Raman shifting of KrF laser radiation for tropospheric ozone measurements William B. Grant, Edward V. Browell, Noah S. Higdon, and Syed Ismail The differential absorption lidar (DIAL) measurement of tropospheric ozone requires use of high average power ultraviolet lasers operating at two appropriate DIAL wavelengths. Laboratory experiments have demonstrated that a KrF excimer laser can be used to generate several wavelengths with good energy conversion efficiencies by stimulated Raman shifting using hydrogen (H2) and deuterium (D 2 ). Computer simulations for an airborne lidar have shown that these laser emissions can be used for the pecise (less than 5% random error) high resolution (200-m vertical, 3-km horizontal) measurement of ozone across the troposphere using the DIAL technique. In the region of strong ozone absorption, laser wavelengthsof 277.0and 291.7nm were generated using H 2 and D 2 , respectively. In addition, a laser wavelength at 302.0nm was generated using two cells in series, with the first containing D 2 and the second containing H 2 . The energy conversion efficiency for each wavelength was between 14 and 27%. 1. Introduction A number of differential absorption lidar (DIAL) systems have been developed during the past few years to measure tropospheric and stratospheric ozone. Most of these have used wavelengths in the Hartley band of ozone near 300 nm, with tropospheric systems relying on wavelengths shorter than 300 nm, and stratospheric systems relying on wavelengths longer than 300 nm. These DIAL wavelengths have been generated using flashlamp or laser pumped dye la- sers, 1 ' 2 Raman shifted XeCl (308 nm) or KrF (248.4 nm) excimer lasers using H 2 and/or D 2 cells,' 3 - 6 or Raman shifted quadrupled Nd:YAG laser with H 2 and D 2 cells 7 ' 8 or D 2 and HD cells. 9 In this paper, the case is made that the Raman shifted KrF laser can be better than Nd:YAG laser based lidar systems for both daytime and nighttime measurements when the D 2 and H 2 cells are used in series and the combined first Stokes shifts of D 2 and H 2 (D 2 + H 2 ) are used, giving radiation shifted 7146 cm-' to 302.0nm. With this approach, two pairs of lines are available for DIAL measurements of ozone: one pair at 277.0 and 291.7 nm for lower ozone densities and shorter ranges, and one at 291.7 and 302.0 nm for higher ozone densities and longer ranges. Both lab- The authors are with NASA Langley Research Center, Atmo- spheric Sciences Division, Hampton, Virginia 23665-5225. Received 18 June 1990. oratory measurements of Raman shifting and comput- er simulations of lidar performance are presented in support of this claim. The original motivation for placing two cells in series was twofold: first, that it should be possible to gener- ate a useful amount of 302.0-nm radiation via com- bined Raman shifting from D 2 and H 2 ; and, second, that the pump beam could be used more efficiently for generating the Stokes shifts of the individual gases. The latter follows from the fact that the entire pump beam is used in the first cell, and then the sizable residual pump beam (50% of the input energy) is sent to the second cell. Note that generation of 302.0-nm radiation using a KrF laser and H 2 and D 2 was first reported in Ref. 10,in which both gases were contained in one cell. II. Laboratory Setup The laboratory setup used is shown in Fig. 1, with the parameters of the lasers, Raman cells, and lenses given in Table I. A dichroic beam splitter that reflect- ed 90%of the radiation of S 1 of D 2 and 30%at S 2 of D 2 was placed near the exit of the first cell at an angle of incidence of 450. (For the field system, a dispersing prism will be used to separate the various wavelengths after the D 2 cell, since dichroic beam splitters do not have a sharp enough cutoff.) For the measurement of the Raman conversion energies, a fused silica window was used to divert -4% of the beam after the second cell to a lens followed by a dispersing prism, and the energies at the separate wavelengths were measured using a pyroelectric energy monitor. 2628 APPLIED OPTICS / Vol. 30, No. 18 / 20 June 1991
Transcript
Raman shifting of KrF laser radiation for troposphericozone measurements
William B. Grant, Edward V. Browell, Noah S. Higdon, and Syed Ismail
The differential absorption lidar (DIAL) measurement of tropospheric ozone requires use of high averagepower ultraviolet lasers operating at two appropriate DIAL wavelengths. Laboratory experiments havedemonstrated that a KrF excimer laser can be used to generate several wavelengths with good energyconversion efficiencies by stimulated Raman shifting using hydrogen (H2) and deuterium (D2). Computersimulations for an airborne lidar have shown that these laser emissions can be used for the pecise (less than 5%random error) high resolution (200-m vertical, 3-km horizontal) measurement of ozone across the troposphereusing the DIAL technique. In the region of strong ozone absorption, laser wavelengths of 277.0 and 291.7 nmwere generated using H2 and D2, respectively. In addition, a laser wavelength at 302.0 nm was generatedusing two cells in series, with the first containing D2 and the second containing H2. The energy conversionefficiency for each wavelength was between 14 and 27%.
1. Introduction
A number of differential absorption lidar (DIAL)systems have been developed during the past few yearsto measure tropospheric and stratospheric ozone.Most of these have used wavelengths in the Hartleyband of ozone near 300 nm, with tropospheric systemsrelying on wavelengths shorter than 300 nm, andstratospheric systems relying on wavelengths longerthan 300 nm. These DIAL wavelengths have beengenerated using flashlamp or laser pumped dye la-sers,1' 2 Raman shifted XeCl (308 nm) or KrF (248.4nm) excimer lasers using H2 and/or D2 cells,' 3-6 orRaman shifted quadrupled Nd:YAG laser with H2 andD2 cells7' 8 or D2 and HD cells.9
In this paper, the case is made that the Ramanshifted KrF laser can be better than Nd:YAG laserbased lidar systems for both daytime and nighttimemeasurements when the D2 and H2 cells are used inseries and the combined first Stokes shifts of D2 and H2(D2 + H2 ) are used, giving radiation shifted 7146 cm-'to 302.0 nm. With this approach, two pairs of lines areavailable for DIAL measurements of ozone: one pairat 277.0 and 291.7 nm for lower ozone densities andshorter ranges, and one at 291.7 and 302.0 nm forhigher ozone densities and longer ranges. Both lab-
The authors are with NASA Langley Research Center, Atmo-spheric Sciences Division, Hampton, Virginia 23665-5225.
Received 18 June 1990.
oratory measurements of Raman shifting and comput-er simulations of lidar performance are presented insupport of this claim.
The original motivation for placing two cells in serieswas twofold: first, that it should be possible to gener-ate a useful amount of 302.0-nm radiation via com-bined Raman shifting from D2 and H2; and, second,that the pump beam could be used more efficiently forgenerating the Stokes shifts of the individual gases.The latter follows from the fact that the entire pumpbeam is used in the first cell, and then the sizableresidual pump beam (50% of the input energy) is sentto the second cell. Note that generation of 302.0-nmradiation using a KrF laser and H2 and D2 was firstreported in Ref. 10, in which both gases were containedin one cell.
II. Laboratory Setup
The laboratory setup used is shown in Fig. 1, withthe parameters of the lasers, Raman cells, and lensesgiven in Table I. A dichroic beam splitter that reflect-ed 90% of the radiation of S1 of D2 and 30% at S2 of D2was placed near the exit of the first cell at an angle ofincidence of 450. (For the field system, a dispersingprism will be used to separate the various wavelengthsafter the D2 cell, since dichroic beam splitters do nothave a sharp enough cutoff.) For the measurement ofthe Raman conversion energies, a fused silica windowwas used to divert -4% of the beam after the secondcell to a lens followed by a dispersing prism, and theenergies at the separate wavelengths were measuredusing a pyroelectric energy monitor.
Fig. 1. Schematic diagram of the experimental setup used to studyRaman shifting of a KrF laser using two cells in series; see Table I for
the parameters used.
Table 1. Parameters of the Components Used In the Experiment
Laboratory Field
KrF laserWavelength 248.4 nmLinewidth >50 cm-'Pulse energy to 300 mJ 450 mJ
in far fieldPulse rep. freq. 10 Hz 49 Hz
Beam size 6 mm X 24 mmBeam divergence 0.3 mrad X 0.6 mrad
D2 Raman cellLength 1.0 m 1.7 m
H2 Raman cellLength 1.5 m 1.5 m
Lens focal lengthsInput to D2 cell 2.0 m 1.5 mCoupling between cells 0.75 m 0.75 m
The Raman cell windows (50.8-mm diameter) wereslightly canted to avoid back reflection and were de-signed to be thick enough (20 mm) so that they wouldnot rupture from a pressure 10 times the 650 psi oper-ating pressure. The maximum pressure allowable in-creases as the square of the thickness11'12:
W = t2 (87rm Smax)/(9m + 3), (1)
where W is the uniformly distributed load (lb), t is thethickness (in), m = 1/,u, where g is Poisson's ratio, andSmax is the maximum allowable stress (psi).
The load W is assumed to be spread over a diameterhalfway between the clear aperture (25 mm) and theedge of the window blank (50 mm). The Smax is takenhere as one-tenth of the tensile strength. For fusedsilica, Smax = 710 psi and , = 0.17 (Ref.13). Thus, fora 1000-psi cell, the window thickness should be 0.97 in.(2.5 cm).
The Raman shifting energy conversion efficiencywas determined by dividing the shifted energy by the
IS~
_a)._
Ld
0._
C00
a)CLW.
60
50
40
30
20
10
00 200 400 600 800 1000
Hydrogen Pressure (psi)
Fig. 2. Energy conversion efficiency vs H2 pressure for the firstStokes lines of H2 and D2 and for S1 of D2 incident in the cellcontaining H2 for the two-cell arrangement shown in Fig. 1. Forthese measurements, 220 mJ of KrF laser radiation was transmittedinto the D2 cell, which had a pressure of 315 psi, and 50 mJ of S1 of D2
was delivered to the H2 cell. Also plotted is the transmitted pumpradiation.
transmitted laser energy measured when the cells werefilled with helium. Thus, any losses due to the opticalcomponents were normalized out (see Fig. 2).
111. Experiments
A number of measurements were performed to de-termine the dependence of the energy conversion effi-ciency for Raman shifting of H2 and D2 in individualcells. With the lens focused in he center of the cell, thefocal length (f.l.) of the lens (1 or 2 m) did not affect thefirst Stokes (1) values for H2, although the secondStokes (S2) was somewhat better with the 1-m f.l. lens(21 vs 18%); (see Ref. 14 for slightly different results.)For D2 , a maximum conversion efficiency of 29% wasobserved with a 2-m f.l. lens, but only 22% with a 1-mf.l. lens. When the cell is too short, there is the dangerof damaging the windows with long focal length lenses,so that if shorter focal length lenses are used, theconversion efficiencies tend to be lower. Significantincreases in the S1 conversion efficiencies were notedwith increases in pressure for both gases: for H2, theincrease was from 25% at 200 psi to 32% at 650 psi; forD2, the increase was from 26% at 200 psi to 29% at 500psi. The S2 radiation remained constant to slightlydecreasing with pressure above 200 psi for both bases.Similar results have been observed by others (e.g, Refs.10, 15) with the explanation that the effect arises fromthe combined action of cascade stimulated Ramanshifting and four wave mixing.15
The next set of experiments led to the conclusionthat the first cell should contain D2 and the second, H2.When H2 was placed first, Raman shifting in the sec-ond cell does not occur for H2 pressures above -15 psi.The reason may be that in H2, which has much highergain than does D2, the divergence of the shifted radia-tion may be increased significantly above that of thepump beam16 -19 because of four-wave mixing.19 Inaddition, the pump beam may be depleted morestrongly where the beam has a higher power density,thereby affecting its divergence as well. Unlike H2, D2
does not increase the divergence of the Raman shiftedradiation, 9 and this may result from the lower gain ofD2.20
For the combined D2 + H2 Raman shift at 302.0 nm,the maximum conversion efficiency found was 14%.The conversion efficiency for S1 of D2 was optimizedfor a pressure of 300 psi, while the conversion efficien-cy for S1 of D2 was still increasing at a pressure of 650psi, which was the maximum pressure allowable in thecell with 20-mm thick windows with a factor of 10safety margin. The corresponding conversion effi-ciencies (related to the initial KrF laser pulse energy)for D2 were 27% for S1 (268.3 nm) and 17% for S2 (291.7nm), and for H2, 145 for S1 (277.0 nm), with minimalS2 (313.0 nm). (Slightly improved conversion effi-ciency for 302.0 nm in H2 was measured for a pressureof 1000 psi.) Note that 14% conversion efficiency a302.0 nm implies -50% conversion from 268.3 nm.This value can be compared with the 30% conversion to277.0 nm of the depleted pump in the H2 cell. Thereason for the difference is not understood.
Note that the transmitted pump energy increasesfor H2 pressures above 250 psi. This is probably be-cause the conversion efficiency for higher-order Stokesgeneration is reduced because of an index-of-refrac-tion mismatch that affects the coherence length forfour-wave mixing process (e.g., Refs. 15, 19, 21, 22).
If He buffer gas is added to the cell containing hy-drogen, it might be possible to improve the conversionefficiency for S119,23,24 and reduce the divergence'9 ofthe shifted radiation.
Attempts were made to generate 313.1-nm radiationin the second cell using the depleted pump beam thatpassed through the first cell containing D2. While thegeneration of 277.0-nm radiation (S1 of H2) was rea-sonably efficient (20%), it was found not to be possibleto generate 313.1-nm radiation with an efficiencyabove a few percent. It is thought that the beamdivergence limits S2 generation in the second cell con-taining D2 due primarily to the SRS process in the firstcell and possibly to the large number of optics withone-tenth wavefront distortion at 633 nm in the path.
Note that 302.0 nm is in the quasi solar blind spec-tral region,825 which should permit daytime measure-ments to be made using conventional interference fil-ters or a grating spectrometer in the receiver.
It was experimentally determined that for our setup,high quality fused silica windows sustained opticaldamage when laser pulses of 280 mJ passed trough a 2-m lens 1.5 m from the window. This represents anaverage energy density of -3 J/cm 2 or a power densityof 100 MW/cm2. In the future, to use long focal lengthlenses with the D2 Raman cell to obtain higher energyconversion efficiencies, we may use either speciallyprepared fused silica with colloidal silica AR overcoat-ings2 6-2 8 or MgF2 with the same AR coatings if colorcenter formation in fused silica is found to be a prob-lem. 2 9
It has been demonstrated in the laboratory that theuse of two cells in series, the first containing D2 and thesecond, H2, can be used to shift the radiation of the
KrF laser to 277.0, 291.7, and 302.0 nm. For an air-borne DIAL system, we expect to have pulse energiesof 45 mJ each at 291.7 and 302.0 nm and 36 mJ at 277.0nm using a single KrF laser operating at up to 49 Hz.This assumes that we have 450 mJ/pulse from the KrFlaser, along with 17% energy conversion efficiency for291.7 and 302.0-nm radiation and 14% conversion effi-ciency at 277.0 nm, and that these values are degradedby 40% to account for reflective and absorptive lossesdue to the optical components. While there is someconcern about the degradation of Raman shifting effi-ciency at higher laser prf (above -20 Hz),30 31 it ap-pears that if the focal length of the input lens is longenough, the problem is not important. We plan to usea 1.75-m f.l. lens which should lead at most to a fewpercent degradation.
IV. Lidar Simulations
The ozone absorption cross section at 263 K de-creases from 155.2 X 10-20 to 111.7 X 10-20 cm 2 in goingfrom 289.0 to 291.7 nm,32 while the cross sectionchanges from 43.4 X 10-20 to 28.9 X 10-20 cm 2 in goingfrom 299.0 to 302.0 nm. The differential cross sectionfor the 289.0/299.0-nm wavelength pair is 111.8 X 10-20cm2, while that for the 291.7/302.0-nm wavelength pairis 82.8 X 10-20 cm2, representing a decrease of 26%when using the Raman shifted KrF laser rather thanthe Raman shifted Nd:YAG laser for troposphericozone measurements using these wavelengths. Be-cause of the differences in differential absorption crosssection and atmospheric scattering, it would requireabout twice the average laser power to make measure-ments with the Raman shifted KrF laser compared tothe Raman shifted quadrupled Nd:YAG laser with thesame statistical uncertainty, assuming that solar back-ground and detector noise sources are small. The KrFlaser DIAL system should be able to generate at leasteight times the average power of a Nd:YAG laser DIALsystem at the 289.0-nm line using one Nd:YAG laser(or four times that for a dual laser system), making it abetter choice than a Raman shifted quadrupledNd:YAG laser system for both daytime and nighttimetropospheric DIAL systems when a 10-nm wavelengthseparation is used. In addition, the smaller absorp-tion cross sections of KrF produced laser wavelenthswould permit DIAL measurements to be made overeven longer ranges.
To further evaluate the capabilities of the proposedultraviolet (UV) DIAL system, simulations were run toestimate random errors for the two pairs of wave-lengths for the DIAL system operating in a nadir direc-tion from an aircraft flying at a 10-km altitude (Figs. 3,5) and for the system operating both in the nadir andzenith directions from an aircraft lying at a 5-km alti-tude (Figs. 4, 6).
The lidar system parameters for the enhancedNASA LaRC airborne DIAL system used in the simu-lations are given in Table II. The method used forcalculating random errors is similar to that using Eq.(2)33:
Earth surface reflectance: 0.1Filter bandwidth: 1.7%Field of view: 1.5 mradReceiver optical efficiency: 30%PMT quantum efficiency: 25%Aircraft speed: 200 m/s
RANDOM ERROR
Fig. 3. Simulations of random error vs altitude for the proposed UVDIAL system operated from an aircraft flying at an altitude of 10 km,operating at 49 Hz, and with a vertical resolution of 300 m and ahorizontal resolution of 12 km (this represents -1 min. of flying timeor 2900 laser pulses). The ozone profile used was 150% of the U.S.Standard34 (60 ppb at the ground). Nighttime background condi-
tions were assumed.
RANDOM ERROR
Fig. 4. Similar to Fig. 3 with the lidar at 5-km altitude, both up anddown looking, for 200-m vertical resolution and 3-km horizontalresolution, and with the tropical wet season ozone profile of 10 to 30
ppb. 3 6
0 A ,n,) E-.egy (J) 49H2
* 277/292.0 30----------------- 292/302.0 30AZ-300 .- A XI 2k-
E __
2 -\
0 0 - I 0 15 20
RANDOM ERROR
Fig. 5. Same as Fig. 3 using the U.S. Standard ozone profile.
i=2
An_ 1 = (P+B+D),
n N"/22an(R - R2 ) E p2jj=1
(2)
where N is the number of shots averaged, a the differ-ential absorption cross section, R, - R 2 is the differen-tial range of the measurement, P is the lidar signal, irefers to on or off, Jrefers to the range cell, B is the daybackground signal, and D is the detector dark currentsignal.
The molecular model atmosphere used is the U.S.Standard 1976 model,34 and the aerosol profiles werederived from the Deirmendjian water haze Model L35
with a 23-km ground visibility. Two ozone modelprofiles have been used in these simulations: a lowozone model for the troposphere from observationsover the tropical rain forest during the wet season36
and a high ozone model using the U.S. Standard mod-el34 plus 50%. The 50% limit for the high ozone modelwas selected on the basis of the midlatitude ozonemodel34 which exhibits about 50% variability in themid and low troposphere. For these simulations atwavelengths <302.0 nm, it was found that the daybackground had insignificant influence because of acombination of the extinction by atmospheric ozone25
and the background rejection by the grating spectrom-eter planned for use in our field system.
It is necessary to average lidar signals along thevertical (z) and horizontal (x) directions to achieve thedesired precision in DIAL measurements. In the ab-sence of large uncorrectable systematic errors, the ran-dom errors determine the useful range and the wave-length pair suitable for DIAL measurements. Ingeneral, the 277.0/291.7-nm pair is more suitablewhere low atmospheric ozone concentrations are mea-sured (Fig. 4), and the 291.7/302.0-nm pair is moresuitable where moderate to high ozone levels are en-countered (Fig. 3) and when long-range (5- to 10-km)measurements are needed.
The systematic errors in the measurement of ozoneusing the proposed system arise mainly from the rela-tively large separation (10 to 15 nm) between the onand off wavelengths. Because of this separation, themolecular and aerosol scattering and extinction in theatmosphere at the on and the off wavelengths are
Fig. 6. Same as Fig. 4 using the U.S. Standard ozone profile.
different, giving rise to small systematic errors whichhave been previously discussed in the literature.28 37
The correction due to molecular extinction can bereadily made using a model atmosphere as in the U.S.Standard atmosphere. In regions of inhomogeneousaerosol distribution, systematic effects due to atmo-spheric backscatter can be corrected by using the Ber-noulli method.837 Other systematic effects could arisefrom interference by SO2 absorption. However, gen-erally, the SO2 concentrations are very low (<1 ppb)compared with ozone and a correction for SO2 interfer-ence is not required.
V. Concluding Remarks
Laboratory measurement results have been present-ed showing that a KrF laser in combination with Ra-man shifting using D2 and H2 cells in series can gener-ate UV radiation at 277.0, 291.7, and 302.0 nm withenergy conversion levels between 14 and 27%. Simula-tions show that using this source in the proposed KrFlaser-based airborne DIAL system, precise measure-ments (5%) of ozone are expected over a 5-10-km rangein the troposphere with high spatial resolution suitablefor atmospheric studies involving production, loss, andtransport of ozone in the atmosphere.
For stratospheric ozone DIAL measurements, a Ra-man shifted KrF laser could be used with the pair oflines at 302.0 nm (D2 + H2) and 313.1 nm (S2 of H2),although the initial KrF laser beam will have to be splitbefore it enters the H2 and D2 cells.
The authors would like to thank A. F. Carter, N.Barnes, J. Williams, J. Bradshaw (Georgia Institute ofTechnology), and H. Komine (Northrop Corp.) formany helpful discussions; and K. Benton, N. Mayo, J.Moen, J. Siviter, and J. Williams for their technicalsupport; S. Kooi for preparing some of the graphics;and two anonymous reviewers for their helpful com-ments.
References
1. G. J. Megie, G. Ancellet, and J. Pelon, "Lidar Measurements ofOzone Vertical Profiles," Appl. Opt. 24, 3454-3463 (1985).
2. E. V. Browell, "Differential Absorption Lidar Sensing ofOzone," Proc. IEEE 77,419-432 (1989).
3. 0. Uchino, M. Tokunaga, M. Maeda, and Y. Miyazoe, "Differen-tial-Absorption-Lidar Measurement of Tropospheric Ozonewith Excimer-Raman Hybrid Laser," Opt. Lett. 8, 347-349(1983).
4. Ts. Shibata, T. Fukuda, T. Narikiyo, and M. Maeda, "Evalua-tion of the Solar-Blind Effect in Ultraviolet Ozone Lidar withRaman Lasers," Appl. Opt. 26, 2604-2608 (1987).
5. W. Steinbrecht, K. W. Rothe, and H. Walther, "Lidar Setup forDaytime and Nighttime Probing of Stratospheric Ozone andMeasurements in Polar and Equatorial Regions," Appl. Opt. 28,3616-3624 (1989).
6. I. S. McDermid, S. M. Godin, and L. 0. Lindqvist, "Ground-Based Laser DIAL System for Long-Term Measurements ofStratospheric Ozone," Appl. Opt. 29, 3603-3612 (1990).
7. G. Ancellet, A. Papayannis, J. Pelon, and G. Megie, "DIALTropospheric Ozone Measurement using a Nd:YAG Laser andthe Raman Shifting Technique," J. Atmos. Oceanol. Technol. 6,832-839 (1989).
8. A Papayannis, G. Ancellet, J. Pelon, and G. Megie, "Multiwave-length Lidar for Ozone Measurements in the Troposphere andthe Lower Stratosphere," Appl. Opt. 29, 467-476 (1990).
9. D. A. Haner and I. S. McDermid, "Stimulated Raman Shifting ofNd:YAG Fourth Harmonic (266 nm) in H2, HD and D2," IEEE J.Quantum Electron. QE-26, 1292-1298 (1990).
10. T. R. Loree, R. C. Sze, D. L. Barker, and P. B. Scott, "New Linesin the UV: SRS of Excimer Laser Wavelengths," IEEE J.Quantum Electron. QE-15, 337-342 (1979).
11. S. Timonshenko, Theory of Plates and Shells, (McGraw-Hill,New York, 1940).
12. A. Pope and K. L. Goin, High-Speed Wind Tunnel Testing,(Krueger, Huntington, NY, 1978, reprint of edition published byWiley, New York).
13. Heraeus Amersil, Inc., "Fused Quartz and Fused Silica for Op-tics," p. 28.
14. P. Falsini, R. Pini, R. Salimbeni, M. Vannini, A. F. M. YusufHaider, and R. Buffa, "Simple and Efficient H2 Raman Conver-sion of a XeCl Laser with a Variable Numerical Aperture Cou-pling Geometry," Opt. Commun. 53, 421-424 (1985).
15. Z. Chu, U. N. Singh, and T. D. Wilkerson, "Multiple WavelengthGeneration by Stimulated Raman Scattering in Molecular Gas-es, submitted to Appl. Opt.
16. B. Bobbs and C. Warner, "Absence of Second Stokes in a RamanGenerator with no Four-Wave Mixing," Opt. Lett. 11, 88-90(1986).
17. K. J. Druhl, S. Shakir, and M. Yousaf, "Approximate Theory ofStokes Amplification and Conversion Valid at Large Gain," Opt.Lett. 11, 446-448 (1986).
18. J. L. Carlsten, J. Rifkin, and D. C. MacPherson, "Spatial ModeStructure of Stimulated Stokes Emission from a Raman Gener-ator," J. Opt. Soc. Am. B 3, 1476-1482 (1986).
19. D. Diebel, M. Bristow, and R. Zimmermann, "Stokes ShiftedLaser Lines in KrF-Pumped Hydrogen: Reduction of BeamDivergence by Addition of Helium," Appl. Opt. 30, 626-628(1991).
20. J. J. Ottusch and D. A. Rockwell, "Measurement of RamanGain Coefficients of Hydrogen, Deuterium, and Methane,"IEEE J. Quantum Electron. QE-24, 2076-2080 (1988).
21. G. V. Venkin, L. L. Kulyuk, and D. I. Maleev, "Investigation ofStimulated Raman Scattering in Gases Excited by Fourth Har-monic of Neodymium Laser Radiation," Sov. J. Quantum Elec-tron. 5, 1348-1351 (1976).
22. T. R. Loree, R. C. Sze, and D. L. Barker, "Efficient RamanShifting of ArF and KrF Laser Wavelengths," Appl. Phys. Lett.31, 37-39 (1977).
23. A. Luches, V. Nassisi, and M. R. Perrone, "Improved ConversionEfficiency of XeCl Radiation to the First Stokes at Pump Ener-gy," Appl. Phys. B 47, 101-105 (1988).
24. Q. Lou, "Research on the Characteristics of H2 Raman Conver-sion Pumping by a 1-J XeCl Excimer Laser," J. Appl. Phys. 66,2265-2273 (1989).
25. M. Maeda, T. Shibata, S. Ichimori, T. Ono, and H. Akiyoshi,"Solar-Blind UV Ozone Lidars," submitted to Appl. Opt.
26. J. G. Wilder, "Porous Silica AR Coating for Use at 248 nm or 266nm," Appl. Opt. 23, 1448-1449 (1984).
27. I. M. Thomas, "High Laser Damage Threshold Porous SilicaAntireflective Coating," Appl. Opt. 25, 1481-1483 (1986).
28. F. O'Neill, I. N. Ross, D. Evans, J. U. D. Langridge, B. S. Bilan,
and S. Bond, "Colloidal Silica Coatings for KrF and Nd:GlassLaser Applications," Appl. Opt. 26, 828-832 (1987).
29. G. C. Escher, "KrF Laser Induced Color Centers in CommercialFused Silicas," Proc. Soc. Photo-Opt Instrum. Eng. 998, 30-37(1988).
30. V. Yu. Baranov et al., "Conversion of X = 308-nm Radiation atPulse Repetition Frequencies up to 600 Hz by Stimulated Ra-
man Scattering in Compressed Hydrogen," Sov. J. QuantumElectron. 18, 1272-1277 (1988).
31. Q. Lou, T. Yagi, and H. Saito, "KrF/H 2 Raman Conversion atHigh Repetition Rate using a Hydrogen Gas Circulating Sys-tem," J. App. Phys. 67, 6591-6591 (1990).
32. L. T. Molina and M. J. Molina, "Absolute Absorption CrossSections of Ozone in the 185- to 350-nm Wavelength Range," J.Geophys. Res. 91, 14501-14508 (1986).
33. E. E. Remsberg and L. L. Gordley, "Analysis of DifferentialAbsorption Lidar from the Space Shuttle," Appl. Opt. 17, 624-630 (1978).
34. U.S. Standard Atmosphere, 1976, NOAA, NASA, and USAF,(U.S. GPO, Washington, DC, NOAA-S/T 76-1562, 1976).
35. D. Deirmendjian, Electromagnetic Scattering on Spherical Po-lydispersions (Elsevier, New York, 1969).
36. V. W. J. H. Kirchhoff, I. M. 0. da Silva, and E. W. Browell,"Ozone Measurements in Amazonia: Dry Season vs. Wet Sea-son," J. Geophys. Res. 95, 16,913-16,926 (1990).
37. E. V. Browell, S. Ismail, and S. T. Shipley, "Ultraviolet DIALMeasurements of 03 Profiles in Regions of Spatially Inhomoge-neous Aerosols," Appl. Opt. 24, 2827-2836 (1985).
