1318 OPTICS LETTERS / Vol. 21, No. 17 / September 1, 1996

Range-resolved bistatic imaging lidar for themeasurement of the lower atmosphere

K. Meki

Faculty of Engineering, Shinshu University, 500 Wakasato, Nagano-shi 380, Japan

K. Yamaguchi

Meisei Electric Company, Ltd., 249-1 Moriya-kou, Moriya-cho, Kitasouma-gun 302-01, Japan

X. Li, Y. Saito, T. D. Kawahara, and A. Nomura

Faculty of Engineering, Shinshu University, 500 Wakasato, Nagano-shi 380, Japan

Received March 12, 1996

A bistatic imaging lidar system using a cooled CCD camera as a detector has been developed for the observationof aerosols, fog, and clouds in the lower atmosphere, especially within several hundred meters of the Earth’ssurface. Theoretical discussion showed that the received signal does not depend on the measured range. Thepotential of the bistatic imaging lidar was confirmed through results of nighttime observations of atmosphericphenomena up to 300 m. A range-resolved profile was obtained without scanning and with a very short timeresolution, within 1 min. 1996 Optical Society of America

Since the development of the first lidar system byFiocco and Smullin1 in 1963, various lidar systems havebeen applied to atmospheric research. The lidar sys-tem generally used has a monostatic conf iguration inwhich a laser transmitter and a receiver are located atthe same place. However, a monostatic lidar system isnot effective for measuring the lower atmosphere be-cause the transmitting beam does not overlap the tele-scope’s field of view in the near range. This problemcan be overcome by changing the monostatic conf igu-ration to a bistatic one in which the receiver and thetransmitter are separate from each other. Althougha remote optical sensing system with a basic bistaticconfiguration using a searchlight was reported byElterman2,3 in 1950, the difficulty of adjusting the ele-vation angle of the transmitter and the receiver forrange resolution has prevented widespread use of sucha bistatic lidar system. So only a few such systemshave been designed,4 – 7 and successful implementationof bistatic lidar awaited the experiment of Welsh andGardner in 1989.8 They succeeded in detecting themesospheric sodium layer as an image ,90 km abovethe Earth by using a bistatic imaging lidar system.They also suggested the feasibility of a simple bistaticlidar that uses a CCD camera as an imaging device.

We are applying the bistatic imaging lidar tech-nique for lower-atmosphere research; our interestfocuses especially on observation of atmospheric phe-nomena near the Earth’s surface with a system withadequate range resolution and no need for difficult ad-justment.

We made a theoretical estimation of the performanceof the bistatic imaging lidar by deriving an expressionof the measured image data as a function of a densityprofile. The geometry of the bistatic lidar is shownin Fig. 1. The laser emits in the vertical direction,

0146-9592/96/171318-03$10.00/0

and the receiver, which is separated from the laserby a distance L (baseline), is set at an elevation anglea. The received power from a scattered volume withunit angle da, which is the f ield of view of one pixel, isdescribed as

Pr ­ PtKATtTrbsz, ud

Lda ,

where Pr is the received power, Pt is the transmittedpower, K is the optical efficiency of the receiver, A isthe collecting area of the receiver, and Tt and Tr arethe atmospheric transmittance from the laser to scat-terers and from scatterers to the receiver, respectively.bsz, ud is the scattering coeff icient, which is a func-tion of height z and scattering angle u. We simplify

Fig. 1. Schematic of the bistatic imaging lidar.

1996 Optical Society of America

8778 (DLC)

September 1, 1996 / Vol. 21, No. 17 / OPTICS LETTERS 1319

this equation by integrating only in the horizontal di-rection, because we are interested primarily in the ver-tical structure of the atmosphere. By comparing thisequation with the monostatic lidar equation, we findthat the received signal intensity does not depend onthe square of the range. This means that the detec-tor is not required to have a wide dynamic range ofsignal detection, and therefore the measurable rangeexceeds that of the monostatic lidar. This is one ad-vantage of the bistatic imaging lidar. The height zand the height resolution dz are given simply by

z ­ L tan a , dz ­L

cos2 ada .

The traditional bistatic lidar has not been populatedbecause many vertical or horizontal scans were neededfor a height or range profile to be obtained, but oursystem with the CCD camera makes it possible toget the profile from one laser beam trajectory imagewithout scanning.

The bistatic imaging lidar system was composedof a second-harmonic Q-switched Nd:YAG laser asthe transmitter and a cooled CCD camera (Hama-matsu C-3140) as the receiver. The experimentalsetup is shown in Fig. 1. The laser beam, with a160-mJypulse, a repetition rate of 10 Hz, and a beamdivergence of 0.6 mrad, was emitted in the verti-cal direction. The laser beam was linearly polar-ized, with the plane of polarization perpendicular tothe scattering plane to eliminate the scattering angledependence of Rayleigh scattering cross section. Thescattered light, collected by a commercial camera lens56 mm in diameter with a focal length of 28 mm anda field of view of 12 deg, passed through a narrow-band interference filter (I.F.) with a bandwidth of3 nm at 532 nm and was detected and imaged by theinterline transfer type CCD camera with 510 (hori-zontal) 3 492 (vertical) pixels and a 25% aperture.The CCD chip was cooled to 230 ±C to reduce thedark current [(nine electronsypixel)ys]. The quantumefficiency of the cathode was 33% at 532 nm. Thesignal was digitized by 12-bit analog-to-digital convert-ers (Amp. AyD) and recorded by a personal computer(P.C.). A preliminary analysis of the image data wasmade by the same personal computer to give real-timeinformation to the operator and to check the systemperformance quickly. The minimum detectable pho-ton number of (90 countsypixel)ys was estimated fromthe optical eff iciency, the quantum efficiency, and thedark-current level of the system. We confirmed thathardware errors such as readout noise, amplifier noise,and dispersion of the sensitivity of each CCD chipwere less than 4%. The baseline length between thetransmitter and the receiver was varied in the hori-zontal plane, depending on the experimental require-ments. We could easily adjust the transmitter andthe receiver by watching the image on the monitor.The image always appeared somewhere on the moni-tor, and by using the CCD camera we solved the trou-blesome adjustment problem. An example of imagedlaser beam trajectory obtained by the system is shownin Fig. 2. As this image contained the backscattered

signal and background noise, the background-noise es-timation and reduction were done as follows: (1) Theimage data were divided into two parts; one containedthe backscattered signal and the background noise,shown in Fig. 2(a), and the other was background noiseonly, shown in Fig. 2(b). (2) We obtained the verti-cal profiles of these two parts by summing all data inthe horizontal direction within the set area. (3) Thedifference of these two profiles was replotted as one-dimensional data. The difference of the two imagesgives the return signal profile that is due to scatteredradiation, and it offers information about the verticalstructure of the atmosphere.

The lidar system was tested several times at ShinshuUniversity, Nagano, Japan (36±40′N, 138±12′ E). Thearea is ,350 m above sea level and is surroundedby mountains. An example of the measurement onDecember 22–23, 1994, is shown in Fig. 3. The lidarsystem was situated on the roof of a university buildingat a height of 24.9 m above ground level. In theexperiment the baseline was 5 m and the elevationangle was 83±. The observed height in one set ofimage data was from 46 m (if the system was on theground, 21.1 m) to 383 m. The accumulating timeof the CCD was 30 s, and a preliminary analysisof the image data was obtained within 30 s. The495 signal profiles were taken over 495 min from22:00 to 06:15 local time. During the experiment thetemperature and the humidity were 3–24.6 ±C and77–100%, respectively, and a foggy sky condition wasencountered. The running mean method was appliedto the three adjacent data, and the laser return signalintensity is shown in the figure by the gray scale. Inthe analysis it was assumed that the Mie scatteringwas independent of scattering angle in the range167–179 deg and that image data gave approximatelythe vertical profile of the backscattering coeff icient.

Fig. 2. Image of the laser beam trajectory taken bythe bistatic imaging lidar: (a) backscattered signal andbackground noise, (b) background noise.

1320 OPTICS LETTERS / Vol. 21, No. 17 / September 1, 1996

Fig. 3. Nighttime variation of the vertical profile of the lower atmosphere at ground surface level extracted from thereceived photoelectron number (December 22–23, 1994). Baseline, 5 m; elevation angle of the CCD camera, 83 deg; laseroutput, 9 mJ, 10 pulsesys; accumulating time, 30 s.

Results clearly showed that the atmospheric conditionat the near surface varied greatly over a short time.This suggests that frequent sampling is necessary foraccurate evaluation of lower-atmospheric phenomena.The bistatic imaging lidar technique seems suitable forthis purpose.

The feasibility of bistatic imaging lidar has been de-scribed. A theoretical derivation of the bistatic imag-ing lidar equation showed that the received power doesnot depend on the measured range, and so the systeminherently does not require the capability of wide dy-namic range detection. Also, the experimental resultsshowed that the system is a valuable and powerful toolfor measurement of the lower atmosphere, especiallynear the Earth’s surface where atmospheric conditionschange continually and in a wide range.

References

1. G. Fiocco and L. D. Smullin, Nature (London) 199, 1275(1963).

2. L. Elterman, J. Geophys. Res. 56, 509 (1951).3. L. Elterman, J. Geophys. Res. 58, 519 (1953).4. J. A. Reagan, D. M. Byrne, and B. M. Herman, IEEE

Trans. Geosci. Electron. GE-20, 229 (1982).5. J. A. Reagan, D. M. Byrne, and B. M. Herman, IEEE

Trans. Geosci. Electron. GE-20, 236 (1982).6. K. Parameswaran, K. O. Rose, and B. V. Krishna

Murthy, J. Geophys. Res. 89, 2541 (1984).7. P. C. S. Devara and P. Ernest Raj, J. Aerosol Sci. 20, 37

(1989).8. B. M. Welsh and C. S. Gardner, Appl. Opt. 28, 82 (1989).