coherent laser radar in europehome.ustc.edu.cn/~522hyl/%ce%c4%d5%c2/11.coherent...remote sensing...

Coherent Laser Radar in Europe

JOHN M. VAUGHAN, KURT OVE STEINVALL, CHRISTIAN WERNER, AND PIERRE HENRI FLAMANT

European work in coherent laser radar with 10 pm and shorter wavelength lasers is reviewed. Fundamental aspects include heterodyne studies of signal statistics and fluctuations, and detailed experimental and theoretical work on signal ampl$cation and autodyne arrangements with light reinjected into the laser cavity. Progress with lasers, detectors, and modulators has led to the development of several compact robust jield systems both continuous-wave and pulsed. Various ground-based programs are described including local wind field measurement and wake vortex investigation at airjields, and study of range, image, and Doppler shift of hard targets. Airborne systems have investigated avionics problems of true airspeed, pressure error, and wind shear warning. Other airborne studies include ground imaging, obstacle warning, terrain following, and a compendium of atmospheric backscattering over the North and South Atlantic. In recent years, the European Space Energy has supported studies and technology developmentfor a space-borne wind lidar in the Atmospheric Laser Doppler Instrument (ALADIN) program.

I. INTRODUCTION Remote sensing with coherent laser radar has been pur-

sued in Europe since the 1970’s. Much of the early activity was stimulated by military interest in topics such as range and velocity measurements, low-level wind and vibrometry. This further stimulated the relevant technologies of lasers, detectors, modulators, optical systems, and signal processing. Nowadays, there is also considerable interest in civil applications such as atmospheric measurements, aircraft wake vortices and a possible space-borne wind lidar. It may be noted that the leading international conference series on coherent lidar has met three times in Europe: at Malvern in 1985, Munich in 1989, Paris in 1993, and alternating with Colorado in the US, provides some appreciation of the interest in, and support for, the subject.

The following pages offer a broad review of work in Eu- rope. Obviously, in a limited space, this cannot hope to be totally comprehensive; inevitably much excellent work will

Manuscript received June 2, 1995; revised October 31, 1995. J. M. Vaughan is with the Defence Research Agency, Gt. Malvern,

K. 0. Steinvall is with the National Defence Research Establishment,

C. Werner is with the DLR, Institut fur Optoelectronik, 82234 Oberp-

P. H. Flamant is with the Laboratoire de Meteorologie Dynamique,

Publisher Item Identifier S 0018-9391(96)00920-6.

Worcester WR14 3PS, UK.

S-58111 Linkoping, Sweden.

faffenhofen, Germany.

Ecole Polytechnique, 91 128 Palaiseau, France.

be omitted or dealt with but briefly. However, it is hoped that the fairly extensive list of references will provide access to further material. In the following Section 11, various fundamental aspects are reviewed, including signal statistics and amplification, signal analysis and system calibration. This leads on to technology developments in Section I11 which outlines work on lasers, detectors, modulators, and optical technique. The following Sections IV and V review, respectively, the development and application of ground and airborne systems including atmospheric and solid target investigations. Section VI then outlines various activities across Europe, notably at the European Space Agency, that are developing the technology and systems appreciation of a possible space-borne wind lidar. Finally, Section VI1 provides a summary with some preview on the future.

11. FUNDAMENTAL STUDIES The principle of laser radar for remote-sensing is of

course well established: a laser beam illuminates the object under study, and information is derived from the resul- tant scattered light. Two regimes may be distinguished employing either incoherent, direct detection, or coherent light-beating techniques. In the former case the laser is used as a powerful bright-light source; some predetection optical filtering may be employed to provide spectral information, but basically the measurement is one of intensity only (which may be allied with time to provide for example a simple pulsed range finder). Coherent techniques on the other hand make use of the full spatial and temporal coherence properties of laser radiation. By mixing the scattered light field with an optical local oscillator beam the full phase, frequency and amplitude information in the signal is made available. In addition, shot noise in the local oscillator beam may be used to dominate thermal noise in the detector to provide quantum-limited detection of the signal beam. For effective operation, very precise control of laser parameters is required together with profound under- standing of optical arrangements, scattering mechanisms, propagation, signal processing, etc. The simple coherent principle thus raises many questions of fundamental science and technology. Much attention in Europe has been given to the fundamentals of laser radar as outlined in the following sections. An early example is the calculation of

0018-9219/96$05.00 0 1996 IEEE

PROCEEDINGS OF THE IEEE, VOL. 84, NO. 2, FEBRUARY 1996 205

the antenna parameters and the maximization of system efficiency due to Rye [lj, who also considered refractive turbulence contributions to heterodyne lidar returns [2j.

A. Signal Statistics and Fluctuations When laser light is scattered from a rough surface (that

is with spatial structure comparable to the wavelength) well known speckle effects arise from the mutual inter- ference of different elements of the scattered field. If the scattering object is in motion the speckle pattern is dynamic and detectors in the field register strong intensity fluctuations. The impact of such intensity fluctuations on system performance, selection of appropriate signal analysis and processing, and the possibility of speckle averaging and suppression are thus important questions.

A particularly detailed study of signal fluctuations and their impact on system performance was made by workers at the Swedish National Research Institute [3]-[5j. In the first paper [3] the design and evaluation of a pulsed imaging CO2 laser radar system was described together with the signal statistics recorded for various targets. The transmitter laser produced 160 ns-full width, half maximum (FWHM) pulses at a pulse repetition frequency (PRF) of 1 kHz; the local oscillator offset frequency was kept at 15 MHz by a digital interlocking system. The received signal was amplified, envelope detected and digitized in a transient recorder. Speckle and turbulence effects on signal statistics were evaluated by measuring the signal amplitude distribution for diffuse and glint targets, as well as for topographical targets. A separate He-Ne laser system provided simultaneous measurements of the atmospheric turbulence parameter (3: in order to separate the turbulence effects on the signal return statistics from the speckle- induced variations. A glint target was employed to measure the turbulence induced signal variations. The return signal distribution in the direct detection mode was found to fit the log-normal distribution. Topographical targets such as trees, grass, and rocks exhibited return signal statistics slightly different from the Rayleigh distribution obtained with diffuse targets. The returns from terrain targets were found instead to fit a Weibull distribution.

In further experiments [4] a monostatic, frequency modu- lated continuous wave (FM-CW) CO2 laser radar was used to record the signal amplitude distributions from various targets. The signal amplitude was sampled at rates of up to 1 kHz by a 12-b A/D converter, thus allowing for the collection of enough data in a short measurement period. A spherical mirror with a radius of curvature of 6 m was used as a glint target; various diffuse targets were investigated as well as different terrain targets during different seasons. A semi-rough target can be considered as a combination of both specular (glint) and diffuse (speckle) contributions as illustrated in Fig. 1. Manmade objects, e.g., vehicles, often show both glint and speckle characteristics. For terrain targets, although mainly diffuse in character, the signal amplitude was best described by a Weibull probability distribution function (PDF). By comparing the average return amplitude from the terrain with the retum from a

0 1.25 I-

SIGNAL AMPLITUDE

(a)

1.5 -

Z

c I)

0 1.25-

m rr 1.0 - vt n

0.75- c- 5 0.5 -

0.25-

+

_I

61 0

0 0.5 1 1.5 2 2.5 0

NORMALIZED AMPLITUDE

(b)

Fig. 1. Signal distributions from a curved semirohgh surface: (a) shows the signal probability distribution from 1 024 pixels across the whole cylinder shaped surface and (b) shows the distribution along the center line giving glints.

calibrated reference target, the diffuse reflectance could be calculated. However, large variations in the reflectance values were observed and attributed to, e.g., angle of incidence, humidity conditions, and seasonal variations. At the time this work was carried out, very few experimental data on target signatures and return signal statistics were available in the open literature. The data collected in t h i s work thus permitted a detailed comparison between experimental and theoretical data.

In a third paper [ 5 ] , the performance of coherent CO2 laser radar systems was derived using signal amplitude distributions that were experimentally recorded from several types of target, as described in previous work. Two applications were investigated, namely remote sensing of atmospheric pollutants and target detection with a shot- noise limited system. In the first case, the performance was discussed in terms of the gas concentration accuracy, whereas the detection probability (PD) is considered for hard target detection. The effects of target movement, beam

206 PROCEEDINGS OF THE IEEE, VOL. 84, NO. 2, FEBRUARY 1996

wandering, range scintillation, carrier-to-noise ratio, and background limited detection were extensively analyzed.

In a contrasting study [6] at DRA Malvern, intensity fluctuations in a CW coherent heterodyne lidar at 10.6 pm were explored, with radiation scattered from a model target consisting of a moving belt. Eight classes of experiment were conducted and the fluctuations examined by techniques of real-time temporal autocorrelation to form intensity correlation functions-the first such application to CW heterodyne signals in the infrared. Rather remarkable changes both of characteristic coherence time and shape were exhibited in the correlation functions. The results were related to theoretical considerations and explained in terms of the balance of three different mechanisms of translation, tilt, and ripple.

In recent work Harris et al. [7] exploited the fact that in heterodyne detection (such as in coherent lidar) the optical local oscillator defines a single mode of the incoming- signal light field; this single-mode selectivity preserves the full fluctuation character of scattered light (this is in contrast with direct-detection schemes, as in photon-correlation spectroscopy, where aperture averaging usually reduces the range of fluctuations). Examples of Gaussian and non- Gaussian statistics in laser light scattered from a moving ground-glass screen were studied with great precision. The simple laboratory experiment was shown to have several advantages over equivalent direct-detection schemes, and gave experimentally the theoretically predicted factorial intensity moments (up to the seventh order) that result from zero-mean, circulo-complex Gaussian statistics. In further work exploiting the power of heterodyne techniques, spectral filtering within the Schawlow-Townes laser linewidth

amplification with both traveling wave and cavity gain media. Novel phenomena have been revealed, including generation of image frequencies and noise cancellation due to four-wave mixing. A second stimulus has been the realization of the truly remarkable advances achieved in erbium-doped fiber amplifiers for telecommunications. At the same time amplification by multipassing of conventional cavities and gaseous media have been explored and reveal gain factors of 10-lo2. Finally optical amplification has actually been employed in a number of lidar systems operated in the field. In direct detection, with generally noisy detectors, very useful increases in signal to noise ratio (SNR) have been demonstrated. The topic of amplification was discussed in some detail at a small workshop held in Malvern in 1992 [ 1 I]. One early technique has been developed in which the return signal reenters the source laser and, after amplification, subsequently beats with the main laser output. This method is now known as “autodyning” and has recently been successfully employed in a differential absorption lidar by Richter et al. [12].

Two other studies of lidar systems have recently been conducted. Rahm et al. [13] considered that every real Doppler lidar system will not be ideal, with respect to electrical noise and the heterodyne efficiency. With this assumption an optical amplifier can improve the SNR and the performance of a Doppler lidar. The degree of improvement depends in this case only on the different noise sources of the lidar and the optical amplifier. The expected improvement V in the SNR is calculated to be

NLO + Nelec

NLO + Nelec + Nu V = G

was demonstrated [8]. Such spectral filtering was also shown to significantly alter the statistics of the fluctuating

filter bandwidth becomes much narrower than the overall spectral linewidth of the scattered light, the distribution

field statistics (i.e., a negative exponential). Most recently the intensity time sequences of Lorentzian-broadened light

where NLO is the shot noise produced by the local oscillator

Nelec describes the electrical noise of the detector and the electrical amplifier, Nu is the noise of the optical amplifier

gain of the optical amplifier. From (1) it can be shown that the improvement of V is high

intensity of non-Gaussian light [9]. In particular when the (not to be with the shot noise Of the

of intensities tends to that expected for complex Gaussian due to the process of spontaneous emission, and is the

(with Gaussian statistics) are demonstrated to exhibit a fractal character, with fast and slow fluctuations over nearly a three decade range of timescale [lo]. This contrasts sharply with the much smoother fluctuations observed for Gaussian-broadened light.

B. Signal AmpliJLication and Autodyning The possibility of gainfully employing optical amplifi-

cation for boosting weak signals has been discussed since the earliest days of laser radar. Nevertheless coherent lidar work has largely been carried out without recourse to amplification, often with self-aligning, monostatic trans- mittingheceiving systems and usually with great efforts to achieve optical isolation of signal and local oscillator paths. Recently, however, the situation has changed con- siderably and there is renewed interest in the possibilities of amplification. Several groups have been examining both theoretically and experimentally the fundamentals of optical

1) if the noise of the optical amplifier Nu is low and 2) if the noise of the lidar NLO + Nelec without the

optical amplifier is high.

Thus when the performance of the lidar is bad without the optical amplifier, for example due to a high electrical noise, or poor heterodyne efficiency, the higher will be the potential improvement. To verify this model a pulsed optical amplifier operating at the wavelength of 10.6 pm with CO2 as active medium has been developed, and tested in autumn 1991 in the Doppler lidar of LMDKNRS in France, which had a reported heterodyne efficiency about 1%. The measured value for the improvement V at atmospheric measurements varied between 24.0 and 22.9. In this case an improvement factor in the range (in the near field) of the lidar of up to 13 has been reached.

In another study, Morley et al. [14] investigated a long path CW CO2 laser system as a predetection amplifier.

VAUGHAN et al.: COHERENT LASER RADAR IN EUROPE 207

n

/q U Ar+ LASER

ACOUSTO-OPTIC $j DETECTORS 6 MODULATORS LENS MIRROR

Fig. 2. Experimental arrangement for studying in transmission (at detector A) and reflection (at detector B) the amplification of frequency shifted radiation returned to the laser cavity [15].

The amplifier signal gain and noise characteristics were measured and the SNR enhancement produced in a Doppler radar system was determined over a range of operating conditions in the amplifier. A maximum SNR enhancement of 12.4 dB was measured for the CW gain cell-in good agreement with the theoretical prediction.

Fundamental questions of noise and gain have been investigated in extensive theoretical and laboratory based studies at DRA Malvern and Essex University and reported in several papers including two Physical Review Letters. In the first of these [15], gain profiles for both reflected and transmitted signals were measured for an argon-ion laser operated above threshold in the experimental arrangement shown in Fig. 2. The observed behavior was interpreted by developing a theory which incorporated four-wave mixing of signal and image frequencies. The gain showed typical characteristics of a second-order phase transition with implications for fundamental laser noise. In a second letter [16] the properties of light emitted into the nearest subthreshold mode on either side of the single lasing mode were investigated. This light beats with the laser line and hence contributes to the laser’s intensity-fluctuation noise spectrum over a range of frequencies centered close to the longitudinal mode separation. The noise contribution from one subthreshold mode was shown to be strongly anticorrelated with that from the other, leading to optical noise cancellation. In other studies gain and noise have been investigated in a traveling wave amplifier [17], subthreshold modes [18], [19] and in two extensive theoretical and experimental papers [20], [21] that consider both class A and class B lasers. In particular, the SNR for both heterodyne and direct detection measurements were evaluated E201 and the conditions were determined under which optical amplification can lead to an enhancement of sensitivity.

C. System Calibration and Weak Signal Analysis Accurate calibration of coherent laser radars has been

of interest for several reasons. One motivation is to verify experimentally the theoretical shot noise detection limit and explain the deviations between the theoretically calculated and measured sensitivities. Of direct practical interest is also intercalibration, for example between reference targets

208

and those of interest, e.g., the backscatter of the aerosols. One of the earliest studies was carried out by Renhorn et al. [22] with an FM-CW laser radar. Heterodyne mixing efficiency was measured by using a glint test target (a spherical mirror on a moveable carriage) that gave a sufficiently strong signal to be measured in both direct and heterodyne detection mode. With this arrangement the FM-CW laser radar showed a measured heterodyne efficiency of 0.16 to be compared with a theoretical value of 0.64 (which might indicate some nonideal optical components in the system). With this and other terms the measured signal to noise ratio (SNR) was shown to be about 6 dB below the theoretical upper limit.

There are many reasons for the discrepancy between measured system performance and the optimal performance that one should expect from an ideal system. Although much attention has been paid to the limiting factors in laser radar system, the important limitation due to stray light has been largely ignored. Letalick et al. [23] have investigated the frequency limiting factors in a coaxial laser radar system. It is shown that the mechanical vibrations of the laser radar structure combined with stray light can have a profound deleterious effect on the performance. Pulsed systems with short pulses (<,us) should, however, be largely immune to this problem.

In another series of studies with CW CO2 systems a detailed comparison of experimental and theoretical SNR was made with three different signal analyzers [24], [25]. Allowance was made for transmission in the receiver optics, the effective quantum efficiency in the detector due to shot noise domination by the local oscillator, and for coherent speckle effects. After correction for evaluation of voltage rather than power spectra, agreement within about 6 dB was achieved for the strong signal limit. The second paper [25] considered the more generally relevant low signal case, and operation of a surface acoustic wave spectrum analyzer and digital integrator. This treatment takes account of spreading of weak signals over several channels and is absolutely rigorous in the low signal case where quantum shot noise dominates; the results were incorporated into a calibration study and applied to the derivation of absolute backscatter values.

Application of these studies, and development of a detailed algorithm for rapid data extraction of weak signals from airborne atmospheric measurements, was reported with NASA supported colleagues [26], 1271. In a typical flight up to lo4 spectra would be generated with SNR varying by over four orders of magnitude. Rapid, flexible data processing was obviously required and this was provided by incorporating aircraft flight parameters into a six stage computer algorithm. In the presence of strong scattering and high SNR, Doppler signals could be located without ambiguity and the measurements were reliable and accurate. However, optimum performance under marginal conditions was very important; extensive studies of diagnostic criteria were made, including the interplay and tradeoff between SNR threshold and size of search window, and their impact on missed signal and false alarm rate.

PROCEEDINGS OF THE IEEE, VOL. 84, NO. 2, FEBRUARY 1996

111. TECHNOLOGY DEVELOPMENT 785 nm pump diode couplin 75 m 0.5 mm 1.5% output with fiber pigtail optics (GFh) etalon tuning) etalon (SLM) coupler , r=100mm 0100 pn \ TE(300K) \

A. Lasers-CO2 and Shorter Wavelength

There has been considerable research and development of lasers for coherent laser radar across Europe and several manufacturers, notably SAT in France and Laser Ecosse and Edinburgh Instruments in Scotland provide high performance CO2 lasers. Much early work on stabilization of CW lasers was done by Hall and Jenkins and their colleagues 1281. Recent CW work has employed slab waveguides with radio frequency (RF) excitation [29], [30]; an output power of 65 W in a stable mode has been obtained from a compact laser head that is -260 mm in overall length. A vast amount of work on the technology of TEA CO2 laser technology has been conducted by Willetts and Harris and their colleagues [3 11-1331 including frequency stabilization, gas catalysis and plasma effects. A particularly important paper [34] provided scaling laws for the intrapulse frequency stability of an injection mode selected TEA CO2 laser. Willetts and Harris showed that the frequency rose as the square of time at a rate varying linearly with energy and strongly dependent on spot size. The results were in accordance with a laser induced medium perturbative (LIMP) model [35], which allows the chirp in any TEA laser system to be predicted. A CO2 waveguide laser with programmable pulse profile was developed by Letalick et al. [36]. Polarization modulation was used with a GaAs coupler and a CdTe electro-optical modulator; the observed Q-switched pulse was in reasonable agreement with the calculated form.

In recent years short wavelength lasers have been investigated. At DLR the activities began with the development of high power, CW single frequency, diode pumped Nd:YAG lasers in 1988 [37]. A polarization rotation, “twisted-mode’’ technique has been applied to obtain single-frequency TEMoo output. On the basis of such lasers a coherent optical communication system was developed [38]. By placing in optical contact the components of a twisted-mode-cavity (TMC) laser, the advantages of the discrete laser resonator structure and the inherent frequency stability of monolithic lasers have been combined [39]. The free running monolithic integrated TMC laser exhibited a linewidth of -10 kHz averaged over 10 ms. A residual frequency drift between 10-500 kHz/s was observed. The frequency of a Nd:YAG laser has been locked to a fixed point on a flank of a Doppler- broadened Iodine absorption line [40] to give long term frequency stability of Because the concept can be realized in a relatively robust, compact, light and inexpensive arrangement, application to coherent free-space communication or lidar systems should be possible.

To extend the laser wavelengths to the eyesafe region beyond 1.45 mm, Er:YAG and Er,YbGlas lasers have been investigated [41]. Using several types of cavity design, single longitudinal mode (SLM) operation has been realized with both materials. Within the frame of the program optical detection of turbulence and windshear in the vicinity

output .I)

Br;w&er j 0 3 mm x 6 mm

Tm:YAG crystal (4 %) Haltwave plate

Fig. 3. Optical arrangement for a Tm:YAG stable laser under development at DLR.

of airports (ODIN), a master oscillator for a heterodyne solid-state Doppler lidar is being developed at DLR. For this system Tm:YAG as the laser material, emitting at the wavelength 2.02 pm has been selected; the optical arrangement is shown in Fig. 3.

At FOA, Sweden, the potential of diode based coherent laser radar 1421 for ranging and vibrometry and other applications is studied. The driving force is cost, low observability, and compactness but also the possibility to combine this source with optical fibers and achieve sensor distributed systems. The key parameter for obtaining range with semiconductor lasers is the ratio Play , the power to linewidth ratio. Diffuse targets have been ranged to 50 m with a 30 mW laser with a 10 MHz short-term linewidth. Ignoring atmospheric loss, the maximum range increases as ( P / A Y ) ~ . ~ which gives about 1 km for a 100 mW and 0.1 MHz laser, and about 10 km for a 1 W and 0.01 MHz laser. The atmosphere will of course reduce these ranges somewhat, but for close range applications there is obvious potential. The semiconductor laser under study was developed for coherent communication [43] and can be described as a three-section, strained layer, quantum well, distributed feedback, (DFB) laser in InGaAs-InP. The CW-power was 54 mW with a 3 dB linewidth of 280 kHz as shown in Fig. 4. Most experiments so far have been done with CW-powers around 20 mW and 1-10 MHz linewidth. Range and Doppler information is obtained by frequency shifting or chirping the laser. For lower resolution this can be done by internal tuning of the three section DFB-laser. Using an external cavity and a grating, substantially larger frequency sweeps with the corresponding larger range accuracy can be obtained.

In a lengthy study for the European Space Agency, Armistead et al. [44] reviewed the best coherent semiconductor laser sources for intersatellite links and related technologies.

B. Modulators, Detectors, and Signal Processing Acousto-optic modulators have been extensively used in

FM-CW systems which are discussed in Section IV-B. An improved modulator for CO2 heterodyne laser systems was described by Hulme and Pinson [45] employing acousto- damping layers of indium. Such a modulator was employed to provide a frequency shifted (55 MHz) local oscillator in a laser velocimeter [46]. Breakthrough due to parasitic scattering at the modulator drive frequency was about 60 dB


0 2

Frequency [MHz]

Fig. 4. Self-heterodyne spectrum of a DEB laser diode showing a 3 dB linewidth of 280 kHz. The CW output power was 54 mW WI .

greater than the local oscillator dominated noise threshold. In recent work at DRA notch filters with greater than 70 dB attenuation over +3 kHz bandwidth and less than 3 dB attenuation outside &15 kHz have been achieved. As an alternative to use of a gross shift (tens of MHz) of local oscillator frequency, a novel technique for determining the sign of the wind vector has been described by Hausamann and Davis [47] using a small additional frequency super- imposed on local oscillator or signal beam.

Detectors for coherent laser radars should fulfill the bandwidth requirements which for Doppler application might exceed 1 GHz. A particularly versatile range of detectors are manufactured by the French Company SAT [48]. Performance studies of a group of nominally identical detectors were carried out by Wilson et al. [49]. For one detector, near ideal behavior with up to 13 dB of local oscillator shot noise was obtained. This was illustrated with a graphical evaluation, which compared expected change of signal to noise with change of local oscillator noise. As shown in Fig. 5 good agreement was obtained. In another study [SO] the performance of a cadmium-mercury-telluride (CMT) detector was examined in order to determine the optimum local oscillator (LO) power and bias voltage. A simple model including frequency dependence of the signal and noise currents was supported by measurements. The dynamic performance of the CMT detector was investigated by mixing the radiation from a blackbody source with a CO2 LO laser and the quantum efficiencies for heterodyne detection and for direct-detection were compared. The heterodyne detection quantum efficiency (including mixing efficiency) was found to be about 65% of the quantum

210

Fig. 5. Change in signal to noise ratio SNR with local oscillator level. The full cnrves show the theorehcal change, A SNR as a function of local oscillator noise. For convenience this change of S N R has been evaluated relative to the SNR observed for a local oscillator noise of 1 dB (as m&cated by the dotted lines showing A SIW = 0 dB for LO,,,,, = 1 dB Curve (a) relates to the upper honzontal scale (0-15 dB) and curve (b) is expanded, and relates to the lower honzontal scale (0-1 5 dB) The individual data points shown as crosses have been measured using one particularly good detector and establish essentially ideal performance, over the range 0.1-13 dB of excess local oscillator noise [49].

efficiency for direct detection given by the manufacturer. A recent theoretical and experimental analysis of the performance of CMT detectors in heterodyne mode has been carried out by Oh et al. [51]. It shows that the optimal local oscillator power and relative quantum yield can be determined a priori when the diode characteristics are known.

Considerable attention has been given to signal processing with surface acoustic wave (SAW) spectrum analyzers combined with fast A D converters and integrators as illustrated in the following sections on applications. Use of such SAW systems has been applied not only to analog signals but dso for digital applications [52]. A fully developed on- line data system, based on a SAW spectrum analyzer, has been described by Kopp et al. [53] utilizing a peak-finder circuit, transient recorder and microcomputer to evaluate a temporal series of radial wind components. A profile of wind vectors up to 750 m height could rapidly be obtained.

C. Scanning Devices

The demands on scanners in laser radar systems vary of course very much with the mission requirements. These can range from imaging, terrain following, obstacle avoidance, tracking, or accurate pointing. The scanning solutions found for laser radars go from fully programmable and stabilized two-axis mirrors to wedge pairs for generating fixed or programmable rosette scans over a fixed fields of regard. Stabilized scanning mirrors were described by Sillitto [54] and a programmable stabilized scanner for a helicopter has been used in a depth sounding [55] laser radar. The above

PROCEEDINGS OF THE IEEE, VOL. 84, NO. 2, FEBRUARY 1996

local volume

::.p+ talibrution

0 V m m h - m

I ioke vortex dispersion

.... .. . . . . . . . . . . . . . . . . _ . _ _ . . . . _ _ _ _ I _ .___ _ _ l l _ _ r _ _ l . _ _ _ _ . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ . _ _ _ . __ . ,_____ . . ...

290m

Fig. 6. Strategy of wake vortex measurements at Frankfurt airport with the LDV placed between the two closely spaced runways 1661.

systems are scanning in front of the receiver optics, i.e., in the object space.

For high pixel rate imaging, scanning is often made in the image space. At FOA in Sweden galvanometer scanning in the image plane has been used with a specially developed laser-optical read out of the mirror position. The advantage of this was to provide very high accuracy angular coordi- nates while the scanning mirror was free running and not commanded as in other schemes. Fast scanning introduces fundamental problems, namely spectral broadening and lag angle losses. The degradation of a Doppler laser radar by a fast scanning mirror was investigated by Knopp [56]; a 0.3 mrad beam was scanned 6 mrad at frequencies from 0-80 Hz. It was shown that severe increase of the clutter level due to frequency broadening occurred above 5 Hz with a Doppler target at 4 MHz.

Renhom and Letalick [57] measured the signal degradation due to lag angle by illuminating a glint target at 2.7 km from the laser. The beam was scanned across the target at 92 Hz scanning frequency. The laser was running in the CW mode and the signal was measured with a boxcar averager. The relative signal degradation as a function of increasing lag distance (0.25 m at 2.7 km corresponded to a lag angle of 90 prad) was measured and showed good agreement with expectation.

Iv. GROUND BASED SYSTEMS

Over the years coherent lidar has been applied to many remote sensing tasks. As discussed in Section IV-A, atmospheric effects, where the laser scattering is from small airborne particles (aerosols), have been extensively studied in Europe. This has largely been with CW CO2 lasers operating at ranges up to 1 km; work at longer ranges with pulsed lasers is under development. Sensing of solid targets

has also been extensively applied to imaging, tracking and ranging, etc. and is discussed in Section IV-B.

A. Atmospheric Measurements An early comparison of different types of laser Doppler

velocimeter (LDV) for remote measurement of wind velocity was carried out by Hughes et al. [58]. They convincingly established the advantages of coherent heterodyne systems (almost exclusively CO2 at that time) compared with real- fringe, visible-wavelength systems based on an Ar+ laser, with further experimental confirmation [59]. In other early work [60], [61] based at Malvern, Brown et aZ. deployed a CO2 laser anemometer at the Drax power station and measured speed and turbulence at the different positions in the smoke plume under a variety of conditions [60]. The feasibility of obtaining the transverse velocity of aerosols in the probe volume of a coherent system was also demonstrated [62], from analysis of the Doppler signal envelope.

Extensive atmospheric studies and wind measurements were made at DLR with CO2 systems based on 30 cm optics and surface acoustic wave spectral analysis and an on-line data processing system [52] incorporating pattern recognition procedures. The influence of cloud and ground fog on CO2 CW laser Doppler wind measurements was discussed and the constraints these establish on the routine application of laser remote wind sensing [63]. In another paper [64] extensive measurements of the boundary layer wind profile up to 750 m altitude were made and compared with balloon sondes. An rms difference of 1.3 ms-' in magnitude and 12' in direction was found between profiles measured by lidar and sonde and attributed to atmospheric inhomogeneity. This question of representativity has been further explored theoretically and experimentally in a recent paper [65] that investigates the question of how many


EO

distance. from LDA [ml

Fig. 7. Propagation of E747 vortices generated on runway 2SL (upper part) and 25R (lower part) toward the parallel runways 1661.

25R

*-I- I - '1"

30 f---l

I

1180

Fig. 8. Polar plots of wind component at 5 heights recorded at night in September 1988 with a conically scanned LDV; three scans at 1 Hz rate were made at each height. The wind strength (1 scale division 2 ms-l) increased steadily with height, and the directional wind shear amounting to - 3 6 O is very obvious [67].

revolutions of a conical lidar scan are required in order to get the mean wind. Calculations of the mean wind velocity measurement errors for the surface layer under different types of thermal stratification and for the boundary layer under neutral conditions have been carried out. The theoretical conclusions are confirmed by the experimental results [65].

The DLR laser Doppler velocimeter work has been incorporated into the German Wake Vortex Program. This is based at the busy Frankfurt airport where additional

Fig. 9. Sequence of Doppler spectra showing the transit of a E757 vortex carried by the wind through a fixed lidar beam. The measurement rate was 2.5 spectra s-l and the horizontal frequency scale is &6 MHz. Note the weak peak from the tangent region to the vortex rising close to the top of the scale (31.8 ms-l) near the core [69], [70].

capacity limitations are caused by the small separation (-520 m) of the two parallel runways. This separation is often too small for operating both runways independently with respect to wake vortices. During several extended field experiments, in the years 1983-1985 and 1989-1990, the LDV has been operated at Frankfurt airport measuring more than 1400 landing aircraft in both heavy and large categories [MI. For investigation of vortices into ground effect the LDV container was positioned between the landing corridors about 850 m from threshold. As shown in Fig. 6 a section of the vertical measurement plane across one of the corridors is covered by a fast elevation scan at fixed range setting. After the vortex has passed through that sensing region, the next one is chosen by changing the range setting. As soon as the vortex has reached the lidar position the measurement plane is turned around and the vortex is tracked toward the parallel runway. The transport of vortices from one runway to the other under a prevailing cross wind is illustrated in Fig. 7. A number of such vortices show a steep ascent toward the parallel mnway. This bouncing effect may enhance the hazard since they tend to cross the parallel runway near the altitude of approaching aircraft. In October 1994 another campaign was carried out at Oberpfaffenhofen airfield for study of noncommercial aircraft. The vortices of fixed wing aircraft of medium and light category were measured-also from rotary wing aircraft.

Recent work at DFU Malvern has emphasized the development of compact and robust systems for wind field measurement [67]. One equipment with conical scanning has been housed in a Land Rover and widely deployed across the United Kingdom and Europe in support of site studies, helicopter landing simulations, comparison with balloon sondes [68], and rocket firing trials. This is a mobile and versatile equipment and typically may be brought into


CAA10057. A3QO. Ek90 dew LDV 7Sm: 3.8mJs 122 deg.

8 1054am 2/12/94.

16 The (6)

Fig. 10. Record of lidar spectra with color coded intensity due to the pair of vortices from a A300 aircraft, showing the characteristic cusp-like features [69], [70]. The Wake Vortex lidar was pointed vertically and the wind measured by a second LDV was 3.8 ms-' at 122' at 75 m height. The vertical scale is 0-6 MHz, corresponding to a flow speed scale from 0-31.8 ms-'.

operation within 5 min of arrival at a site. Various suites of measurement programs are available; Fig. 8 shows a typical polar plot of wind field at five heights completed in less than 30 s, and with immediate read out of horizontal wind speed, direction, and up/down draft.

Another CW LDV equipment at DRA Malvern has been developed for wake vortex trials. In a recent paper [69] the Doppler spectrum for a lidar beam interacting with a rotational flow structure was analyzed. It was shown that the important feature is a small discrete peak, well separated from the main body of the spectrum due to scattering from the immediate tangent region. When suitably plotted, as in Figs. 9 and 10 following, the characteristic cusplike features of such spectra were illustrated with a novel color display. Two campaigns have now been conducted [70] at Bedford Airfield in March 1993 and at Heathrow during the winter 1994-1995. In the latter trial the vortices of nearly 2000 aircraft have been recorded in a wide variety of meteorological conditions and from five different sites with aircraft at heights between 30-150 m above ground. Figs. 9 and 10 show two recent records with vortices carried by the

32

prevailing wind through a static beam. Most of the records have in fact been taken with a scanning arrangement with multiple intersections on the vortices. Fig. 11 shows the reconstruction of an unusual vortex trajectory from a B747 in which the vortex, with undiminished strength, returns close to the glideslope nearly 70 s after original passage of the aircraft.

A ground-based, pulsed coherent CO2 lidar for simultaneous, range-resolved, measurements of atmospheric constituents and wind velocity, has been under development at Laboratoire de Meteorologie Dynamique (LMD) in France and is now operational [71]. Range performance to -12 km in the horizontal and up to the tropopause in the vertical has been demonstrated as illustrated in Figs. 12 and 13. The LMD heterodyne lidar uses a monomode TE-CO2 at 10.6 pm laser. The pulsed output is expanded using a 17 cm diameter off-axis Cassegrain telescope and directed into the atmosphere by a two-mirror scanner. The backscattered field is collected by the same telescope and coherently mixed with a local oscillator. The frequency offset is about 30 MHz, and is controlled by an electronic feedback loop.


110-

100

90.

8 0

70-

60 - Metres

50-

40 -

30-

2 0

10.

I Metres

Fig. 11. Reconstruction from the lidar spectra of a vortex trajectory for a B747 aircraft arriving at Heathrow [69], [70]. The lidar was scanned f l O o at a rate of nearly 3' s-'. Note the initial descent of the near-wing vortex followed by ascent close to the glideslope -70 s later.

The mixing signal is focused on a high speed SAT MCT photomixer (at 77 K), then amplified and digitized at 100 MHz by an 8-b 9400 Lecroy oscilloscope and stored on a PC computer.

Several supporting studies for the LMD system have been conducted [72]-[76] and include analysis of the performance of an adaptive notch filter [72] and simulation in the time domain of lidar performance under nonstationary atmospheric conditions such as wind shear and fine layering of scattering properties [73]. For this a feuillet6 (slice) model was developed with successive thin sections of the atmosphere. In another study 1741 the relationships among heterodyne efficiency y, number of speckle cells M , and the ratio of receiver area to coherence area SE/Sc have been explored through the use of mutual coherence functions. A new equation with M = (1 + SR/Sc) was derived and the optimal number of M N 4 for a monostatic system suggested. Other studies include a semianalytic approach to system efficiency [75] and analysis of multiple scattering in cirrus clouds and heterodyne detection [76].

B. Solid Target Measurements Many coherent laser radar demonstrators, mostly for

military applications in imaging, tracking and ranging have

Centre Line

been built in Europe. These systems have largely been based on CO2 laser technology. A great number of tests have been reported with FM-CW systems at several Euro- pean centers for velocity and ranging applications as well as imaging [4], [22], [77]-[80]. Some systems have also been dedicated to vibration sensing [81], [82]. Several of the pulsed systems have been built for ranging studies with TEA lasers [83], [84]. For imaging applications pulsed systems have also been studied in the form of feasibility/technology demonstrators [3], [23], [36]. An early study of heterodyne detection of CO2 TEA laser pulses was conducted by Taylor and Davies in 1977 [85]. Hulme et aZ., in 1980 [86] demonstrated chirp pulse compression with simultaneous measurement of range (to 10 m accuracy) and radial velocity (with accuracy of better than 1 ms-I), and also reviewed rangefinders and lidar application to solid target measurements [87]-[89].

Three ground system demonstrators have been developed in Sweden. The first, an experimental FM-CW CO2 system, was built for target and background characterization. In another system a pulsed laser radar with a programmable transmitter was built to study primarily the combined capability of Doppler and range imaging. The main parts of the optical system are shown by the block diagram in


2t

5 I . /' 4 6

10 10 10 10 Square range corrected signal

5

1 I

10 5 10 15 0 Range (km)

Fig. 12. Atmospheric measurements with the pulsed CO2 laser radar at LMD. Upper: vertical profile in the troposphere, laser output energy 140 mJ per shot, PRF 4 Hz. The square range corrected signal (average of 400 shots) is plotted as a function of altitude. The planetary boundary layer height (PBL) is -1.7 km and semi-transparent cirrus is observed between 4-8 km. Lower: horizontal profile in the PBL, laser output energy 250 d, PRF 4Hz, 200 shot average signal. The lidar line-of-sight is -30 m above ground.

Fig. 14. These are the transmitter, a local oscillator laser, an interferometric block including a thin film polarizer, a beam expander, a pair of galvanometric scanners, a telescope, and a high precision pointer. The laser was a programmable waveguide [36] with intracavity CdTe electroptical modulator with mean transmitter power of 2.2 W and PRF 0-70 kHz. The image field of view of the lidar system was 24 mrad and this was scanned with 100 x 100 pixels at 1 Hz frame rate and range performance to 3 km.

Extensive studies have been conducted at FOA with these ranging, imaging and Doppler systems. With good range resolution a target is easily detectable from range variations. In this regard a new technique has been investigated [90] for terrain segmentation based on range data. The basis of

Histograms of backscattered power

Gate of 150 m

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5

Gateof 300m

M3.69

' "0 0.5 1 1.5 2 2.5 3 3.5 4 4.5

Gale of 800 m

I i

U - 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5

Normalized heterodyne power at 3 km

Histograms of wind velocity estimation On a gate of 150 m

snC I 4 ""

-4 -2 0 2 4 6

m On a gate of 300 m 1

On a gate of 600 m

-4 -2 0 2 4 6 Estimation of wind velocity at 3 km

Fig. 13. Simultaneous measurement at 3 km range of the probability density functions (PDF) of backscattered power (upper) and radial wind velocity (lower) as a function of processing range gate (150 m, 300 m, 600 m). The histograms represents 200 lidar shots. For the power PDF the solid line is the theoretical x~ function, where M is the number of time speckles. M is linked to the ratio of the processing range gate divided by the atmospheric correlation time. It is an experimental parameter given by: A4 = (P/v.p)', where P is the signal mean value, and c p the standard deviation. The number of speckles increases when the processing range gate increases. For the wind velocity PDF the solid lines represent the theoretical Gaussian function using standard deviation ou = 2.221.64 ms-I when the range gate increases from 150-600 m, respectively.

the approach is to model the range values obtained from a horizontal and vertical scan as a piecewise constant (or linear) signal in random noise. The segmentation problem is then to detect the pixels at which the piecewise constant signal changes value. For the detection of these changes a technique based on Kalman filters is used. Automatic segmentation into a physically meaningful region is ob-


I

I I I I

; Pointing I direction

PL (-,

@)

Fig. 14. Block diagram and photo of the optical part (pointer missing) of the pulsed imager.

tained. The computational requirements of the algorithm are modest since the vertical and horizontal directions are segmented independently, after which the results are combined by post processing. The formalism presented is easily extended to Doppler imaging and multisensor data. Note that the data from the segmentation process can be used for further processing e.g., by adding reflectance and other data. Fig. 15 illustrates automatic target detection after segmentation.

In recent studies at FOA the capabilities of diode lasers have also been investigated. Based upon earlier findings a second and more compact version has been built and is presently under evaluation for ranging and vibration studies. This unit [42] is shown schematically in Fig. 16. The InGaAsP-InP semiconductor laser is the three section, strained-layer, quantum well, distributed feedback (DFB) laser [43] described in Section 111-A. All the parts except for the telescope (not shown) are integrated into a single block. Collimating with minimal phase front distortion is

important in order to improve on the useful laser power. This is done by an aspheric lens such that minimal wave- front distortion is achieved. After passing the polarizing cube beamsplitter and the Faraday element, which rotates the plane of polarization 45", the major portion of the beam exits the radar. A small part (4%) of the beam is reflected by the partial reflector and serves as LO. Before mixing of the target and LO beams on the detector, both pass the Faraday element once more and are reflected by the cube beamsplitter. In this way, the important optical isolation of the laser is accomplished. The detector is a 25 pm InGaAs Schottky photodiode. The telescope expands the laser beam diameter to what approximately corresponds to the lateral coherence length in a normal atmosphere, i.e., 3-5 cm. Range and Doppler information is obtained by frequency chirping or shifting the laser-employing the internal tunability of a three section DFB laser. The simplicity of the method has to be weighted against the limited resolution due to nonlinearity of the sweep. When


Fig. 15. Automated target detection in range imagery using the algorithms in [90]. (a) Video image; (b) video and color coded laser radar data; and (c) automatically detected targets.

aser

beamsplit t er

I 40 mm i

Fig. 16. Schematic of the compact diode laser radar (telescope not shown) under development at FOA in Sweden 1421.

frequency shift keying is being used, the modulation frequency and the corresponding phase shift are measures of the range. Standard processing schemes can be used for resolving ambiguity ranges. The device is now under test for ranging and vibration studies.

At FOA, performance prediction for coherent detection has been done using the work on signal statistics and fluctuations described in Section 11-A. Effects due to glint, pure speckle, and semi-rough targets have been studied, together with atmospheric turbulence, terrain background (i.e., target to clutter ratio) at different carrier to noise ratio (CNR), laser wavelength, and weather conditions [ 5 ] . Such performance calculations may compare 1-2 pm systems with CO2 systems and coherent with direct detection. One example is shown in Fig. 17, where the utility range for a coherent rangefinder for ground applications has been plotted for different wavelengths [91]. A CMR of 16.5 dB has been assumed and certain common parameters for the different lidar systems (e.g., 0.1 J pulse energy, 10 cm optics) .

V. AIRBORNE APPLICATIONS Several airborne systems have been built in Europe or

are currently under construction. These are outlined in the following subsections on technical development, avionic applications, hard target measurement, and atmospheric studies.

A. Technical Development of Airborne Systems The first airbome coherent lidar in Europe was the Laser

True Airspeed System (LATAS) designed and built at the

Royal Signals and Radar Establishment (RSRE) in the late 1970’s and flown in aircraft of the Royal Aircraft Establish- ment (RAE) (both now part of the UK Defence Research Agency (DRA)). LATAS, which has been described in several publications [92]-[94], was designed to be compact, robust, and with the lidar head operating remotely in an unpressurized part of the aircraft. The lidar was based on a 4 W CO2 waveguide laser and comprised a 15 cm germanium output lens, polarizing optics with quarter and half wave plates and a 100 MHz CMT detector cooled with high pressure air and Joule-Thomson minicooler. The laser was maintained on the P20 transition with a solid Ctalon filter and feedback loop to a piezo stack on the laser. The optics head was thermally insulated and a single 100 W heater mat and fan maintained uniform conditions with external temperatures down to -70°C. The weight of the optics head was about 24 kgm and size 70 x 30 x 30 cm. Transmission of the outgoing beam was via a 20 cm diameter Germanium window with diamond-like surface coating designed to withstand abrasion, insects, and rain drop impact. Signal processing was in a 25 MHz bandwidth surface acoustic wave spectral analyzer with four overlapping switchable ranges extending up to 62.5 MHz. Signal recording was in a fast integrator and tape recording modulator data acquisition system. A single data sample in the SAW was 25 ,us long and for 100 integrations or samples the rate was about 160/s; for very low signals up to 64 000 integrations could be accumulated for each measurement sample. In practice, throughout nearly 15 years of flying, the equipment proved extremely reliable during far-flung trials around Europe, the United States, and North and South Atlantic (See Section

Another robust, compact system [95] was built in the early 1980’s by Crouzet SA in France. This too was based on a well established 3 W CW CO2 waveguide laser operating on the P20 line. Detection was at a CMT diode with over 200 MHz bandwidth. Polarizing optics with wave and Brewster plates was incorporated, and the local oscillator path included double passage of a Bragg cell to provide a frequency shifted signal for helicopter operation. The focusing optics was an off-axis Dall-Kirkham telescope with effective aperture of 7.5 cm to allow measurements from 10-100 m range. Signal processing was by SAW spectrum analyzer to provide an airspeed range of -25-+400 ms-’. The weight of the complete system was 250 kgm.

V-B).


75 , 0 I q 60

e-.. . I : '*. : *.

'%--C-=* Coh-vegetation

55 -I I

Wavelength pm

4 ! 1 I I I I 0,oO 230 5,OO 7,550 10,OO 12,50

Wavelength pm

(b)

Fig. 17. Example of calculated mean ranges for a 0,l J rangefinder with 10 cm optics against terram (green vegetationhow) and a 1 m2 target (military white/geen) during (a) summer and (b) winter. Measured atmospheric transmission and turbulence from the middle part of Sweden have been used for performance estimates for different wavelengths using either b t or coherent detection.

This airborne lidar has subsequently been developed by Sextant Avionique into the ALEV3 Air Data Calibration equipment [96] which can be fitted in a movable pallet for transport aircraft or helicopter, or in a pod for fighter aircraft. The output beam is switched into three orthogo- nally directed, beam expanders with 75 m useful optical diameter. A Bragg cell may be introduced into the reference ann of the interferometer to give the sign of the airspeed; the optical unit fits into a cylinder of length 650 mm and diameter 300 mm.

An advanced 10 pm pulse heterodyne system is currently being developed, primarily for obstacle warning and terrain following, in the Coherent Laser Airborne Radar (CLARA) project 1971. In this Franco-British, gouernment- to-government initiative, a consortium comprising Dassault Electronique and the Radar Systems Division of GEC Marconi is producing two technology demonstrators to be flown on rotary and fixed-wing aircraft. In addition to its primary roles, CLARA will also be capable of perfomiing functions such as targeting and true air speed measurement. A pod mounted configuration has been adopted and the laser radar comprises three main subassemblies: the sensor head, the scanner and signal, and data processor. The scannet assembly is required to address several quite different modes of operation and is of relatively complicated design. Detection, classification, and real time display of obstacles including power, telecommunication, and barrage balloon cables, masts, guy wires, buildings, and trees must be achieved in daylight, at night, and in adverse weather. In order to ensure that suitable warnings continue to be provided when the aircraft is turning, a large field of regard, the size of which is governed by the flight envelope of the host aircraft, must be adopted for the sensor.

The advanced airborne system Wind Infrared Doppler Lidar (WIND) for atmospheric studies is currently be-

ing developed in cooperation between CNRS and CNES in France and DLR-Munich in Germany [98]. Technical flights are expected by the end of 1996. Currently, a complete instrumental model has been built to evaluate performance and tradeoffs between the various su The equipment is designed to fly at 10 km alti wind field is sampled by a conical scan at 30' from nadir with a 20 s or 10 s scanning period and a 4 Hz or 10 Hz laser pulse repetition frequency. For the project a compact single-mode TE-CO2 laser has been by Soci6tg Anonyme de T616communications ( laser design has been described by Delville et al. [99] and briefly comprises a short Gaussian reflectivity output coupler to discriminate against high order transverse mo and a large cross section to reduce intrafrequency sweep' Good performance characteristics have been demonstrate with m e a pulse energy of 360 mJ at 4 Hz PW and satisfactory pulse shape, beam profile, and frequency chirp. The plagorm for the WIND instrument is the Falcon 20 research aircraft operated by DLR; the arrangement of the scanning lidar above the floor is shown in Fig. 1 telescope is an off-axis Dall-Kirkhm design, is athermalized and ultralight weight, and has a 20 c aperture and an expansion ratio of x 15. The focusing range is adjustable from 200 m to lidar infinity.

The WIND concept is based to some extent borne CW-Doppler lidar (ADOLAR) developed at DLR in 1994. In this equipment the interferometer optics and laser was mounted on one side of an optical breadboard 900 x 300 mm and the telescope was fixed on th side. Conical scan about the nadir was performed Germanium wedge scanner.

B. Airbome Lidar Measurements 1) Avionic Applications: During the 1980's the LATAS

system was housed in the nose of an HS125 (executive jet)

218 PROCEEDINGS OF THE IEEE, VOL 84, NO 2, FEBRUARY 1996

/

Fig. 18. Arrangement of the lidar components above the aircraft floor window as planned in the Franco-German WIND project [Y8].

aircraft and used extensively for free stream airspeed and wind shear measurements; the potential for pressure error calibration of aircraft was also demonstrated. The signal processing system, with 60 kHz resolution and 30 kHz channel width, provided a potential speed precision of better than 0.2 ms-I (-0.1 kn). For true airspeed measurement and pressure error calibration the lidar was typically set to focus between 30-100 m ahead of the aircraft. For wind shear measurements this was extended to 250-300 m ahead. At this longer focal range lidar sensitivity extends out to 700-800 m and thus strong shear or turbulent structures entering the extended probe volume at longer range are evident. Simulations had shown that a focal range of -300 m provided useful warning [93], [94]. If the wind is measured too far ahead of the aircraft then any shear can change with time and thus potentially provide a significant amount of false or misleading data. Simulations showed that there was a significant advantage in controllling the aircraft (a medium size passenger jet) using the airspeed measured by the lidar at 300 m ahead, but increasing the distance to 600 m produced no further improvement.

In July 1982 the LATAS system in the HS125 took part in the Joint Airport Weather Studies (JAWS) project in Colorado. This project was sponsored by the US Na- tional Center for Atmospheric Research, the University of Chicago and the Federal Aviation Agency. In the course of a three week trial the HS125 flew through many severe microbursts associated with the frequenl thunderstorms

JAWS TRIAL"

DOPPLER S P E m AT

GROUND SPEED "IOmP-' 0.5s INlERVALS.

- 25.5" - (--)

Fig. 19. Sequence of lidar spectra recorded at 0.5 s intervals by the LATAS lidar during passage through a thunderstorm microburst at the JAWS trial in Colorado in July 1982 [Y2], [93].

prevalent at that time of year. Fig. 19 shows a sequence of spectra in a lidar record in which the headwind changed by over 40 kn (-20 ms-') and there was a strong downdraft of 1200 ft/min (-6 ms-l). These measurements contributed to the development of a descending vortex ring model for microburst behavior, in contrast to the more usual vertical jet model. The vortex ring model also explains several features observed in the JAWS flights such as dust curtains rising to over 1000 ft (300 m) around the perimeter of several microbursts. By following the sequence of spectra it is possible to see changing wind elements entering and leaving the extended probe volume of the lidar.

In another short program with colleagues at NCAR, single axis LATAS measurements at -17 m distance ahead were compared with a differential gust probe and showed accurate turbulence spectra at frequencies up to 10 Hz. This study [loo] was aimed at investigating the potential for a conically scanned lidar equipment.

The CW Doppler lidar of Crouzet [95] was similarly used for true airspeed measurement and was flown on a variety of aircraft including a Puma helicopter (in 1984), Caravelle transport aircraft (1985-1986) and Mirage fighter (1986-1987). As configured in the three-axis ALEV3 equipment [96] Aerospatiale Flight Test Division has oper-


ated two such systems since mid 1991, accumulating more than 1200 h of flight in two years on various Airbus aircraft including A320, A340-200, A340-300, A330, and A321. The ALEV3 gives a precise real time measurement of the airspeed vector allowing the calibration of static pressure error and of the angles of incidence and sideslip even during dynamic phases [96]. After proving during a three month test period on an A320, the lidar equipments now serve as a reference for certification of other aircraft; they are thus established as a precise and reliable Air Data Calibration tool for aircraft certification in flight test centers [96].

Another airborne avionics project called Future Laser At- mospheric Measurement Equipment (FLAME) is currently supported by the European Community. Feasibility and benefit of a sensor for active wake vortex detection are being studied. Encounters with dangerous wake vortices generated by a preceding aircraft might be avoided if the following aircraft is equipped with a lidar sensor. A compact airborne Doppler lidar based on 2 pm laser technology is being investigated for timely detection of wake vortices and other parameters relevant to air safety. The FLAME consortium consists of Sextant Avionique (France), GEC Marconi (UK), DLR and University of Ham- burg (Germany), University College of Galway (Ireland), INESC (Portugal), and ALPHA (Greece).

2) Hard Target Measurements: As noted the CLARA system [97] is currently being developed for hard target (cables, ground surface, etc.) measurement. This advanced equipment follows successful trials of the LOCUS (Laser Obstacle and Cable Unmasking System) pulsed CO2 laser radar jointly developed by two groups within GEC Marconi. The LOCUS system was flown on A6-E, HS748, and Tornado aircraft and the technology to detect power cables and display them to aircrew was demonstrated.

In another report [ lo l l SFENA demonstrated an FM- CW laser radar for terrain following and terrain avoidance of combat aircraft. A scanning pattern of 4000 dots was generated by a dual-wedge system at a maximum scanning frame rate of 2.5 Hz. A compact engineering model un- derwent flight evaluation trials on a Puma helicopter at the French flight test center in 1985. Lidar data of the terrain flown over by the aircraft was processed to create 10 x 10 km altimetric maps centered on the aircraft and divided into 40 m square primary cells.

The LATAS airborne lidar has also been used for ground imaging [lo21 with the beam aimed downwards at 60" from horizontal and linear scanning over 4 3 " to the side. The rapid linear scan was achieved with two synchronized mirrors. Processing of the signal envelope gave intensity images of natural terrain and man-made objects with significant features clearly evident.

3) Atmospheric Studies: An airborne lidar provides on ideal tool for widespread investigation of atmospheric properties. In this regard a considerable fraction of the flying time with the LATAS airbome lidar was devoted to backscatter measurements. As noted, much attention had been paid to calibration [24], [25] of LATAS and also the development of algorithms for the rapid processing of

Doppler spectra and calculation of atmospheric backscatter coefficients [26], [27]. During the JAWS trial in Colorado the opportunity was taken to compare measurements from the airborne LATAS CW lidar with those from the ground- based, pulsed, CO2 lidar of NOAA's Wave Propagation Laboratory. Seven pairs made during a 20 day p regions of uniform backscatter the two lidars showed good agreement with differences usually less than -50% near 8 km altitude and less than a factor of 2 or 3 elsewhere, with the pulsed lidar data often lower than tae CW lidar. Near sharp backscatter gradients the two lidars showed poorer agreement, with the pulsed lidar usually higher than the CW lidar. These were explained in terms of atmospheric and in- stmmental factors, e.g., spatial resolution [103]. Backscatter structure was also explored; for quiescent meteorological condition the transition from the planetary boundary layer (PBL) to the background layer was often very sharp with backscatter decreases as large as three decades in -70 m. Sharp gradients were also found at the boundaries of shallow (tens of meters) subvisible cirrus clouds [104].

In 1987, with the assistance of NASA supported colleagues, over 30 h of airborne measurements of atmospheric backscatter were reported for four different regions of the Northern Hemisphere outside the UK [105]. The results exhibited great diversity but supported the general conclusions that: 1) airborne laser radars for the measurement of true airspeed and wind shear detection and warning at low levels would have good reliability and 2) a spaceborne lidar for global wind field measurement would provide reliable information for a large fraction of the time. In further studies of airbome data, the results of 40 vertical profiles obtained over the UK during June 1981-October 1983 were compared with a large number of profiles recorded by the WPL groundbase lidar over Boulder, CO [106]. Histogram distributions of the data were created with the backscatter mixing ratio ,B/p (backscatter coefficient divided by air density) sorted into 0.5 and 1 km height intervals and half and one-third decade classes of loglo ( p / p ) . For both sets of data a well defined, clear background'mode at ~ 1 0 - l ~ m2 kg-' sr-l (note units of backscatter coefficient divided by air density) was apparent in the middle and upper troposphere. At lower levels there was a more variable convective branch determined by convective boundary layer activity and surface aerosol flux. An upper level stratospheric branch varied with volcanic activity and the rise and fall of the tropopause. The existence of a ubiquitous background mode has considerable implications for ghbd scale aerosol models and is also important for the design and performance simulations of prospective satellite-borne lidar systems.

In 1986, the attempt was made to relate such occasi atmospheric backscatter measurements to the regular extinction measurements, ..(A), made by the SAGE I1 limb sounding satellite. On three successive days, April 19-21, SAGE II measurements were made at sunset near the UK [107]. The LATAS system was flown as close as possible to the atmospheric paths seen by the satellite. Analy- sis of the measurements showed that individual ratios of

PROCEEDINGS OF THE IEEE, VOL. 84, NO 2, FEBRUARY 1996 220

@(T, 10.6pm) and 41 .02 p m) were in reasonable accord with previously calculated values. However, the trend with height had a nonlinear relation, probably attributable to a steadily changing size distribution of aerosols, emphasizing the difficulty of assigning likely backscatter coefficient to observed extinction values without further qualification in terms of altitude, season, and geographic location [107].

In 1988 the LATAS lidar was installed in a forward looking pod in the bomb bay of a Canberra B57 aircraft to provide a greater altitude capability (to 16 km) and also greater range. In collaboration with colleagues of the USAF Phillips Laboratory, extensive studies of atmospheric aerosols were made over the Atlantic in the South Atlantic Backscatter Lidar Experiment (SABLE) [lo81 and the Global Atmospheric Backscatter Lidar Ex- periment (GABLE) programs of 1988-1990 (the measurements were thus contemporaneous with the NASA GLOBE measurements over the Pacific). Over 180 h of flight measurement data was recorded out of Ascension Island (8S, 14W), the Azores (38N, 25W), Iceland (63N, 23W), and from the UK over the North East Atlantic. In addition to LATAS the equipment deployed included: airborne particle sounding probes, ground based lidars (operating at 10.6, 0.53 pm, and 0.35 pm), balloon radiosondes, and sun tracking photometer [ 1091. Standard meteorological information was incorporated along with, when appropriate, data from the SAGE I1 limb sounding satellite.

In a lengthy paper [ 1101 the airborne backscatter measurements at 10.6 pm, are gathered together and presented. Plots of backscatter versus altitude are shown for 180 flight hours, distributed in altitude (1 km) and backscatter (half decade) intervals, and presented in tabular and histogram form for six principal regions and seasons: South Atlantic summer, South Atlantic winter, Far-North Atlantic spring, Mid-Atlantic springhummer, and North-East At- lantic winter and summer. Several features of the results are discussed including the incidence of low backscatter and dropouts, and the remarkably high incidence of thin cirrus in the southern tropical winter season-see Fig. 20. The compendium should provide a climatology of atmospheric backscatter at 10.6 pm for regions of the Atlantic during the relatively clean atmospheric period 1988-1990. Other material in preparation from SABLE and GABLE in- cludes comparison with SAGE I1 extinction measurements, fluctuations in backscatter values [ 11 11, incidence and properties of subvisual cirrus, transits through boundary layers, and other occasional, strongly scattering, layers, and comparison with the ground based lidar and sondes.

The CROUZET airborne lidar [95] has also been used for atmospheric backscatter measurements. In 1987 the CO2 laser anemometer was flown on a Mirage I11 up to altitudes of 13 km. During the period of May-October, 10 flights were made over France in different meteorological conditions [112]. During the various flights P(T, 10.6 pm) was larger than 1 x m-l sr-l for altitudes up to 12 km in both polar and oceanic air masses.

As already noted the Franco-German WIND project [98] has a twofold objective: to make a significant contribution

1

1'

1:

11

8 8

!! i 6

4

2

0 -

Fie. 20.

- ,...,, , , ......, . , ...... , . .,.,, , , , , I . -......I

I . . . , . , , . I . , . . . . . . , . , . . . ,

! -11 -10 -9 -8 -7 -6 -5

BACKSCATTER COEFFICIENT p (r ,10.6p") m "sr"

Record of backscatter measwement versus altitude (to 52-800 ft.) made during the SABLE trial, Flight 41, out of Ascension Island on 10/7/1989. On this occasion a very thin veil of cirrus was evident from the ground, but proved to he over 3 km deep extending to 15 km altitude, with P ( T , 10.6 pm) rising above 10W6 m-l sr-l. In the mid levels (6-12 km) scattering levels fell on occasion below the LATAS threshold with p(., 10.6 pm) 2 8 x m-l sr-' [108]-[110].

to mesoscale meteorology, and to act as a precursor for spacebome projects as currently foreseen in Europe. In a recent paper [ 1 131 an Instrument Simulator for the program is described. It considers the parameters of a conically scanned airborne heterodyne CO2 lidar (Fig. 18) and inputs instrumental parameters and atmospheric processes during the pulse duration such as wind speed, turbulence, and shear. Factors affecting representativeness and performance are discussed.

VI. EUROPEAN SPACE LIDAR ACTIVITIES Spaceborne lidars have the potential to make significant

contributions to meteorology and climatology because of their ability to provide global data on a variety of atmospheric observables such as height, horizontal extent and density distribution of scattering layers (e.g., clouds and aerosols), vertical humidity, temperature and pressure profiles, as well as wind fields in the troposphere and lower stratosphere. European lidar technology has made considerable progress in the last decade. There is thus a strong scientific, technological and industrial base within Europe from which space lidar systems can be developed and several workshops, technical studies and papers [114]-[ 1171 have addressed the issues.

VAUGHAN et al.: COHERENT LASER RADAR IN EUROPE 22 1

The European Space Agency has set up working groups (Science Advisory Groups) composed of scientists, users and laser instrumentation specialists, with the task of advis- ing the Agency on requirements and development priorities in the area of laser remote sensing. Four lidar systems have been considered [118] as good candidates for space deployment: 1) a simple backscatter lidar, 2) a differentid absorption lidar (DIAL), 3) a wind-profiling lidar, and 4) a ranging and altimeter lidar.

Priorities were given to Atmospheric Lidar (ATLID), the backscatter lidar, and Atmospheric Laser Doppler Instru- ment (ALADIN), the wind lidar [I 191. For the latter, studies were proposed by the Science Advisory Group. Fig. 21 shows the interrelation between these proposed studies, with the primary objectives being the identification of various parametric requirements for the technical performance of the CO2 laser Doppler system and the identification of parameters to the mission from the user point of view. The results of these studies are to be used as a basis for M e r design. As apparent in Fig. 21, two major categories each with a few working packages were identified: 1) user related studies and 2) technical studies.

Concerning the wind profiling lidar, two pre-phase A studies for ALADIN were initiated at the end of 1993 [117], [120], [121] and have now been completed. ESA is &Q

supporting the development of key technologies. Signifi- cant progress has been made with a breadboard electron- beam sustained COZ-laser for this application (Study 3 0 in Fig. 21) [122] and also a switchless TEA-CO2 laser [123]. This work has encompassed a variety of technological disciplines including electron-beam generation, laser-head engineering, advanced resonator design, low temperature gas catalysis, and high-voltage engineering. ESA has also funded a study in which laser and receiver specifications were determined for both DIAL and Doppler wind. For such combined use the technical solution suggested for the transmitter laser was for a laser-diode pumped TmHoYAG or YLF laser emitting at -2 pm wavelength in an injection seeded master oscillator/amplifier configuration.

For ALADIN a simulation study (study A3 in Fig. 1) on the representativity of a single line-of-sight component was also carried out using data from the NOM-WPL ground based Doppler lidar. The main question to be answered was how representative is a single line-of-sight component? It was shown [ 1241 that representativity depends mainly on the atmospheric aerosol loading and the laser stability. Suf- ficient aerosols and laser pulse power produce a processable signal. Averaging and/or selection of useful information out of a speckle signal is necessary.

A B USER RELATED TECHNICAL

STUD1 ES STU DIES

Fe. 21. Block diagram of the studies proposed in the ESA report [118]. A1 Doppler Lidar Performance, A2 OSSE, A3 Components, A4 Sampling Strategies, A5 Data Dissemination, A6 Synergism, BO Laser, B1 Lag Angle, B2 Accuracy, B3 Parametric Analysis, B4 Instrument Requirements.

the 9-12 pm band have historically offered high efficiency and been widely used. It is, however, gratifying that solid- state and diode lasers at wavelengths greater than 1.5 pm are now becoming available with the required coherence properties. They should provide a valuable extension to lidar capabilities [ 1251-notably in their potential for pow- ering extremely compact systems. This is particularly true for hard target applications where there is strong motivation to reduce size and costs. However, the greater atmospheric turbulence degradation at the shorter wavelengths, and greater loss due to scattering, remain partly unresolved issues; CO2 also has some advantage of compatibility with thermal viewers. For the future, combined laser diode and detector arrays also offer exciting prospects, as oes the

coherent laser radar has developed strongly in Europe during the last two decades and, with many active programs, should continue to do so.

possibility of optical synthetic aperture lidar. In 2 summary,

VII. SUMMARY AND CONCLUSIONS The strength of coherent laser radar in Europe has been

illustrated with programs that extend from fundamental physics to support for spaceborne technology, and include several advanced ground and airborne systems for civil and military applications. For lidar operation outside the laboratory, eye safety is a paramount concern; CO2 lasers in

ACKNOWLEDGMENT The authors are greatly indebted to their many colleagues

worldwide for their stimulus and cooperation over the years, and express particular thanks to Michael Harris, Christopher Hill, and Guy Pearson of DRA (Malvern) and Milton

PROCEEDINGS OF THE IEEE, VOL. 84, NO. 2, FEBRUARY 1996 222

Huffaker of STC (Boulder, CO) for their helpful comments on this manuscript. Dec. 1983.

heterodyne systems,” Appl. Opt., vol. 22, no. 23, pp. 3787-3795,

[25] J. M. Vaughan, R. D. Callan, D. A. Bowdle, and J. Rothermel, “Spectral analysis, digital integration, and measurement of low backscatter in coherent laser radar,” Appl. Opt., vol. 28, no. 15, DD. 3008-3014. AUB. 1989. REFERENCES

B. J. Rye, “Antenna parameters for incoherent backscatter heterodyne lidar,” Appl. Opt., vol. 18, no. 9, pp. 1390-1398, May 1979. - , “Refractive-turbulence contribution to incoherent backscatter heterodyne lidar retums,” J. Opt. Soc. Amer., vol. 71, no. 6, pp. 687491, June 1981. H. Ahlberg et al., “Imaging Q-switched CO2 laser radar with heterodyne detection: Design and evaluation,” Appl. Opt., vol. 25, no. 17, p. 2891, Sept. 1986. D. Letalick, I. Renhom, and 0. Steinvalll, “Measured signal amplitude distributions for a coherent FM-cw CO2 laser radar,” Appl. Opt., vol. 25, no. 21, p. 3927, Nov. 1986. -, “Target and atmospheric influence on coherent CO2 laser radar performance,” Appl. Opt., vol. 25, no. 21, p. 3939, Nov. 1986. J. M. Vaughan et al., “Correlation analysis of fluctuations in coherent lidar signals at 10.6 pm,” J. Mod. Opt., vol. 38, no.

M. Harris, G. N. Pearson, C. A. Hill, and J. M. Vaughan, “Higher moments of scattered light fields by heterodyne analysis,” Appl. Opt., vol. 33, no. 30, pp. 7226-7230, Oct. 1994. M. Harris, G. N. Pearson, and J. M. Vaughan, “Electrical method of spectral fiItering within the Schawlow-Townes laser linewidth,” Elec. Lett., vol. 30, no. 20, pp. 1678-1679, Sept. 1994. G. N. Pearson, M. Harris, E. Jakeman, and D. Letalick, “Spec- tral filtering of light processing non-Gaussian statistics,” J. Mod. Opt., vol. 41, no. 11, pp. 2067-2077, 1994. M. Harris, G. N. Pearson, C. A. Hill, and J. M. Vaughan, “The fractal character of Gaussian-Lorentzian light,” Opt. Comm.,

4, pp. 623-648, 1991.

- vol. 116, pp. 15-19, 1995. Proc. 1st Workshop on Opt. Ampl$ers for Lidar Applic. (unpub- lished), J. M. Vaughan and C. L. Werner, Malvern, UK, Nov. 1992.

- vol. 116, pp. 15-19, 1995. Proc. 1st Workshop on Opt. Ampl$ers for Lidar Applic. (unpub- lished), J. M. Vaughan and C. L. Werner, Malvern, UK, Nov. 1992. P. Richter, I. Peczeli, and S. Borocz, “Coherent infrared lidar with enhanced optical heterodyne detection,” J. Mod. Opt., vol. 41, no. 11, pp. 2079-2084, 1994. S. Rahm, C. Loth, A. Dabas, and S. Elouragini, “Improvement of heterodyne detection with optical amplifiier and pulsed coherent Doppler lidar,” J. Mod. Opt., vol. 41, no. 11, pp. 2145-2152, 1994. R. J. Morley et al., “Enhancement of coherent laser radar performance by predetection amplification,” Appl. Opt., vol. 33, no. 18, p. 3951, June 1994. M. Harris, R. Loudon, G. L. Mander, and J. M. Vaughan, “Above-threshold laser amplifier,” Phys. Rev. Lett., vol. 67, no. 13, pp. 1743-1746, Sept. 1991. M. Harris, R. Loudon, T. J. Shepherd, ;and J. M. Vaughan, “Noise cancellation in laser emission,” Phys. Rev. Lett., vol. 69, no. 16, pp. 2360-2363, Oct. 1992. -, “Optical amplification and spontaneous emission in an Ar+ discharge,” J. Mod. Opt., vol. 39, no. 6, pp. 1195-1203, 1992. -, “Noise and gain in sub-threshold laser modes,” Opt. Comm., vol. 91, pp. 383-389, 1992. M. Harris, C. A. Hill, and J. M. Vaughan, “Experimental investigation of sub-threshold transverse laser modes,” Elec. Lett., vol. 29, no. 11, pp. 997-998, May 1993. R. Loudon, M. Harris, T. J. Shepherd, and J. M. Vaughan, “Laser-amplifier gain and noise,” Phys. Rev. A, vol. 48, no.

R. Loudon et al., “Gain and noise in sub-threshold longitudinal laser modes,” Phys. Rev. A, vol. 50, no. 1, pp. 658-674, July 1994. I. Renhom et al., “Performance study of a coherent laser radar,” in Proc. Coherent Infrared Syst. and Applic. 2, vol. SPIE-415, pp 39-50, 1983 D. Letalick, I. Renhom, 0. Steinvall, and .I. H. Shapiro, “Noise sources in laser radar systems,” Appl. Opt., vol. 28, no. 13, p. 2657, July 1989. R. Foord, R. Jones, J. M. Vaughan, and D. V. Willetts, “Precise comparison of experimental and theoretical SNR’s in CO2 laser

1, pp. 681-701, July 1993.

i.’ Rothermel et al.,o“Calculation of aerosol backscatter from airbome continuous wave focused COa Doppler lidar measurements 1. Algorithm description,” J. Geophys. Res., vol. 96, no.

J. Rothermel, D. A. Bowdle, and J. M. Vaughan, “Calculation of aerosol backscatter from airbome continuous wave focused CO2 Doppler lidar measurements 2. Algorithm performance,” J. Geophys. Res., vol. 96, no. D3, pp. 5299-5305, Mar. 1991. D. R. Hall, R. M. Jenkins, and E. K. Gorton, “A frequency- stabilised cw waveguide carbon dioxide laser,” J. Phys. D: Appl.

K. M. Abramski, H. J. Baker, A. D. Colley, and D. R. Hall, “Single-mode selection using coherent imaging within a slab waveguide CO2 laser,” Appl. Phys. Lett., vol. 60, no. 20, pp. 2469-2516, May 1992. R. J. Morley, E. F. Yelden, H. J. Baker, and D. R. Hall, “Mode and frequency control of compact CO2 slab lasers through intracavity coherent imaging,” Appl. Opt., vol. 34, no. 3, pp. 418-425, Jan. 1995. D. V. Willetts and M. R. Harris, “An investigation into the origin of frequency sweeping in a hybrid TEA CO2 laser,” J. Phys. D: Appl. Phys., vol. 15, pp. 51-67, 1982. -, “Acoustic phenomena associated with a TEA laser discharge,” J. Phys. D: Appl. Phys., vol. 16, pp. 125-133, 1983. -, “A plasma effect in injection-mode selection of TEA CO2 lasers,” Appl. Phys. B, vol. 33, pp. 91-93, 1984. -, “Scaling laws for the intrapulse frequency stability of an injection mode selected TEA CO2 laser,” IEEE J. Quant. Electron., vol. QE-19, pp. 810-814, May 1983. V. G. Roper, H. M. Lamberton, E. W. Parcell, and A. W. J. Manley, “Laser induced medium perturbation in a pulsed CO2 laser,” Opt. Commun., vol. 25, pp. 235-240, 1978. D. Letalick, I. Renhom, and A. Widen, “CO2 waveguide laser with programmable pulse profile,” Opt. Eng., vol. 28, no. 2, pp. 172-179, Feb. 1989. K. Wallmeroth and P. Peuser, “High power, cw single frequency, TEMoo, diode-laser pumped N d YAG laser,” Elec. Lett., vol. 24, no. 17, pp. 1086-1088, 1988 K. Wallmeroth et al., “Toward a coherent optical free-space communication system,” Elec. Lett., vol. 26, no. 9, pp. 572-573, 1990. K. Wallmeroth, “Monolithic integrated Nd:YAG laser,” Opt. Lett., vol. 15, no. 16, pp. 903-905, 1990. R. Heilmann and J. Kuschel, “Absolute frequency locking of a diode pumped Nd:YAG laser for application in free-space optical communication,” Elec. Lett., vol. 29, pp. 810-811, 1993. S. Nikolov and L. Wetenkamp, “Lasing characteristics of a 1.644 nm Er3+ YAG monolith pumped at 647 nm, 787 nm, and 964 nm,” in Proc. 11th Int. Congress Laser 93, Munich. Berlin: Springer, 1993, pp. 112-115. I. Renhom, B. Broberg, and K. Vilhelmsson, “Coherent laser radar based on semiconductor technology,” in Proc. CLEO Europe, Amsterdam, Sept. 1994. U. Olander et al., “Three-section GRIN-SCH SL-MQW 1.55 pm SI-DFB laser for optical communications,” in EOCOC 92, Berlin, Germany, 1992. C. J. Armistead et al., “High-power coherent semiconductor lasers-state of the art, Part I,” ESA J., vol. 13, pp. 271-301, 1989; “High-power coherent semiconductor lasers-State of the art, Part 11” ESA J., vol. 14, pp. 41-65, 1990. Noordwijk, The Netherlands: ESA, ESTEC. K. F. Hulme and J. T. Pinson, “Improved acousto-optic modulators for CO2 heterodyne laser systems,” Elec. Lett., vol. 19, no. 7, pp. 254-256, Mar. 31, 1983. W. R. M. Pomeroy, J. M. Vaughan, and D. V. Willetts, “Performance study of an acousto-optic frequency shifter in a CO2 laser velocimeter,” J, Phys. E: Sci. Instrum., vol. 13, 1980. D. Hausamann and B. W. Davis, “Sign of the wind vector: A simple method for its determination with a homodyne cw laser Doppler velocimeter,” Appl. Opt., vol. 29, no. 19, July 1990. M. Royer and M. Bourdillot, “HF IR detectors for heterodyne wind speed lidar and spacebome applications,” in Proc. 7th

D3, pp. 5293-5298, Mar. 1991.

Phys., vol. 11, pp. 859-869. 1978.

VAUGHAN et al.: COHERENT LASER RADAR IN EtJROPE 223

Con$ on Coh. Las. Rad., Appl. & Tech., (LMD/CNRS), Paris, 1993, Paper Th D10, p. 314.

[49] D. J. Wilson, G. D. J. Constant, R. Foord, and J. M. Vaughan, “Detector performance stuhes for CO2 laser heterodyne systems,” Infrared Phys., vol. 31, no. 1, pp. 109-115, 1991.

[50] D. Letalick and I. Renhom, “Performance studies on a HgCdTe detector for heterodyne systems,” in Proc. 6th Con$ on Coh. Las. Rad., Appl. & Tech.; OSA Tech. Dig. Ser., 1991, vol. 12.

[51] D. Oh, P. Drobinski, P. Salamitou, and P. H. Hamant, “Op- tical local oscillator power for CMT photo voltaic detector in heterodyne mode,” Infrared Phys. & Tech.,, 1995.

[52] M. Alldritt, R. Jones, C. J. Oliver, and J. M. Vaughan, “The processing of digital signals by a surface acoustic wave spect” analyser,” J. Phys. E: Sci. Instmm., vol. 11, pp. 116-1 19, 1978.

[53] F. Kopp, F. Bachstein, and C. Wemer, “On-line data system for a cw laser Doppler anemometer,” Appl. Opt., vol. 23, no. 15, pp. 2488-2491, Aug. 1, 1984.

[54] H. G. Sillitto, “Advanced optics and scanning systems for airbome laser radar,” in 7th Con$ on Coh. Las. Rad. Appl. & Tech., (LMDKNRS), Paris, 1993, Paper FB6.

[55] R. Axelsson, P. Sundberg, and 0. Steinvall, “Programmable scanner for laser bathymetry,” Int. Hydr. Rev., vol. 67, no. 1,

E561 R. Knopp, “Degradation of laser radar by fast scanning mirror,” Coh. Las. Rad.: Tech. & Appl., vol. SPIE-1181, 1989.

[57] 1. Renhom and D. Letalick, “Laser radar performance degradation due to lag angle phenomena,” in 7th Con$ on Coh. k s . Rad. Appl. & Tech., (LMDICNRS)), Paris, July 1993.

[58] A. J. Hughes and E. R. Pike, “Remote measurement of wind speed by laser Doppler systems,” AppZ. Opt., vol. 12, pp.

[59] A. J. Hughes et al., “Remote measurement of wind speed by laser Doppler systems,” Opt-Electron., vol. 4, p. 377, 1972.

[60] A. Brown, E. L. Thomas, R. Foord, and J. M. Vaughan, “Mea- surements on a distant smoke plume with a CO2 velocimeter,”

[61] J. M. Vaughan, “Remote wind measurements in the atmosphere using laser Doppler methods,” in Laser 79 Opto-Electronics, W. Waidelich, Ed. New York: IPC, 1979, pp. 461-469.

[62] J. O’Shaughnessy and W. R. M. Pomeroy, “Single-beam atmospheric transverse velocity measurement,” Opt. & Quanturn Electron., vol. 10, pp. 270-272, 1978.

[63] C. Werner, F. Kopp, and R. L. Schwiesow, “Influence of clouds and fog on LDA wind measurements,” Appl. Opt., vol. 23, no. 15, pp. 2482-2484, Aug. 1984.

[64] - , “Remote measurements of boundary-layer wind profiles using a CW Doppler lidiu,” J. Clim. anddppl. Met., vol. 23, no. 1, pp. 148-154, Jan. 1984.

[65] V. A. Banakh, I. N. Smalikho, F. Kopp, and C. Werner, “Representativity of the wind measurements by a CW Doppler lidar in the atmospheric boundary layer,” Appl. Opt., vol. 34,

[66] F. Kopp, “Doppler lidar investigations of wake vortex transport between closely spaced parallel runways,” AiAA J., vol. 32, pp.

[67] J. M. Vaughan and P. A. Forrester, “Laser Doppler velocimetry applied to the measurement of local and global wind,” Wind Eng., vol. 13, no. 1, 1989.

[68] G. D. J. Constant et al., “Intercomparison of 10 pm laser Doppler wind measuring equipment with a series of balloon cable mounted anemometers,” SPIE, vol. 1181, pp. 2-7, 1989.

[69] -, “Coherent laser radar and the problem of aircraft wake vortices,” J. Mod. Opt., vol. 41, no. 11, pp. 2153-2173, 1994.

[70] Much of this work has been undertaken under contract for the Chief Scientists Division of the UK National Air Traffic Services as part of their R&D program to support the safe improvement of airport capacity.

[71] X. Favreau et al., “Heterodyne detection for simultaneous range resolved measurements of atmospheric constituents and wind velocity,” in Coherent Laser Radar Topical Meet., Keystone, CO, USA Tech. Dig. Ser., vol. 19, 1995.

[72] J.-L. Zarader et al., “Performance of an adaptive notch filter for spectral analysis of coherent lidar signals,” J. Amos. Ocean. Tech., Jan. 1995.

[73] P. Salamitou, A. Dabas, and P. H. Flamant, “Simulation in the time domain for heterodyne coherent laser radar,” Appl. Opt., vol. 34, no. 3, pp. 499-506, Jan. 1995.

pp. 161-170, Jan. 1990.

597-601, Mar. 1973.

J. Phys. D: AppZ. Phys., Vol. 11, pp. 137-145, 1978.

pp. 205.5-2067, 1995.

805-812, 1994.

[74] -, “Characterisation of pulsed coherent Doppler lidar with the speckle effect,” Appl. Opt., vol. 33, no. 21, pp. 6524-6532, Sept. 1994.

[75] P. Salamitou, F. Darde, and P. U. Flamant, “A semi-analytic approach for coherent laser radar system efficiency-The nearest- Gaussian approximation,” J. Mod. Opt., vol. 41, no. 11, pp.

1761 F. Nicolas, P. H. Flamant, and A. Dabas, “Heterodyne detection and multiple scattering in cirrus clouds,” in Colorado Laser Radar Topical Meet., Keystone, CO, OSA Tech. Dig. Ser., vol. 19, 1995.

[77] J. L. Meyzonette, B. Remy, and G. Saccomani, “A coherent laser radar with chirp compression,” Laser Radar II vol. SPIE- 783, 2987.

[78] P. H. Davies, D. W. Brown, R. Foord, and J. M. Vaughan, “An FM-CW laser radar with electro-optic intra-cavity modulation,” Coherent Laser Radar: Technol. and Applicat., vol. SPIE-1181,

[79] W. Alexander, D. L. Clark, and C. A. Stewart, “A CO2 laser radar configured for Doppler imaging,” Active Infrared Syst. and Techn., vol. SPIE-806, pp. 102-109, 1987.

[80] G. Sepp et al., “Moving target detection with compact laser Doppler radar,” Coherent Laser Radar: Techn. and Applic., vol.

[U] G. Constant, R. Foord, and J. Wilson, “Imaging and vibration studies with a chirp laser systems,” in Proc. 3rd Topical Meet. on Coherent Laser Radars and Applic., Gt. Malvern, 1985, pp. 3 w .

[82] W. Kranz, “Target classification by laser vibration sensing,” Coherent Laser Radar: Technol. and Applicat., vol. SPIE-1181, 1989.

[83] J.-J. Boichot, R. Charesse, and T. Midavaine, “CO2 laser rangefinders made in Soci6t.6 Anonyme de T616ccommunications,” in 7th Con$ on Coherent Laser Radar Applicat. and Techn., (LMDKNRS), Paris, 1993, Paper FB2.

[84] N. J. D. Martin, “Experimental 10.6 pm pulsed coherent laser rangefinder systems,” Active Infrared Syst. & Techn., vol. SPIE- 806, pp. 119-121, 1987.

[85] M. J. Taylor and P. H. Davies, “Preliminary studies of heterodyne detection of CO2 TEA laser pulses,” RSRE Memo. No. 3092, 1977.

1861 K. F. Hulme, B. S. Collins, G. D. Constant, and J. T. Pinson, “A CO2 laser rangefinder using heterodyne detection and chirp pulse compression,” Opt. and Quant. Electron., vol. 13, pp. 3545, 1981.

[87] P. A. Fomester and K. F. Hulme, “Review: Laser rangefinders,” Opt. and Quant. Electron., vol. 13, pp. 259-293, 1981.

[88] K. F. Hulme, “Advanced CO2 laser rangefinders,” Opt. and Laser Techn., pp. 213-215, 1982.

[89] __, “CO2 laser heterodyne rangefinders, velocimeters and radars,” Infrared Phys., vol. 25, no. 1/2, pp. 457-466, 1985.

r901 D. htalick. M. Millnert. and I. Renhom. “Terrain sezmentation

2101-2113, 1994.

pp- 285-290, 1989.

SPE-1181, pp. 296-300, 1989.

. > - using laser radar range data,” Appl. Opt., vol. 31, no. 15, pp.

r911 0. Steinvall, “Performance of coherent and direct detection laser 2883-2890, 1992.

. - radars for hard target applications,” in Coherent Laser Radar Topical Meet., Keystone, CO, July 1995.

[92] A. A. Woodfield and J. M. Vaughan, “Airspeed and wind shear measurements with an airbome CO2 CW laser,” AGARDo- graph, no. 272, pp. 7.1-7.17, 1983.

[93] -, “Airspeed and wind shear measurements with an airborne CO2 CW laser,” Int. J. Aviation Safely, vol. 1, pp. 207-224, 1983.

[94] -, “Using an airbome CO2 laser for free stream airspeed and windshear measurements,” in AGARD Con$ Proc., no. 373, pp.

[95] B. Morbieu and J. Mandle, “True airspeed measurement with a coherent laser radar,” in 3rd Int. Con$ on Advanced Infrared Detectors & Syst., London, June 1986, pp. 154-157.

[96] B. Morbieu, H. Combe, and J. Mandle, “ALEV3, a 3-axis CO2 CW anemometer for aircraft certification,” in Proc. 7th Con$ on Coh. 121s. Rad., Appl. & Tech., (MDKNRS), Paris, 1993, Paper Rw1.

E971 N. Martin, R. Ganini, and X. Chazelle, “CLARA-A Franco- British development of airbome laser radar,” in Proc. 7th Con$ on Coh. Las. Rad., Appl. & Tech., Paris, July 1993, Paper FB1.

[98] F. Kopp, C. Wemer, P. H. Flamant, and C. Loth, “The German-

22.1-22.18, 1984.


French project WIND for the development of an airborne lidar,” in Proc. 7th Con$ on Coh. Las. Rad., Appl. & Tech., Paris, July 1993, Paper MA4.

[991 P. Delville et al., “A new TE-CO2 laser for coherent lidar and WIND applications,” in Coh. Las. Rad. Con$, Keystone, CO, July 1995, OSA Tech. Dig. Ser., vol. 19.

[loo] R. J. Keeler et al., “An airborne laser air motion sensing system. Part I Concept and preliminary experiment,” J. Atmospheric & Oceanic Techn., vol. 4, no. 1, pp. 113-127, Mar. 1987.

[loll B. Stephan and P. Mktivier, “Flight evaluation trials of a heterodyne CO2 laser radar,” Active Infrared Syst. and Techn.,

[lo21 R. D. Callan, D. A. Huckridge, C. R. Nash, and J . M. Vaughan, “Active airborne linescan experiment,” Airborne Reconnais- sance XVIII, vol. SPIE-2272, pp. 183-191, 1994.

[lo31 D. A. Bowdle et al., “Aerosol backscatter measurements at 10.6 micrometers with airborne and ground-based COz Doppk; lidars over the Colorado high plains 1. Lidar intercomparison, J. Geophys. Res., vol. 96, no. D3, pp. 5327-5335, 1991.

[lo41 D. A. Bowdle, J. Rothermel, J. M. Vaughan, and M. J. Post, “Aerosol backscatter measurements at 10.6 micrometers with airborne and ground-based CO2 Doppler lidars over the Col- orado high plains 2. Backscatter structure,” J. Geophys. Res., vol. 96, no. D3, pp. 5337-5344, 1991.

[lo51 J. M. Vaughanet al., “Atmospheric backscatter at 10.6 pm; a compendium of measurements made outside the United King- dom by the airborne LATAS coherent laser radar velocimeter,” RSRE Malvern Rep. No. 87002, 1987.

[lo61 J. Rothermel, D. A. Bowdle, J. M. Vaughan, and M. J. Post, “Evidence of a tropospheric aerosol backscatter background mode,”AppZ. Opt., vol. 28, no. 6, pp. 1040-1042, 1989.

1071 J. M. Vaughan et al., “Comparison of SAGE II solar extinction data with airborne measurements of atmospheric backscattering in the troposphere and lower stratosphere,” Nature, vol. 332, no. 6166, pp. 709-711, 1988.

1081 S. B. Alejandro, G. G. Koenig, J. M. Vaughan, and P. H. Davies, “SABLE A South Atlantic aerosol backscatter measurement program,”Amer. Met. Soc., vol. 71, no. 3, pp. 281-287, 1990.

1091 S. B. Alejandro et al., “Atlantic atmospheric aerosol studies-1. Program overview,” J. Geophys. Res., vol. 100, no. D1, pp.

[ 1101 J. M. Vaughan et al., “Atlantic atmospheric aerosol studies-2. Compendium of airborne backscatter measurements at 10.6 pm,” J. Geophys. Res., vol. 100, no. D1, pp. 1043-1065, 1995.

[ l l l ] J. M. Vaughan et al., “Fluctuation and change in measured backscatter from atmospheric aerosols,” Proc. Top. Mtg. on Coherent Laser Radar, OSA Tech. Dig. Series, vol. 12, pp.

[112] J. Mandle et al., “Aerosol backscatter measurements at 10 pm using a CO2 laser aneomometer developed for aircraft certification,” in Coherent Laser Radar: Techn. & Applic., vol.

[113] J. Perin et al., “Retrieval of Meso-scale wind field Simula- tion and performance for WIND,” in Coh. Lax Rad. Conk, Keystone, CO, 1995, OSA Tech. Dig. Ser., vol. 19.

[114] Proc. ESA Workshop on Space Laser Applic. and Techn. (SPLAT), Les Diablerets, ESA, SP-202, May 1984 (see [MI).

[ 1151 J.-L. Fellow, Ed., “BEST-Tropical system energy budget,” in Le Service des Publications du CNES, Oct. 1988.

[ 1161 J. M. Vaughan, “ALADIN-Atmospheric laser Doppler instrument,” in Conf Proc.: Earth Observations Satellites-The Techn. Behind the Image, Royal Aeronaut. Soc., London, Paper 7, Jan. 1990.

[117] A. Marini, Ed., Proc. ESA Doppler Wind Lidar Workshop, ESTEC, Sept. 1995 (see [44]).

[118] H. Lutz, E. Armandillo, and B. Battrick, Eds., “Laser sounding from space,” in ESA Technol. Working Group Rep. on Space Laser Sounding and Ranging, ESA SP-1108, Jan. 1989 (see W1).

[119] P. Betout, D. Burridge, and C. Werner, “ALADIN-Atmospher- ic laser Doppler instrument,” Doppler Lidar Work. Group Rep., ESA SP-1112, June 1989 (see [44]).

[120] A. Marini and M. Rast, “The ESA atmospheric laser Doppler instrument (ALADIN) requirements for a coherent Doppler wind lidar,” in Proc. 7th Con$ on Coherent Laser Radar, (LMDKNRS), Paris, France, 1993.

[121] C. Norrie, A. Marini, and E. Armandillo, “ESA technology activites for spaceborne Doppler wind lidar,” in Proc. Satellite

vol. SPIE-806, pp. 110-118, 1987.

1035-1041, 1995.

288-289, 1991.

SPIE-1181, pp. 246, 1989.

Remote Sensing, vol. 2310, 1994. [122] D. Willetts et al., “A COz-laser for spaceborne Doppler wind

lidar applications,” in Proc. 7th Con$ on Coherent Laser Radar, (LMDKNRS), Paris, France, 1993.

[123] R. G. Harrison and P. K. Bhadani, “Switchless discharge TEA- CO2 -laser technology for spaceborne Doppler wind lidar,” in Proc. 7th Conf on Coherent Laser Radar, (LMDKNRS), Paris, France, 1993.

[124] C. Werner, J. Streicher, and G. Wildgruber, “Representativity of wind measurements from space,” in Proc. Central Symp. Int. Space Year Con$, Munich, MadApr. 1992, ESA SP-341.

[125] J. M. Vaughan, “Discussion of the relative merits and difficulties of 10 pm laser Doppler wind measuring systems,” Coh. Laser Rad.: Tech. & Appl., vol. SPIE-1181, pp. 226-230, 1989.

John M. Vaughan graduated with a degree in physics and the D.Phi1. in atomic spectroscopy from Oxford University, Oxford, UK, in 1959 and 1962, respectively.

After four years at the Clarendon Laboratory as a DSIR Research Fellow and Research Officer, he spent two years at Princeton University, Princeton, NJ, as a Lecturer with rank of Associate Professor. From 1968 to 1970 he was with Birmingham University as a Senior Research Fellow. He joined the RSRE (now the Defence Research Agency), Malvern, Worcestershire, UK, in 1970. He is presently Head of Laser Applications at the DRA, where he holds an individual merit post of Deputy Chief Scientist Officer. His main research interests lie in optics, spectroscopy, lasers and applications, as well as the development of coherent laser radar for remote sensing. He has authored over 150 papers and reports on these subjects, together with a definitive volume on the Fabry-Perot interferometer.

Dr. Vaughan was the Conference Chair for the 3rd International Meeting on Coherent Laser Radar, held in Malvern, UK, in 1985.

Kurt Ove Steinvall received the M.S. degree in physics from the University of Uppsala, Sweden, in 1969, and the Ph.D. degree in lasers and electro-optics from the Chalmers Institute of Technology, in 1974. In 1969 he has been employed by the National Defence Research

Establishment (POA), Linkoping, Sweden, where he has worked on fast gas discharges for lasers and other applications. Since 1977 he has been leading the Laser Group at FOA, and in 1991 he was appointed Research Director. His research interests include lasers, lasers for countermeasures, laser warning, laser radadlidar, fiber optics, and oceanic and atmospheric optics. He is the author or coauthor of more than 80 internal reports and more than 40 articles and contributions to international journals and conference proceedings,

Dr. Steinvall is a member of SPIE, the Society for Photo-Optical Instrumentation Engineers, the Optical Society of America, the Swedish Physical Society, and the Royal Academy of Military Sciences. He received two national rewards for laser work. He will be the conference chair for the next International Topical Meeting on Coherent Laser Radars in Linkoping, Sweden, in June 1997.

Christian Werner received the M.S. and Ph.D. degrees in physics from Munich University, Germany, in 1966 and 1981, respectively.

From 1966 to 1980 he was a scientist in the German Aerospace Research Establishment (DLR), Institute of Atmospheric Physics. Since 1980 he has been a Section Head in the Institute of Optoelectronics, responsible for laser techniques.

Dr. Werner is a member of the ESA Science Advisory Group for ALADIN, and a LASER Science Committee (Munich) member. He organized and chaired the 5th Coherent Laser Radar Meeting held in Munich in 1989 and well as several SPIE meetings and conferences.


Pierre Henri Flamant received the Doctorat in physics from the Univer- sity of Paris, France, in 1979.

He is presently Head of the Lidar team and Director of Research at LMDKNRS. He is currently involved in the space borne lidar program at the French Space Agency (CNES) and the European Space Agency (ESA). His main interest is in atmospheric dynamics, earth radiaiton budget, and water cycle. he is currently workmg on data analysis, lidar methodologies, and instnun” development to retrieve the atmospheric properties relevant to climate and mesoscale meteorology. He is a member of the ESA ALADIN advisory group for a space based wind Doppler lidar. He has more than 50 papers published in peer review journals.

Dr. Flamant is a member of the International Radiation Commission. He organized and chaired the 7th CLRC conference held in Paris in 1993, as well as the International Laser Radar conference.

226 PROCEEDINGS OF THE IEEE, VOL 84, NO. 2, FEBRUARY 1996

coherent laser radar in europehome.ustc.edu.cn/~522hyl/%ce%c4%d5%c2/11.coherent...remote sensing...

Documents