Radio Propagation and Remote Sensing of the Environment - Chapter 11 ppsx

11.1 SOME CHARACTERISTICS OF ANTENNA SYSTEMSAll radio devices for remote sensing are equipped with antenna systems for thereception of radiowaves and for their radiation, in the case of

11.1 SOME CHARACTERISTICS OF ANTENNA SYSTEMS

All radio devices for remote sensing are equipped with antenna systems for thereception of radiowaves (and for their radiation, in the case of active instruments).This is the reason why we will consider again the properties of an antenna We havedirectivity and pattern The definition of directivity was based onthe angular distribution of the radiated power relative to the power at the antennainput The energy losses (e.g., taking place in the feeding systems) were not takeninto consideration; therefore, we need to include in our analysis the antenna effi-ciency:

which reflects the ratio of radiated power P i to input power P t So, the gain,

, often figures into engineering computations

The effective area of the antenna was used to characterize the receiving properties

of the antenna Normally, this notion relates to the geometrical antenna area via theaperture efficiency :

G( , )θ ϕ =η θ ϕD( , )

ηa e

A A

= maxTF1710_book.fm Page 303 Thursday, September 30, 2004 1:43 PM

established already in the Chapter 1 that transmitting antenna are characterized by

304 Radio Propagation and Remote Sensing of the Environment

Here, the maximum effective area occurs in the numerator Equation (1.114) helps

us to define the aperture efficiency as:

The side lobes of both kinds of antennae (transmitting and receiving) are ofinterest, as very often they are sources of interference One of the main reasons forthe high side lobe level is the sharp change of apparatus function at the edge of theantenna; therefore, uniform apparatus function leads to a high level of the side lobes.The first side lobe of the circular antenna, in this case, is 17.6 dB less compared tothe major one; therefore, in order to lower this level, we want the apparatus function

at the edge of the antenna to tend to zero As a modeling example, we can choosethe apparatus function for the circular antenna in the view Theantenna pattern will be:

and the first side lobe level will be 24.6 dB;4 however, in this case, the beamwidth

is whereas it is for the uniform apparatus function The utilizationfactor is 0.75.4 The stray factor:

(11.5)

is applied for the side lobe role general assessment Here, Ω0 is the solid angle ofthe major lobe estimated by the first zero line of the antenna pattern In the case ofthe patterns represented by Equation (1.118), it is reduced to , where

is the second root of the first-order Bessel function, and we have the

ηa A

2 2

2 2

rr

F( )r = −1 (r a)2

θ

( )= ( ) ( )

4 22

4

J ka ka

sinsin

0 64 λ a 0 51 λ a

β

θ ϕ

θ ϕπ

ΩΩΩ

β = J j0 2

1 2( , )

j1 2,

TF1710_book.fm Page 304 Thursday, September 30, 2004 1:43 PM

Radio Devices for Remote Sensing 305

corresponding quantity This value is defined by a similar formula as inEquation (11.4), except that in this case the root of the second-order Bessel functionwill be the zero-order function argument j2,2 = 5.136 and β= 0.017

The level of the side lobes, as a rule, is not very important for radar Moreimportant is the signal-to-noise ratio; therefore, the standard approach is to havemaximal aperture efficiency and, correspondingly, maximum directivity For radi-ometry, however, a low level of side lobes is one of the most important parameters

So, we may have to sacrifice sensitivity due to some deterioration of the spatialresolution

11.2 APPLICATION OF RADAR DEVICES FOR

The main purposes of such research are analysis of radiowave intensity andpolarization, the study of the shape and time delay of signals, etc These data areused for radar mapping, altimetry, scatterometric research, and subsurface sounding.The main areas of application are geology (e.g., structure and lithology), hydrology(e.g., determining soil moisture; mapping river basins, inner reservoirs, floods, snow-cover), farming (e.g., mapping crops, monitoring vegetation growth and maturation,determining field borders, observing soil moisture dynamics), forestry (e.g., estimat-ing merchantable wood, monitoring deforestation, detecting forest fires, assessingfire hazard situations), oceanology (e.g., sea-wave zone monitoring, measuring near-water wind velocity, determining the direction in which a pollution area is moving,studying sea currents and geoid forms), atmospheric research (e.g., sounding cloudsand rain areas, determining temperature height profiles, defining atmospheric turbu-lence, detecting minor gaseous constituents), study of ionospheric dynamics (e.g.,altitude profiles of electron concentrations, turbulence motion, electron and iontemperature measurement), cartography (e.g., topographical survey of regions thatare difficult to access, land use mapping), polar area observations (e.g., monitoringand investigating sea ice, mapping continental ice covers, observing iceberg forma-tion and their motion, researching changes in glaciers)

Generally, four types of radar devices are used for remote sensing: panoramicradars, scatterometers, radio altimeters, and subsurface radars Panoramic radarinvolves sectoring or a circular scanning view and side-looking radar; the latter typeincludes radar with a real aperture and synthetic-aperture radars (SAR) Scatterom-eters are classified according to the methods for selecting a given area The mainapplications are angle–angle (the antenna pattern is narrow on both planes),angle–time (the pattern is narrow only on the azimuthal plane, thus the signal is

β = 0 16.TF1710_book.fm Page 305 Thursday, September 30, 2004 1:43 PM

306 Radio Propagation and Remote Sensing of the Environment

modulated by narrow pulses), angle–frequency (the pattern is narrow on the muthal plane and is directed along the platform flight line; signal discrimination isachieved by Doppler selection), frequency–time (SAR with pulse modulation).Altimeters are classified by the modulation method: pulse, frequency, or phasemodulation Subsurface radars differ in where the antenna is mounted: remote orapplied (the antenna is put on the surface of the studied object)

azi-11.3 RADIO ALTIMETERS

The radio altimeter is in theory one of the simplest radar systems Applications ofremote sensing using radio altimeters include measuring water levels, defining geoidforms, estimating seawave intensity, monitoring glacier growth, mapping land sur-face topographic, determining upper cloud levels, etc Pulse-modulated or chirpsignals are normally used for altimetric research For the traveling time of the signal(tr–t) over the path, transmitter–surface–receiver serves as a device for measuring thealtitude of the altimeter platform above the studied surface The common relation is:

where v is the mean velocity of radiowave propagation in the atmosphere of Earth.The frequency spectrum of altimetric systems is chosen in such a way that theinfluence of the ionosphere is small Usually, the C, X, and Ku-bands are preferablefor these devices

The error in the altitude determination is:

where δtr–t and δv are errors in the delay time and mean velocity, respectively Thefirst summand contribution is defined by the leading edge steepness of the optimallyprocessed pulse and the signal-to-noise ratio at the altimeter output Generally,

where τi is the leading edge duration after reflection (scattering), and S/N is thesignal-to-noise ratio.89 Changes in the leading edge as a result of scattering by aThe second term in Equation (11.7) depends on the propagation conditions Inthe case of spaceborne altimeters, the contributions of both the troposphere andionosphere must be taken into account The tropospheric contribution, corresponding

to Equation (4.57), can be written as:

H=t vr-t2

δH= vδt− +t− δv

2 r t

r t2

rough surface are described in Chapter 6

Radio Devices for Remote Sensing 307

(11.9)

Equation (4.69) permits us to estimate the ionospheric contribution as:

For the X-band, it will be about 10 cm For greater accuracy when defining altitude,

it is necessary to develop procedures for correcting the radio propagation In thecase of the troposphere, atmospheric models are used Microwave radiometers areadded to measure water vapor content, which gives us a more accurate calculation

of the mean value of the refractive index Correcting for ionospheric influencesrequires the design of a two-frequency altimeter or the use of very high frequencies,which generates new problems (scattering by clouds, for example) In order toovercome these, continuous analysis of signals is needed using, for example, astrobing technique

The accuracy of the altimeter also depends on the antenna beamwidth and itspositional stability These effects become most apparent in the case of strong reflec-tion from the surface edge coinciding with the footprint board As was noted insignal.89 Examples of backscattered signals are given in Figure 11.1 This is onereason why estimating the time delay using the arrival of the leading edge of thepulse gives us a more accurate altitude measurement Simultaneously, pulse distor-tion analysis allows the study of the roughness parameters In particular, this tech-nique is used for sea waves intensity monitoring

FIGURE 11.1 The approached pulse shape of normalized altimeter impulses: (1) salt desert; (2) and (3), agricultural holdings; (4) surface of lake.

r t2

− δ =1 21 10. ⋅2 21

0.4

2 4

1 3 4

0.2

0

t

0.8 1 u

TF1710_book.fm Page 307 Thursday, September 30, 2004 1:43 PM

Chapter 6, surface roughness leads to a change in the duration of the scattered

308 Radio Propagation and Remote Sensing of the Environment

The first spaceborne altimeter was tested during the SKYLAB mission It ated at a frequency of 13.9 GHz, and the pulse duration was of the order of 25 nsec.The measurements taken showed that we can determine the topographic relief underthe satellite trace with an accuracy of ±15 meters for the case of a relatively smoothsurface without mountains and steep slopes The altitude above a quiet water surfacecould be determined with an accuracy of about ±1 meter.89

oper-The investigation of ocean problems requires a more accurate altimeter Researchhas shown that errors of altitude determination must be less than 10 cm Very shortpulses are needed in this case, which explains why chirp modulation is useful inmodern altimeters In particular, pulses with a duration of the order of 100 µsec and

a frequency deviation inside ∆F = 320 MHz are used in modern altimeters Thecompression procedure provides the altitude definition with an accuracy that issimilar to pulses of 3.1-nsec duration

A spaceborne altimeter of very high accuracy is applied in the U.S.–Frenchmission TOPEX/POSEIDON, which is designed to research water surface charac-teristics.145 The C- and Ku-bands (frequencies of 5.3 and 13.6 GHz, respectively)are used The two-frequency system allows us to correct the ionospheric influences.Microwave radiometers, operating at frequencies of 18, 21, and 37 GHz, providethe tropospheric correction A good system of 1.5-m antenna orientation minimizeserrors connected with main lobe deviation from the nadir There is a system based

on the ground for the satellite coordinates determination with heightened accuracy

A complicated procedure of corrections has been developed for altimeter dataprocessing and interpretation Depending on the seawave intensity and the dataaveraging time, the accuracy of the mean level of the oceanic surface definition lieswithin a range of several centimeters

11.4 RADAR SYSTEMS FOR REMOTE SENSING OF THE

ENVIRONMENT

A different type of panoramic radar is commonly used for remote sensing of theenvironment These radar systems differ in their principles of operation and in thetechnique of observation The received pulses are distinguished by arrival time,which allows determination of the distance to the illuminated object (in the case ofmonostatic radar), and amplitude, both of which allow us to define the scatteringproperties (reflectivity) of this object In this process, the direction of the antennamajor lobe is fixed, and the direction of the investigated target is determined Thescattering properties of targets can depend on polarization of illuminated waves andscatter the waves with a polarization different from the polarization of incidentwaves So, the polarization matrix or Stokes matrix are the subject of interest inmore complicated systems These data open the way for mapping the scatteringproperties of the environment

The radar systems applied to remote sensing are separated roughly into twoclasses: radar for atmospheric research (troposphere and ionosphere) and radar formonitoring the surface of Earth (land and water) Ionospheric radar will be discussedTF1710_book.fm Page 308 Thursday, September 30, 2004 1:43 PM

the targets themselves, as we saw in Chapter 5 (Section 5.6) The targets usually

Radio Devices for Remote Sensing 309

which is used primarily for the investigation of hydrometeors Scanning space ismade possible with the help of a pencil-beam antenna The main parameter ofmeasurement is the reflectivity, which is determined via the intensity of the receivedpulses The power of a signal received from a point scatterer is defined by the radarequation:

Here, W is the received power; P is the pulse transmitted power; D and Ae are theradar antenna directivity and its effective area, respectively; σ is the target differentialcross section of backscattering; L is the distance between the radar and the target;and λ is the radiowavelength In contrast to the traditionally applied formulae, weuse the physical definition of the cross section, which is different from the radar one

by the factor 4π

different engineering coefficients describing the feeder system efficiency We alsoassume the absence of wave attenuation in the environment In the future, we willgenerally restrict ourselves to the case of clouds as the subject of the study Theyare distributed targets; therefore, the power scattered by a cloud layer of thickness

l can be defined as:

, (11.12)

where is the differential cross section of the backscattering per unit volume

of the cloud The coordinate integration in Equation (11.12) is realized over theplane transverse to the wave propagation direction This integration is easily trans-formed to integration over the solid angle To calculate this integral, we can use themodel antenna formula (Equation (1.123)) to obtain:

( )= eσ = eπ

σλ

P L l L

( )= σ π( , ) ∫ ( ) = ( , )

λ

σ πd

2 π

W L P L l A

L

( )= σ πd0( ) ( )e

202

,

TF1710_book.fm Page 309 Thursday, September 30, 2004 1:43 PM

(see Chapter 5) Among other factors, we do not take into account

in Chapter 13; here, let us describe briefly tropospheric radar, or weather radar (WR),

310 Radio Propagation and Remote Sensing of the Environment

where N is the drop concentration Equation (11.15) is valid on the assumption thatall the drops are similar In reality, their radius a is the stochastic value, and weneed to apply to the main value:

2 0

εd

212( , )L =N k a −

λ

εε

l=cτp2TF1710_book.fm Page 310 Thursday, September 30, 2004 1:43 PM

Radio Devices for Remote Sensing 311

Different distribution functions are used for calculating the 39 One of themost universal is the distribution proposed by Deirmendjian:86

(11.20)

Here, am is the modal radius (i.e., most probable one) The parameters α and γdepend on the type of hydrometeors The averaged value is:

for various types of hydrometeors Spectral analysis of the scattered signals allows

us to study the internal motion in clouds Similar formulae can be used for snowand hail clouds It is sometimes necessary to use more exact formulae for thescattering cross section, especially in the case of hail, whose size can be comparablewith the wavelength The reflectivities of precipitation are connected with its inten-sity J (mm/hour) by the empirical formula:89

Parameters A = 200 and b = 1.6 are used for rain, and A = 2000 and b = 2.0 areused for snow in moderate latitudes.89 Weather radar applications are used in mete-orological services and aviation Recently, they have also found application in spaceresearch along with the help of so-called rain radars.91,92

Radar is sometimes applied to tropospheric turbulence research The specificcross section of the backward scattering is, in this case (accordingly to Equation(5.173)):

One can see that in the case of Kolmogorov–Obukhov turbulence.The calculations show that this cross section is small, and powerful radar is neededfor the detection of tropospheric turbulence.89

Pencil-beam antennae are seldom used for remote sensing of land, but they dofind application in airborne navigation systems; however, they have poor spaceresolution due to difficulties encountered by having a large-size antennae on board.This is the reason why side-looking radar (SLR) and synthetic aperture radar arewidely used We will first discussSLR An antenna with a narrow pattern in one

αγ

αγγ

β α

6 66

β

γα

γ m

The data represented in Table 12.2 allow us to determine combinations of parameters

312 Radio Propagation and Remote Sensing of the Environment

direction can be rather easily mounted on the flying platform — for example, a leakywaveguide antenna mounted along an airplane fuselage So, the spatial resolution

of the radar along the flight direction can be written as:90

where d is the horizontal antenna size, and R is the slant range to the illuminatedelement of the land surface The radar radiates short pulses which permit discrimi-nation of the reflections from the land elements separated by distance:

where θ is the angle between vertical and the direction to the illuminated element(see Figure 11.2a) Equation (11.25) determines the spatial resolution across thedirection of flight More exactly, lines of similar ranges are circles and lines ofsimilar angle positions are lines (see Figure 11.2b) Thus, it can be said that Equation(11.24) defines the resolution by azimuth and Equation (11.25) by one-in range.The SLR antenna looks sideways, thus its name The operation of such radar isbased on backward scattering of rough surfaces; for example, SLR does not “see”smooth water surfaces The backward scattering depends on inclinations of the landThe latter one is realized during the platform flight when the range selection givesthe scan line, and the frame scan is realized in the flight process It is similar toformation of a television image In the first airborne SLR, the radar image wasdisplayed by an electron-ray tube and then photographed on film The film served

as the memory for information storage Now, such storage is a function of thecomputer and associated devices

or water elements (see Chapter 6) and is the basis of mapping a landscape by SLR

FIGURE 11.2 Side-looking radar operation.

Y h

Radio Devices for Remote Sensing 313

Equation (11.11) can be used for the calculation of the received signal power

It is necessary only to substitute slant range R for distance L The target cross section

is now defined by the sizes of the space resolution element (pixel):

Note the antenna area dependence on the nadir angle, θ, which means that themagnitude of the reflected signal depends on the pixel position Certainly, the specificcross section σ0 also depends on the pixel position This dependence is the basis ofthe radar image

The swath of the radar image is determined by the pulse time repetition (T):

where θ0

is assumed, then, that YSW << Rmin The antenna beam angle in the vertical plane has

to be of the order of:

sin

2

222

kd kd

23

Ysw≅ cT

2sinθ0TF1710_book.fm Page 313 Thursday, September 30, 2004 1:43 PM

is the nadir angle (see Figure 11.2b) of an SLR observation near line It

314 Radio Propagation and Remote Sensing of the Environment

In this case, all of the observed points will be sufficiently illuminated On the other

hand, the spurious signals corresponding to reflections from points outside the swath

will be essentially weakened These signals appear due to the periodic repetition of

the pulses radiation

The time repetition cannot be too long It is necessary to illuminate any pixel

at least once during the flight; therefore, the following inequality has to be performed:

where v is the platform velocity

The analysis of SLR images has some peculiarities One of them is connected

with the fact that the pixel position is fixed in the slant range scale, which leads to

the distortion of images To correct this, we must remember the relation

, which is valid when we assume the land to be smooth and plain

This formula needs to be changed slightly to take into consideration the spherical

character of the surface of Earth The relief heights also distort scales and form radar

shadows (Figure 11.3)

The spaceborne SLR for the Russian Kosmos-1500 mission114 and subsequent

Ocean satellites is an example of the effective application of such instruments for

remote sensing This 3.1-cm SLR operates at a 650-km orbit and provides images

FIGURE 11.3 Artifact of radar image.

R

cT R

2

T X v

R vd

introver-Radar image

reduction Shadow Shadow

Shadow Low signal

TF1710_book.fm Page 314 Thursday, September 30, 2004 1:43 PM

Radio Devices for Remote Sensing 315

with a δX resolution of 2.1 to 2.9 km and a

δY resolution of 0.6 to 0.9 km within a swath

that is 475 km wide The wide swath is

advan-tageous for ice patrols This SLR can prepare

a map of the ice situation in the Arctic zone

in a period of 3 days This period is short

enough to suggest that the ice situation

remains the same throughout Equation

(11.29) is not sufficiently correct in this case,

so we must use more exact relations to clarify

our geometrical understanding

The spatial resolution of the spaceborne

SLR Ocean is close to limit resolution and it has no difficulty in realizing good

range resolution when wide-band signals are applied For example, a range resolution

of 10 m requires pulses with a duration of about 5 · 10–8 sec or signals with a

bandwidth of about 20 MHz; however, to achieve the same azimuth resolution for

a spaceborne SLR we must increase the 3-cm antenna to a length of about 3 km! It

is not feasible to have such an antenna, but the problem can be overcame with the

help of so-called synthetic-aperture radar (SAR)

Let us explain the principles of this kind of radar operation Figure 11.4 shows

the geometry of the problem The radar platform (P) moves with velocity v in the

x-direction The considered elements of illuminated surface are situated in point

(X,Y) and, at time moment t, is removed from the SAR by distance:

Due to the Doppler effect the signals, scattered backward by the surface elements,

will have the frequency shift (see Equation (2.98)):

The coefficient appears twice because of the doubled Doppler shift for the radio

propagation forward and backward This frequency shift depends on the x-coordinate

of the surface element So, different elements can be resolved by spectral analysis

of the received signal The spatial resolution is related to the spectral resolution by

2 ,

-X V

316 Radio Propagation and Remote Sensing of the Environment

However, the signal can appear at the filter output after time interval

This means that signal accumulation takes place during this interval The platform

will pass the distance during this cumulative interval As a result, the

frequency signal filtration is similar to the coherent summation of the signals by the

antenna with the synthetic aperture of length dsynth Equation (11.33) can be rewritten

in the form:

which is equivalent to Equation (11.24) The difference between physical and

syn-thetic apertures is that the physical aperture provides a parallel summation of the

signal, whereas the synthetic aperture relies upon a sequential summation technique

It might appear that an unlimited increase of the synthetic aperture (or the time

of synthesis) would allow a very high resolution to be achieved; in fact, though, this

is not true We have to keep in mind that the process of coherent summation takes

place Only the surface elements lying within the Fresnel zone scatter the signals

with the phases The signals of elements that are outside this zone get an additional

phase shift because of the spherical wave diversity The Fresnel zone size is

in our case and

This resolution can be improved by correcting the spherical diversity This operation

is called focusing and is realized by signal summation with weighting factors of the

form:

These weighting factors compensate the summand vt in the bracket of Equation

(11.32), which provides the constancy of Doppler shift for the chosen surface

element It is similar to the phase correction in optical lenses by the law

(thus the term focusing) In the case of the focusing procedure, themaximum synthetic aperture length is equal to the real antenna footprint on the

2 0

2 2 0

δ Φ = kx2 2F

d synth R d

max = λ 0

δ X= d2TF1710_book.fm Page 316 Thursday, September 30, 2004 1:43 PM

Figure 11.5 provides a comparison of similar radars with and without focusing

Radio Devices for Remote Sensing 317

It would appear that an unlimited decrease in the real antenna size would allow

us to realize high azimuth resolution, as much as we want; however, it does not

First, a small antenna will not produce an adequate signal-to-noise ratio Also, the

synthesized aperture is similar to the antenna array, with sources separated by

distance vT, where T is the pulse repetition time It is well known4 that the pattern

of such an array has, besides a major maximum in the direction ϕ = 0, similar

maxima in directions determined by the equality (n = 1, 2,…) The

signals scattered from these directions interfere with the main one and are

indistin-guishable from it This means that the angular width of the major lobe of the real

antenna has to be sufficiently narrow for suppression of this noise; thus, we can

conclude that the inequality must be performed This last inequality means

that the distance passed by the platform at the time interval between two sequential

radiations must be less then the best azimuth resolution

The resolution limit is defined by the time interval while the

point of interest is illuminated by the SAR This is correct if the SAR antenna pattern

is fixed; however, the limit imposed by Equation (11.37) can be overcome if the

antenna beam is directed toward the point of interest over a time interval during the

platform flight This mode of operation (i.e., spotlight) requires an antenna with a

changeable azimuth beam direction This operation can be realized, for example, by

using a phase antenna array The time of illumination can be greater than

Correspondingly, the length of the synthetic aperture will be greater than , and

the spatial azimuth resolution improves compared to that given by Equation (11.37)

The range resolution of SAR is realized by short pulse radiation This method

is similar to the one for SLR and was used, for example, in the Russian Almaz

mission.146 Today, use of the compressing technique has become more common

This technique has been applied for the European Remote Sensing satellites

(ERS-1,2), the Japanese Earth Resources Satellite (JERS-1), the Canadian RADAR

FIGURE 11.5 Linear resolution in a direction of synthesizing for radar with a wavelength

of 4 cm and with the antenna size of 2 m (1) without synthesizing the aperture and without

focusing; (2) with synthesizing of the aperture but without focusing; (3) with synthesizing of

the aperture and with focusing

10000

δ , m

1000 100 10 1

318 Radio Propagation and Remote Sensing of the Environment

SAT, and others.146 The range resolution in this technique is determined by therelation:

where ∆f is the radar signal bandwidth Equation (11.29) can be used, as before, to

define the swath In order to extend the observation band, the technique of switchingthe antenna beam direction in the vertical plane can be used (for example, inRADARSAT)

A typical property of SAR images is the speckle structure The origin of thiseffect is the stochastic process of scattering by the surface Usually, the scatteredsignal satisfies Gauss’s law, which leads to Rayleigh’s distribution of its amplitude

In this case, the intensity of the signal power fluctuation equals the mean poweritself Intensive fluctuations are very undesirable for the numerical analysis of image,

so to overcome this problem the process of averaging is used We could see (Equation(6.166)) that the spatial correlation coefficient of the scattered signal is determined

by the degree of overlap of the antenna footprints This means, in our case, thatsignals of neighboring pixels are statistically independent; therefore, summation ofthese signal intensities leads to smoothing of the speckle structure This is a commonmethod for improving the quality of SAR images The other method involves partlyoverlapping the neighboring image scenes Of course, the speckle structure leads to

a loss of spatial resolution

The radar formula for SAR can be obtained from Equation (11.28), in which it

is necessary to substitute for d When the pulse compression procedure is

used, the factor:

(11.39)

must be introduced Here, τp is the duration of the uncompressed pulse and ∆f is its

bandwidth In the following discussion, it will be useful to take into account themean radiated power:

θ σ δλ

∆