for Mil- .J
0 'i milia-Durgasarovar link is shown in Fig. 6(a). The
SUMMER POS T- WINTER PRE -curve represents the average fade dura~on against MONSOON MONSOON each fade depth level. A distribution curve de-
~ MIL MILIA ~ MAOPET 0 LAOPANI fi~~~ by t= 550 L can ?e r~ei~ed .tr°m the plot.
Similarly the fade duration dlstnbution curves for Maopet-Durgasarovar and Laopani-Habaipur links are also obtained and these are defined by Fig. 5 -Seasonal variation of fade occurrence. t = 514 L and t = 246 L respectively.
SHARMA et ai.: 6/7 GHz FADINGS OVER ASSAM & lROPOSPHERIC PARAMETERS 173
Y
100 -100
-; t=550L ~ t =514L
z 0 z
0 ~,
i::; "<{.',
~ 1 Q: 10
"
:0.Q: :J'
:J C
C "
W "
C 0
'"'t w «
C u..
l"f: 1 1
0-00 20.00 40'.00 60-00 0.00 20.00 40.00 60.00
FADE DEPTH (dB) FADE DEPTH (dB)
(a) (b)
100
t = 246 L
-f .:!!..
-
-0- -~o- Experimental Curve~ 1 Best Fit Curve
~ a.
:J C
CW
~ u.. 1
"""1 0.00 20.00 40.00 60.00
FADE DEPTH (dB)
(c)
Fig. 6-Fade duration distribution over (a) Milmilia-Durgasarovar link, (b) Maopet-Durgasarovar link, and (c) Laopani-Habai- pur link.
6 Discussion ground. The possibility of the obstruction of first
The probability distribution of fades over all Fresnal zone during K = 2/3 situation as seen ( the links studied here stlows that the multipath from Fig. l(d) suggests that this terrain also fa-
occurrence factor is highest over the Laopani- vours obstruction-type fadings.
Habaipur link. Frequent (short lived) link cutoff is There are reports showing association of micro- an interesting feature over this link. One of the wave fades with atmospheric variables and also possibilities, apart from the partial reflection from with atmospheric boundary layers (ABLS)19.20. To layers, is the interference between di~ect wave examine the association, if any, of the microwave and different wave components such as specular fades in the Milmilia-Durgasarovar link with at- and non-specular components reflected froJ n mospheric parameters, the ABL pattern received
174 INDIAN J RADIO & SPACE PHYS, AUGUST 1995
m
400 200
o '---~---~r_~----~~---~~~---~---J
23 21
19 17TIME ( nrs )
15
Fig. 7-Ground-based inversion layers as observed through sodar over Milmilia-Durgasarovar link.
01
23
400
200 "
o
°c
26 22
18~~.~.~--
e21
19 1719
18
17 16 14TIME (hrs )
15
Fig. 8- Development of ground-based inversion layer and the corresponding temperature record during that time.
through two sodar systems will be taken. Before proceeding to this analysis, let us review here (in short) the basic features of microwave fades over this link, which are:
(1) Large nocturnal fading with high fade depth.
(2) The favourable time for development of fadings ispost-midnight.
(3) The nocturnal high level fades disappear af- .ter sunrise and fast- or scintillation-type fades
with low depths develop after a couple of hours of sunrise and may continue a little beyond noon hours local time (LT).
(4) Fadings are mostly absent during 14 to 16 hrs.
Two types of inversions in the study of near earth environment are often detected through so- dar echograms as well as from radiosonde data analysis.These are:
(1) Ground-based inversions
(2) Elevated inversions \.
The ground-based nocturnal inversions (as seen
on the sodar echograms) are developed due to the effect of cooling of the earth's surface21•23• A typi- cal echogram of nocturnal ground-based inversion is shown in Fig. 7. The depth of this inversion layer does not remain constant and we observe the depth to vary between 50 and 1-50m, de- pending on the atmospheric conditions. Ground- based inversions are generally developed after a couple of hours of sunset when ground tempera- ture reaches a steady low -value, an evidence of this can be seen in Fig. 8. Here the temperature is recorded by an electronic thermograph developed by us. The inversion layer develops at around 20 hrs when the temperature is approaching the steady state. This layer generally dissipates after sunrise (not shown in the figure).
After discussing the basic features of ABLs, we examine the association of microwave fades with ABLs and structures. The analysis shows that during the situation when high level nocturnal in- versions are present, deep fadings are often de- tected. Figure 9 shows a representative sodar.
echo gram when - 30 dB fades were detected over Milmilia link. This can probably be explained
SHARMA et al.:617 GHz FADINGS OVER ASSAM & TROPOSPHERIC PARAMETERS 175
04
o
m
200
o
7 6
03 02
5
43
T!ME ( hrs
I
2 o
Fig. '}- Representative sodar echogram corresponding to deep microwave fadings over Milrnilia-Durgasar- ovar link.
-10
o
200: ,
o
02 24 22
TIME (hrs )
20 18
Fig. IO-Sample record of shallow microwave fadings corresponding to low level inversion layer.
through interaction of the first Fresnel zone with a high rise ABL. In a situation when the Fresnel zone clearance at the mid path is affected by a boundary layer, a part of signal energy that passes through the ABL will propagate with a differen- tial propagation velocity resulting in fades of the microwave signal. But in a situation when inver- sion layer goes low (50 m) as marked in the sodar echogram, a Fresnel zone clearance is well main-
tained and the consequent effect of ABL on mic- rowave fades is expected to be low. In view of this we note that for low level nocturnal inver- sions, signal undergoes only shallow fades (Fig.
10).
In an attempt to. explain relatively large fade in winter over this link, we present in Fig. 11 the temperature inversion events at different seasons and at two different times of the day. For this in-
176 INDIANJ RADIO & SPACE PHYS,AUGUST 1995
70 .1i
60 G
50 ~ ~ EXPERIMENTAL CURVE '1
\/) 40 \I:J: 1 -BEST FIT CURVE
>- ...
~ Q.
0 30 '8 1
u. 2 w
0 0
.~ .u.; 0.1
0 1
z w
~
...SUMMER POST- WINTER PRE- u. 0'01
MONSOON MONSOON 0
.0'001 -! 1
0.00 20.00 100.00 60.00
~ 00.00 GMT hrs FADE DEPTH(dB)
(0 )
l\\1 12 .00 GMT hrs ~
Fig. II-Occurrence pattern of temperature inversion over ~0
Milmilia-Durgasarovar link. ~
~ 1.1'00
version analysis, radiosonde data and temperature ~ information received through our tower have ~ been utilized. The relatively large fade occurrence Z events and a number of temperature inversion!
events during winter are to be noted and these in- w 21'00 versions are likely to cause significant winter post- IzI
midnight! early morning fades. ~
We have noted that though the overall fade oc- ~ currence is maximum during winter, the link suf- ~ fers severe cutoff during post-monsoon months. 0 1'00
. b th ..1 10 100
This can be shown y worst mon s statIstIcs.
The wors mon t
th .
IS e e d fin d (as per CCIR)as 10 -s: I EFFECTIVE EARTH RADIUS FACTOR (K)lows: "If P i( z) be the time percentage exceeding a ( b )
threshold level (z) in the ith month of the year Fig. 12(a)-Worst month ~tatistics of fade depth distribution then the month for which the PAZ) is highes~
th
12 th.
th th" Th (b)-Occurrence over~filmilia-Du~gasarovarlink. of different values of Kover Milml-..among e mon s IS e worst mon .e lia-Durgasarovar link.
worst month (November) statistics of fade depth .,
distribution over Milmilia-Durgasorovar link is
shown in Fig. .12(a). The link cutoff (45 dB below summer maximum26. The significant difference in normal level) is 0.007% of the total data time of seasonal variations in the occurrence of fades the month, which is 0.004% higher than the aver- over P&T links (MiImilia-Durgasarovar and Mao- age link cutoff value of the year. The relatively pet~Durgasarovar) and Railway link (Laopani- large occurrence of signal black out can be corre- Habaipur) is rather interesting. The former links lated with the large occurrence of ducting condi- show the low latitude coastal region characteris- tions during the post-monsoon period as shown in tics v,:hile the latter link follows the low latitude Fig. 12(b), where the occurrence of different va- interior region fade characteristics. This indicates
lues of K are plotted for winter and post-mon- that links near Guwahati experience more fades in .
soon periods. winter, which can probabily be associated with )
Seasonal variations in fadings over LOS links very thick fog rising up to about 300 m and have been reported for the last two decades24;25. above over river Brahamaputra and the associat- These workers showed through experiments that ed beel systems. Thus parts of the paths (MilmiI-
seasonal fade characteristics may vary significantly ia-Durgasarovar and Maopet-Durgasarovar links) depending on the terrain conditions. The coastal are covered with thick fog.
region fade characteristics can be defined by win- The relation of average fade duration over the ter maximum and fades over land-locked paths by above mentioned links shows that Milmilia-Dur-
SHARMA et al.: 6/7 GHz FADINGS OVER ASSAM & TROPOSPHERIC PARAMETERS 177 gasarovar and Maopet-Durgasarovar links experi- 6 Reddy B M, Physics of the troposphere (Handbook on ro- ence fades of large durations with coefficient fac- dio propagation for. tropical and subtropical countries), -y tor 550 and 514 respectively while the Laopani- (INSDOC,NewDelhi),1987,59.
H b
.
l'nk ff f d ' f all d t .7 Gera B S & Sarkar S K, Indian J Radio & Space Phys, 9 .a ropur ~ su ers a es 0 s~ ura IO~S (1980)86.with coefficient factor 246, suggestmg that doffil- 8 Rao D N & Reddy K K, Prolonged signal fade out a,zd nant fading m~hanism over two districts of As- effect of antenna height on the perfonnance of line-of sight sam valley are different. This suggests that fadings micro~ve link in a .hilly terrain, ~per present~d at the
.over Kamrup district are basically caused by InternatIonal SymposIum on InternatIonal GeoscIence and
t bl I d t tu h . N d .Remote Sensing held during 18-21 August 1993 "at
S ~ e ~yer~ s roc res, w ereas m a~aon IS- Tokyo, Japan, 1993.
tnct fadmg IS developed due to the movmg struc- 9 BuIlingtonK,BellSystTechJ(USA),50(6)(1971)2039.
J tures. 10 Barbara A K, Devi M, Timothy K I & Sharma S, A Few
Aspects On Microwave Propagation Characteristics Over Assam Valley, Tech Rep. 2 (Department of Physics, Gau-
r ~ Acknowledgements h~ti University, G~wahati), 1991, 5. ...
The authors are thankful to the Department of II TImothy K I, M,crowave P~opagatlon CharacterIStIc: A
..Study Over Assam Valley WIth Respect to Hydrometeors,
Electr()l1lcs, Government of india, for the finan- Ph D thesis Gauhati University Guwahati 1993.
cial support to carry out this research work. They 12 VigantsA,BellSystTechJ(UsA),150(3)(i971)815.
are also thankful to the authorities of P&T and 13 LinSH,BellSystTechJ(USA),50(10(1971)3211.
N F Railway for giving permission to use their 14 Pearson K W, Proc lEE, (UK), 112(7) (1965) 1291.
microwave links and to the Director India Mete- 15 KaylorRL,BellSyst TechJ(USA),32(5)(1953) 1187.
.' .'.. 16 RiseSO,BellSystTechJ(USA),38(3)(1958)581.
orologlcal Department, Guwahatl, for provldmg 17 StephansenET,RadioSci(USA), 16(5)(1981)609.
necessary data. 18 Barnet W T, Bell Syst Tech J( USA), 51(2) (1972) 321.
19 Das J, De A K & Majumdar D D, In! J Remote Sens (UK), 11(6)(1990) 1033.
References 20 SchiavoneJA,RadioSci(USA), 17(5)(1982) 1301.
1 Webster A R, IEEE Trans Antennas & Propag (USA), 21 Singal S P, Aggarwal S K & Gera B S, Indian J Radio &
AP-31 (1983) 12. Space Phys, 11 (1982) 23.,
~" 2 RuthroffCL,BellSystTechJ(USA),50(7)(1971)2375. 22 WyckoffRJ,BeranDW& HallF FJr,JApplMeteorol 3 SchiavoneJ A, Bell Syst TechJ( USA), 60(6)(1981) 803. (USA), 12 (1973) 1196.
4 Feher Kamilo, Digital Communication Microwave Appli- 23 GossardEE,RadioSci(USA), 12(1)(1977)89.
cation (Prentice liaII Inc, New Jersey, USA), 1981. 24 Hay D R & fuaps G E, Can J Phys (Canada), 37 (1951) 5 Mon J P, Weill A & Martin L, Effect of tropospheric dis- 313.
turbe"ce on a 4.1 and 6.2 GHz line of sight path, Paper 25 Morita K & Kakita K, Rev Electr Commun Lab (Japan), presented at Commission F Symposium, URSI, Lennox- 6-6 (1958) 352.
ville, Que, 1980. 26 SchiavoneJA,RadioSci(USA),16(5)(1981)1301.
..~
,,
.~
