nasa multipurpose airborne dial system and...

NASA multipurpose airborne DIAL system and measurementsof ozone and aerosol profiles

E. V. Browell, A. F.and W. M. Hall

Carter, S. T. Shipley, R. J. Allen, C. F. Butler, M. N. Mayo, J. H. Siviter, Jr.,

An airborne differential absorption lidar (DIAL) system has been developed for the remote measurementof gas and aerosol profiles in the troposphere and lower stratosphere. The multipurpose DIAL system canoperate from 280 to 1064 nm for measurements of ozone, sulfur dioxide, nitrogen dioxide, water vapor, tem-perature, pressure, and aerosol backscattering. The laser transmitter consists of two narrow linewidth Nd:YAG pumped dye lasers with automatic wavelength control. The DIAL wavelengths are transmitted witha 00-,usec temporal separation to reduce receiver system complexity. A coaxial receiver system is used tocollect and optically separate the DIAL and aerosol lidar returns. Photomultiplier tubes detect the back-scattered laser returns after optical filtering, and the analog signals from three tubes are digitized and storedon high-speed magnetic tape. Real-time gas concentration profiles or aerosol backscatter distributions arecalculated and displayed for experiment control. Operational parameters for the airborne DIAL systemare presented for measurements of ozone, water vapor, and aerosols in the 290-, 720-, and 600-nm wavelengthregions, respectively. The first ozone profile measurements from an aircraft using the DIAL technique arediscussed in this paper. Comparisons between DIAL and in situ ozone measurements show agreement towithin 5 ppbv in the lower troposphere. Lidar aerosol data obtained simultaneously with DIAL ozonemeasurements are presented for a flight over Virginia and the Chesapeake Bay. DIAL system performancefor profiling ozone in a tropopause folding experiment is evaluated, and the applications of the DIAL systemto regional and global-scale tropospheric investigations are discussed.

1. Introduction

A multipurpose airborne differential absorption lidar(DIAL) system has been recently developed at theNASA Langley Research Center to remotely measurethe profiles of various gases and aerosols in diverse at-mospheric investigations. The airborne DIAL systemhas the flexibility to determine the spatial distributionof gases such as ozone, water vapor, sulfur dioxide, andnitrogen dioxide with simultaneous measurements ofaerosol backscattering at multiple laser wavelengths.This capability provides the opportunity to examine thecomplex interactions between atmospheric dynamicsand chemistry with high spatial resolution in the tro-posphere and lower stratosphere.

Initial Langley development of the DIAL techniquewas aimed at ground-based investigations of watervapor and sulfur dioxide.2 These experiments werethe first to demonstrate the DIAL technique in actual

M. N. Mayo is with Kentron International, Inc., Hampton, Virginia23666; R. J. Allen and C. F. Butler are with Old Dominion UniversityResearch Foundation, Norfolk, Virginia 23508; the other authors arewith NASA Langley Research Center, Hampton, Virginia 23665.

Received 13 July 1982.

atmospheric measurements using the flexibility andenergy efficiency of laser-pumped dye lasers. Theknowledge gained from these ground-based DIAL ex-periments was used in development of the airborneDIAL system. The airborne DIAL lidar has the flexi-bility to operate in the UV for measurements of ozoneor sulfur dioxide, in the visible for nitrogen dioxide, andin the near-IR for water vapor, atmospheric tempera-ture (using water vapor or oxygen absorption lines), andpressure (using oxygen lines). Aerosol backscatter in-vestigations in the UV, visible, and near IR can beconducted simultaneously with the DIAL measure-ments. The capabilities of the airborne DIAL systemare functionally the same as those proposed for an earlyphase of the NASA Shuttle Lidar Program.3 -5 Theexperience gained with this system will be useful for apreliminary evaluation of several potential Shuttle lidarinvestigations.3

Recent emphasis with the airborne lidar has been inmeasurements of ozone, water vapor, and aerosol pro-files. An understanding of the tropospheric ozonebudget is essential to establishing a firm knowledge oftropospheric photochemistry and the potential impactof pollutants upon the photochemical system. Highspatial resolution measurements of water vapor profilesare important for applications such as the initialization

522 APPLIED OPTICS / Vol. 22, No. 4 / 15 February 1983

of numerical weather forecast models in regions notcovered by radiosondes and in studies of latent heat fluxand troposphere/stratosphere exchange, to name a few.Aerosols can be used as a tracer of atmospheric dy-namics. Aerosol distributions can provide informationon the boundary layer mixing depth, condensation level,cloud top altitude, cloud statistics, and the altitude ofstable layers above the boundary layer.6 7 In additionto describing the multipurpose airborne DIAL system,this paper discusses the first 03 profile measurementsmade with an airborne DIAL system in the troposphere.Aerosol profile measurements obtained during a re-gional tropospheric flight experiment are presented, andDIAL system performance for the measurement ofozone profiles during a tropopause folding experimentis examined.

II. Airborne DIAL System

The airborne lidar system uses the DIAL techniquefor the remote measurement of gas profiles. Thistechnique has been discussed in detail by numerousauthors 8 -12 ; thus only a brief review is presented here.The DIAL technique determines the average gas con-centration over some selected range interval by ana-lyzing the difference in lidar backscatter signals for laserwavelengths tuned on and off a molecular absorptionline of the gas under investigation. The value of theaverage gas concentration N between range R1 and R2can be determined from the ratio of the lidar signals atthe on and off wavelengths. This relationship is givenby

N = 1 l Pff(R2)P..(Rl)2(R2 - R1) (on - off) Pff(Rl)Pon(R 2 )

where Con -uoff is the difference between the absorptioncross sections at the on and off wavelengths, and Pon (R)and Poff (R) are the signal powers received from rangeR at the on and off wavelengths, respectively. Thisrelationship assumes that the aerosol and molecularoptical properties are equal at the on and off DIALwavelengths. If there is an interfering gas which doesnot have the same absorption coefficient at thesewavelengths, the concentration of this gas must beknown or determined by a separate measurement.

A block diagram of the DIAL system is shown in Fig.1. Two frequency-doubled Nd:YAG lasers are used topump two high conversion efficiency tunable dye lasers.All four lasers are mounted on a rigid support structurewhich also contains the transmitting and receiving op-tics. The dye laser on and off wavelengths that are usedin the DIAL measurement are produced in sequentiallaser pulses with a time separation of 100 ,4sec or less.This close spacing insures that the same atmosphericscattering volume is sampled at both wavelengthsduring the DIAL measurement. The output beams areseparated and steered using dielectric coated optics.They are transmitted out of the aircraft through a 40-cmdiam quartz window, coaxially with the receiver tele-scope.

The wavelength of the two dye lasers is determinedusing a 1-m monochromator and a spectral reference

Fig. 1. Airborne DIAL system schematic.

lamp. The monochromator output is displayed in realtime by an optical multichannel analyzer (OMA), andsimultaneous operation with the spectral referenceprovides the laser wavelength to an accuracy of less than±10 pm. When more accurate wavelength control isneeded, such as for H20 DIAL measurements, it is ac-complished using a closed-loop wavelength controlsystem. This system uses a stepping motor to controlthe dye laser grating angle, and it provides wavelengthcontrol to better than ±0.3 pm.

The receiver system consists of a 35-cm diam Casse-grain telescope with optics to direct the received signalsonto the detectors, which are gateable photomultipliertubes. As many as three photomultiplier tubes can beaccommodated. When the system is operating in thevisible or near IR, only one tube is needed for the on andoff lines, with the off line also providing an aerosolmeasurement. Frequency-doubling crystals (KDP) areused to double the visible radiation into the UV whenmaking measurements in this spectral region, and theresidual off-line visible wavelength is transmitted andused to measure atmospheric aerosols. Two photo-multiplier tubes are used, one optimized for the UVwavelength region and the other optimized for thevisible wavelength. Three 10-bit transient digitizers,operating at a 10-MHz conversion rate, sequentiallydigitize the on- and off-line DIAL and aerosol returnsignals. The data are then stored on a 1600-bpi high-speed magnetic tape unit by means of a PDP 11/34minicomputer. Gas concentration profiles can be cal-culated in real-time and displayed on a video system orhardcopy printer for real-time operator experimentcontrol.

The DIAL system installed in the NASA WallopsFlight Center Electra aircraft is shown in Fig. 2(a). Thereceiving telescope and receiver optics housing areshown at the end of the laser support structure. Theenclosed portion of the structure contains the Nd:YAGpump lasers, and the monochromator and OMA aremounted underneath. The dye lasers can be seenmounted on top of the structure along with the trans-mitting optics housing. The control electronics rackis at the far end of the structure. Figure 2(b) shows thedye lasers with the covers removed. The data systemcan be seen beyond the control electronics rack.

15 February 1983 / Vol. 22, No. 4 / APPLIED OPTICS 523

ON LINE LASER SYSTEM

QUARTZWINDOW

DNTROL *I alOl lCONTROL SYSTEM

~TERMI3-T~cR * NEAR IR DIAL ONLYSYSE ** UV DIAL ONLY

Fig. 3. Overall DIAL transmitter configuration.

a Side view of DIAL system oriented for nadir measurements

b Top view of DIAL system looking forward in aircraft

Fig. 2. Airborne DIAL system installed in Wallops Flight CenterElectra aircraft.

111. DIAL Transmitter

The airborne DIAL system transmitter configurationis shown in Fig. 3. Two Quantel model 482 fre-quency-doubled Nd:YAG lasers are used to pump twoJobin Yvon high-power high-resolution tunable dyelasers. The output beams are directed by dielectriccoated optics coaxially with the telescope through a40-cm diam quartz aircraft window. For coarse tuningof the dye lasers, a small fraction of each output laserbeam is directed through a 1-m monochromator si-multaneously with the output of a spectral referencelamp. The output of the monochromator is detectedby an OMA, which provides a real-time display of thelaser line position. This method provides a wavelengthpositioning accuracy of better than 10 pm. For H20vapor concentration measurements in the 720-nmwavelength region, the on-line dye laser wavelength

must be maintained as close as possible to the center ofthe absorption line. This is accomplished by a Fizeauinterferometer wavelength control system which pro-vides wavelength control to better than 0.3 pm. Amore detailed discussion of the various laser transmittersubsystems is presented in the following paragraphs.

As shown in Fig. 4, the Nd:YAG pump laser operateswith a TEMOO oscillator and two amplifiers to producea near diffraction-limited beam at 1.06 Aim with 1 J/pulse at 10 Hz. Frequency doubling is accomplishedwith 40% efficiency by an angle-tuned temperature-stabilized KDP-II crystal giving 400 mJ/pulse with a12-nsec pulse length at 532 nm. The two wavelengthsare spectrally separated by two polarization-sensitivedichroic beam splitters (R, 99% from 525 to 535 nm;Tp 90% from 1.06 to 1.07 m).

The dye laser13 utilizes a tuned oscillator, preampli-fier, and two amplifiers operating near saturation. Theoscillator uses a high-efficiency grazing-incidence ho-lographic grating which is optimized for Littrow oper-ation at 724 nm (H20 DIAL) with an angle of incidenceof 84.50 (2748 lines/mm). For UV DIAL measurementsof 03 and SO2, the dye lasers operate between 571 and600 nm. The same grating is used, but it operates withan angle of incidence near 530 giving a dye laser line-width of <8 pm. The dye solution for operation near572 nm is rodamine 6G (chloride) in water and Am-monyx. The dye concentrations for the oscillator andamplifiers are 4.5 X 10-4 and 1.75 X 10-4 M/liter re-spectively. The dye solution for operation near 600 nmis rhodamine B (chloride) in water and Ammonyx. Theconcentrations are the same as for rhodamine 6G. Theconversion efficiency of the dye lasers at these wave-lengths is typically 30% giving 120 mJ/pulse at 10 Hz.For doubling into the UV (286- and 300-nm wavelengthregions), angle-tuned temperature-stabilized KDP-Icrystals are used with conversion efficiencies of 25%giving 30 mJ/pulse with a linewidth of <4 pm.

For H20 DIAL operation, the holographic grating inthe dye laser oscillator is used near its optimum angleof incidence (84.5°). The dye laser produces 72 mJ/pulse at 10 Hz, which represents an 18% conversion ef-ficiency from the pump laser output at 532 nm to thedye laser output at 724 nm. The dye solution used forwavelength generation near 724 nm is Carbazine 720 ina solution of water, Ammonyx, and NaOH. The con-centrations for the oscillator and amplifiers are 3 X 10-3and 1.25 X 10-3 M/liter, respectively. The output beam


PUMP LASER- QUANTEL MODEL 482MIRROR POLARiZER OSCILLATOR HEAD ETALON M

I - IPk -CELL PIN HOLE LN

X Mwl

x1- UP DOUBLINGP AMPLIER COLLIMATOR AMPLIFIER CRYSTAL

1 …l_~1 _ ___ __ ___ ___

I-

M MIRROREs. = BEAMSPLITTER

WIS. = DICHROIC BEAM SPiTTERDC. = DYE CELLCL. = CYLINDRICAL LENSDS. DISPERSIVE SYSTEM

P = PRISM

DYE LASER - J.Y MODEL HP - KR

* DOUBLING I CI C.L.; X PLATE ICRYSTAL I LENS OTU 4 PIN-HOLEI

T MsPMIRROR GRATING

COLLIMATOR _ _ _ _I

E LATE *ONLY FOR UV DIAL OPERATION

Fig. 4. Nd:YAG pumped dye laser layout.

is -10 mm high by 2 mm wide with a divergence of 0.2X 0.6 mrad, respectively.

A Fabry-Perot interferometer (Tropel model 360)and a Princeton Applied Research OMA were used toverify the specified dye laser linewidth. The interfer-ometer plates were separated by 20 mm to obtain a freespectral range (FSR) of 13.1 pm at 724.5 nm. TheFabry-Perot mirror reflectivity is 97.5%, which gives areflectivity finesse Fr of 124. The mirror pair is flat toX/100, which gives a flatness finesse Ff of 50. The totalfinesse Ft of 46 was computed from the relationship1/F2 = 1/F2 + 1/Ff. The OMA detector was positionedin the focal plane of a 1.1-im lens to observe the fringes.Figure 5 shows a typical fringe pattern. This photo-graph was averaged over 80 shots with an intracavityetalon located in the dye laser oscillator. The linewidthis obtained from the approximate relationship AX(FWHMA/SA,B) FSR, where FWHMA is the width offringe A (several fringes out from the central fringe) at50% of the peak value, and SA,B is the wavelength sep-aration between fringes A and B. The fringes in Fig.5 indicate a laser linewidth of 1.57 pm. Measurementsof the dye laser linewidth were also made with no in-tracavity etalon, and a linewidth of 2.5 pm was ob-tained.

Figure 6 shows the optical arrangement for laser beamseparation and steering. For UV DIAL operation [(Fig.6(a)] dichroic beam splitters are used to separate the UVand visible dye laser wavelengths (R, 2 99% from 280to 305 nm; Tp 90% from 570 to 605 nm). Energymonitors are used in each of the transmitted beams.The insertion loss for the energy monitor is <2%, sincereflection near the Brewster angle is used to direct someof the transmitted energy through a diffuser onto aphotodiode. Optical filters are also used in the energymonitors to eliminate unwanted background light.Pellicles direct a small percentage of the undoubtedlaser beams to a monochromator for wavelength mon-itoring. The optical configuration for near-IR opera-

A BFig. 5. Typical fringe pattern for the dye laser beam using aFabry-Perot interferometer. The trace represents an 80-shot

average.

tion is shown in Fig. 6(b). The energy monitors areagain used in each dye laser beam. A small amount(<1%) of each dye laser beam is transmitted through atotal reflector (R 2 99% from 710 to 770 nm), and it isthen directed to the monochromator for wavelengthmonitoring. Due to the requirement for wavelengthcontrol of the on-line laser with an accuracy of betterthan +0.5 pm, a residual portion of the on-line laseroutput is directed to a wavelength control instrumentwhich is described below. Figure 6(c) shows a diagramof the transmitting optics. The UV or near-IR DIALbeams are directed by common total reflectors (R 299%) coaxially with the receiving telescope. In the caseof UV DIAL measurements, the 600-nm dye output isdirected by a separate set of total reflectors. Perfor-mance characteristics of the laser transmitter aresummarized in Table I.

For wavelength monitoring of the dye lasers, smallportions of the on-line and off-line dye laser outputs are


OPTICAL BEDTOP VIEW

TO MONOCHROMATOR

ON LINE D.B.S.572 nm ______CL

fP LLC

D.. 6•

RS.

OFF LINE

~ D.B S. QB..S ICHR

D.B.S. t DICHRiE.M. = ENER

(a) UV DIAL WAVELENGTH SEPARATION,

SIDE VIEW

LE : DUMP I

E.M.I

IE. M.

OIC BEAM SPLITTER;Y MONITORS I

OPTICS

OPTICAL BEDTOP VIEW

ON LINE E.MT.R.

_____ .. TO X CONTROL

724nm (ON LINE)TR

TOMONOCHROMATOR

724 nm (OFF LINE)R.

OFF LINE E.M. O

-.T - -- -' MONOCHROMATORTR.

UV OR NEAR-IRDIAL BEAMS

TPR

TELESCOPE

T JR.

QUARTZWINDOW

T.R. = TOTAL REFLECTORS

(c) TRANSMITTING OPTICS

TR. = TOTAL REFLECTORSE.M.= ENERGY MONITORS

(b) NEAR-IR DIAL BEAM COMBINING OPTICS

Fig. 6. Airborne DIAL transmitter optical system configuration.

Table 1. Airborne DIAL Transmitter Characteristics

Two pump lasers Quantel model 482Pulse separation 100 ,usecPulse energy 400 mJ at 532 nmPulse length 15 nsecRepetition rate 1, 5, or 10 Hz

Two dye lasers Jobin Yvon model HP-HR

Fundamental Doubled fundamental Near IR(near 600 nm) (UV near 300 nm) (near 720 nm)

Dye output energy 120 mJ/pulse 30 mJ/pulse 72 mJ/pulseTransmitted laser energy 80 mJ/pulse 29 mJ/pulse 62 mJ/pulseLaser linewidth <8 pm <4 pm <3 pm


at 1

I

l

coupled into a light pipe which directs them into asimple optical system attached to the entrance apertureof 1-m McPherson monochromator. The optical sys-tem combines the beams of the dye lasers with theoutput of a reference lamp and focuses them onto themonochromator entrance aperture. A 2400-line/mmholographic grating is used in the first order to producehigh dispersion at the exit focal plane of the mono-chromator. The OMA employs a silicon intensifiedtarget detector head (SIT model 1205D) to detect thespatially distributed laser and reference lines. Theoutput from the SIT detector is displayed on a scope.The SIT is enhanced to provide a wide-range wave-length sensitivity from the UV to the near IR. Thereare 500 channels across the SIT detector with 2 5 -Amspatial separation between each channel. The spectralresolution of this system is 6.3 pm/channel at 600ni.

DIAL measurements of 03 are made in the Hartleyband with the on-line nominally set near 286 nm and theoff-line set near 300 ni. Wavelength monitoring isdone in the visible at the undoubled wavelengths. Aneon lamp provides reference lines in the wavelengthregion of the undoubled dye laser outputs. Using theOMA-monochromator system, the dye laser wavelengthcan be adjusted to within about ±3 pm of the desiredneon line.

Coarse wavelength adjustment for H20 DIAL mea-surements is made with the OMA, which has a spectralresolution of 4.1 pm/channel near 720 ni. DifferentH2 0 absorption cross sections are required to optimizethe optical depth in a DIAL measurement for enhancedsensitivity or reduced signal dynamic range. The neonor krypton spectral line nearest the selected H20 ab-sorption line is used for the reference wavelength.Positioning of the dye laser wavelength relative to theselected reference wavelength can be made to within±12 pm provided that their separation is <1.5 ni.This dye laser adjustment technique is sufficiently ac-curate for the off-line wavelength. Since the pressurebroadened H20 line has a nominal linewidth of 10 pm,this method must be supplemented by additional finetuning of the on-line wavelength. Wavelength coinci-dence between the on-line laser and the H2 0 absorptionline is accomplished by slowly tuning the laser throughthe absorption line and minimizing the lidar returnsignal from a distant range. Enhanced sensitivity canbe obtained using an optical depth of at least 1. In thiscase, relatively large changes in optical depth or signalattenuation are observed as the absorption line isscanned in wavelength.

For atmospheric constituents that have spectrallynarrow absorption linewidths (AX), the DIAL datareduction algorithm assumes that the differential ab-sorption cross section for the gas under investigation isa function of pressure only. This condition is approx-imated when a nearly monochromatic laser transmitterhaving a linewidth of <AXa/4 is tuned to the center ofa temperature insensitive absorption line. Any trans-mitter wavelength instability or drift from the absorp-tion line center results in a change in the absorption

Fig. 7. Wavelength stabilization system schematic.

cross section which must be accounted for in the DIALcalculation. Accordingly, some form of wavelengthstabilization is usually employed for the on-line laser,while the off-line laser wavelength, which is consider-ably less critical, is not automatically controlled.

A wavelength control system developed under thedirection of M. L. Chanin at the CNRS, Verrieres leBuisson, France, is used in the airborne DIAL systemto provide a means for maintaining high-precisioncontrol of the on-line dye laser wavelength.14 Thesystem is shown schematically in Fig. 7. A Hewlett-Packard stabilized He-Ne laser (model 5501A) oper-ating at 632.914 nm is used as a reference wavelength.With the aid of a beam chopper, the reference and dyelaser beams are alternately passed through a Fizeauinterferometer. The interferometer consists of an inputup-collimator, a wedge, and output cylindrical lenswhich focuses the interference fringes onto a diodearray. The wedge thickness of 4.3 mm provides a freespectral range of 57.6 pm at 720 nin. With a reflectancefinesse of 50, the spectral resolution of the wedge is 1.1pm. A pair of dye laser fringes at 724.3 nm is typicallyseparated by 340 of the total 512 elements in the diodearray. Thus the element spacing represents 0.17 pmat this wavelength. The fringe analyzer electronicscomputes the position of the reference and dye laserfringe centroids. The number of elements betweenthese fringes is compared to thumbwheel inputs on thecontrol panel. The difference between these readingsis used to reposition the holographic grating throughcontrol of the grating stepping motor. The system isinitialized by nulling the difference before the motorcontrol is turned on. Wavelength control of the on-linedye laser can be maintained to within ±0.3 pm of itsinitialized wavelength.

IV. Receiver SystemThe receiver system is designed to collect the back-

scattered laser light and optically process the lidar re-turn signals before directing them to the photomulti-plier tube (PMT) detectors. A 35-cm diam Ritchey-Chretien Cassegrain telescope is the primary lidar col-lector with an effective focal length of 400 cm. Tele-scope mirrors are aluminum coated with a protectiveMgF 2 coating optimized for maximum reflectance at 300ni. A variable field stop is located in the focal plane


of the telescope. The field of view (FOV) of the receivercan be adjusted from -0.8 to 4.0 mrad. A nominal FOVof 1.5 mrad is used with a laser divergence of -1.0 mrad.After passing through the field stop, the UV DIAL re-turns are reflected 90° to the telescope optical axis bya dichroic beam splitter (R 2 95% from 280 to 305 nm).The visible aerosol lidar return is transmitted throughthe beam splitter (Tp > 83%) and turned in an orthog-onal direction by a total reflector. Both reflected beamsare independently collimated to be compatible withinterference filters and PMT photocathode areas. Aquartz lens is used to collimate the UV return into a20-mm diam beam, and a conventional glass lens is usedin the aerosol beam. Two PMTs can be used to detectthe DIAL returns at different sensitivity levels by re-flecting a portion of the return beam with an uncoatedquartz plate to a second PMT.

Angle-tuned narrow bandpass interference filters areplaced in front of each PMT to reject background lightand any simultaneous lidar return at an unwanted laserwavelength. The UV interference filter consists of alow-pass UV filter and a visible blocking filter, whichresults in a nominal transmission of 62% from 285 to 308nm with blocking of >103 above 330 nm. A typicalvisible and near-IR interference filter has a peaktransmission of 45% with a bandwidth of 0.5 nm. Peaktransmission can be moved to shorter wavelengths byfilter rotation from normal incidence to a maximumangle of incidence of 15°.

As many as three PMTs can be used simultaneouslyto detect multiwavelength lidar returns. Each tube wasselected for its spectral response characteristics and theability to gate the tube electronically. The UV lidarreturns are detected by an RCA model 7268 PMT,which has a bialkali photocathode with a quantum ef-ficiency (Q.E.) of 28% at 285 nm. The visible andnear-IR returns are directed onto RCA model 7265PMTs having standard multialkali (Q.E. of 7.3% at 600nm) and ERMA-III (Q.E. of 3.9% at 720 nm) photo-cathodes, respectively. These tubes have high gain (4X 106 to 2 X 108 at 2400 V) and gating characteristicswhich make them well suited for lidar applications.The airborne DIAL receiver characteristics are sum-marized in Table II.

Table II. Airborne DIAL Receiver Characteristics

DoubledFundamental fundamental Near IR(near 600 nm) (UV near 300 nm) (near 720 nm)

Area of 0.086 m2 0.086 m2 0.086 m2

receiverReceiver 26% 28% 29%

efficiencyto PMT

PMT quantum 7.2% 29% 3.9%efficiency

Total receiver 1.9% 8.1% 1.1%efficiency

Receiver <2 mrad <2 mrad <2 mradfield of view(selectable)

V. Timing, Control, and Signal ProcessingElectronics

The primary consideration in the airborne DIALsystem timing concept is the synchronization of thesequential laser firing, the PMT control functions, andthe PMT signal digitization. A master control unit wasdeveloped for precise control (10 nsec) of laser flash-lamp firing, lase control, photomultiplier gating, anddigitizer triggering. All critical frequencies are con-trolled by a single crystal time base. The two Nd:YAGpump laser pulses and, therefore, the on- and off-lineDIAL pulses are temporally separated by 100 usec.The PMT gates are activated at a selectable time delayafter each laser pulse is transmitted. Since both DIALbackscattered returns are detected sequentially by thesame PMT, independent control of the PMT gain forthe on- and off-line lidar returns is required for opti-mum digitization of both signals. The maximum PMTgain is controlled by manually adjusting the high voltagesupplied to the PMT dynode network. Adjustment ofthe gain between the on- or off-line returns can be au-tomatically controlled by two techniques: (1) supplyingalternating gate voltages to the PMT focusing electrodeor (2) supplying a 0-28-V signal to the even numbereddynodes to control the gain over a range of 0-42 dB. Athree-channel PMT dynode and focus electrode driverunit is used to drive three independent PMT devices,providing options for variable PMT gain and PMTcutoff. Each PMT dynode chain is configured forstep-gain operation with up to four variable-gain vari-able-width gain increments during the lidar return. Alight emitting diode (LED) based laser-return simulatoris available for production of synthetic signals from eachPMT without the requirement for actual atmosphericlaser returns.

The three independent PMT signals can be smoothedin the analog domain by linear phase low-pass videofilters. The variable selection video filters are providedfor removal of very high-frequency noise and signalcomponents which are present in the PMT anode signaloutputs. A noise compensation system is under de-velopment for elimination of large-background levelvariations which decrease the effective dynamic rangeof the transient digitizers. The return signals are dig-itized into 10-bit words at a 10- or 5- MHz sequentialword rate (2.5-MHz bandwidth maximum) by threeBiomation 1010 Transient Digitizers. Both the on- andoff-line DIAL returns are processed sequentially duringone 2048-word sweep of the first transient digitizer. Toimprove the dynamic range for digitization of the lidarreturns the DIAL returns can be simultaneously digi-tized at different sensitivity levels using a second PMTand/or a second Biomation. The third Biomation 1010unit is reserved for digitizing the aerosol lidar return.Details of the timing control and signal processingelectronics are given in Ref. 15.

VI. Data Acquisition System

The airborne DIAL data acquisition system (DAS)is based upon the Digital Equipment Corp. (DEC)PDP-11/34 processor with 28K words of 16-bit MOS


Fig. 8. Components of airborne DIAL data acquisition system.

RAM memory. The overall flow chart for the DIALDAS is shown in Fig. 8. In general, data input/output(I/O) functions are shown on the left-hand side of thefigure with data storage and operator I/O functionsshown on the right. The operator communicates withthe software operating system through a ruggedizedAnn Arbor 400S console terminal. Data are presentedto the operator using a Lexidata 3400 Video GraphicsDisplay Terminal. Hard copy graphics are obtainedeither through Polaroid photography or by softwarecopy to a Trilog T100 Graphics Line Printer. TheDIAL data are stored in real time using a 1600-bpi PE45-ips magnetic tape unit on 731.5-m (2400-ft) reels of1.27-cm (0.5-in) wide magnetic tape. Storage and re-trieval of program information are accomplished usingDEC RX01 dual floppy disk drives, which were chosenfor reliability in the aircraft environment.

As discussed in the previous section, the acquisitionof DIAL data is accomplished using three Biomationmodel 1010 transient digitizers. The internal memoriesof these transient digitizers are made available to thePDP-11/34 CPU through a direct memory access(DMA) interface of custom design.1 5 The DMA in-terface addresses the three Biomation 1010 transientdigitizers in parallel, and it performs both data transferand add-to-memory functions at PDP 11 memory cyclerates.

Supplemental information is acquired in real timethrough additional channels of parallel I/O. A DR11-Cbased 16-bit parallel interface is used to obtain navi-gation information from either LORAN-C or InertialNavigation devices. The serial navigation signal isdecoded by circuits of custom design, and simple

modifications will allow operations with the OMEGAnavigation aid. In addition, eight analog-to-digitalconversion (ADC) and two digital-to-analog conversion(DAC) channels are available for system diagnosticapplications through a DEC ARi analog real-timesubsystem module. Diagnostic applications includereal-time monitoring of laser energy and beam charac-teristics and automatic laser alignment. In situ mea-surements of atmospheric pressure, temperature, dewpoint, and IR radiometer measurements of surface andcloud top temperatures can also be recorded simulta-neously with the lidar data.

The DIAL operating system software is designed totake full advantage of the computer hardware and in-terface capabilities. Highest priority is given to theacquisition and compression of lidar signal informationwith subsequent storage on digital magnetic tape. Theoperating system software is interrupt driven by thetransient digitizer interface. Data transfer operationsfrom each of the three Biomation 1010 transient digi-tizers (2048 words each) require -2 msec, and the entiredata transfer operation is readily accomplished withinthe 10-HIz operation time envelope. The primary lim-itation to repetition rate and data volume magnituderesides in the digital magnetic tape transport. Doublebuffering is implemented to speed the average datatransfer rate from CPU memory to magnetic tape, anda third buffer is maintained to provide a complete dataset for real-time display operations. Operator inter-vention is interrupt driven by keystroke at the lowestinterrupt priority level. Operator commands are de-coded from a single line of character input to providecontrol over data acquisition, magnetic tape, and


plotting operations.' 5 The real-time display of raw andprocessed DIAL signal information is performed on atime available basis.

Four basic modes for DIAL data display are available,each provided with a variety of display options. Thefirst and most fundamental display mode presents rawdata from each of the Biomation 1010 transient digi-tizers. This display mode allows operator viewing ofall DIAL data as it exists in each of the Biomation 1010memories. The display can be adjusted to present the10-bit signal magnitude with variable magnification inall 2048 words of the three transient digitizers. DIALsignals are typically recorded in two Biomation unitswith different gain settings to enhance signal dynamicrange. A second display mode presents the raw datasignals simultaneously or individually from the tran-sient digitizers with or without background subtractionin an overlapped format. Display options are availablefor background signal subtraction, data smoothing overany specified range interval (running mean), and cor-rection for range-squared lidar signal dependence.Each display option or combination of display optionsmay be activated or deactivated in real time to observesignal features in the most useful format.

A 16-level gray scale display format is available forpresentation of the spatial distribution of aerosol scat-tering. In processing the aerosol lidar return, thebackground signal level is subtracted from the lidar-plus-background signal, and the geometrical rangesquared lidar signal dependence is eliminated. Theresulting lidar backscatter profile is indicative of thedistribution of aerosols along the lidar line of sight. Thevertical resolution of the aerosol data is 15 in. Thenominal horizontal resolution is 10 m for aircraft op-eration at a 10-Hz repetition rate. The backscattersignal level is converted into a 16-level gray scale displayline where stronger scattering is indicated by a higherbrightness on the monitor or a darker pixel on theprinted version of the display. Sequential gray scalelines are used to construct a real-time picture of theaerosol vertical distribution under the Electra flightpath. Each of the gray scale displays can contain 300individual or horizontally integrated aerosol profiles.At a laser pulse repetition rate of 1 or 10 Hz, the 300individual profile displays correspond to a nominalhorizontal traverse of 30 or 3 km, respectively. Thishorizontal scale assumes a nominal ground speed of 100msec' for the Electra aircraft. The gray scale displayformat shows the terrain profile, and it clearly identifiesthe distribution of aerosols in the boundary layer andthe free troposphere.

A final display mode is provided for presentation ofDIAL measured gas concentration mixing ratios as afunction of altitude or vertical range. For each DIALreturn, the background signal level is integrated overa 5-gsec interval after the ground return. This averagebackground is subtracted from each return signal pro-file, and these signals are then smoothed using a runningaverage over a range interval corresponding to thechosen range cell size. This smoothing technique canbe used only for those atmospheric conditions where the

aerosol scattering is not changing rapidly along theDIAL measurement path. The DIAL equation dis-cussed previously is evaluated using the smoothed lidarreturns over a specified range cell size, usually 210 m.Ozone mixing ratios are determined by dividing eachrange cell concentration by the corresponding standardatmospheric number density at that altitude. A cor-rection factor of 6.7 ppb is subtracted from the ozonemixing ratio to compensate for Rayleigh extinctiondifferences between 286 and 300 nm. Water vapormixing ratios are determined by dividing each range cellconcentration by the standard number density at sealevel, since the product of the water vapor absorptioncross section at line center and the atmospheric numberdensity is independent of pressure. Each DIAL pulsepair produces a mixing ratio profile. Any number ofDIAL measurements can be averaged together to reducethe profile statistics at the expense of increased hori-zontal range for the measurement. The standard de-viation for the resulting profile is computed at incre-ments equivalent to the range cell size and displayed onthe mixing ratio profile.

VII. 'Ozone and Aerosol Profile Measurements

The first remote measurements of ozone profiles withan airborne DIAL system were made during flight testsbetween 22 May and 6 June 1980. The DIAL systemwas operated in the UV for ozone measurements withthe on wavelength at 285.95 nm and the off wavelengthat 299.40 nin. The differential absorption cross section,i.e., the difference in the absorption cross sections, forthis pair of wavelengths is 17.4 X 10-19 cm2.' 6 TheDIAL system was operated in a nadir-directed modefrom the NASA Wallops Flight Center Electra aircraftat a nominal altitude of 3.2 km. The initial DIAL sys-tem measurements of ozone and aerosols were per-formed on four flights in the vicinity of the ChesapeakeBay. An instrumented Cessna 402 aircraft was used toobtain in situ profiles of ozone concentration, temper-ature, dew point, and aerosol total scattering crosssection. Ozone was measured by the ozone-ethylenechemiluminescent technique (Monitor Lab model8410), which has an absolute accuracy of ±10% or ±5ppbv (lower limit), a precision of ±3 ppbv, and a re-sponse time of 3 sec to 90% of the actual ozone concen-tration. Characteristics of the other in situ instrumentsare discussed in Ref. 17.

Examples of the intercomparison between the DIALand in situ ozone profile measurements are given in Fig.9. Digitization of the UV returns was optimized atdifferent altitudes by adjusting the PMT gating timedelay after laser firing and by adjusting the digitizersensitivity. The DIAL measurements represent a100-shot average or 2-km horizontal resolution at a 5-Hzpulse repetition frequency with a vertical resolution of210 in. The horizontal bars on the DIAL ozone profilesrepresent the standard deviation of the calculated av-erage ozone concentrations. Average values of theozone concentration obtained on the Cessna for con-stant altitude traverses over a 32-km leg below the El-ectra aircraft, and the standard deviation of those


5,s PT DELAY 10 SIOT I15.40 35 EDT 210m RAN(

C'N. -

I

AVERAGE lOres PMT DELAYGE CEU 15 30 28 EDT

I \ .59I I

\ I I

a I I I I I I

3000 F-

--- CESSNA IN SITU SPIRAL DATA* CESSNA CONSTANT ALTITUDE DATAA TETHERED BALLOON DATA

AIRBORNE DIAL DATA

- 3952m AIRCRAFT ALTITUDE (15:23:12 EDT)--- 3048m AIRCRAFTALTITUDE (15 10:05EDT)

100 SHOT AVERAGE210m RANGE CELL2500 F-

B

- 2000

. A

60 70 80 90 100 Ito ' 60 70 90 90 100 110

OZONE CONCENTRATION, ppbv

Fig. 9. Comparison of DIAL and in situ ozone profile measurementson 22 May 1980.

--- CESSNA IN SITU SPIRAL DATA* CESSNA CONSTANT ALTITUDE DATA

DIAL DATA (10: 58:21 EDT)

RCRAFT ALTITUDEXAVERAGENGE CELL

1 500 -

10001_

500 F-

I I I I I I I I I I I Iu0 10 20 30 40 50 60 70 80 90 100 110 120


Fig. 11. DIAL, Cessna, and tethered balloon ozone data on 5 June1980.

5001

_ l -i AIRBORNE

ax>_ ~~~~~~~3200 AFe 100 SHOT

IL-' 210 .. RAI

- ~ ~~~~ - - - - -r

.5s .; --

I I I I l I70 80 90 100 110 120 130 140 150


Fig. 10. Ozone profile comparison of DIAL and in situ measure-ments on 29 May 1980.

measurements is also shown in the figure. The Cessnaobtained spiral data at each end of the leg, and thesedata are given in the figure as an envelope of valuesbetween 200- and 1725-m MSL. The in situ data in-dicate a mixed layer height of -600 m with an enhancedaverage ozone concentration of 100 ppbv in the mixedlayer compared with -80 ppbv in the free troposphere.The agreement between the remotely determined ozoneconcentration profile and the in situ data is within 10%,which is comparable to the accuracy of the chemilumi-nescence instrument. The magnitude of the DIALmeasurement uncertainty is due primarily to photonstatistical errors which can be reduced by averaging overa larger horizontal or vertical extent; however, thenatural spatial variability of ozone also contributes tothe standard deviation of the average DIAL ozoneprofile. An example of another ozone profile compar-ison is shown in Fig. 10. These measurements weremade near Wallops Island, Va, on the morning of 29May 1980. The mixed layer had a height of -500 mwith an average ozone concentration of <115 ppbv. Anozone enriched layer was found above the mixed layerwhich had concentrations exceeding 130 ppbv. Thisstable layer extended up to 1500-m MSL with lowerozone values toward the top of the layer. The Cessnaspiral data above 1700-m MSL and below 800-m MSLwere obtained at the opposite end of the 37-km Cessnaleg from where the 800-1700-m MSL data were ob-

tained. The DIAL and in situ measurements of ozoneshow agreement to within the uncertainties of the twotechniques.

The DIAL system was flown on 5 June 1980 at twodifferent altitudes over Wallops where both the Cessnaand a tethered balloon were making in situ measure-ments. These data are presented in Fig. 11. A com-parison is shown between two DIAL profiles made fromdifferent aircraft altitudes in the overlap altitude regionbetween 1000- and 1800-m MSL. The agreement be-tween these profiles is within the DIAL measurementuncertainties. This agreement indicates that the av-erage DIAL ozone profile obtained at the higher aircraftaltitude would approach that determined at a loweraircraft altitude given a larger number of DIAL mea-surements. The Cessna and tethered balloon datadiffer consistently by -10 ppbv. In the altitude regionswhere the DIAL ozone profiles show a <5-ppbv uncer-tainty, the DIAL data agree more closely with thetethered balloon data than the Cessna data. The DIALmeasurements follow the tethered balloon data near2000-m MSL, which at this altitude are 20 ppbv lowerthan Cessna data. This trend exists up to -2900-mMSL where there is again agreement with the Cessnaprofile data. Differences in the ozone calibrations ofthe Cessna and tethered balloon ozone instrumentscould account for this consistent disagreement. Also,since the Cessna flight track was located 5 km upwindof the tethered balloon/DIAL measurement site forsafety reasons, differences in the ozone concentrationsat the two locations could have existed. Evidence ofozone horizontal variability can be seen in the sequentialairborne DIAL measurements discussed below.

The spatial variability of ozone can be seen in Fig. 12,where a sequence of five DIAL profiles are shown overa horizontal distance of 10 km. Very low values ofozone (25-30 ppbv) persisted at 2700-m MSL, where theair was clean (0.014-km-' extinction coefficient) and dry


2000

--- CESSNA IN SITU SPIRAL DATA* CESSNA CONSTANT ALTITUDE DATA

-AIRBORNE DIAL DATA

500

2500_

2000

1500B

v NnX

3500 r

, 0001

15:21:33(A) 15

TIME, EDT

15:22:14 (C)i21:53 (B) 15:22:35 (D) 15:22:56(E)

zi . i

3952m AIRCRAFT ALTITUDE100 SHOT AVERAGE210m RANGE CELL

LOCATION OF ORIGIN FOR ABOVE PROFILES

L08IB

1000_

500 _

OALOC

l

LOD

L LI I I I I I I I I

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200OZONE CONCENTRATION, ppbv

Fig. 12. Sequential ozone profiles obtained by the airborne DIAL system on 5 June 1980. The zero ozone concentration level is shifted by30 ppbv for each subsequent ozone profile. The subscript identifies the appropriate profile.

1453

en_-TeMD. < _--- Dew Pt.

I I I I-10 0 10 20

TEMPERATURECENTIGRADE

30

1455 1500

I , , I38°30.8 N 38°2E76 45 0'W 76°2E

L.8'N6.0' W

VIRGINIA k-12km -. CHESAPEAKE BAYFig. 13. Intensity modulated display of aerosol data taken on 24 July 1980 at about 1500 EDT over Virginia and the Chesapeake Bay. Thetemperature and dew point profiles were obtained over land by the 1400 EDT radiosonde released by Dulles airport (near Washington,

D.C.).

with 8% relative humidity at 5 0C. The ozone concen-tration increased to 60-65 ppbv in the mixed layer, theheight of which increased from -1800- to 2250-m MSLover the 10-km measurement leg. Enhanced ozoneconcentrations are often found in the mixed layer as aresult of photochemical reactions involving NO. andhydrocarbons.'8 Variability in the mixed layer heightcan be readily seen in the aerosol lidar returns that areobtained simultaneously with the DIAL ozone mea-surements. The small-scale vertical structure in theDIAL profiles of ozone may be actual localized varia-tions of ozone concentration similar to the variabilitydetected by in situ instruments on the Cessna duringa spiral or constant altitude flight. An airborne DIALsystem provides the capability to obtain in real time the2-D gas distribution below or above the aircraft.

Aerosol backscatter profiles at 600 nm are obtainedsimultaneously with the dual-channel UV DIAL data.The aerosol lidar return is corrected for backgroundoffset and range-squared effects before it is convertedinto a 16-level gray scale display line. An example ofthis display technique is shown in Fig. 13, wherestronger scattering is indicated by higher brightness.This picture was produced from 600 laser shots alonga 90-km flight track of the Electra during a flight to theeast over the state of Virginia. The ground reflectionappears as a bright line at the bottom of the picture.The left side of the picture was obtained over land nearWashington, D.C. and the right side over the Chesa-peake Bay. As shown on the left side of the picture, thelidar detects the presence of clouds and provides a directmeasurement of cloud top height. When these clouds


3000

2500

2000

1500

B)

.9

E

I

I-J

3 -

2 -

I -

0-

1503 EDT

3

-O

I

200 m RANGE CELL100 SHOT AVERAGE

B

7

6

5 6 8 ' ~~~1 12 ' 14 ' 16 18 20 ' 2OZONE CONCENTRATION, pphm

Fig. 14. Simulation of DIAL measurement uncertainties for tropopause folding investigation with an on-line wavelength at 290 nm and anoff-line wavelength at 300 nm. A horizontal resolution of 1 km (100-shot average at 10 Hz), a vertical resolution of 200 m, and an aircraft altitude

of 4 km was assumed in this calculation.

are optically thick, signals are obtained to limited rangesin the cloud, and ground returns are not detected. Theabrupt increase in aerosol scattering at the level ofmaximum cloud top height indicates a mixed layerheight of -2.5 km over land. There is substantially lessvertical mixing over the Chesapeake Bay. The aerosolsabove the strong stable layer over the Chesapeake Baywere advected by a northerly wind from the highermixed layer existing upwind. The condensation levelor height of cloud base over land at an altitude of 1.2-kmMSL corresponds to the mixing condensation leveldetermined by a radiosonde released by Dulles airportnear Washington, D.C. at 1400 EDT. An estimate ofthe maximum cloud top height from the radiosondedata agrees with the lidar values determined over land.Mixed layer and tropospheric dynamics can be readilystudied using the aerosol distribution informationavailable from the airborne lidar system.6'7

VIII. Simulation of DIAL Tropopause FoldingExperiment

Simulations have been conducted to evaluate thecapability of the airborne DIAL system for investigatingwater vapor and ozone distributions in the troposphereand lower stratosphere. Since the potential for airbornewater vapor DIAL measurements has been thoroughlydiscussed in Ref. 19, this section will address the capa-bility of the DIAL system for investigating a tropopausefold event, which is an important type of strato-sphere-troposphere exchange mechanism. A simula-tion of a tropopause folding investigation is shown inFig. 14. The profile for the tropopause folding eventrepresents a vertical cross section of ozone data obtainedin 1978 by Shapiro over California.2 0 DIAL systemparameters from Tables I and II, an aircraft altitude of4 km, and a zenith pointing DIAL system were assumedin these calculations. The standard deviation uncer-tainty for the DIAL measurement with a vertical andhorizontal resolution of 200 m and 1 km, respectively,is also shown in the figure. This spatial resolution is

compatible with this type of investigation because thevertical halfwidth of the intrusion is -1 km, and thehorizontal extent can be as large as 250 km. Uncer-tainty of the DIAL measurements is better than ± 10%through the lower ozone layer and increases to ±30% inthe lower stratosphere. To investigate the lowerstratosphere, a maximum Electra altitude of -8 kmwould be used along with a lower absorption cross sec-tion for the on DIAL wavelength. It is expected thatmeasurements to altitudes of 15 km could be made withuncertainties of less than ±10% and vertical and hori-zontal resolutions of 500 m and 2 km, respectively.Since the DIAL system also has an aerosol channel, thedistribution of condensation nuclei can be used as atracer of tropopause dynamics along with ozone.

IX. Summary

This paper has presented a detailed description of theNASA multipurpose airborne DIAL system. Thissystem is functionally the same as the first-generationlidar system which has been proposed for operationfrom the Shuttle in the late 1980s.3,4,5 The lasertransmitter consists of two Nd:YAG pumped dye lasers,which are tunable with frequency doubling from 280 to1000 nm. The lasers are fired sequentially to producethe necessary on and off wavelengths for the DIALmeasurements, and the absorbed laser wavelength isactively controlled to maintain coincidence with theabsorption line of the gas under investigation. Whenconducting UV DIAL measurements, laser output near600 nm is used simultaneously to obtain a profile of theaerosol scattering distribution along the lidar line ofsight. A collocated receiver system collects, opticallyseparates, filters, and detects the backscattered DIALand aerosol lidar returns. The analog data from thephotomultiplier tubes are digitized and stored onmagnetic tape. The data can be processed in real timeby a minicomputer to produce DIAL gas profiles oraerosol lidar displays for experiment control.


The first airborne DIAL measurements of ozone andaerosol profiles have been presented, and comparisonshave been made with in situ measurements. Agree-ment has been attained between the ozone measure-ments to an accuracy of 10% in and above theboundary layer. High spatial resolution synoptic dataon ozone and aerosols provided by the airborne DIALsystem will dramatically increase our ability to studytropospheric and lower stratospheric ozone on a regionalscale and provide the large-scale coverage required forinvestigating global budgets of ozone and aerosols. Inaddition, measurement of water vapor profiles with thissystem will provide important data on latent heat flux,air mass modification over bodies of water, water vaportransport into the stratosphere, and initialization con-ditions for weather forecast models. The airborneDIAL system also has the capability to measure sulfurdioxide (near 300 nm), nitrogen dioxide (near 440 nm),2

and simultaneous water vapor and temperature (withthe addition of a third laser).21 22 The development anddemonstration of the DIAL system represent importantsteps toward developing a capability to study large-scaletropospheric processes from space. A Shuttle-bornelidar system will utilize many of the same laser tech-niques that are now being used in the airborne DIALsystem to conduct high spatial resolution atmosphericinvestigations on a global scale.3 -5

References

1. E. V. Browell, T. D. Wilkerson, and T. J. McIlrath, Appl. Opt. 18,3474 (1979).

2. E. V. Browell, Opt. Eng. 21, 128 (1982).3. E. V. Browell, Ed., "Shuttle Atmospheric Lidar Research Pro-

gram-Final Report of Atmospheric Lidar Working Group," NASASpec. Publ. 433 (1979).

4. R. V. Greco, Ed., "Atmospheric Lidar Multi-User InstrumentSystem Definition Study," NASA Contract. Rep. 3303 (1980).

5. J. E. Harris and E. V. Browell, "Evolutionary Shuttle Atmo-spheric Lidar Program," in Conference Abstracts, Ninth Inter-national Laser Radar Conference, Munich, Germany, 2-5 July1979.

6. E. V. Browell and S. T. Shipley, "Lidar Meteorology," in Pro-ceedings, International Geoscience and Remote Sensing Sym-posium, Washington, D.C., 8-10 June 1981.

7. S. T. Shipley and E. V. Browell, "Airborne Lidar Measurementsof Mixed Layer Dynamics," in Conference Abstracts, EleventhInternational Laser Radar Conference, Madison, Wisc., 23-25June 1982.

8. R. M. Schotland, "The Determination of the Vertical Profile ofAtmospheric Gases by Means of a Ground Based Optical Radar,"in Proceedings, Third Symposium on Remote Sensing of theEnvironment, Oct. 1964 (U. Michigan, Ann Arbor, 1965).

9. R. M. Measures and G. Pilon, Optoelectronics 4, 141 (1972).10. R. L. Byer and M. Garbuny, Appl. Opt. 12, 1496 (1973).11. R. M. Schotland, J. Appl. Meteorol. 13, 71 (1974).12. R. T. Thompson, Jr., "Differential Absorption and Scattering

Sensitivity Predictions," NASA Contract. Rep. 2627 (1976).13. F. Bos, Appl. Opt. 20, 1886 (1981).14. C. Cahen, J. P. Jegou, J. Pelon, P. Gildwarg, and J. Porteneuve,

Rev. Phys. Appl. 16, 353 (1981).15. C. F. Butler, S. T. Shipley, and R. J. Allen, "Investigation of Po-

tential of Differential Absorption Lidar Techniques for RemoteSensing of Atmospheric Pollutants," Old Dominion U., Norfolk,Va., Tech. Rep. GSTR-81-8 (1981).

16. E. C. Y. Inn and Y. Tanaka, "Ozone Absorption Coefficients inthe Visible and Ultraviolet Regions," in Advances in Chemistry,No. 21 (American Chemical Society, Washington, D.C., 1959),p. 263.

17. G. L. Gregory, S. M. Beck, and J. J. Mathis, Jr., "In Situ Correl-ative Measurement for the Ultraviolet Differential AbsorptionLidar and the High Spectral Resolution Lidar Air Quality RemoteSensors: 1980 PEPE/NEROS Program," NASA Tech. Memo.83107 (1981).

18. T. E. Graedel, "Urban Precursors and Their PhotochemicalProducts," in Man's Impact on the Troposphere, J. S. Levine andD. R. Schryer, Eds., NASA Ref. Publ. 1022 (1978).

19. E. V. Browell, A. F. Carter, and T. D. Wilkerson, Opt. Eng. 20,684(1981).

20. National Center for Atmospheric Research, Annual Report, FiscalYear 1978, Boulder, Colo.

21. E. V. Browell, S. T. Shipley, A. Rosenberg, D. Hogan, and T. D.Wilkerson, "An Airborne Lidar for Simultaneous Measurementsof Temperature and Water Vapor," in Conference Abstracts,IAMAP Third Scientific Assembly, Hamburg, Federal Republicof Germany, 17-28 August 1981.

22. A. Rosenberg and D. B. Hogan, Appl. Opt. 20, 3286 (1981).

The authors express their appreciation to L. W.Overbay, N. L. McRae, W. J. McCabe, and W. W.Midgette, Jr. of NASA Langley Research Center (NASALaRC) for their skilled technical support in the devel-opment and operation of the airborne DIAL system.We also recognize T. D. Wilkerson and R. Khanna of theUniversity of Maryland, College Park, Maryland, andM. L. Chanin of the CNRS, Verrieres le Buisson,France, for their technical assistance in the developmentof the DIAL system for water vapor measurements.Appreciation is expressed to M. L. Chanin for supplyingthe laser wavelength control system and to G. L. Greg-ory (NASA LaRC) for providing the correlative in situozone measurements. NASA Wallops Flight Center isrecognized for their cooperation in providing the Electraaircraft for our flights. This research was supported inpart by the Environmental Protection Agency.


nasa multipurpose airborne dial system and...

Documents