Tropospheric Scatter Propagation

Ian Roberts ZS6BTE *

 

Reprinted as published in Radio ZS, December 1984

 

Introduction

During the last 20 years or so, with the appearance of high power UHF amplifiers and low noise signal amplifying devices, a wide-band propagation mode capable of conveying VHF signals over distances of 800 km or more has become increasingly important in high priority commercial and military links.

The mode is loosely referred to in the industry as “tropo” or “tropo scatter”.

In recent years the pages of Radio ZS have recorded interesting long distance VHF and UHF phenomena as noted by local radio amateurs. It is evident, though, that many of the reports are rendered “tongue in cheek”, without much understanding of the propagation mechanism witnessed and it is commonplace to see E, sporadic E, F2, E/F2 backscatter, tropospheric ducting and tropospheric scatter confused.

The first five modes depend entirely upon solar radiation of the upper ionospheric layers for success, the latter two have nothing to do with solar activity. Tropospheric ducting is a freak occurrence involving inversions or peculiarities in the moisture content, pressure, and temperature domains in the vicinity of the ground and hence may be detected by antenna systems. The mode is obviously unpredictable. Accordingly, with solar activity presently at a low level, the only long distance mode left for the VHF enthusiast is tropo scatter. Radio amateurs, with their unique talents and privileges, are in a particularly good position to add greatly to the existing knowledge of tropo scatter.

 

Background and history of tropo scatter

Marconi described tests in 1933 at 550 MHz over a 270 km path between Rocca di Papa, Rome and Cape Figari, Sardinia.

In 1949 the USA froze the issuing of television broadcast station licences because of propagation beyond anticipated boundaries and co-channel interference on a massive and unexpected scale.

By 1953 Bell Telephone Laboratories, primed by much theoretical speculation and increasing empirical evidence put forth their “Polevault” VHF over the horizon communications system.

Armstrong, in work previously unreported, verified independently propagation at ranges up to 500 km.

The US Air Force, in about 1955, took the plunge and commissioned a link over hostile territory, thereby obviating the need for numerous conventional line-of-sight links.

And that is where the mysteries of tropo scatter propagation have been largely hidden, in classified material, generally not available to the radio amateur. Additionally the precise methodology of tropo scatter remains ill-understood even in professional circles and most performance evaluations are based on empirical data collected during field testing.

 

Concepts and Parameters

Various important parameters, peculiar to the mode, need further examination.

The K-factor, generally K4/3 radius of the earth. Much as light passing through a prism refracts towards the denser medium, so a VHF beam passing along the surface of the earth tends to refract along the denser air at the surface to achieve a distance considerably more than the true line-of-sight condition. This accounts for the fact that radar, operating for example at L-band (1100 MHz), can “see” a target below the visible horizon (which is itself below the physical horizon) Fig 1.

Fig. 1: Bending of antenna beam due to refraction (True earth radius, a)

Typically K is described as K4/3 at frequencies below about 1 GHz; meaning over the horizon propagation, but under severe conditions may fall to below unity.

 

Generally, K =effective radio earth radius/true earth radius and is greatly dependent on the surface refractivity index of the terrain over which the VHF beam is passing. In South Africa a ball-park value for this variable would be 280. The VHF-UHF Manual (RSGB) has an interesting description of this refractivity index and derivation.

 

There is a non-correlation

1)     in the signals received by two adjacent antennas from a dual polarisation transmitting site when the receiving antennas have opposite polarisation, e.g., horizontal/vertical, Fig 2.

2)     in the signals received (same polarisation) by two antennas spaced a finite distance apart, e.g.  100 wavelengths, Fig. 2.

3)     in the signals received (same polarisation) by two antennas receiving signals widely separated in frequency, e.g.  10 MHz, Fig.2.

4)     in the signals received by two antennas with slightly different beam headings, Fig.2.

 

Fig. 2: Non-correlation between the signals received by two antennas with 1) opposite polarisation 2) physical separation of 100 wavelengths 3) slightly different beam headings 4) wide frequency separation

 

 

These characteristics are put to good use in professional systems. For example, a tropo link with “quad diversity” would often use parameters 1) and 2) and be capable of receiving both horizontal and vertical polarisation on each of the two antennas spaced apart as above. Each antenna, similarly, would transmit horizontal and vertical polarisation on a common frequency.

FM is currently the preferred mode. The various signals are combined at IF (pre-detection combination diversity). Since the respective noise inputs add in random fashion and the signals linearly, a higher signal to noise ratio is obtained. Typical signal to noise ratios (with psophometric weighting) are plus 40 dB – good enough for a good quality telephone line or medium speed data with error correction.

A representative tropo link uses quad diversity, 27m parabolic antennas, 10 kW c.w. at 900 MHz, carries 132 FDM telephone channels, distance 500 km. A link of this nature would otherwise require 10-15 line-of-sight microwave stations.

 

Geometry of Tropo Scatter Path

Fig.3: Geometry of Tropo Scatter Path

 

R is 4/3 earth radius (8446 km)

D is great circle path distance

h1 and h2  respective antenna heights above sea level

h11 and h12  height of radio horizons above seal level

d1 and d2 great circle distance between radio horizons and respective antennas

 

The scatter angle    Ө = Ө0 - Ө1 - Ө2 radians

where                    Ө0 =   d/R

                             Ө1 =   ((h1- h11)/d1 + d1/2R)

                             Ө2 =   ((h2- h12)/d2 + d2/2R)

Typical scatter angles are up to 4 degrees.

Each 1 degree increase in scatter angle introduces an additional 10 dB path loss and high value scatter angles are avoided in professional systems. This is easy when one can choose a mountain top site.

In Fig. 3 the zone where the beams intersect is called the scatter volume and the properties of this volume define the quality of the scatter path.

 

Amateur application of tropo scatter

Inspection of a standard K4/3 path profile indicates that one may not expect a local radio horizon (d1 and d2) of more than 30 km assuming 20 m antenna height and level ground.

Under these conditions could one expect a tropo scatter path to exist between Johannesburg and Port Elizabeth assuming typical ham radio gear?

In order to address this question it is necessary to calculate or estimate the following:

a)     distance

b)    scatter angle

c)     path loss

d)    system noise temperature

e)     signal to noise ratio, which would give an indication of the signal to be expected.

 

a) The distance is calculated from the great circle path distance equation by assuming JHB to be the point of departure and using the respective latitudes and longitudes. So d = 872 km

b) The Scatter angle

The take-off from the PE end is particularly advantageous with the beam passing over the Sundays River valley and establishing a radio horizon in the Patterson area at about 60 km.

So, retaining the nominal 30 km radio horizon at 20 m antenna in JHB and remembering to use the same units in the equation: h1 in JHB about 2 km with horizon at 1.38 km (hills south of Alberton) h2  in PE at 0.5 km with horizon at 0.48 km (60 km out) then scatter angle:

                                      =       872/8448 – ((2.0-1.38)/30 + 30/16896) – ((0.5-0.48)/60 + 60/16896)

                                      =       0.0769 radians

Ө       =       4.4 degrees

c) Path Loss

The median path loss LP  consists of three components, viz.

LP          = LFS +  LS – 0.2(NS – 310) dB

LFS        = free space path loss, 92.4+20log dkm +20 log FGHz

where d       = distance in km, F = frequency in GHz

i.e. LFS         = 125.18 dB

Ls           = is the all year median scatter loss normalized at a surface refractivity index NS  310

          = 57 + 10log(0-1) + 10log(F/0.4)                                   

          = 81.96 dB

The factor NS in terms of C.C.I.R. recommendations and is mapped globally. In South Africa the value of NS varies between about 310 and something much less (e.g. the generally taken 280) depending on water vapour content, pressure and temperature to name a few components.

So LP = 213.14 dBi (this is an EME -  type path loss).

d) System Noise temperature    

TSYS     =       α(Tα) +  To(1-α) + T1 + Tm/gm-1

where  α      =       transmission line coefficient as a factor, 0.8 for 1 dB loss

          Tα          =       temperature of transmission line, k, equal to To for uncooled lines, (= 290k)

            To          =       ambient temperature k, typically 290

            T1          =       temp of 1st RF amplifier stage, 150k (preamp used)

            Tm         =       temperature of 2nd RF amplifier stage (the receiver 600k )

gm-1       =       gain of stage T1 = 32 (15 dB)

(The terms were explained in reference 3)

 So              TSYS                     =       about 460k

The receiver noise power ratio   Pn  consists of the “pure” KTB noise modified to incorporate the receiver noise figure, i.e. FkTB, where F is the receiver’s noise figure. If one assumes the receiver’s RF stages to be T1 and Tm  with filter losses of 1 dB, then F1 is about 2 dB. In a bandwidth of 1000 Hz, Pn turns out to be -168 dBW

e) Signal to noise ratio

SNR  =       Po +Gt + Gr – LP - Pn   (see ref 3)

Po          =       output power of TV TX, 100 W (20 dBW)

Gt          =       gain of Tx antenna, dBi (12.0)

Gr          =       gain of Rx antenna, dBi (12.0)

LP          =       tropospheric scatter path loss, dB (213.14)

Pn          =       noise power ratio of Rx = 10log FkTB, -169, k is Boltzman’s constant 1.38 x 10-23 and B = 1000 Hz in CW, T= TSYS , F is Rx noise factor = 2

So SNR       =       20+12+12-213.4-(-169)

=       -0.40 dB

However, since an isotropic path was used about 5 dB should be added to this. The ear should have no trouble tracking a beacon-like signal at this sort of SNR, indeed it should be continually audible with signal levels changing in sympathy with changes in the surface refractivity index. For example, an increase in this quantity from 280 to 300 would reduce the path loss by 4 dB and increase the SNR accordingly.

 

General

As a matter of interest, the typical heights of the scatter volume (assuming unobstructed paths) are listed below:

distance

150 km

300 m

2000 m

 

300 km

 600 m

3000 m

 

600 km

3000 m

20000 m

The shorter paths are characterised by deep, fast fading. Longer hops show a steadier path loss consistent with the median path loss for that month. It is suggested (in classified literature) that the best tropo conditions prevail during a hot, summer afternoon, while the worst conditions occur during winter nights.

Much remains to be researched, or remains unreported. For example, what is the effect of a thunderstorm on the scatter volume? What happens when a tropospheric duct intercedes? Is the north/south path more favourably propagated as in F2 /TEP propagation? ** See p.s. at end of this article.

Numerous high power RF sources exist in South Africa, namely the SABC’s FM and TV broadcast signals. The photograph is of reception by the writer of Nelspruit (ch 24) TV transmitter over a path of 270 km. This signal is continually detectable at the QTH in Pretoria which has an inferior radio horizon in all directions. Fading on this signal is in excess of 15 dB, with several cycles per second being typical over this distance.

Tropospheric Scatter reception: the SABC’s Nelspruit ch. 24 TV transmitter received over a scatter path of 270 km, the shadowing is typical of a camera with focal plane shutter.

Conclusion

A method has been illustrated whereby VHF signals can be propagated much further than the normal line-of-sight, point-to-point, condition.

References

1)     VHF-UHF Manual (RSGB)

2)     Tropospheric Scatter (Point to Point Communications, Feb; 1964)

3)     System Noise Temperature and System Performance (Radio ZS, Sept; 1982)

4)     Radio Relay Systems (Thomson-CSF 1981)

* P.O. Box 32564, Glenstantia 0010

** p.s. added 19 January 2013:

1.     Thundershowers in the scatter volume increase the refractivity index and the S/N ratio by around 20 dB. Lightning bursts in that volume further increase the S/N ratio 10 dB or so by providing a more solid reflection sheet. The carrier frequency is much more dispersed, by as much as 100 Hz at UHF under severe thunderstorm/unsettled weather conditions.

2.     The north/south path does not show noticeable enhancement.

3.     The effect of a tropospheric duct remains untested.