ItADC-TR-78-34JN POUSI kEPORTFEISUARY 1978

ItSA Follage Penetration Summary

" .- AUGUST GOLDEN, Capt, USAF

It, ROME AIR DEVELOPMENT CENTER

AIR FORGE SYSTEMS COMMAND

A GRIFTIrS AIR FORCE BASE, NEW YORK 13441

Best Available Copy 06 0 9 0574,

1'1

This report has been revlmed by the RADC Information office (01)and is releasable to the National Technical Service (NTIS). At NTISIt will be releasable to the general public, Including foreign nations.

This technical report has been reviewd and approved for publication.

APPROVED:

LIPC KSMITHChief, Electroargnetic Systems Concepts Branch

APPROVED:

WALTER ROTtVMN

Acting ChiefElectromagnetic Sciences Division

FOR MHE COINEPI(;

Plans Office

• I..

I..

I II I I#'~*

.. . .. ,t ,I.. . ,,4., ' " - . ....... .. . .... 1 A

Unclassifiedj,* ECURITY CI.Alit Fit*AIION OF THIR 0409 (le~p le Eetsin.eod) ________________

RZAD INE1TRUC11ONS14REPORT DOCUMENTATION PAGE WEoltz COMPLEFING FoRMi

---- 'S"VT NCES ON I. I- ARCIPItNVIS CATALOG NUM8ER

flut (ani Submis)5. Typt or RESOAT 6 11,91IO0D COV11RED

A FOLIAGE _ENETRATION SUMMAR In ousN. 0911RPORhIINQ 01R16 ROEPORT NUIMBER

AUTHR # . 0041 AAC T OA ORANllY SIUPiIR7sAugust Golden Capt, USAF

Deputy for Electronic Technology (RADC/EEC) 4600W1602MNNU

Masachsets 073 PE62702F

Aepprove for pEletoic Teheole g disributio uniiEd.

Foliagen pronetrationn t

Apreoenedfrpbi ees;ditiuinulmtd

energy111U10 ii ARWN at(note mhorAtet etrapdl In floliage, at h~leigherý frqunie.AloWsg

d Actin of sualrvdeyfeitiong frte andflig attenuation within flaerepsen tedy -is-

energ to ttenutedSoreUrpiTl inBICTO folag at5 hiADe freuenies DAlso Btig-

ne elcin rmtefAaoaritrfc r hw ob infcn ol

mechanism.........a..,.....................................................................................-

B.41si Ni lo (o

FIT "JOf 0'$ I ;,". t. .

..........,, ...... . .. ..... • ,w

fly

Contents

1. INTRODUCTION 5

2. FOLIAGE ATTENUATION REVIEW 5

3. FOLIAGE ATTENUATION FORMULATION 9

4. TRANSMISSION COEFFICIENT FORMULATION 20

REFERENCES 27

Illustrations

1. Foliage Attenuation vs Frequency -Averages or Measured Data 7

2. Foliage Attenuation vs Frequency - Plot of Various Formulations 83. Foliage Attenuation vs Frequency - Multilateration Study Report 0

4, Problem Geometry 10.5, Foliage Attenuation per Unit of Forest Height vs Frequency -

Constant Conductivity >I0- 4 S/m 140. Foliage Attenuation per Unit of Fprest Height vs Frequency -

Constant Conductivity <5)< 10"5 S/m 167. Foliage Attenuation per Unit of Forest Height vs Freqluency-

"Variable Conductivity 188, Foliage Attenuation per Unit of Forest Height vs Frequency -

Variable Conductivity e = 1. 0 199. Foliage Attenuation per Unit of Forest Height vs Frequency -

Vuriable Conductivity e 1.05 19

fi ~3

:,

S .................................-...-....... .,.y,,h

** .. *

I.

Illustratlons

10, Foliage Attenuation per Unit of Forest Height vs Frequency -Variable Conductivity c = 1. 1 20

11. Foliage/Air Interface Transmission Coefficient vs Frequency -Normal Conductivity 25

12. Foliage/Air Interface Transmission Coefficient vs Frequency -Increased Conductivity 25

13. Foliage/Air Interface Transmission Coefficient vs Frequency -Constant Conductivity 26

Table

1, Total Foliage Attenuation Example 26

4I

0i

. °4

.........................................

•I

A Foliage Penetration Summary

1. INT1i0I)ITTON

SucceRsful airborne radar design requires the consideration of many inter-related factors, When that radar must detect fixed and slow-moving ground targetsimbedded in a foliage environment, the factors requiring consideration increase innumber and complexity due to the increased radar performance required. Someof these factors are the percentage of energy transmitted into the foliage region,

the attenuation of electromagnetic energy in the foliage, the target backscattercross-section, and the effects of multipath. We shall be concerned only with thefrequency dependence of the first two factors: the transmission of RF energy Into

and its subsequent attenuation within foliage. Our approach shall be to reviewexisting, relevant data and theories and then to generate curves useful for the pre-

diction of system performance tn foliated environments.

II.• ~~2 'iOLII.IAGEI, A'riTNth\TION HlViEV

We begin by presenting existing data from various sources on the foliage

attenuation of RV energy, L. V. Surgent references many sourcem for

(Received for publicotion 15 February 1978)

1. Surgent, L.V, (1974) Foliage Penetration Radar, Hiatory and DevelopmentTechnology, Army Land rWarfare Lab, Aberdeen Proving Ground, Md.

-- ' . .. .' .. . .. ""

.'7 i II I I I I II I I

electromagnetic-fieid attenuation data in a report wr'itten for the Army Land

Warfare Lab, Data from these sources are shown in Figure 1, which was extractedfrom Surgent's report. The numbers of the figure pertain to the references listed

below the figure. Numbers that appear twice refer to maximum and minimum data

values,Chudleigh and Moulton2 made a straight-line approximation for field attenuation

from an essentially different set of data sources than Surgeut's. Their relationship

is

* a 0,06 (f/l00)0, 82 dB/m f in MHz

Nathanson 3 summarizes the attenuation In tolitige via the following approximation:

a - 0.25 f 0 7 5 dB/m f in 0Hz

Li 0,044 (Y/10o)00') dB/m , f in MHz

Nathanson references the work by Saxton and Lane that is displayed as curve ( on

Figure I and was one of Chudleigh's references.

Lincoln Laboratory4 saw fit to describe the relationship between attenuation

and frequency asI,

a, 0.18 + 0, 3 logl 0 (f/100) dB/m f in MHz

log, 0 [l. 412 (f/ oo0) '03

which in obviously completely different from the Chudleigh ov Nathanson results,Lincoln Laboratory used data from Jansky and Bailey (curve 7 of Figure 1) as part

of their data in arriving at a result.

Over the frequency interval of 100 to 1000 MHz, these two formulations differlittle, However, as the frequency continues to increase, a divergence in calculated

attenuation develops. The Lincoln Laboratory prediction results in an attenuation

proportional to the logarithm of frequency, while the other approximations indicate

a variation proportional to some power of frequency less than unity.

2. Chudleigh, W., Moulton, S. (1973) Long-Range Standoff Radar SurveillanceStudy, AFCRL-TR-73-0145.

*3. Nathanson, F. E. (1909) Radar Design Principles, McGraw Hill, p 19.

4. Lincoln Laboratory (1909) TR 472, pp 92-93.

I.

i(

w 0.0500

0,-l

2 5 10 20 50 100 200 500 1000 2000

FREQUENCY, f WMHO)

G S. Krevsky, "lil and ViF' Radio Wave Attenuatinn ThroughJungjle an(I Woods," IMr.t Trans on Antennas and Propagation,

Vol. AP.II, lip 506-507, July 63.G J.A, Sdxton and J.A, lanp, "VUr and UlIF Recepttoii-Effocts

uf Trees and Other Obstarles," ilrles. World, Vol. 51,

lip 229-232, NAy 1965.01 L.G. Sturqij I lnd Staff, "tropical Prnpalnqatin lRosearch,"

J.insky and 1lt ley Engijleerlag Dopt, , Atlantic Research,

Corp. , Aloxandria, VA,, final Report, Vol. 1. June 1966,G I.. Parker and GA.iL. 11lean. "Feasibility Study of the

lUio of an Upani.Wti'ro TrAnti, nslon Line, COpacItor And

CLvities to MNoisure Elocti'lcal Prolperties of Ve(litotton,"

- - Stanfoid Research lnst,, Special Tech Repoýt 13, AugustS ' - 1966.

WIL, Parker onn W,. 1ikarahliroaiya, "Elrctrical ConstentlMeasured Iii Vegetation and in Earth at rive Sites inMalilard," Stunford Research Inst., Special lech. Repeort

43i Dec 1967.

Figure 1. Foliage Attenuation vs Frequaency -Averages or Measured Data

II

II* •5Oli~i.•,.AliqMlkMiaIiJ•it."l. ,,ir Ii~ l~il ,i .- , ...

These various approximations to the attenuation factor are all ohown in Fig-ure 2, and it is obvious that Lincoln Laboratory is the most conservative over thefrequency interval shown. This conservatism is reasonable when frequencies be-low 100 MHz are to be considered, due to the spread in data shown in Figure 1,

Finally, Figure 3* suggests that foliage-induced losses do not Increase withradar operating frequencies beyond 300 MHz when the incident ray grazing angle

is greater than 6 degrees.In spite of the divergent measurements depicted on Figure 1, the preponderance

of data supports the conclusion that RF signal attenuation per unit of path lengthincreases as the frequency increases,

1.0

0, -4.,,011I

/. SAXTON- LANES/ •:) JANSKY" BAILEY

S. / / : / ,• . . .. .N A T H A N S O N.,. .,. CHUDLEIGH

"-'•LINCOLN LAB

0,01 - i I i i i I i i i . .

10 100 1000FREQUENCY, f(MHZ)

Figure 2. Foliage Attenuation vs Frequency - Plot or Various

Formulations

t

'Appeared as Figure 21 in Appendix B of,,RADC-TS-74-275, Vol II, Multilateration

Radar 9urveillance/Strike System (MRSO) Study,

2.0-

00

GRAZING ANGLE

z L.O

I-..

S~120

10 I00 IOOQ

FREQUENCY, WMHz)

Figure 3. Foliage Attenuation v.s Frequency - MuIttlateratlon Study Report

3. V0,'1,AI;E AT"TENIATION M1011I1,010IA'II

How then can we explain Figure 3, which run' counter to this measured trend?

Possibly through R purely mathematical development, which we shall now pursue.

Assume that a forest region, or tree height h, is a region of lossy dielectric

material with a constant relative permittivity und conductivity given by el and.0

respectively. According to Tamir, this model may not be realistic for frequen-

cies above 100 Ml-ir, based on data taken by JIang und Parker of SRI in 19110,

Tamir concludei; that for lower, frequencies the model is good. SuegentI notes that

the dielectric slab model for the forest has been shown to be valid for WeFe up to

L Band, it the modpl Is utilized only for a"'background configuration," This con-

clusion, Surgent indicates, results from Lincoln Laboratory's work, 0 However,

we shall cuntillue and apply this model for frequencies above 100 MHz to the prob-

lem of RF fieli attenuation within the slab region.

5. Tamir, r. (1907) On radio-wave propagation in forest environments, APS-15

(No. 0),

I6. Lincoln Laboratory (1910) Tactical Radar Program, ESD-TR-89-354, QuarterlyProgress Report.

9*I.

I

The forest I:: 1 egion II of Figure 4, The incident RF energy from region I has

the E field porpandicular to the plane of incidence and its propagation vector makes

the angle 01, with the normal to the semi-infinite forest region. In this develop-

ment, the finite thIckness of region II is not accounted for.

1 1

__Ii

Figure 4. Problem Geometry

From Maxwells Equations, assruming et time dependence in both regions, in

region II we have

Vxrl 2 2 2" (*J)2

J•LJ( (C . j- r2

J: -' w %E22I.'

with c" a/lwe

We know that the wave propagation in each region can be written as

a Exe' In I k W 2 M C

-2 221 I 'k : W oE2

Utilizing this complex propagation constant, we can solve for the fields in the two

regions in the same manner as for lossleas dielectrics, arriving at the following

equations resulting from the requirement for the continuity of the electric field at

the interface between regions I and II:

"* 10

• Ii

5'1

. l "EjklsinOy E e JklsinnY .+k 2 yI 2

This requires that k, sin 01 ,k 2 , which is real, and then kz 2 -y2,

Thus we have waves propagating in the +y direction and the -z direction. The +y

directed wave is unattenuated, while the -z wave is attenuated as

jkx~

where

k72 [u o c2 2 k2 sin2 r) /2 ki Ic' -, C , sin2 Il 1/2

We can solve for the attenuation, (r, arriving at

2 lk1 ~f~'sin~ o\2 / sin 2 a /

Thus, the one-way power attenuation through a forest of height h (meters) is

ATTEN - -8,(B86 a h df ,

It is obvious that for a fixed conductivity and rixed real part to the dielectric

constant, r', that as the frequency increases 0'/c approaches zero. Thus for U

large enough

-2 '2 sin 0,l

and we can write

kI e"

1 o 0 1 nepers/m (1)

s- •in 2 O. 4' - sin2 I

which is seen to be independent or frequency. At some sufficiently large frequency

and normal Incidence

t; ',1

"a"

4 a nepers/m i11 377 ohms

At some sufficiently small trequency c" >> ct e sin 2 91 C C02 0 where

#1 is called the grazing angle. Then 0

11/2

____ .r"-e 2 + 12 2

-2 --2. = M 0' oC (2)

The latter is the form for attenuation of a plane wave in a conducting medium, as

given by Ramo, Whinnery, and Van Duzer, 7 This form also holds If c' I and0

Equations (1) and (2) suggest that the attenuation car. be written -s

a auonstant X f0 ' dB/m

over some frequency range, after which it transitions to a constant value.This critical frequency can be found as follows- The constant attenuation con-

dition applies when

(o 2 2(e(HC02

or

f >> a=fc

2ffeo(C' - eog2

As the condition pertains to the square of F, f 1- 0 to should yield the constantattenuation condition

7. Ramo, Whinnery, and Van Diuzer (1965) Fields and Waves in CommunicationElectronics. John Wiley & Sons, N. Y.

12

Si

71 0

For low frequencies,

or

0

*Again, f should be sufficient for c ,t i Taiir~ indicates that the rangeof values for e' and a should be less than the limits he used, These were, for fre-

*quencies between 850 MHz and 100 MHz, 1. 05 < ' 1, 15 and 0. 05 < a < 0. 15ms/ni, i~ppnman used the following values, considered good at 2 MHx:

C, =C: 0. 02< a <0. 05Sfm ..

Surgent suggested that at 190 MHz

It Is hard to conclude that a Is constant with frequency from this sampling of data.

However, we can make this assumption and proceed, as our Intent is to derive

total attenuation in foliage curves consistent with the values on Figures 1, 2, and/or3. Therefore, assuming some representative dielectric constant values, let usobserve the resulting values for a in dE/ni for varying radar ranges and heights aspresented on Figure 5. It is obvious that high dielectric constants reduce the forestpath length and the resultant attenuation rate in foliage for a constant conductlilty.Figure 6()most clsl approximates Figure 3 in terms of the trend of the curveswith frequency. The specific curve, however, could not be reproduced.

How can the curves of Figure 5 be compared to those of Figure 2 '? Figure 2contains measured attenuation versus frequency and has no hint of a plateau on th"7

8. Lippman, B.A. (1965) Jungle as a Communication Network, ARPA Report.

13

o . .

F) in

w b b

to 0 Ila I

*~~~&i~" b Nvb_ _ _

*1QP IDNIvn~i

14.

Iwos

attenuation. Figure 5 in turn is based upon an assumption of constant dielectric

constant and conductivity with frequency. It was shown that for a sufficiently low

frequency the attenuation can be written asI'

ai a 8 ,88 647 v? dB/m . (3)

This form agrees well with measured rates, with the frequency dependency varying

from log, 0 I0 ' to logl 0 P3/ 4 . However, this form for the attenuation should be

good only for frequencies well below 1. 8 MHz, as this is the previously defined

critical frequency for values of a = 10-4 S/m and e' ' 1. This critical frequency

would suggest a constant value of a above 20 MHz, which is not supported by meas-

urement.

Also, the attenuation would vary as

a -= 0, 173,M'H d13/m ,

which doesn't agree with Figure 2 (that is, at 10 MHz, a 0. 55 dB/m, not

0.01 dB/m).

To make Eq. (3) approximate the data of Figure 2 at 100 MHz, we must choose

a and el' to be of the following orders of magnitudet el'* 1.0 and a im 1. X 107 S/re.

The latter quantity is 2 orders of magnitude below Tamir's smallest value. For

various fixed conductivities in this range, curves of attenuation per meter of forest

height are shown in Figure 6.

From Figures 5 and 8, it can be concluded that the constant complex conductLv-

ity with frequency model is incorrect.

A better model can be derived. The measured data of Figure 2 should hold for

a wave in the forest that was generated at normal incidence, In this case,

,- neper/m

It we assume t' e constant and e"/c' 4c 1, then

where q c wave impedance in forest (real) - 0/e 120W ohms. Then, using

Nathanson's approximation for the wave attention, we find

15

W ' ' W J1&i~ ims"' " ....... ....... ..

4on

IOU) 1

00I

o 0 1!

o x

0 1

(4 In-0 N - N -

CY N cy 0

(w/q~~2 m)Ntvn3

16

I,

a(f) 2a ' 00.86 O"75 PS/m r finlMHz I' 1I. 'It seems resonable then to use this variation in a(f) in evaluating the signal atten-

* uation. This has been done, and the results are shown In Figures 7a through 7d.These results are very much different from those obtained by using constant

conductivity, but the results for normal incidence agree with measured data. In

Figures B, 9, and 10, these curves are plotted in log-log fashion to show theiragreement with Figure 3. By design, the attenuation agrees at normal Incidence

with Nathanmon's estimate, Au the incidence angle departs from normal incidence* (900), the attenuation Increases. This agrees with the model in which the attenua-

tion per meter is constant at a given frequency, but the path length changes with

incidence angle. Since at near-zero-degrees incidence the propagation constant k

is nearly parallel to the dielectric surface, and the component of k parallel to this

interface is real, all the attenuation must be absorbed by the propagation factor

normal to the dielectric interface, k5 .A physical model to explain this assumed variation of conductivity is as followsI

The forest it filled with one type of material of constant conductivity and permittiv-

Ity. However, this material is formed into numerous, assorted, rod-shaped bodies

of various lengths, diameters, and length-to-diameter aspect ratios. Also, farmore space is filled with twig and branch-size objects than is filled with trunk-msie

objects. This variation in density of object size results in a greater currentdensity being produced at higher frequencies via there being more resonant objectsat the higher frequencies. of course, greater current density implies a higherconductivity, and greater attenuation over a given path length.

If the above it true, the difference between 400 and 1000 MHz at 30 incidence

is 5. 6 dB/m vs 1.2 dB/m one-way nttenuation (e' 1. 05).Queries were made to locate substantiation of Figure 3 at HADC and Lincoln

Laboratories, but no originator could be found, It is thus advisable to use Figure 0as a more realistic estimate of attenuation in foliage. Figure 9 to selected over

I: Figures 8 and 10 as a dielectric constant of 1.05 it more consistent with Tamir 8

and Surgentts conclusions. - -

17

.1

"/, • mW•W v~w ,w•.,,.*., • . ............... ... . .17

• I I I I I I I I I I I I I I

T 10

-A - 4)

18

. ... . . ....

06r

(121,0 *0*

01" °..........

0.01 10

"0.0' a a .II

10Ol 100 1Q00t0 fEQUENCY'f (MHi)

Ftgure 8. Foliage Attenuation per Unit of Foreit Height vII Frequency -

Variable Conductivity 9 1'. 0

1.0 - 0*

0.

0,11' I '

to 1o0 I000FREOUENCY ,I IIAHuI

!, If

Figure 9, Foliage Attenuation per Unit of Forest height v- Frequency -

Variable Conductivity e - 1.05

19

V ,.., .. ... . . . . .

1.0 *0

S0.11-

Z

10 100 1000FR EOUENCY, fMWHO

Figure 10, Foliage Attenuation per Unit of Forest Height ve Frequency -Variable Conductivity e - 1. 1

4. 'I'RANSIS118,4N ('0EFFICIFNTFIORS1I1I, 0I'ON

Foliage attenuation, however, is not the only contributor to signal loss. Atthe smaller grazing angles, 0,, a significant portion of the incident energy is re-flected from the dielectric discontinuity. The relationship governing the trans-mission into and reflection from the forest-air interface will be developed next.

For our horizontally polarized electric field, again from Figure 4, the fl(,ldsmust satisfy the following conditions at the boundary of regions I and Ili

" + X + " -]y + ÷ = -yIX Ex1- Ex 2 0 , Y y1Y 1 2

that in, continuity of fields at the boundary.I ~It the fields in the two regions can be defined an planar with a real or complex

i propagation constant (e1kz as field travelling in -r, direction),

t Jlkyy+kzZ)

20

:'I

I.I:i: Then from

:Ex 0 0

In region I for the incident field

- 0 -1 . -'I~ ~ 1z;

atnd for the reflected flrlr

Ex• wp•o

H ky1Itl

Similarly, in region 11

a 2

H Y; "'2

Inserting these ratios into the boundary conditions yields

* j-* 1f•-- + tran rmtidion coefftclentEi•I• '1 _

Si 2 Z2-7

•.1.kz, 'k2" Y'2 ' Y2 kYj ,mn9 k~lkCs0

21

'I II I III I l II II l m II II I IIII I I IIH

and

k2 2 jAA (I e"l k2 (C ' J , Iwe have

Ex2" 2 cos

E05 o C 9 + ý C J. -.I. 2 n i

11 we allow 01 ,r •2, -. -. 0 unless el - je" - sin2 01 -. 0 properly, But this cannot

be if e" # 0, Thus, for either e' > 1 or c" 0 0, as grazing incidence is approached

the transmission into medium two goes to z.ero. This loss must be added to the

actual attenuetion in medium two,

The incident power In region I Is given by

Ii He I x n*) 'E X.2 / n

P 1 1 1 -- I xI t n

The power flowing through region II at the interface Is .

As

+ jk + k

H2H WAA 0 2 Y 2

we can write

I x Ex F k +2- 2 woo z2 Y2

+'~~x (-k k i).. +:- '+WMo Z(2 Y2 WMo Z2

22

Then

I. II.. I ,

2;2

with

Y2 + "2(2R + 'z21

LI ~2

As

k y 2 x k , s in 0,

the real part of Isk2

k 2- 2 n 2] 1/2kzS 3 • con € + 4(el Colo. f) ell,,

and

1~1/2P 2 1 + 112( 4 + •'J•1, /2

For small grazing angles, ! -, 0, what is the relative size of e" to2Col2 81? Using the formulation for volume conductivity developed previoualy,

,.,a(f) s,0.272 X 10-10 P3/' 9/

ell cog 2 Ox for all cases where ' 1, 105

I 23

.1

Therefore

"2 . 1 12 'P

This transmission coefficient is plotted in Figure 11 for various values of

dielectric constant and a variable conductivity. For the range of values choson,

it is obvious that the conductivity plays no role, If the conductivity is Increased

by a factor of 10, then the transmission coefficient is only weakly dependent upon

frequency, as shown on Figure 12, This is due to e" . -. 0 as w .*. ®. For allwecases it is obvious that for grazing angles less than 40, th• transmitted signal is

severely attenuatedi that is, 10 log 1,P"12 > 3 dB, even at 13' grazing angle. It thus

seems that overcoming the specular reflection from foliage requires a significant

radar transmitter power increase,If it is argued that both the conductivity and permittivity are constant with fre-

quency, the percentage of power penetrating the foliage iemaiiis the same at fre-

quencies above 100 MHz, as the tranumission coefficient is relatively insensitive

to the conductivity, In fact, using a cy :; 0, 15 mS/m, r. decreases for frequenciesbelow 100 MHz, as shown in Figure 13,

In researching the subject of the fraction of the incident power that enters

region 1I, I found another formulation for the transmission coefficient that has been

utilized by others, 0 This formulation is

sin 201 sin 202S sin2 (t -1 . )2)

with 0i and 02 as defined in Figure 4. This equation was, in turn, extracted from

Stratton, A review of Stratton shows that this equation gives the fraction of power

normally incident on region II that enters region II, This is not the same as thetotal power in region It relative to the total power incident from region I, as given

by

''2__ 2_ coso -0 sin01TO 2 sin 02con 01+ Vs-i" -s0i

9, Labitt, M. (1970) VisibtlitZ of Targets as Seen Prom a Long-Range AirborneSLincoln Lab Memo 43M-885.

10, Stratton, J,A, (1941) Electromagnetic Theory, p 496, McOraw Hill.

24

. 20 - (a)I I.,,

-I0 102 l0r'104fa -.05 FREQUENCY, I (MHz)

POWER TRANSMITTED vs FREQ

S#-OGRAZING ANGLEV1_ -40 -4/

-300

-20 I',• -20

~-I0 __ __ _ __ _ - _____, __,_ 4_-- 4. _'1"0 - 4- 1

10 100 100• .. .. 0 100 , .00.... i , , , , i.IFREQUENCY, f (MHz) FREQUENCY, f (MHZ)

Figure II. Folrige/Air Interface Transmi~witon Coofficiont vs Frequency -

Normal Conductivity

-40 E'I.01 -40 #,.l.On

&,8.6;AS/m - 8.6 S/m3- O -30C

-20 -20_. _.

-10 10 4

17-1" i n7-1310 100 1000 10 100 1000

FREQUENCY, f (MHz) FREQUENCY, f (MHz)

F'igure 12. Foliage/Air Interface Transmission Coefficient vm. Frequency-Increased Conductivity

25

t 1'.0 5 l t

S aO.1 5mS/m CONSTANT a 0.15 mSl/m

920-4..C

,'• _ _ - . ..

tOO 1000 100 1000FREQUENCY, f (MHZ) FREQUENCY, f (MHO)

Figure 13. Foliage/Ai' Interface Transmission Coefficient vs Prequency -Constant Conductivity

If 0"= , we can write

sin 2ol 2cos 0l sin 022 +sin (01 02)

and the results differ by

T, COS i 01T1 cos 02 sin 511

Thus at small grizing angles, a larger transmission loss is calculatea herein

than that derived in Labitt,

Combining now the tranamission lose and the attenuation loss we can derive

the two-way signal loss experienced by an AMT1 radar looking for targets under a

foliage canopy. of course, this excludes the other loss mechanisms previously

mentioned. Choosing for parameter el = 1.05, a % 0.86 S/ni at I MHz, and a

""ore.st height or 50 ft (15, 24 in), the results are as tabulated in Table 1,

Table 1. Total Foliage Attenuation Example

41 rceq, G~lz e d13/m 2 Total Atten, dB

40 0.4 0,5u 7 31.1

1.0 1.2 7 .... 50 1

120 0,4 0.42 2 16.8S1.0 0A. 2 28.4

26

.2

References

1, Suraent, L. V. (1974) FoliaPe Penetration Radarz H13to1 and Deve0opmentTechn1oio , Army Land warfare Lab, Aberdeen proving GrounId, td.

2. Chudleigh, W., Moulton, S. (1973) Lo.n-Range Standoff Radar Surveillanbe

Study, AFCRL-TR-73-0l4 5,

3, Nathanson, F. E. (19B9) Radar Design Principles, McGraw Hill, p 19.

4, Lincoln Laboratory (1969) TH 472, pp 92-93.

5. Tamir, T. (1967) On radio-wave propagation In forest environments, APS-15(No. 6).

e. Lincoln Laboratory (1969) Tactical Radar Program, ESD-TR-g9-354 ,

Quarterly Progress Repsrt.

7. Ramno, Whinnery, and Van Duzer (1965) Fields and Waves in Communtcation

Elect cs. John Wiley & Sons, NY.

8. Lippman, B.A. (1965J Jungle as a Cornmuntcatlon Network, ARPA Report.

9. Labitt, M . (1976) Visibility of Targets as Seen.From a Lon g-Range Airborne

iad.r, Lincoln Lab Memo 3M-B8.

"10. Stratton, J.A. (1941) Electrornagfetic,, Theon, p_496, McGraw Hill.

*.

MISSION

RomAirDevelopment Cente?'

MDC plane MM conftcts research, .xfploraeory and adfranoeddeveoqgaiet prozaw in commnd, oamitzol, and 00MswICOetGcn(C3) activites~, anM in the 03 area of W~oZSati0. gwjeww

and int.221gence. rho principal tweliial miorsian areasare coanlmcationam, electromagrnetic guidanae add m'n~tml,surveillancie of ground and &*riopace objaeea, IflC*'3iUUAQA4ta collection and handllng ,Inforuatlcm myste teoh0*lgyIonospheric prcpagatlon, aolid state scoISw* de ixoviv@

pfigxicx and electronic reliability, inaintainability liad

cvAmpa:011ilty.

Best Available Copy

Mal4 toed 1.. Air ForatH..... OR', Mesa. 01131