Airborne Measurements of Forest and Agricultural Land Surface ...

.

Airborne

Measurements

of Forest

Emissivity

at

Y70, Tel:+

DERA

Met.

Fax:+

(Remote

Sensing)

44-1252-515523

GU14

Uppsala surements

in

radiometers have been same (N OPEX) area of

September have allowed

1995 the

and March calculation

of

operated Sweden,

1907. the

on near

Their emissivity

meaof

the ment

land, where it is approximately 0.96. These differences examined and a model is presented to parameterise surfaces by use of a Debye-like effective permittivity

cannot

reflection

coefficients.

such

The

launch

crowave

This

range

measurements

provide

between

small-scale,

sat ellite

images.

This

paper

made

by

This

forest

open

scattering

term.

surface

three

flight

land,

and

18 March

heavy

made rain.

paper

of

near

which

considers

in

its high

frequency.

way,

flights

colder

the

r;

the

gap

terms

of e~, the

effective

limit,

v,,

The

ionic

as its contribution

and

conductivity

in

is insignificant

vegetation’s

wetland

and

ume

vegetated

in

on

took

place

a warm on 14

and

volume

this

extreme

the

observed

as a

in terms

scattering

quency

angular

in the variation

parameter, surface

This

infrared

This

paper

main

is para-

radiometers,

has been

Airborne

neglected,

cludes

various

20GHz.

and

of two

instruments

the

in the

proposed of the

of its

[3], which

vol-

are fre-

absorbed

into

polarisation model.

to In

by the

ratio

the

mixing

The

vegetation,

measurements on the

in

this

effective

as measured

Scanning both

views

onboard

with

black

by the

Hercules.

are

and

MARSS

the

UK

MeThe

microwave (Microwave

System).

These

are

a 3 second

scan,

which

downward

are described

made C-130

study

as Deimos

Radiometer radiometers

the

is also

setting

is that

used

known

power

due

canopy.

is determined

albedo,

(UKMO)

instruments

relaxation

the

INSTRUMENTATION

Office

of effec-

or dry)

radiometry.

presents

teorological

within

process

short

wavelengths,

coefficients

and

Q = 0.5,

sparse,

emissivity

to extinction

temperature

very

scattering

permittivities

form

permittivity,

Qr”

+

miUimetre

scattering

dependent.

v, which static

Q)I’h

when

at

effective

single

by thermal

system

(1 -

(except thick

III.

above

the

Q, was

is included

= (I – qru + Qrh

optically

absorption

bo-

the effective

term

mixing,

[2] and

(2)

vegetation

appears

15%

can be represented

frequency

into

to represent

OR

Most

NOPEX

1995

on a Debye-like

of variable

absorbed

be tuned

for polarisation

=

rj

conditions.

y, c at frequency

are

of emissivity:

about

surface/canopy

based

may

Choudhury

to explain

MODEL

the surface

permittivit

meterised

model

of dense

with was

6 September

effects

cover

[1], of a

0-50°

Uppsala

areas pine

and

large-scale,

the

as lakes,

further

EMISSIVITY

reflector,

relative

on

1997 in much

In this

of a specular

cm,

as well

some

Two

II.

tive

and

by Wang

effective

was

after

This

extensive

the

snow

addition

variables. parameter

Airborne

measurements

over

area,

by spruce trees,

and

Office

contains

and

to span

of airborne

of these

which

the

as

exhibiting

as dry

without that

such

scattering,

surfaces

such

vegetation

parameters,

additional

added

presented and

vOIUme

However, spectra,

It is proposed

roughness

surfaces,

and

EW,

Em.

comple-

of the sur-

GHz

link

as




emissivity

y, c, and

(e) is the

6)].

represent E,

permittivit

emissivity

1 – I’(v,

setting

as sea ice by

relative

d. The

can

by

non-monotonic

An

generation

provides

microwave

incidence

next

temperature

as sea.

face’s

the

instruments

atmospheric well

of

of its complex

a scheme

changes INTRODUCTION

meto.gov.uk

[c(v, 0) =

water,

these I.

UK.

of incidence,

of this

Such open

Fresnel’s

in terms

angle

boreal forest and agricultural land surfaces at 24, 50, 89 and 157 GHz over a range of incidence angles and polarisations. These results show consistent differences between dense forestry, where the emissivity is close to 1 and open are these and

OLX,

Email:tjhewison@

surface microwave over the

Surface

Wavelengths

Hampshire,

Abstract—

Passive an aircraft

Land

J. Hewison

Office

Farnborough,

44-1252-395781

Agricultural

Millimetre Tlm

UK

and

(and

body

upward

for

calibration

in full

in

targets.

total in-

MARSS) These

[4], [5].

(1) IV.

The tical

fresnel and

formulae,

horizontal

define

polarisations,

the power rti

and

reflectivity

in ver-

I’~ of a specular

Measured surface

EMISSIVITY

brightness

emissivity

0-7803-4403-0/98/$10.00 (c) 1998 IEEE

to

CALCULATION

temperatures extend

them

must to

be converted

general

application.

to

2

The

following

using

only

formula

aircraft

is used

to

calculate

the

TABLE

emissivity COEFFICIENTS

data:

OF

Category

Z(V, q – C(v,(9)=

~z(~,

where

e is the emissivity

O, T.

and

TZ are the

peratures,

The

There brightness

craft’s

assumed of

T5,

(~

10mm)

at

the

surface

temperature

satellite

the

mil.limetre

the

aircraft.

The

conditions

by

radiative

wavelengths. available

the

airThis

at

era

from

was

neous

studied

and

to

very

This

is typical

of

schemes

for

ples

and

Other

Forest

1.57

1.22

grease should

low

bias

10–50

than

seconds,

Both each

of its

five

Ice

decrease

with

“Open”

to track

which

and

horizontal

to

canopy

142 sam-

ranging

lengths two

views,

vertical

to

flights,

periods

astype

refers

refers

three

measure

downward

was

vegetation

scan

channels

angle.

sation

The

rotates

—40°.

tates,

but

from

of l–5km.

rotate

also

the

fact

curves

scan

Fig.

to

effective

11, shown

tal polarisations. the average

of these mixing

The

Lake

calm

following

I,amkaouchi

the

et al.

mean

lines

for

lines

polarisations, parameter,

the [6],

and

view,

view

roand

vertical

calculated

the

Soil

standard together

erwith given

shows

larisation,

Other

Forest

and

rocky

especially

to the

emissivities

was measured in this

is characteristic surface,

of

though

this

here,

as their

nadir

be differentiated.

This

group

to bare

soil,

and actu-

contrast.

observed

optical

in

March

frequencies, depth

all remaining winter.

was somewhat

at the

fit

1997 with

of the

and

no po-

canopy

and

categories

ob-

albedo.

and

outcrops,

polarmixing

footage.

y at all

combined

group

of this

indicate

on these

horizontal

was even greater

curves

was

scattering

in summer

horizonthe

single

use in op-

average

polarisation

emissivit

served

and

more

Forest

high

lower

combined

not

due to the large

its low

graphs

by setting

Conifer

very

inthese

for

temperature

video

emissivity

to have

which

curves

a reasonable

at 24GHz

been

could

similar

po-

roughness

a polarisation

of ice on the

on the

spectra

and

the variance

have

be

cannot

reproduce

dashed

provides

shows

behaviour

evident

little

model,

could

with

the polarisation

Crops/Grass

very

expensive

the surface

C. However, and

to

157 GHz.

surprisingly where

be attribut-

behaviour

as the

of vertical

This at

the

apparent

thick

scattering

The

be obtained 0.5.

may

Such

computationally

except

though The

emissivity,

Bragg

an average

Q =

very

of the

slab.

observed.

a simple

is too

appears

curves

reflection

sufficiently

has a high

was

as would

was not

ally

dielectric

sat ellit e ret rievals.

samples

shows

not

model

specular

re-

of the

in these

to

at 24 GHz

was

optics

Frozen

so these

validation

lower,

resulting

lower

The in

forest average

canopy

exposing

some

a decrease

in

density bare

soil

emissivity,

frequencies.

Q = 0.50.

behaves

as a specular

double-Debye shown

resolv-

coefficients

on these

0.50

behaviour,

A geometric

emissivity

for each of the samples

in each

dashed

Watep

also

radiometrically

permittivity

as solid The

polarisation

closely

1 shows

87.3

is insignificant.

to unity

but

Close

into

0.42 \ 0.50

evident

tends

parameter,

polari-

to vertical

channel

in the nadir

calculated

measured

the

view

with

extremes.

grouped

emissivity

fitted

Table

were

in each

so the

+20°

89 GHz

138

1997,

due

specular

it

Soil

by

represent

than

bias

emissivity

erational isations,

March

100),

emissivity

that

Bare

to be >0°

forward

rotates

is aligned

horizontal

scan

video,

categories.

lines

results,

the

angles

contrast the

in

(
0.70

?:88

0.70

.
.: 0.60

0.95

0.60

Other

1.00

Forest

1,00

0.98 0.96

0.90

2 0 ~

Conifer

i ,00

0.95

0

Frozen

Soil

1.00

0.94 0.92

n

, ?:88

0.90

0 ; 0.50 0 m m 0.40

0.50 0.85

0.40

0.92 R

$28 0.70 % .
.: 0.60

0.95

‘?:88

?38

0.98

0.98

0.96

0.96 0.94

\

0.90 0,50

0.50

0

0.40

~

?:38 0.98

0.60

al :

188

0.40

0.94

0,94

0.94

0.92

0.92

0.92

v

0.85

0.92

I

888 0.70 > .-

838

%88

0.70

.? .; 0.60

0,95

0.60

‘?88

?:88

?38

0.98

0.98

0.98

0.98

0.98

0.96

0.96

0.96

0.96

0.96

0.94

0.94

0.94

0.92

0.92

0.90

a90L-----J o

—

0.50

0.94 0.85

0.40

0.30

0.30 50

a Incidence

0.80 0

Angle

50

Incidence

Fig.

radiometers.

They

parametric these

!38

0.90

a ; 0.50 0 * ~ 0.40

gles

?:88 ““

have

model

surfaces needed

the

the

range

50

o Incidence

development

y. A:

of a fast,

and

view

of satellite

\

❑ MB

50

a

Angle

the emissivity

of frequencies

generation

0.90

of Emissivit

of representing

next

0.92

0.90 Angle

Variation

the

0.94

0.92

5a

Incidence

Angular

allowed

capable

over for

1.

0

Angle

m

Incidence

Vertical,

edges

0

Angle

5

Incidence

O: Mixed

the

of the

support

of

Air

Force,

an-

ical

Research

the scientists

model

presented

is based

of effective

tions

can predict

from

20-200

the

GHz. by

high

to represent land

cause

variation With

and

with

extensive

ground

truth to

Debye

Fresnel’s

of emissivity

a more

density

polarisation the

observed

surfaces.

It

dependent, of

equa-

frequency

dataset, data,

it

physical

[2]

may

be

the

mixing

86,

is likely

limitations

of

is necessary

in polarisation

that

this

air

crew

of the

of the

Royal

Meteorolog-

cannot

the

instruments

is fre-

be confirmed used

Field

[4]

T. Hewison, with

1994,

[5]

S. J. English,

Hz

Frequency,”

J.

Geophys.

M.Owe, “Microwave Vegetation Modelling of Soil Emissivity Observations,” MeteoroLAtmos.Phys.,

Design

C. Guillou, of

Journal

this

1.4G

of

Deimos:

A

Microwave

at 23.8GHz and 50.3GHz for C-130 Aircraft,” Pro. IGARSS,

Flight

surements

at

OpUsing

pp.225-239. “The

Channels

search

bein

Bare

pp.5277-5282.

A. A.van de Griend and tical Depth and Inverse Nimbus/SMMR Satelfite

contrast

parameter

this

over 1981,

[3]

variables,

parameter

decrease

though

of

C. Prigent

water the

vapour

Royal

and

UK

1995,

D. C. Jones,

continuum

Radiometer

the

Met.

“Aircraft

absorption,”

Meteorological

Society,

Re-

pp.2261-3. mea-

Quarterly VO1.120,

1994,

pp.603-625. [6]

VII.

l?rocesses

and

technicians

Flight.

Content

Res.,

K. Lamkaouchl, data

framework

ground

J. Hewison, S. J. English, “Analysis of a Microwave Airborne Campaigm over Snow and Ice (MACSI)”, Proceedings of EMAC94/95 Fhal Results workshop WPP-136, ESTEC, 1997. J.R. Wang and B. J. Choudhuxy, “Remote Sensing of Soil Moisture

com-

experiment.

This

Angle

[1] T.

VO1.54,

An additional

quency

Incidence

sounding

on a 3 parameter

permittivity,

possible to relate these parameters ~iuch as biomass or sod moisture.

over

50 --

a

Anqle

REFERENCES

formulation

plemented

0.90 50

Incidence

Polarisation

and

instruments. The

Angle

❑ : Horizontal,

0.92

n

work of

has

NOPEX

land-surface

been

carried

- a NOrthern Experiment.

[7]

out

within

hemisphere The

author

the

climate

A. Balana

Sea

TEC/ESA

ACKNOWLEDGMENTS partly

for

Water Contract

W. J. Ellkon, C. Guillou,

No

A. Balana, C. Prigent,

tric

properties

No

11 197/94/NL/CN,

acknowl-

0-7803-4403-0/98/$10.00 (c) 1998 IEEE

of

and

W. J. Ellkon,

(3o-1OOGHZ),”

sea

Report

“New on

11 197/94/NL/CN, G. Delbos, “Study

water,” 1996

and

permittivity

extension

K. Larnkaouchi, measurement

ES TEC/ESA

to

ES-

1996

Report

L. Eymard, of

the on

dielecContmct