EDL-G266 463413

Directional Reflectance
and Emissivity of
= _an Opaque Surface

FRED E. NICODEMUS

")Y11 L\NI\, ELECTR NIC SYSTEMS

.'ELGENERAL TELEPHONE &ELENTRONIC.

II

7! r, PREP A RE D FOR THE UNITED STATES ARMY

 EDL-G? 66

ELECTRONIC DEFENSE LABOR-P9-% S

Mountain View, Califor--i-a

) TECHNICAL MEMORWIW .

i'19 May I_64,

DIRECTIONAL REFLECTANCE AND EMISSIVITY

OF AN OPAQUE SURFACE

Fred E. Nicodemus8

Approved for Publication .... John Barton

Head
OR Analysis Group

R. H. Krolick
Manager
Applied Science
Laboratory

Prepared for the .J.S. Army Electronics Research

and Development Laboratories.

SYLVANIA ELECTRIC PRODUCTS INC,

 EDL-G266

C ONT E N T S

Section Title Page

1. ABSTRACT .......... .......................... 1

2. INTRODUCTION ....................... 2

3. DIRECTIONAL REFLECTANCE ............ 5

4. DIRECTIONAL EMISSIVITY (AND

ABSORPTANCE) ................... 13

5. EXAMPLES .......................... 16

6. SUM M ARY . ......................... 20

7. APPENDIX A -- Reciprocity in an Isothermal

E nclosure . . . . . . . . . . . . . . . .. . . . . . . . 23

8. REFERENCES ............................... 24

-ii1 -
 EDL -G(66

ILLUSTR ATIONS

Figure Title Page

Geometry of Incident and Reflected Elementary

Beams. (Z-Axis is Chosen Along the Normal
to the Surface Element at 0) ....................... 7

TABLES

Table Title Page

! Radiometric Quantities, Symbols, Definitions

and Units. ................................. 4

- 1ii -
 DIRECTIONAL REFLECTANCE AND EMISSIVITY
OF AN OPAQUE SURFACE

Fred E. Nicodemus

1. ABSTRACT.
\

Concepts, terminology, and symbols are presented

for specifying and relating directional variations in

reflectance and emissivity of an opaque surface elernent.

Their relationship to more familiar concepts, including

those of perfectly diffuse and specular rcflectancc, is

given, and they are applied to illustrative examples. It

is shown that, when the usual reciprocity relationship

holds, the reilectance for a ray incident on an opaque

surface element is related by Kirchhoff's Law to the

emissivity of that element for a ray emitted along the

same line in the opposite sense.

 Z. ~IN UROLtJC)D• L'ION.

Reflectance and emissivity of the surface of an opaque body

are considered as properties of the surface material and of its micro-

scopic configuration (roughness) but not of its gross configuration

(curvature). This distinction between microscopic and gross details

of the surface configuration, which is to some extent an arbitrary one,

will be discussed fuither below. But reflectance and emissivity are

commonly defined or specified in ways which include an implicit

(and often overlooked) dependence on the geometry of the radiation

beam (including incident, emitted, and reflected rays and the effects

on those rays of the gross surface features). Even when this depen-

dence is recognized, the specified reflectance or emissivity is usually

applicable only to situations which reproduce the same geometry. On

the other hand, it is possible to specify the reflectance and emissivity

of an opaque surface (i.e., of any planar surface element) concisely

and unambiguously as functions of direction (with reference to the

orientation of the surface element) which can be applied quite generally.

The purpose of th!s paper is, first, to describe such a way of

specifying the directional reflectance and emissivity of an opaque surface,

to recommend appropriate terminology and symbols, and to relate them

to those in common use. Second, a relationship will be established

between the directional reflectance of a surface element (I e , its

reflectance for a ray incident from a particular direction) and the

directional ernissivity of the surface element for radiation emitted

in that same direction. In other words, the related quantities are

the reflectance for a ray incident along a line which intersects the

surface elerrent and the emissivity for a ray emitted along the same

line in the opposite sense.

The radiometric quantities used are listed in Table I, repro-

1
duced from an earlier paper The radiometric relations will be

analyzed below primarily in terms of the basic quantity radiance (N).

1
In the earlier paper it was shown that when radiance is defined, as

in Table I, as the radiant flux or power (P) per unit solid-angle ( a ) -

in-the-direction-of-a-ray per unit projected-area (Acos6)-perpendic-

ular-to-the-ray, it has the same value at any point along this ray with-

in an Isotropic medium, in the absence of losses by absorption,

2
scattering, or reflection. More generally, the quantity N/n (where

n is the index of refraction) in the direction of a ray was shown to be

invariant along that ray, even across a smooth boundary between

different lossless media.

1 Se list of references in ,½ection ,

-3-
 l i IInI 3ad nWc It I -L(I IAltti'51 SN.I'll k1,d(Iet tiit'i ...
s.and ullli ts.

QnaroitI Ss mnbo.
I Deftinig relations eits
Raiant nlc'rgv U J
) UI
i I i•cI,•,g
t l.cdi 5i a t, Jt .cIn

aU
Radiadnt posn er P P -
-,waott (IN)
0t

ta•,dmant hOintenit, J
oil
JI--- XV•: '-S

Radiant estittance If, lv o.'V i

Irradiance H Hj 0.4

al'l
Radiansce X NV -W.cm-XV sr-I
eosOOA 80

Wavelength X micron (y)

op
Spectral radiant power Px Pi- W.-
ON

Spectral radiant intensity Ax A-, -Vsr-',Ct-

ax
oW
Spectral radiant enshttance WX w'x -- W Acm--1
ax

OH
Spectral irradiance HX Hm-

ON
t
Spectral radiance NX IVcm-I.sr" u-I

Radiant enissivity Ratio of "emitted" radiant power to that

from an ideal blackbody at the same tem-
perature.
R ,.liant absorpra,,ce aRatio) of "absorbed" radiant power to incident
radiant power.
-A-e. •acn 0 Ratio of "reflected" radiant power to incident
radiant power.
Radiant tranismittanle e Ratio of "transmitted" radianl power to
incident radiant power.
nre
The ipe, tral radiant emissivity .eX) !{x 'Lea#0eilaX. Hence, the subscript notation ef, which
,:,!,;I1 be c-fused with a'/aX, is tnt recommended, although itis often used. Similarly, it is recommnended
ihaa the spectral absirptance, spectral reflectance, and spectral tranesnittanc- be written as nI.X, p(M,
a,,x r respectivel-.

-. ci,,-I ... 4 -- n* -
 ED L- G2 ,(,

In Table I. the definitions given for radiant err. ssivity dnd

radiant reflectance take no account of the effects of the geometry of

the radiation beam. The following treatment will refine these

definitions to recognize explicitly the way in which these quantities

may vary with orientation (relative to the surface). Only opaque

surfaces (of zero transmittance) and the geometrical ray optics of

incoherent radiation will be considered.

3. DIRECTIONXAL REFLECTANCE

Consider a radiation field, where the radiance N.1 is a function

of both position and direction, incident on the surface of an opaque body

where some of the radiation is absorbed and the rest is reflected (as

used here, "reflected" includes diffuse reflectance or scattering) to

iorm a second radiation field, where the radiance N of the reflected

r

radiation is also a function of position and direction. N r is directly

proportional to N. in the sense that, if the value of N. is multiplied by

a constant that is independent of position and direction, the resulting

values of N will all be multiplied by the same constant factor. It will

be seen below that the interdependence of the spatial and directional

distributions of N and N . is more complex.

Next, consider only the radiant power incident on a particular'

 1."
1) L-GC',

element 8A of a reflecting surface through an elementary beam of

solid angle 6 i from a direction (, 1 p0), where 61is the angle from

the normal to 8A and y i is the azimuth about that normal (see Figure

1). This incident radiant power is given by 1

6P ( t , y = N1 (0i T) cos 6i 6A 60n

= N1 (0t, oti) 61't 6A I ], (1)

where 61f'i = cos 86 60

= sin e cos eid 1dicp1

is the "projected solid angle" 1, 4 of the elementary beam. Corre-

spondingly, the irradiance at 6A is

6H11 (0, Pi ) = r" (oiYI) 6 'i [w. cm- 2 ]. (2)

Then the radiant intensity of the surface element 6A, due to reflection

(scattering) of radiatlon from this incident elementary beanm, in the

direction (e r, Cr) is

8J1 (6r Tcp) = p'(i, PVToi )cos 6 6Pi (61,cT) 1

[w.sr-a (3)
r r r ' 'rr r Ii

or, by dividing both sides of Equation (3) by IAcos8r, we obtain the

reflected (scattered) radiance

6N r (9r' 'Pr) = P'(il i. 0 r. 8Hr)

6H(8, ) w. cm-2 .r-] (4)

- 6-
 z

6q-Q

-- er

Figure 1 Geometry of incident and reflected elementary

beams. (Z-axis is chosen along the normal to the
surface element at 0.
 1
whre n (91¢I ' NrI(•r~.L.W) eNr(9Spt) sr (i

is the partial retfect,tice or "reflection- (distribution function' of the

surface element5A for radiation incident from the direction ( ni, o

and reflected (scattered in the direction (@r, Pr). Furthermore, by

6
a reciprocity theorem of wide generality ',7 first enunciated by Helmholtz,

wc may write

P, (e , O p, ) = P' (e C
2 '', ' ) [l ,. (6)

Thus p' (e1, R, z, cpo) is ordinarily the partial reflectance between the two

directions (9, pT ) and (8•', T) where either direction. may be that of

the incident elementary beam and the other that of the reflected (scattered)

elementary beam.

Hence, we can write the expression for the radiance at a point

of the reflecting surface (taken as the origin for spherical coordinates)

in the direction (Gr' Tr) due to reflection (scattering) of all beams of

incident radiation as

* A search for a proof (in English) of this important theorem also turned

up a number of authors who referred to,or made use of,the theorem in

various ways without giving a proof' 9, 10, ii, 12 Including Von Helmholtz

although Planck14 states, without specific citation, that Von Helmholtz

7
"proved" the theorem. DeHoop not only gives a proof (essentially the

same as that of Kerr 6).but also includes an explicit statement of the

requisite conditions.

-8-
 N r rr ' Ti, 9r' rP Nii
i1d'i, i)'h
i 1 1 d
0 0

J p'(ecp , yr) N, ( 9,,c) d o [w-cm ".sr-s l (7)

hr

where we adopt the following notation to designate integration over a

hemisphere:
Ff (6,c) dO fJ f Ce, ) sin de d0
"h o
STT
and f(e, ) do' = J f Me, cp) sin a cos 6 d 6 d cp.
h o o

This relation -- Equation (7) -- is for a particular point, or for the

surface element 6A at that point. For a more general expression, we

must also establish the reflected radiance from other points. When p'

and N are expressed as functions of spatial location (as well as direction)

for all points on the reflecting surface, Equation (7) gives the reflected

radiance N r as a function of position for these points on the reflecting

surface, as well as for direction ( r'1 r ) at each such point. However,

it is important in that case to recognize that Equation (7) is written above

in coordinates which, for convenience, are specially oriented with

respect to the surface element 8 A. Appropriate adjustments must be

made when dealing with Irregular surfaces where the direction of the

normal, with respect to fixed coordinates, changes in going from one

surface element to another.

-9-
 Whether su rfac e I rreg ula rities are treated as rll' ros copic,

(in the sense that their effects are integrated or averaged in the dis-

tribution function or partial reflectance C, 1::.9g., y2)), or as

macroscopic (in the sense that they may be analyzed into smaller

surface elements 6A for treatment as In the preceding paragraph) can

be arbitrary, depending on the degree of resolution desired, or can be

dependent on circumstances limiting achievable resolution. For example,

in examining the reflectance of a highly irregular surface containing

deep cavities, such as a piece of volcanic scoria, or a coarse, blackened

cellulose sponge in the laboratory, it may be possible to consider the

reflectance of different portions of the walls of single cavities (which

are ther regarded as macroscopic Irregularities). But when studying

the possible effects of similar surfaces which may exist on the moon,

where such flne detail cannot possibly be resolved by the best telescopes

on earth, these are necessarily treated as microscopic Irregularities. 15, 16

Still more complicated considerations are introduced when microscopic

irregularities are small enough to have dimensions of the order of, or

less than, the wavelength of the incident light or other electromagnetic

radiation. 17, 18, 19, 20

The total reflectance. p of a surface element 6A is defined in

general as

p - 8P r / PI [dimensionless], (8)

-10-
 bI)[ - ;2.€

whe re FPis the total ra diant power iticidt'elt ( tr [- all directions) on

A. and P £ the total re'iltittng vet'lected radiant power (in all
""

directionsi. As stated above, the value of p depends upon the geometry

and spectrum of the incident beam of radiation, which may be different

in each particular case. Here for the moment we are concerned

prrniarllv with the geometrical relations. Hence, for the remainder

of this paper, except where otherwise stated, we will eliminate spectral

considerations by restricting the spectrum of the incident radiation

to a region over which p does not change significantly with wave-

length. It is then useful to consider some special cases of incident-

beam geometrv.

If the incident radiation is well collimated, within a small

element of solid angle 6,2i = sin 8 deP dTi from the direction(O8 i,0 the

total radiant power incident on 6A is

8P, = 51HI (C i'IA [w] . (9)

Then. from Equation (5),

IN (v )' Q T"-, er Ic ) 5-H1 (8 (q)

r r r Ii r r i ii
SP ( 8a rt r 6PiF/6A [w. cm--.sr-1] (10)

- iil-
 1,:1) L ~-(G

But ,P = 5A $N (? , ) dO
V . r r r

= 5Pt' P (9t? Or' Qr d2'r

I di L J)
[w],

where d ( e) is the (total) directional reflectance, for a well-

collimated incident beam, given by

- Fh p. (9 1, Ti, Gr' (P r [dimensionless] (12)

ýdi (i)

For isotropic surfaces, there is no dependence on the azimuth CPand

Ec-uaton (12) simplifies to the frequently recognized dependence on 0:

Pdi (9i' 1) = Pdi (P). If the well-collimated beam is incident perpen-

dicularly on a plane surface, we have the commonly-reported normal

reflectance, pn = Pdi (0). If a point on the surface of a solid is uniformly

irradiated from all external directions, i. e., if N is a constant, the

reflected radiance in the direction (Or I ýPr), from Equation (7) is given by

0 0
Nr(0r'
( r) - Ni fh p, ( V
Di, r' Tr) dC I

[w. cm-2.sr1]; (13)

= Pdr (9r'pr) Ni

where Pdr (ýr'Tr) =- h P' (1 i' 0r'

rr) d' r [dimensionless] . (14)

But, from the reciprocity relation, Equation (6), and Equations (12)

and (14) we can write

Pdif 1t ' 1
G Pd, (
= Pdr ' D = Pd (01, P) [dimensionlessl . (15)

- 12 -
 Thus. the (tota I) dir ect iona I r eflec ta nc e Pd (0 '.) for a welll-colimated

beam iocrdent from the direction (01, •I) is also the ratio hetween the

reflected radiance N (@l, ý, ) in that same direction ant. the incident

radiance N. when the surface is uniformlv irradiated from all directions

(Ihemisrpherical irradiation). This relation -- Equations (13) and (15) --

8
is the basis for a reflectometry technique described by McNicholas.

4. DIRECTIONAL EMISSIVITY (AND ABSORPTANCE)

More important, Equations (11), (13) and (15) are the basis for

evaluating and equating the directional absorptance and directional

emissivity of the surface element 6A in a simple relation which has the

same form as the Kirchhoff's-Law relation -- see Equation (18) below.

If, in Equation (13), the uniform incident radiance N.i is equal to Nb (T),

the blackbody radiance (either total or spectral, i. e., in a small wave-

length interval at a given wavelength) in an isothermal enclosure at

T°K, and if, in fact, the reflecting surface forms the wall of such an

enclosure so that it, too, is at this same temperature, then the radiance

in the direction (8_, ,4 ) from the element of wall surfP :e 6A is made up

of an emitted radiance and a reflected radiance, as follows:

Ne + N = E d (891' 1 ) Nb (T) 4. Pdr (81 ,pl) Nb (T)

SNb (T) [w. cm-2, sr-] . (16a)

Similarly, of the rad.ance Nb (T) incident on the element 6A from a direction

(elI',1'I) , a portion N is absorbed and the remainder N. is reflected

(scattered) in all directions (into a hemisphere):

- 13 -
 121)1,-(2366

N' Nit d q' 06 Nb (T) 1 edi (01, ý I) NT (T)

Nb (T) Iw c -• - sr-' (16b)

Here, Zd ( i the directional emissivity (at temperature T) of the

element IA for radiEation emitted in the direction (01 (P, ) and od.Y( CO
,

is the absorptance (at T) for radiation incident from that direction.

Consequently, from Equation (15),

= - Odi (9',, 4) = ad (e8,1' [dimensionless]. (17)

Note that equilibrium maintenance with conservation of energy (Kirchhoff's

Law) by itself would justify only each line of Equation (17) independently,

and the Helmhotz Reciprocity Law (which is the basis for Equation (6)

and, in turn, Equation (15)) must also be invoked in order to equate

them to each other and so to relate emissivity for radiation emitted into

a given direction to the absorptance for radiation incident from that same

direction (See Appendix A concerning a contrary position.)

In the more familiar form of Kirchhoff's Law,

= 1I - P = C [ dimensionless], (18)

directional quantities are not considered. Instead, the total emissivity

for radiation emitted in all directions (into a hemisphere) is related to

the total reflectance (in all directions into a hemisphere) for uniform

incident radiance (from all directions, i. e. , from a hemisphere) and to

the total absorptance for uniform incident radiance (from all directions,

i.e. , from a hemisphere). The total reflectance a in Equation (18), for

uniform incident radiance (N. = a constant independent of direction) is then

- 14-
 Ai, Nr N. rhd(0, C) di

=P Qp = _________
rA h N 2T d2'

P
.h Pd , ,) d2' [dimensionless] (19)

The quantities in Equation (18) are those involved in heat-transfer

computations where the interest is in the not flow of energy across a

bounding surface, involving radiation received, emitted, or reflected

in all directions.

Equations (17) and (18) apply In all cases to spectral radiation

(I. e., the radiation in a very small wavelength Interval about a specified

wavelength) and hence also to any spectral interval In which p or Pd (and

therefore also C or d and o, or ord) do not change significantly with

wavelength. When thermal equilbrium exists (I.e.. when N = Nb (T) =

ýI
; Nxb(T, X) d X, where N1b (T, %) is the spectral radiance of a blackbody
0
a T K), they also apply to total radiation (all wavelengths), even though

the spectral reflectance varies with wavelength. However, if the spectral

reflectance is not a constant and the spectral distribution of the incident

radiation is arbitrary (non-equilibrium condition), Equations (17) and (18)

do not necessarily hold for the total (all wavelengths) reflectance,

absorptance, and emissivity.

-15-
 _____ %I P11. ES.

'--de :o ."t.- i v the ioregoing treatment of reflectance, it

:nav be helpful to apply it to some f" quently encountered situations.

in reflectance measurements. it is a common practice to make compar-

!Sons with standard surfaces which approximate the limiting cases of

perfectly diffuse reflectance (MgO is often used) and specular reflectance

(a highly polished mirror).

irst, a perfectlv diffuse reflector Is characterized bf a

constant value o: partial reflectance p' in all directions . If such a

-ur-ace s di..usely irradiated (N"I constant over a hemisphere) and the

.eflected radiation in a well-collimated beam in any particular direction

.s measured, or. in the reverse situation, if well-collimated incident

radiation is reflected into a hemispherical receiver (e. g.. an integrating

sohere). the ratio of- reflected power (flux) from a given surface area

to the incident power on that area is given in either case by the direc-

tional ref-lectance pd which, by Equation (12). is also a constant:

Pe = P'd' = [dimensionless]. (20)

h
Hence, from Equation (13), the total reflectance for any arbitrary

configuration of incident radiation is given by

-16-
 r L1,"N
d2'dA p ci
Pr - •1 -"P h NdI'd A

P N L,
d2'dA r NI d•'dA

= Pd -- fd•lcnsionlce] , (u1)

where the integration with respect to dA is carried out over the same

area in both numnerator and denominator, and N 1 may be any function

of direction and position. When the incident radiation is uniformly

distributed over the surface (even though it is not necessarily uniform

with respect to incident direction), the reflected radiance N r in any

direction is related to the irradiance HI by the partial reflectance p'

defined in Equation (5), as follows:

1
P' = Nr/HN = p/v [sr- ]. (22)

Second, a perfectly specular refler:r , .aracterized by the

relation

Nr(9, . " - Pd ( , p) N, (e,8 ) 1w. sr-] (23)

By comparing this with the general relationship between incident and

reflected radiances, it can be seen that Equation (23) will result if the

partial reflectance p' in Equation (7) has the form

Pt (91, (;, ar', r) = 2 Pd (ei'ct) 6(sins Or - sin'2 9)

Pi ±)
W(Tr- [sr-] (24)

where t (sIn 2 i -sin~g ) and ,((pr -CIi a) are Dirac delta-functions which

satisfy the defining relations

- 17-
 (it) = 0 for U t 0,
Siu) du = 1, ,rnd

f kit) iu) du = f (0),

when the integration is carried out over the full range of the variable,

0 < a -/• and 0 <p 2- , in each case.

Sometimes attempts are made to state apparently simple

relationships between the output of a given reflectometer for a diffuse

standard surface and a specular standard surface. This is not a simple

matter. It depends critically upon the configuration, and a wide variety

employed.
of configu rations are

As an illustration, assume that a sample surface is uniformly

irradiat'd by a well-collimated beam of uniform radiance Ni within a

small solid angle Li incident from the direction (p, p) . Assume also that

a detector is placed with appropriate optics (stops and, if necessary,

focussing elements) to insure that it receives radiation only from a well-

defined portion of the irradiated surface, of area A A, through a well-

defined solid angle, AZ'r < A•C, in the direction (0, cr± ,). First, if the

reflecting surface is perfectly diffusing, the reflected radiance is con-

stant in all directions and is related to the incident irradiance. H, = N A

cos 9,= "N . by Equations (5), and (21). The total power (flux)

received by the detector is then

Pd = cr 'r

9
-p/) Ni r cos r' do r' dA

= (P-) N1 A 1 A'r I?' AA (25)

- 18-
Note that if there is vitnettin,.z, - o that the solid aingle A 0 through which

the detector receives radiation is not exactly the sane for each point of

the surface A A. it may be difficult to evaluate the integrals.

Nex.•, if a specular standard surface is substituted for the

diffuse surface (and if it is carefully aligned to insure that the solid angle
I2s completely filled with reflected radiation), then, from Equation
r

(23), the total power (flux) received by the detector can be written as

P rj jNrd . dA

) N1 " A [w] . (26)

= Pd (' ,

If these were ideal standards, with reflectance values of unity in each

case (P = Pd = 1), the ratio of the detector outputs (proportional to

received power) for the two surfaces under the described conditions

would then be

P 7

=- = [ dimensionless] . (27)
Pd e n? I QLCos 0i

Itis obvious that this relation depends directly (inversely) on the solid-

angle spread of the incident collimated ieam. The dependence on the

other factors in the configuration -- the alignment, the solid angle of

acceptance of the detector, etc, -- is clear from the foregoing discussion

and specification of the conditions for which this relation was derived.

-19-
 ,ilv what nmiht be terned the evternal radionietric relations

have been considered in the foregoinE treatment and no attempt has

been made to deal with the deeper theory relating reflectance, emissiv-

ity. and absorptance to the optical constants of the materials, A good

summary of the most important aspects of that approach is given In

r. SUMMARY.

The partial reflectance of a surface element p' (6k,

@P1 ', 82' Va)

defined in Equation (5) as the ratio between the reflected radiance

in the direction (a=, a) and the incident irradiance from the direction

6. , a) •which produces it. Integration of this quantity over the solid

angle of a hemisphere in Equation(iZ) yields the directional reflectance

Pd ý I' I ), which is the fraction of the radiant power incident from the

dcrection ( ) that
t is reflected In all directions (into a hemisphere).

Furthermore, if the reciprocity theorem--Equation (6) -- is applicable,

as it ordinarily is, at least to a good approximation, then this direc-

tional reflectance is also the ratio between the radiance in the given

direction and the incident radiance when the surface element is

uniformly irradiated from all directions, as indicated in Equation (15).

-20 -
 "hoheeo ,i jiitv o-' III opaI.que sutrf1t-e eh, m:iIt In the dire't oii

, 1. A. t eC 1 t ot I10 def r e.C t i 0( J, I 0,V) IS S.hown

-u: 7quaticon 4 171. \hen the reocprocity theorent is applicable (as Is

usuallv the case), the emnis ivttv tn a given direction Is equal to the

absor'tivitv tfot radiation incident from that direction, which is also

ec.a 1 . to one nt an s the directional reflectance for that same direction.

A perfectly diffuse reflector is characterized by uniform

re, ectance in all directions. It is shown In Equation (21) that this

;5 ea.u al to pi times the partial reflectance. The relationship

retween the partial and directional reflectances for a perfectly

specular reflector involves Dirac delta-functions, as given in Equation

The completely general expression, relating the reflected

radiance of a surface element in a given direction Nr (9r, yr) to the

incident field of radiation, specified by expressing the incident

radiance as a function of direction N1 (9l ) is given in terms of the

partial reflectance p'(81 , PC ep)by Equation (7). This holds true

regardless of the geometrical configuration of the Incident beam. The

radiant power in a reflected beam Is then computed by integrating

the resulting value of reflected radiance, as a function of direction,

over the appropriate projected area and solid angle as indicated in

-21 -

!
 E•1)1,- G..,t

,-e first line of Equation (25) and discussed in greater detail in

Reference 1.

As with many idealized physical quantities. partial and

d.rectional reflectances can never be measured exactly. even with

perfect instrumentation. Since a measurement reqLuires a beam

o. radiation of non-zero cross section and solid angle, the measure-

ment at best can only yield average values over these intervals of

projected area and solid angle. However, the concepts, terminology,

and symbols presented here make it possible to specify explicitly

and unambiguously the interrelationships and approximations

involved in dealing with real situations. Also, the application of

Kirchhoff's Law -- Equation (18) -- to the directional quantities can

be stated explicitly, as in Equation (17).

- 22 -
7 APPENDIX A -- Reciprocitv in an Isothermal Enclosure

A paper by Bauer presents an alleged proof of the Helnrihoh.o

rec~proc.:.- law for diffuse reflection as a consequence of equilibrium

condctions '-- an -sothermal enclosure. The argument hinges on the

statement that the second law of thermodynamics requires that there

be no net exchange of ener-" by radiation between any two individual

elements of the itern-al (opaque) surface of an isothermal enclosure.

H!Zowever. it seems to me :hat the requirements of the second law apply

to the total flow of energy, taking in:.' account radiation emitted into,

and absorbed from, all directions (full hemisphere) by such a surface

element, as the basis for Kirc!-hoff'- Law( E:*quation 18).

-23-
 EDI,-G.6

S REFERENCES.
R

Fr Ed Nicodr::. "Radi-:ace." A:n J Phys.. v 3 1. n . p {&S;

May -3 (UNCL-ASSIFIED .oublicai.on)

2. G. Kehon. e: a!..
"Infrared Target and iBackground Radionetric
'ejs-:rcen:ts % 3. n 3. p 13): Sept. 1-63;
" infrared Physics.
A!.-:,, by~-' U. Mi..-. under ONR Contract NOnr-
.224 (12) as Report No. 2S'-'4-T. AD 275-310; January 1962.
(RNCLASS=.T.--D . ..- ti-.

.--h" E. el. "R-ciz--:rc Quan::.es. Syv-bols, and Units,

Pro: RE. v - - 1432: September 1050. (UNCLASSIFIED
u":a:-on)

4 R Clark -Tones. "'I--ersed Radiation Detectors, " Appi Opt,

•". n :;. o 007: Sep:c.ber 1o62. (UNCL.SSIFIED publication)

KD.
. :dw- -. .. T. Gier. K. E. Nelson and R. D. Roddick
"'ntegrating Sphere f-.r T-perfectly Diffuse Samples, " J Opt Soc
-................ 270; November 1,061. (UNCLASSIFIED

oDna1.d Kerr. "ADpica:on ot the Lorentz Reciprocity Theorem

"S.... ....z.. .Ao:)endix A (p. 693) of "Propogation of Short Radio
Wa--_" Vol. 13 ' t'he M. . T.. Radiation Laboratory Series,
- -- " Ed.. .:Graw-:-!!. New York; 1951. (UNCLASSIFIED
p_ ..ic ..i n)

A DeHz=. "- Reciorozi:v TheoreD :or ,e Electromagnetic

ield S:at:ered an Obstacle, 'App! Sci Res. Sec. B, v S. n 2,
a'-
5 1.
-, ;-0. (UNC L-LSS=-=-• TD m_0,.,-=,.o..).
' :-;

. ....
N:hola. Ab.oute tethods in Reflectometry, " J Res
-7 Stand . Zý, (RP-3); 192•. (Submitt~d to Johns
o-n-. n:-.&er .- S Ph. D. Diisertation. 1928). (UTNC LASSIFIED

C. DeVos, "E----.
-.. n o- the Quality of a Blackbody, " Physica,
20. ., 194 (UFCL-SS./IED
S publication)

i.0 Max B-rn and E.mil 1.. ol• , "Principles of Optics. " Pergammon
Pres-t . New *-r*..*-', ) 380; .959. (UNCl.ASSIFIED publication)

i i. C n.--.r.ttee :n Colori.etr,, Optical Society oi America, "The

:ierze cClohr." Thoma.s Y. Crowell Company, New York,
P 7E; 5 (UNCLASSIFIED p•blication)

- 24 -
 :2 T ~~de Ia Perreie. T. S. Moss~. and H. HeItrbert. T1
.%lasurenents o: Absorpt-vity and Rczfle, t~v-tv." Infrare~d PhysiCs.
v 3..n 1. o35. Jan,:a-VNlarC- 1903. (UNCLASSIFIED publication)

13.Vo \ W.
H-n-olrz. "Helm;holtzs Treatise on Physxological
Optics." trans. fronm T-- rd Gernian Ed~tton. J. P. C Southall
(ed. ). Optical Soc~ety of .cernca. v 1. p 2131; 1924. (UNCLASSIFIED
~aon)
14. M. Planck. "1n trocuct--on to Theareti cal Phys ic s."1 t rans. by
HL. 3 rose. Macn-illan Company. New York. v 5. "Theory
o: Heat." 1193; 1957. (UN.CLASSiFIED publication)

is. Bruce W. Hap'-e and Hu!;; Va- Horn. 'Photomnetric Studies ,.f
Cornolex Sur.`aces. with Anolicat-.ons to the Moon, " J Geophys
Res. v oS. n 15. p 4545; 1 Aug.ust 1963. (UNCLASSIFIED
po.bzica~ion)

10. 3-.:ce W. :4anke. "A Theoretical Photometric Function for the

Lunar Surtace." Geonhvs Res. v oS. n 15, p 457l1. Au..ust
l9a3. tUNC LASS. -E!'D publ1-cation)

17. iC-or Twerskv. "On. Scattering and Reflection of Electromagnetic

WNaves by Rouz' S.rfacei. " I'R-- Trans on Antennas and Propagation,
nA-S
n. -3 S1; January 1937. (UNCLAkSSIFIED publication)

Is. %ic-:o. Twers'ky. "On -MultipleScattering of Waves," J Res

'\a:! 3-.r Stand. v 6-*D. ~ 15; 1960. (UNCLASSIFIED publication)

Victor wes-."Mulzinle Scattering o: Waves and Optical

Phenomena." J7On: Soc Amner, v 32. n 2, o I'5; February 1962.
tUNCLASS:Fe--._zD publi-cation)
20. :z -r ennet-, and 0 Porteus. "Relation Between Surface
0.
Rou hness and Spec-ar Re:-lectance at Normal Incidence. " J
Opt Soc Arner- 51, n 2. p 123; February 1961. (UNCLASSIFIED
publ-.cat.ion)

2> -chard K McDonald, "1Techn;iques for Measuring Emnissivities."1

Procc:i. . n 3. p 153; July 1960. (UNCLASSIFIED article
 !.:0t' ,-G' 2.(~'0

in SECRET V\0''C . t, iI'shci ,t1d dtst rttibtCd by Oft icc ot Naval

Re srch.', 4t15 Suý.11ninir Strcet. Bostotn)

21 en:r:v H Bl'au. )r - John L Miles and Lelhnd E. Ashmnan.

"R d'.+t'n Characteristics- of Solid M1atcrials - A Review,'
S,":e:'n"c Reusort No. 1. AFORCO-TN- ;8- 132. Contract No. AF-19
c04e1 -13O. AD 1MO-S$3. Arthur D. Little, Inc. ; 31 March 19i8.
KUNCOLASSIFIED publication)

27 0. Da'.cer Reflectiou Nieasurenient' on Open Cavities" Optik,

"vIl IS, no 12 P V0 ,
3
De"cnt, 'r 1061.

6 -- (tin + Ž;6+)
 ,.-

4 tz ' - ,C*+~ ,'4

i

:- 'z •- f .• •• - .

_ . • • •4' -. - . • - • .. , ', <

 V

(4

- "C.--

I .N'4C

<C-'- z,- 4 -

- 4- -4'-- - I '>

cc, .a-& > ,EZ

4 >- 4 4 0. o-
I <
N
4.
, 4 :g'
...
V>C

-'C.'
a I 1>-. - E-' '
04

'C'-
4 1 <> 4- 0

-1- ----- 1
z

4]

- 4N
>> 0
44 0
4 0

4 I *-'' **0 ',.

4'

z'
4
Z I -4:
aNCE
o-
E

E
4 .D -
040'

'-ye
04 - .. 0.44
4.344 0.
4 > S 0
7* Cr U2 0
5r-
I.---- 4. 4,
>4.4:44

----- I-

Ii