o '795 Raman iidar profiling of water vapor and aerosols over the Southern Great Plains Richard A. Ferrare NASA Langley Research Center, Hampton, Irtrginia, 23681, USA r.ferrare,_Jarc, nasa.g9_ David D. Turner Pacifu: Northwest National Laboratory, MS K9-38, Richland, Washington, 99352, USA Lorraine A. Heilman Science Applications International Corporation, Hampton, Virginia 23681, USA Tim P. Tooman, John E. M. Goldsmith Sandla National Laboratories, L|vermore, California, 94551, USA Oleg Dubovik Science Systems Applications Incorporated NASA GSFC, Greenbelt, Maryland 20771, USA John A. Ogren NOAA Climate Monitoring and Diagnostics Laboratory, Boulder, Colorado 80303, USA Abstract: Raman lidar profiles of aerosol extinction and water vapor are used to characterize the vertical variability of aerosols and water vapor over northern Oklahoma OCIS codes: (010.3640) Lidar, (010.3920) Meteorology 1. Introduction The U.S. Depamnent of Energy Atmospheric Radiation Measurement (ARM) program has developed and maintained an operational Raman lidar at its Southern Great Plains (SGP) Cloud and Radiation Testhed (CART) site in north-central Oklahoma (36.61 N, 97.49W). The CART Raman Lidar (CARL) was specifically designed to continuously measure profiles of water vapor and aerosols throughout the diurnal cycle [l]. We have developed and implemented automated algorithms which use the Raman lidar data to retrieve profiles of water vapor mixing ratio, relative humidity, aerosol extinction and aerosol backseatter coefficients, linear depolarization ratio, and a cloud mask, Additional integrated products, such as total preeipitable water vapor and aerosol optical thickness, are also routinely computed. We discuss the lidar system, the methodology for deriving water vapor and aerosol profiles, and how these profiles have been used to characterize the clear sky above the SGP site. 2. Raman Lidar System CARL uses a tripled Nd:YAG laser, operating at 30 Hz with 350-400 millijoule pulses to transmit light at 355 nm A 61-crn diameter telescope collects the light backsealtered by molecules and aerosols at the laser wavelength and the Raman scattered light from water vapor (408 run) and nitrogen (387 nm) molecules. A beam expander reduces the laser beam divergence to 0.I mrad, thereby permitting the use of a narrow (0.3 mrad) as well as a wide (2 mrad) field of view. The narrow field of view, coupled with the use of narrowband (--0.4 mn ban@ass) filters, reduces the background skylight and, therefore, increases the maximum range of the aerosol and water vapor profiles measured during daytime operations. The signals from the Rayleigh/Mie (355 nm), Raman N2 (387 run), and Raman H20 (408 nm) channels are detected by photomultiplier tubes and recorded using photon counting with a vertical resolution of 39 meters. In normal operation, data are recorded as l-rain profiles corresponding to the accumulation of signals from approximately 1780 laser shots. During post-processing, the vertical and/or temporal resolutions can be degraded to improve the signal/noise ratio. Water vapor mixing ratio profiles are computed using the ratio of the Raman water vapor signal to the Raman nitrogen signal. These water vapor profiles are integrated with altitude to derive precipitable water vapor (PWV). The CARL water vapor mixing ratio profiles and PWV retrievals are automatically calibrated using the coincident nighttime measurements of precipitable water vapor 0aWV) from the microwave radiometer (MWR) [2]. Relative humidity profiles are computed using water vapor mixing ratio profiles from CARL and temperature profiles Optical Remote Sensing of the Atmosphere Topical Meeting, February 5-8, 2001, Coeur d'Alene, Idaho derived using a physical retrieval from the high resolution downwelling longwave radiance dataobserved by the Atmospheric Emitted Radiance Intefferometer (AERI) [3]. When averaged into 10-minute profiles, the lidar water vapor profiles are usually limited by the solar background to altitudes below approximately 3-4 km during the daytime, but typically extend to 10-12 kmatnight. Profiles of aerosol scattering ratio, which is the ratio of aerosol+molecular scattering to molecular scattering, are derived using the Raman nitrogen signal and the signal detected at the laser wavelength. This profile is automatically calibrated by normalizing it to unity in a clear-air (cloud and aerosol free region) between 6-9 kin. Aerosol volume backscattering cross section profiles are then computed using the aerosol scattering ratio and molecular scattering cross section profiles derived from AERIretrievals of atmospheric density. For altitudesabove 800 m where the overlap function for the nitrogen channel is unity, aerosol extinction coefficient profiles are computed from the derivative of the logarithm of the Raman nitrogen signal with respect to range. For altitudes below 800 m, aerosol extinction profiles are derived from the aerosol extinction/backscatter ratio datafrom 800- 1000 m together with the backscatter profile, which extends to 60 mabove the surface. Aerosol optical thickness values are then computed by integrating the extinction coefficient profile from the surface to 7 km (or cloud base, whichever islower). 3. CARL Measurements Figure la shows average aerosol extinction profiles that were computed as a function of optical thickness to characterize the vertical distribution of aerosols. Figure lb show7s .wate.r va+p..or m..ixin.ig.ra.t. iop..r7ofile..s co.mput.ed f.rom. the Raman lidardata. The scale height of the aerosol extinction profiles varies with both season and with APT and shows more variability thandoes the scale height forwater vapormixing ratio. :I +iL z_[,.......... _*_01o,-........,,'=...............T...[........ _,o[• =,=',,,_r' 000,. °f 1+0. 10 O.3-OA. 0_._ .__ ..,_,-,-_,_. O! 0.00 Aero0Aso0l Ext0l.n2c0tlofi 0.3(k0in -_) 0.+0.00 Aero0sAol0 Ex_nc2t0ion 0(.k3i0n -_) 0.40 {_N;_er2Vapor4 Mixi4n1g R8atio 1(0g/kg)12 0Water2 Vapo4r Mfxing_stio!?g_gI_ Figure la (left) Seasonal mean profiles, derived from data collected by the Raman lidar from April 1998 - January 2000, of aerosol extinction coefficient as afunction of aerosol optical thickness (APT). The inset plots indicate the fraction of the total number of samples forthat season that was in the given bin. The range of APT isindicated for each profile. Profiles were not plotted for bins with less than 5% of the total samples for the season. Black dots represent scale height (i.e. where distribution falls to 1/o of total integrated value.) 1b(right) Same except for water vapor mixing ratio profiles as afunction ofprecipitable water vapor (PWV). Profiles of the aerosol extinctionfoackscattering ratio, S_, were derived from the CARL data acquired during 1998 and 1999. Mean values of S_at355 nmtypically ranged between 60 to70 st. Large (>15%) variations in the vertical profile of S, occurred about 20% of the time, which implies that significant variability in the vertical distribution of the aerosol size distribution anti/or composition often occurs. The Raman lidar measurements of S, were compared withestimates of particlesize and refractive indexderived from analgorithm that usesground-based Sun photometer measurements of Sun and skyradiance [4]. Figure 2ashows that,duringcases when S, did notvary greatly with altitude,the lidarand Sunphotometer measurements of S, showed good agreement. • , - = • • - , - ! --, - ,ZDI SIo_l.¢lg ...... _m 1 "'[',' t-i _2 ::_'.I0 '+0.+11"0._12"I0.:15"0.:14 " 0.+15"O.+_l"0. Elfeceve Rdu= (Ym,emode) O,m) Figure 2a (left) Comparison of S, derived from CARL data and retrieved from Sun and sky radiance inversion. 2b. (right) Correlation between S=and effective radius of the aerosol free (accumulation) mode retrieved from Sun and sky radiance inversion. Optical Remote Sensing of the Atmosphere Topical Meeting, February 5-8, 2001, Coeur d'Alene, Idaho We have also begun comparing profiles of aerosol extinction and relative humidity derived from the CARL data with profiles derived from in situ sensors on a light aircraft flown periodically above the SGP site. These In-situ Aerosol Profiling (IAP) sensors include a nephelometer and particle soot/absorption photometer to measure dry (RH <= 40%) aerosol scattering and absorption, respectively, and a Vaisala humicap hygrometer to measure relative humidity. Figure 3a shows a comparison of relative humidity and Figure 3b shows a comparison of aerosol extinction. For the aerosol extinction comparison, the CARL profile at 355 nm was scaled to 550 nm using the aerosol scattering wavelength dependence measured by the nephelometer. The nephelometer aerosol scattering and extinction profiles were scaled to the ambient relative humidity using a humidification (i.e. hygroscopic growth) factor derived from measurements of light scattering as a function of relative humidity made at the surface. Both the lidar measurements of S, and the aircraft measurements of aerosol scattering and absorption are being used to examine the vertical variability of aerosols above the SGP site. ..... , .... , .... o,.. "----_7.,.-m.7_-4- 2 Fig. 3a) (left) Comparison of relative humidity derived from CARL data and measured insitu onIAP aircraft between 21:00-22:30 UT on May 4, 2000.3b) (right) Comparison of aerosol extinction profiles derived from CARL and insitu aircraft data during this same period. 4. Summary Profiles of water vapor mixing ratio, relative humidity, aerosol backscattering and extinction derived from CART Raman lidar data have been used to help study the vertical variability of water vapor and aerosols. These products have begun to demonstrate how this automated lidar system can be used to remotely characterize the clear sky over the ARM SGP site. 5. Acknowledgements SGP CART Raman lidar, CIMEL Sun photometer, and AERI data were obtained from the Atmospheric Radiation Measurement (ARM) Program sponsored by the U.S. Department of Energy, Office of Energy Research, Office of Health and Environmental Research, Environmental Sciences Division. The CIMEL Sun photometer is also part of AERONET, a network of Sun photometers managed by B. N. Holben, NASA/GSFC. The IAP program is a joint effort between the ARM program and the NOAA Climate Monitoring and Diagnostics Laboratory. The DOE ARM and NASA EOS Validation Programs provided funding for this work. We thank Wayne Feltz for providing the temperature profiles retrieved from the AERI. 6. References Goldsmith, J.E.M., F.H. Blair, S.E. Bisson, and D.D. Turner: Turn-key Raman lidat for profiling atmospheric water vapor, clouds, and aerosols. Appl. Opt, 37, 4979-4990, (1998). Turner, D.D., and J.E.M. Goldsmith: Twenty-Four hour Raman lidar measurements during the Atmospheric Radiation Measurement program's 1996 and 1997 water vapor intensive observation periods. ,1"A.tmos. Oceanic Technol., 16, 1062-1076, (1999). Feltz, W.F., W.L. Smith, R_O. Knuteson, H.E. Revercomb, H.M. WoolK and H.B. Howell, 1998: Meteorological applications of temperature and water vapor retrievals from the ground-based atmospheric emitted radiance interferometer (AERI). J.Appl. Meteor., 37, 857-875, (1998). Dubovik, O. A. Smimov, B.N. Holben, M.D. King, Y.3. Kaufrnan, T.F. Eck, and I. Slutsker, Accuracy assessment ofaerosol optical properties retrieved from AERONET Sun-and sky-radiance measurements, £ Geophys. Res., 105, 9791-9806, (2000). Optical Remote Sensing of the Atmosphere Topical Meeting, February 5-8, 2001, Coeur d'Alene, Idaho