JOURNAL OF RESEARCH of the National Bureau of Standards-D. Radio Propagation Vol. 67D, No.3, May- June 1963 Reception of Skywave Signals Near a Coastline

J. Bach Andersen

Contribution from Laboratory of Electromagnetic Theory, The Technical University of Denmark, Copenhagen, Denmark

(Recei\'ed November 5, 1962) An experimental investigation has been made on t he influence of gt'ou nd inhomoge­ neities on the reception of skywave sig nals, especiall y t he in fluence of the conductivity contrast ncar a coastline. This gives rise to a r apid decrease in fi eld strength ncar til(' coastline as is "'ell known from g roundwave mixed path theory. Comparison with Lheory is given. Infl uC'llce of dilhts(' reflection from t he ionosplwre is also cOIl :; idereri .

1. Introduction conditions, i. e., a s Lraight boundary lin e, n,lt ionosphere could be f\ccoun ted for. ground- thi s is, 1'01' exa,rnple, the case at f1 coast­ To makc t he study IlS complete as possible th e lin e- a nd also that they may vary witli Lhe depL h. corresponding curve fo r the groundwave mixed pllth This paper deals with the influence of the ground was also m easured , lh e transmiL Ler b ein g siLuated on t he reception or skywave signals at a n actual flt ground level far from Lh e COflstlin e over the sea. site ror a receiver stn,tioll close to the seashore at Finall~T the aLtenuaLion of t.h e groundwave on Lhe R eersf), D enmark. homogeneo us ground was measured in order Lo find Part of the a ntenna station which will be erected t he degree of in homogeneity of the ground a nd to at R eers0 will receive skywave signals from Godthab estim ate . the value or t he ground co nstants. in Greenland. In the d ay tim e t he ionospheric re­ The eff ect of diffuse reflection from t he iO ll osphere fl ections will take pl:tee in t he E -Iayer a nd during and t he effect of t he Clll'vature of the ear th is also ~ t he ni ght in th e F-laycr, a two-hop path via th e mentioned. E-layel' and a one-hop p ath via the F-Iayer; the geometry is such that t h e angle of arrival in both 2 . Theory for a Two Section G round cases is very small, about 4°. It is well known that it is difficult to m ake an tennas which operate well In a previous paper [Andersen , 1962] a solution at such small angles du e to the absorption in the was given for the radiation field from a ver tical dipoJe ground. situated on an inhomogeneo us ground co nsistin g o f One method of obtaining an improvement in the two or three sections with arbitrary electrical pl'Op­ recep tion at low angles is to use an extended ground­ elties.! Numerical calculations were given in the wire system in the direction of propagation (or case where one or two sections wero perfectly CO Il ­ reception) [A. C. Wilson, 1961]. Another method ducting. The result for a t wo section ground is would b e to choose the a n tenna site n ear a coast- given here for reference where a t ime factor I lin e in order to utilize the high conductivity of sea exp (-iwt) is employed. A(f , 1'0) is the radiation r wa ter wherever it is possible. However, the antenna pattern for a ver tical dipole placed on section 1 at would still be situa ted on a lossy ground a nd it a distance To from seetion 2, and f is Lhe angle of r would b e of interest to know the influence of the elevation . various sections on the radiation pattern. A theo­ r etical solution to this problem has previously been I A more ge nom! so lution has heen given hy W ait [1961 :for th e case ofa spher· given [A nd ersen , 1962] but due to the idealized ical earth. J lowev('r, it is no t in as con venient fo rm as (1) here. 325 (1)

(nth medium)

2(l -cosy.,) to c iui (1-ui) Po to= kl'o(l-cos y., ) (lc= wave number)

(numerical distance to boundary)

(Fresnel in tegral)

(So mm erfeld's attenufLtion function)

R. is the Fresnel reflection coeffi.cient for vertical The horizon tal profiles of the paths are shown on polariza tion. fi gur es 4 and 2 respectively for the skywave and the vVhen medium 2 is perfectly conducting, '/)2 = 0 groundwave measurement. UnJortunately, the two and R v(2) = 1 and paths could no t b e made the same. The paths are nearly perpendicular to the coastline. For both the skywave path and the groundwave path there is a -

LAN D ISEA --

- FIG UH E '2 .- GI· ollndwave mixed path, sea-land. '( - 7 0 0 , l\!.ca surcd fi eld intensi t y - , Th eoretical cun'c . E= lO ,u= 2.10- 3mho/m f- ~ ~ f = 25.218 _\ I Cis -80 r---.r-. r- "- f 0 15.663 MelS - 90 i't" high conductivity soil is entered. The small hill I- seems also to have some eff ect on the received fi eld ~ ~.n -1 00 strength but the unknown ver ticftl distribution of I- different layers in the ground nny be tbe true cause for the r apid variation. One hunch ed ftll d fif ty -11 0 ~ 10 2 0 30 40 5 0 100 20 3 00 4 00m m eters from the coastline Lh e ground is homogeneous, and the correlation b etween exp erim en t ft nd th eory F I GUHE I.- Groundwave fi eld strength over land. seems to be good.

4. Groundwave Mixed-Path Measurement 5 . Skywave Measurement

A small transmitter U = 25.21 8 M c/s) was erected The experimental setup consisted of two identical at Funen about 20 km from R eers0 over the water receivers with identical dipole antennas and two in a direction almost perpendicular to the coastline. separate recording units. The receivers were high­ Though there was some attenuation of the ground­ quality communication receivers. The two sets wave over the water this has a minor inf-luence on were separated 100 to 400 m in differen t distances the groundwave mixed path attenuation curve as from t he coastline. The skywave signal from a measured on land; the situation is approximatively transmitter at Godthab, Greenland, sending a t a that of an incoming plane wave over a plane perfec t frequency of 14.447 Mc/s, was recorded simul­ conductor m eeting a conductivity contrast when taneously at the two receiving places. Because only due corrections are made for the inverse distance the relative mean fi eld strength was desired, no dependence. absolute value was found. By calibrating the t wo The fesul t of the measurement is shown in figure 2 sets t o a common logarithmic scale, the m ean differ­ where the points denote the measured fi eld strength ence between the signals at the t wo differen t places in comparison with the theoretical curve for a ground­ was easil y found. One day's recordings were in w[w e transmission from a ped ectly conducting general sufficient to provide a reliable measure. ground t o a lossy ground with ground constants The rapidly varying component was somewhat £= 10 and (J = 2. 10 - 3 mho/m when the transmit.ter suppressed by applying a time constant of the order is infinitely far ftway . According to the theory a of a few seconds, and the slowly varying components rapid decrease in field strength should take place at the two antennas were highly correlated, because right after the coastline, the so-called "converse" the optical path is nearly the samo. The distribu­ eff ect. This is seen to be true for the first two tions of the signal have not been st udied as only the points but then a sharp increase follows when the mean value is of interes t here.

672865- 63--5 321 27.5

23.2

18.8

14.2

9 .0

> ,::t. .0 'U

23. 1 ,_.- -,- f-- -~ f:::", 19.8 ,- .. 1--' 16.4 - 1-- --f--I-- 12·6 l~- 9 .0 ttl~ :1·· ·Nf · 4.7

FIGU RE 3.-Example oj recordings.

Simultaneous recordings for To 72 m (upper curve) and TO 420 m (lower cun-e) from coastli ne.

An example of the recordings is shown in figure 3. The upper curve is for the receiver set close to the 1m ~~~~EPROFIL EOF PATH coast (r = 72 m), and the lower curve is for the Om receiver far from the coast (r= 420 m). The mean /0. difference between the two signals is 7.6 db. The 0 / q results for other positions of the antennas are shown \ in figure 4 as points together with a suggested curve ·1 ~: \ connecting the points Two theoretical curves are \, n\ p- ..... 2 . q , shown for two different conductivities, rr = 10- mho/m \ ..... , 3 , and u= 2·10- mho/m, of which the latter should be ~~ ~/,_/ typical according to the results from the ground­ -6 " f-~ wave measurement. It is seen that there is not db ~ - ...... ~ -- 0 such a good agreement in the skywave case as in the -8 --- groundwave case, though the general form of the ~ r--E- 10, "-10- 1 mho/m curve is correct. ·10 ~ The rapid fall in the beginning is present, but the ~ ( high conductivity section gives an increase in field -12 ~ t- IO , ~_ 1 '1 0 ' l mho/m strength which displaces the curve in the vertical direction. This is similar to the recovery effect ·1 4 known from groundwave mixed-path theory. Fur­ I I thermore, it is suggested that the caraway field gives o 20 100 200 300 400 a similar attenuation of the skywave as it does to fO ' meter the groundwave. Due to the inhoHlOgeneous ground, there is a discrepancy between theory and experi­ ment, but the form of the curve can be explained by FIGV R E 4.- S kywave measurement near a coastline. applying the mixed path concepts known from 0 , Measured fi eld intensity ground wa ve theory, together with the theory re­ ~ , Suggested curve connecting tbe points ~ . 'J'hcol'cticaJ CUrVeS ferred to in section 2. If the ground had been f ~ 4 0 . f = 14.447 Mels 328 homogeneous a loss of 13 db co uld havc been expected 60 is the standard d eviation of the distribution. 400 m from the coast and even farther away, in­ It can easily be shown that if stead of the 8 db m easured. Apart from the influence of the inhomogeneous sec tions, three more reasons could be given to explain some of the discrepancy. These arc the fini te theLl conductivity of the sea, the difruse reflection from the ionosphere and the curvature of the earth. B o(if;o) = P o' A(if;o) . [1 + a265 + highel' order terms].

5 .1. Finite Conductivity of the Sea Assuming A (if;) = C ~B v cos if; y,L he radiaLion pat­ By applying formula 1 instead o( formula 2 the tern for a "ertieal dipole over lossy ground, LheH finite conductivity of the sea would be accounted fol'. A rougher estimate is found by simply finding cos if;o a2= cot if;o-tgif;o- · ,I, + the magnitude of the proper refl ection coeffIcient Slll 'f' 0 U for the sea and displacing the asymptotic value (/'0---7 00) by an a mount equallo - [ 4+ 2 cos if;o cot if;0- 3 sill if;o 2 cos2 if;o ] sm if;o+u (sill if;0+U)2 . t.A= 20 log 1 R" db ~ Thus the influence o( the diffuseness o( the iono­ sphere is diffcrent fo r a n a nLenna on Lhe sea (u = O) because 0 db is referred to as the field in LensiLy at and on the ground (u r!' 0). The e([ecL Lu rns out Lo bc the sea (fig. 4). The effect diminishes the coasLline small though a can be seen frolll Lable 1, B is Lhe effect. ratio (l + az65),. =o For the cas e o( if; = 4 0, f= 0 and (T = 4 mho/m (1 + az65) ",.

t.A= l.2 db. measured in db, if; 0 = 40, (T = 2· 10 - 3 milO/ill , f= 10, j = 15 Me Is. The eO'ect diminishes the ('01tstlinc cff ect. 5.2. Diffuse Reflection From the Ionosphere TABLE 1 It is generally believed that the vertical reception pattern and the vertical radiation pattern for an oo· antenna arc identical. When dealing with ionospheric propagation it must be remembered that the wave H db O. 1 O. 4 0.8 1. 3 1. 9 refl ected from the ionosphere is no t a plane wave but a wave distributed around some preferred direction due to the diffuse reElection from the 60 co uld be computed if the distribution of the ionosphere. Thus a sharp zero in the radiation difference bctween the signals arriving at the paced pattern would be fUl ed with power coming from antennas was known [Bramley, 1951]. This has not neighboring angles, and the effective reception been done here. Bramley reports 60 aro und 1 to 2° pattern would not have a zero. For further refer­ for vertical incidence first order refl ections, and Page ences see Bramley [1951] and Page and Monteath and :Monteatb repor t 60 of the order of 10° for a [1955] . lower frequency. Some observations have also been The effect is treated here, too, because diffuse made on second order reflections. These show llluch reflection will diminish the coastline effect presented wider angular distributions and indicaLe Lhat the in this paper. Power will enter under elevation intermedi ate ground reflection m ay give rise to a angles greater th an 4°, and the coastline effect is greater spread in the direction of alTiv al t han t hat less pronounced for higher angles. produced by the ionosphere. If P (if; ) is the angular distribution oJ incoming For such a low angle of arrival as used here power distributed around some preferred direction (if; 0""4°), the distribution of the power is probably if;o and A(if; ) is the vertical radiation pattern (power), somewhat skew so more power is concentrated at then an effective reception pattern can be defin ed as higher angles; this effect would also increase B.

5 .3. Effect of Curvature of Earth For small angles of ,m 'ival the plane ear th assump­ tion is no longer valid. The field pattern based on A usual assumption is that the power distribution the Fresnel reElection coefficient h as a zero for if; = 0, is normal 0' but the solution JOl' a curved lossy surface has a finite value. Numerical res ults covering all prac­ P (if; )= ,P o e (Po total po\yer) -20~ tical cases h ave been given by Wait and Conda '\ 27r60 [1958], and the mixed path over spherical ear th has also been treated by Wait [196 1]. 329 In our case >/; = 4°, und it turns out that for j = 15 This study has been supported by Air Force NIc/s the curvature of the earth can be neglected. Cambridge R esearch Center of the Air Research I The geometrical optics approximation is valid, and and Development Command, United States Air the result for t he curved earth is the same as for a Force, through its European Office, under Contract plane earth. For >/; < Yz 0 IVait's results should be No. AF 61(052)- 503. The Danish Post and Telc­ tLpplied. graphs Administration furnished t he author with 6. Conclusion the necessary experimental apparatus as well as ' instructions and help in how to use it. The author The skywave field strength from a distant trans­ expresses his gratitude to these organizations. mitter has been m easured near a coastline in order to find the ratc of attenuation when passing from a well conducting (perfectly conducting) medium to a 7. References lossy medium. For small angles of arrival much Andersen, J . Bach (June 1962), The radiation field from a power is lost in thc ground so it is important to know vertical dipole on an inhomogeneous ground, Proc. Sym­ how close to the coast one should be in order to posium on Electromagnetic Theory and Antennas, Copell­ utilize the well conducting sea water. Previous hagen (in press). theoretical work suggested that when the coastline Bramley, E. N . (Jan. 1951), Diversity eff ects in spaced a erial reception of ionospheric waves, Proc. lnst. Elec. Engrs was in the first Fresnel zone in relation to the receiver, 98, pt 3, 19- 25. the rate of decay of the groundwave and the skywave Clemlllow, P. C. (1953), Radio propagation over a flat earth were nearly the same. This has partly been verified across a boundary separating two different media, Phil. here and suggests that a mixed-path groundwave Trans. Roy. Soc. London A246, 1- 55. Feinberg, E . L. (1959), Propagation of radio waves along a n measuremen t should give valuable information inhomogeneous surface, Nuovo Cimento, Suppl. 11, abou t the ground when a specific ground has to be Serie x. evaluated for HF transmitting or receiving purposes. Page, H., and G. D. Monteath (May 1955), The vertical For the ground studied here it turns out that radiation patterns of medium wa ve broadcasting a erials, Proc. lnst. Elec. Engrs. B102, No.3, 279- 296. there is some disagreement with simple t wo-medium Wait, J. R., and A. M. Conda (Oct. 1958), Pattern of an theory due to inhomogeneities of the ground. A antenna on a curved lossy surface, IRE Trans. on Ant. well conducting section near the coastline has the Prop. AP 6, No. 4, 348- 359. effect of displacing the "effective" coastline, the ·Wait, J . R. (196 l), On the theory of mixed path groundwave propagation over a spherical earth, J. R es. NBS 65D field strength does not decay as rapidly as expected, (Radio Prop.) No. 4, 401- 410. but th e effect is the same on the groundwave and the vVilson, A. C. (1961) Measurements of low angle radiation skyw:we. The ground consists of several sections of from a monopole, J. R es. NBS 65D (Radio Prop.) No. different conductivities, and they give rise to recovery 6, 641- 645. effects and converse-effects. These concepts are easily extended to the skywave in order to explain the deviations from t heory in a qualitative manner. It is shown that the effects of finite conductivity of the sea, diffuse refiection from the ionosphere, and curvature of the earth are of minor importance. The first two may give rise to a decrease in the above mentioned coastline effect by a total amount of 3 db. (Paper 67D3- 266)

']

i:-'

330