1170 LIDAR / Atmospheric Sounding Introduction

a distant large telescope for the receiver. In thisconfiguration, now known as bistatic, the range ofthe scattering can be determined by geometry. In thebistatic configuration, shown in Figure 1, the field ofview of the receiver is scanned along the transmittedbeam in order to obtain an altitude profile of thescattered light.

The first results obtained using this principle werereported in the late 1930s when photographic record-ings of light scattered from a searchlight beam weremade. .

Typically, modern lidar systems are monostatic inconfiguration, with the transmitter and receiver co-located.' Monostatic systems can be subdivided intotwo categories: coaxial systems, where the laser beamis transmitted coaxially with the receiver's field ofview, and biaxial systems, where the transmitter andreceiver are located adjacent to each other. Monostaticlidar systems use pulsed light sources, thereby enablingthe range at which scattering occurs to be deter-mined from the round-trip time of the scattered light(Figure 2).

By the early 1950s, refinements in technique andimproved instrumentation, including electrical re-cording of the intensity of the backscattered light,allowed the measurement of atmospheric densityprofiles up to altitudes of around 67 kill. Thesemeasured density profiles were then used to derivetemperature profiles using the Rayleigh-lidar tech-nique, which is described later.

The invention of the laser in 1960 and the giantpulse, or Q-switched, laser in 1962 provided apowerful new light source for lidar. The first use of alaser in a lidar system was reported in the early 1960sand since then developments in lidar have been linkedclosely to advances in laser ter' ')logy.

~

ftotal = fup+ fdown = 2 fJ.zlc

t!.z = (flolal' c)/2

Figure 2 Schematic illustrating the process of ranging based ontiming the returned signal.

detection and recording systems. Figure 3 is a blockdiagram of a generic lidar system, which shows howthese subsystems work together to form a completelidar.

Transmitter

The transmitter generates light pulses with the re-quired properties and directs them into the atmos-phere. Pulsed lasers, with their inherently lowdivergence, narrow spectral width, and short, intensepulses are ideal as the light sources for lidar systems.

In addition to a laser, the transmitter of a lidar oftenincludes a beam expander, whose purpose is to reducethe divergence of the beam being transmitted into theatmosphere. This allows a reduction in the back-ground measured by the lidar. At night, the back-ground is due to light from the Moon, stars, airglow,and artificial lights. During the day, background ispredominately due to the Sun. Background can enterthe lidar receiver either directly or after scattering inthe atmosphere. A reduction in the divergence of the

Llgnt transmllteCinto atmosphere

Figure 3 Schematic of a generic lidar.

transmitted beam allows the field of view of thereceiver to be reduced, resulting in a lower back-

ground.The narrow spectral width of the laser has been used

to advantage in a variety of ways in lidar systems. Itallows the spectral filtering of light by the lidarreceiver. A bandpass filter tuned to the laser wave-length selectively transmits photons backscatteredfrom the laser beam, while rejecting photons at otherwavelengths, thereby enabling a reduction in thebackground by several orders of magnitude. The pulseproperties of pulsed lasers allow ranging to beachieved by timing the backscattered signal, thusallowing the simpler monostatic configuration.

The major influence on the type of laser used in alidar is the parameters the lidar is being designed tomeasure. Some measurements require a very specificwavelength and/or tunability, i.e. resonance-fluores-cence and differential-absorption lidar (DIAL). Thesetypes of lidars can require complex laser systems toproduce the required wavelengths, while other simplerlidars, such as Rayleigh, Raman, and aerosollidars,can operate over a wide wavelength range. Although itmay be possible to specify the exact performancecharacteristics of the laser required of a particular lidarmeasurement, these characteristics often need to becompromised in order to select from the types of lasersavailable.

Receiver

The receiver of a lidar collects and processes thescattered laser light before directing it onto thedetector. The first optical component, the primaryoptic in the receiver usually has a large diameter,enabling it to collect a large amount of the scatteredlaser light.

Lidar systems typically utilize primary optics withdiameters ranging from about 10 cm up to a fewmeters in diameter. Optics at the smaller end of thisscale are used in lidar systems that are designed towork at close range - a few hundred meters - and maybe lenses or mirrors. Optics at the larger end of thisrange are used in systems designed to probe the middleand upper atmosphere and are typically mirrors.

After collection by the primary optic, light is usuallyprocessed in some way before being directed to thedetector system. Processing can be based on wave-length, polarization, and/or range, depending on thepurpose for which the lidar has been designed.

As described previously, the simplest form ofprocessing based on wavelength is the use of anarrow-band interference filter to reduce the back-ground. Much more sophisticated spectral filtering

LlDAR / Atmospheric Sounding Introduction 1171

schemes are employed in Doppler and high-spectral-resolution lidar systems.

Signal separation based on polarization is a tech-nique often used in the study of atmospheric aerosols.Information on aerosol properties can be obtainedfrom the degree to which light scattered from apolarized laser beam is depolarized.

Processing of the backscattered light based on rangecan be performed in order to protect the detector fromthe intense near-field returns of high-power lidarsystems. This protection is achieved by using a fastshutter that closes the optical path to the detectorwhile the laser is firing and for a short time afterward.The shutter opens again in time to allow transmissionof light backscattered from the altitude range beingstudied.

Detection and Recording

The signal detection and recording section of a lidartakes light from the receiver and produces a permanentrecord of the measured intensity as a function ofaltitude. In the first lidar systems the detection andrecording system comprised a camera and photo-graphic film. Today detection and recording isachieved electronically. The detector is a device thatconverts light into an electrical signal and the recorderis an electronic device, often involving a microcom-puter, which processes and records this electrical

signal.Photomultiplier tubes (PMTs) are devices used as

detectors for incoherent lidar systems working in thevisible and ultraviolet. PMTs convert an incidentphoton into an electrical current pulse large enough tobe detected by sensitive electronics. Other devices thatare less commonly used as detectors in lidar systemsinclude multianode PMTs, micro-channel-plates(MCPs), and avalanche photodiodes.

There are two ways the output of a PMT can berecorded electronically; the pulses can be countedindividually (photon counting) or the average currentdue to the pulses can be measured and recorded(analog recording). Which method is the more appro-priate depends on the rate at which the PMT producesoutput pulses, which is proportional to the intensity ofthe light incident on the PMT. If the average timebetween PMT output pulses is much less that theaverage pulse width, then individual pulses can beeasily identified and photon counting is the moreappropriate recording method. However, if the aver-age time between PMT output pulses is close to, orgreater than, the average pulse width, then it becomesimpossible to distinguish overlapping pulses, and soanalog recording becomes the more appropriatemethod.

~(~

1172 LIDAR / Atmospheric Sounding Introduction

Coherent DetectionThere is a class of lidar systems that determine windspeed by measuring the Doppler shift of backscatteredlight. There are two ways these measurements can beachieved, namely incoherent and coherent detection.Incoherent systems measure the wavelength of thetransmitted and received light independently, usinga spectrometer, and determine the Doppler shiftfrom these two measurements. Coherent detectionsystems use a local oscillator, a narrow-band contin-uous-wave laser, to set the frequency of the transmittedpulses. Systems incorporating coherent detection use alocal oscillator on a photomixer. This arrangementresults in the output of the photomixer being a radio-frequency (RF) signal whose frequency is the differ-ence of the frequencies of the local oscillator and thebackscattered light. Standard RF techniques are thenused to measure and record this RF signal. Themeasured RF signal is used to determine the Dopplershift of the backscattered light and thus the windspeed.

The Lidar Equation

The lidar equation is used to determine the number ofphotons detected by a lidar system. The lidar equationtakes into account both instrumental parameters andgeophysical variables. The general form of the lidarequation includes all forms of scattering and it can beused to calculate the signal strength for any lidar.

The number of photons detected as pulses at thephotomultiplier output, per laser pulse, is

r PS(A)Tt(A)Tr(A){2(A)J.1.)'

dO'.xL ~ (A)Nj(r) drdA

I

A

In eqn [1] A is the area of the telescope; PS(A) is theconvolutionofP(A) andS(A), whereP(A) is the numberof photons emitted by the laser in a single laser pulseand S(A) is a function which takes into account anywavelength shift during scattering, including Dopplerand Raman shifts; LlA is the wavelength range forwhich PS(A) is nonzero; 't"t(A) and 't"r(A) are the opticaltransmission coefficients of the transmitter and re-ceiver optics respectively; Q(A) is the quantum effi-ciency of the photomultiplier; r is the range and R 1 andR2 are the minimum and maximum ranges for a rangebin; ~ (A) is the overlap factor which takes into accountthe intensity distribution across the laser beam and thephysical overlap of the transmitted laser beam and thefield of view of the receiver optics; 't"a (r, A) is the optical

~ .,;

transmission of the atmosphere along the laser path;(dO";/dQ)(A) is the backscatter cross-section for scat-tering of type i; and Nj(r) is the number density ofscattering centers, which cause scattering of type i.

The general form of the tidar equation, as expressedin eqn [1], can usually be greatly simplified whenapplied to a particular lidar system.

Rayleigh Lidar

Rayleigh lidar is the name given to the class of lidarsystems that measure the intensity of light backscatterby molecules from altitudes between about 30 and100 km. The intensity profiles measured by Rayleighlidars are used to calculate relative density profiles,which are in turn used to calculate absolute temper-ature profiles. The terms Rayleigh scattering andmolecular scattering are often used interchangeably,as are the terms Mie scattering and aerosol scattering.Rayleigh theory named after its founder, Lord Ray-leigh, describes the scattering of light by moleculesthat are small compared with the wavelength of theincident radiation; Mie theory describes scattering byaerosols that are not small compared with the wave-length, so there is a strong connection between thesetwo pairs of terms.

Rayleigh scattering explains the color, intensitydistribution, and polarization of the blue sky in termsof scattering by atmospheric molecules. For objectswith dimensions greater than about 0.003 times theincident wavelength, the more general Mie theorymust be used to calculate scattering effects.

The Rayleigh backscatter (0 = n) cross-section forthe atmosphere below 90 km can be expressed as

dUR(O = 1t) - C m2 sr-ldQ - l4 [2]

[1]where the value of C is between about 4.75 x to-57and 5.00 x to-57, depending on the value used forindex of refraction of air. Above 90 km altitude, theconcentration of atomic oxygen becomes significant,causing the refractive index of air to change, resultingin eqn [2] becoming less accurate with increasingaltitude. The Rayleigh backscatter cross-section, eqn[2], can be used in conjunction with the lidar eqn [1] todetermine the intensity of the backscatter that can beexpected for a particular Rayleigh lidar system.

The Rayleigh lidar technique relies on the measuredsignal being proportional to the atmospheric density.This is not the case in any region that containsaerosols. From the surface to the top of the strato-spheric aerosol layer, about 25-30 km, the atmospherecontains a significant concentration of aerosols, thusthe Rayleigh technique cannot be directly applied to

~