• Ansmann, A., , Riebesell M. , , and Weitkamp C. , 1990: Measurement of atmospheric aerosol extinction profiles with a Raman lidar. Opt. Lett, 15 , 746748.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ansmann, A., , Riebesell M. , , Wandinger U. , , Weitkamp C. , , Voss E. , , Lahmann W. , , and Michaelis W. , 1992a: Combined Raman elastic-backscatter lidar for vertical profiling of moisture, aerosol extinction, and lidar ratio. Appl. Phys, B55 , 1828.

    • Search Google Scholar
    • Export Citation
  • Ansmann, A., , Wandinger U. , , Riebesell M. , , Weitkamp C. , , and Michaelis W. , 1992b: Independent measurement of extinction and backscatter profiles in cirrus clouds by using a combined Raman elastic-backscatter lidar. Appl. Opt, 31 , 71137131.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Arshinov, Y., , and Bobrovnikov S. , 1999: Use of a Farby-Perot interferometer to isolate pure rotational Raman spectra of diatomic molecules. Appl. Opt, 38 , 46354638.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Behrendt, A., , and Reichardt J. , 2000: Atmospheric temperature profiling in the presence of clouds with a pure rotational Raman lidar using an interference-filter-based polychromator. Appl. Opt, 39 , 13721378.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bösenberg, J., and and Coauthors, 2001: EARLINET: A European aerosol research lidar network. Advances in Laser Remote Sensing: Selected Papers Presented at the 20th International Laser Radar Conference, A. Dabas, C. Loth, and J. Pelon, Eds., Ecole Polytechnique, 155–158.

    • Search Google Scholar
    • Export Citation
  • Cairo, F., , Donfrancesco G. Di , , Adriani A. , , Pulvirenti L. , , and Fierli F. , 1999: Comparisons of various linear depolarization parameters measured by lidar. Appl. Opt, 38 , 44254432.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Campbell, J. R., , Hlavka D. L. , , Spinhirne J. D. , , Ferrare R. A. , , and Turner D. D. , 2000: Automated aerosol retrieval algorithms for ARM micro pulse lidars. Preprints, Symp. on Lidar Atmosphere Monitoring, Long Beach, CA, Amer. Meteor. Soc., 71–74.

    • Search Google Scholar
    • Export Citation
  • Cooney, J. A., 1972: Measurement of atmospheric temperature profiles by Raman backscatter. J. Appl. Meteor, 11 , 108112.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dubovik, O., , and King M. D. , 2000: A flexible inversion algorithm for retrieval of aerosol optical properties from Sun and sky radiance measurements. J. Geophys. Res, 105 , 2067320696.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Feltz, W. F., , Smith W. L. , , Knuteson R. O. , , Revercomb H. E. , , Woolf H. M. , , and Howell H. B. , 1998: Meteorological applications of temperature and water vapor retrievals from the ground-based atmospheric emitted radiance interferometer (AERI). J. Appl. Meteor, 37 , 857875.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Feltz, W. F., and and Coauthors, 2000: Retrieval and visualization of AERIPLUS temperature and moisture profiles for assimilation into ARM single-column models. Proc. 10th Atmospheric Radiation Measurement (ARM) Science Team Meeting, San Antonio, TX. [Available online at http://www.arm.gov/docs/documents/technical/conf_0003/feltz_wf.pdf.].

    • Search Google Scholar
    • Export Citation
  • Ferrare, R. A., , Melfi S. H. , , Whiteman D. N. , , and Evans K. D. , 1992: Raman lidar measurements of Pinatubo aerosols over southeastern Kansas during November-December 1991. Geophys. Res. Lett, 19 , 15991602.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ferrare, R. A., , Melfi S. H. , , Whiteman D. N. , , Evans K. D. , , Schmidlin F. J. , , and O'C. Starr D. , 1995: A comparison of water vapor measurements made by Raman lidar and radiosondes. J. Atmos. Oceanic Technol, 12 , 11771195.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ferrare, R. A., , Melfi S. H. , , Whiteman D. N. , , Evans K. D. , , and Leifer R. , 1998: Raman lidar measurements of aerosol extinction and backscattering: Methods and comparisons. J. Geophys. Res, 103 , 1966319672.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ferrare, R. A., and and Coauthors, 2000a: Comparisons of LASE, aircraft, and satellite measurements of aerosol optical properties and water vapor during TARFOX. J. Geophys. Res, 105 , 99359948.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ferrare, R. A., , Turner D. D. , , Heilman L. A. , , Feltz W. F. , , Tooman T. P. , , Dubovik O. , , and Halthore R. , 2000b: Characterization of the atmospheric state over the SGP using Raman lidar and AERI/GOES measurements. Proc. 10th Atmospheric Radiation Measurement (ARM) Science Team Meeting, San Antonio, TX. [Available online at http://www.arm.gov/docs/documents/technical/conf_0003/ferrare-ra.pdf.].

    • Search Google Scholar
    • Export Citation
  • Ferrare, R. A., , Turner D. D. , , Heilman L. A. , , Feltz W. F. , , Dubovik O. , , and Tooman T. P. , 2001: Raman lidar measurements of aerosol extinction-to-backscatter ratio over the Southern Great Plains. J. Geophy. Res.,106, 20 333–20 348.

    • Search Google Scholar
    • Export Citation
  • Gobbi, G. P., 1998: Polarization lidar returns from aerosols and thin clouds: A framework for the analysis. Appl. Opt, 37 , 55055508.

  • Godin, S., and and Coauthors, 1999: Ozone differential absorption lidar algorithm intercomparison. Appl. Opt, 38 , 62256236.

  • Goldsmith, J. E. M., , Blair F. H. , , Bisson S. E. , , and Turner D. D. , 1998: Turn-key Raman lidar for profiling atmospheric water vapor, clouds, and aerosols. Appl. Opt, 37 , 49794990.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Grund, C. J., , and Eloranta E. W. , 1991: University of Wisconsin high spectral resolution lidar. Opt. Eng, 30 , 612.

  • Harrison, L., , and Michalsky J. , 1994: Objective algorithms for the retrieval of optical depths from ground-based measurements. Appl. Opt, 33 , 51265132.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Harrison, L., , Michalsky J. , , and Berndt J. , 1994: Automated multifilter rotating shadow-band radiometer: An instrument for optical depth and radiation measurements. Appl. Opt, 33 , 51185125.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Holben, B. N., and and Coauthors, 1998: AERONET—A federated instrument network and data archive for aerosol characterization. Remote Sens. Environ, 66 , 116.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kato, S., and and Coauthors, 2000: A comparison of aerosol optical thickness derived from ground-based and airborne measurements. J. Geophys. Res, 105 , 97919806.

    • Search Google Scholar
    • Export Citation
  • Keckhut, P., , Hauchecorne A. , , and Chanin M. L. , 1993: A critical review of the database acquired for the long-term surveillance of the middle atmosphere by the French Rayleigh lidars. J. Atmos. Oceanic Technol, 10 , 850867.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Klett, J. D., 1981: Stable analytical inversion solution for processing lidar returns. Appl. Opt, 20 , 211220.

  • Melfi, S. H., , Whiteman D. N. , , and Ferrare R. A. , 1989: Observation of atmospheric fronts using Raman lidar moisture measurements. J. Appl. Meteor, 28 , 789806.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Menzel, W. P., , Holt F. C. , , Schmidt T. J. , , Aune R. M. , , Schreiner A. J. , , Wade G. S. , , and Gray D. G. , 1998: Application of GOES-8/9 soundings to weather forecasting and nowcasting. Bull. Amer. Meteor. Soc, 79 , 20592077.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Miller, E. R., , Wang J. , , and Cole H. L. , 1999: Correction for dry bias in Vaisala radiosonde RH data. Proc. Ninth Atmospheric Radiation Measurement (ARM) Science Team Meeting, Tucson, AZ. [Available online at http://www.arm.gov/pub/docs/documents/technical/conf_9903/miller-er-99.pdf.].

    • Search Google Scholar
    • Export Citation
  • Murayama, T., , Okamoto H. , , Kaneyasu N. , , Kamataki H. , , and Miura K. , 1999: Application of lidar depolarization measurement in the atmospheric boundary layer: Effects of dust and sea-salt particles. J. Geophys. Res, 104 , 3178131792.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Murayama, T., and and Coauthors, 2001: Lidar network observation of Asian dust. Advances in Laser Remote Sensing: Selected Papers Presented at the 20th International Laser Radar Conference, A. Dabas, C. Loth, and J. Pelon, Eds., Ecole Polytechnique, 169–172.

    • Search Google Scholar
    • Export Citation
  • Nedeljkovic, D., , Hauchecorne A. , , and Chanin M. L. , 1993: Rotational Raman lidar to measure the atmospheric temperature from the ground to 30 km. IEEE Trans. Geosci. Remote Sens, 31 , 90101.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Penney, G. M., , and Lapp M. , 1976: Raman-scattering cross-section for water vapor. J. Opt. Soc. Amer, 66 , 422425.

  • Pilinis, C., , Sandis S. N. , , and Seinfeld J. H. , 1995: Sensitivity of direct climate forcing by atmospheric aerosols to aerosol size and composition. J. Geophys. Res, 100 , 1873918754.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Revercomb, H. E., 2000: Science plan for the 3d ARM water vapor IOP and the ARM-FIRE water vapor experiment (AFWEX) at the ARM SGP CART site, Lamont, OK. [Available online at http://www.arm.gov/docs/iops/2000/sgp2000afwex/afwex.html.].

    • Search Google Scholar
    • Export Citation
  • Russell, P. B., , Swissler T. J. , , and McCormick M. P. , 1979: Methodology for error analysis and simulation of lidar measurements. Appl. Opt, 18 , 37833797.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sassen, K., 1991: The polarization lidar technique for cloud research: A review and current assessment. Bull. Amer. Meteor. Soc, 72 , 18481866.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sherlock, V., , Hauchecorne A. , , and Lenoble J. , 1999: Methodology for the independent calibration of Raman backscatter water-vapor lidar systems. Appl. Opt, 38 , 58165837.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shibata, T., , Sakai T. , , Hayashi M. , , Ojio T. , , Kwon S-A. , , and Iwasaka Y. , 1996: Raman lidar observations: Simultaneous measurements of water vapor, temperature, and aerosol vertical profiles, part II. J. Geomag. Geoelectr, 48 , 11371144.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shipley, S. T., , Tracy D. H. , , Eloranta E. W. , , Trauger J. T. , , Sroga J. T. , , Roesler F. L. , , and Weinman J. A. , 1983: High spectral resolution lidar to measure optical scattering properties of atmospheric aerosols: 1. Theory and implementation. Appl. Opt, 22 , 37163724.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Smirnov, A., , Holben B. N. , , Eck T. F. , , Dubovik O. , , and Slutsker I. , 2000: Cloud screening and quality control algorithms for the AERONET database. Remote Sens. Environ.,73, 337–349.

    • Search Google Scholar
    • Export Citation
  • Smith, W. L., , Feltz W. F. , , Knuteson R. O. , , Revercomb H. E. , , Howell H. B. , , and Woolf H. M. , 1999: The retrieval of planetary boundary layer structure using ground-based infrared spectral radiance measurements. J. Atmos. Oceanic Technol, 16 , 323333.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Spinhirne, J. D., , Reagan J. A. , , and Hermann B. M. , 1980: Vertical distribution of aerosol extinction cross section and inference of aerosol imaginary index in the troposphere by lidar technique. J. Appl. Meteor, 19 , 426438.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stokes, G. M., , and Schwartz S. E. , 1994: The Atmospheric Radiation Measurement (ARM) Program: Programmatic background and design of the cloud and radiation testbed. Bull. Amer. Meteor. Soc, 75 , 12011221.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Turner, D. D., , and Goldsmith J. E. M. , 1999: Twenty-four-hour Raman lidar measurements during the Atmospheric Radiation Measurement program's 1996 and 1997 water vapor intensive observation periods. J. Atmos. Oceanic Technol, 16 , 10621076.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Turner, D. D., , Shippert T. R. , , Brown P. D. , , Clough S. A. , , Knuteson R. O. , , Revercomb H. E. , , and Smith W. L. , 1998: Long-term analysis of observed and line-by-line calculations of longwave surface spectral radiance and the effect of scaling the water vapor profiles. Proc. 8th Atmospheric Radiation Measurement (ARM) Science Team Meeting, Tucson, AZ. [Available online at http://www.arm.gov/pub/docs/documents/technical/conf_9803/turner-98.pdf.].

    • Search Google Scholar
    • Export Citation
  • Turner, D. D., , Feltz W. F. , , and Ferrare R. A. , 2000: Continuous water vapor profiles from ground-based operational active and passive remote sensors. Bull. Amer. Meteor. Soc, 81 , 13011317.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Turner, D. D., , Ferrare R. A. , , and Brasseur L. A. , 2001: Average aerosol extinction and water vapor profiles over the Southern Great Plains. Geophys. Res. Lett., in press.

    • Search Google Scholar
    • Export Citation
  • Vaughan, G., , Waring D. P. , , Thomas L. , , and Metev V. , 1988: Humidity measurements in the free troposphere using Raman backscatter. Quart. J. Roy. Meteor. Soc, 114 , 30683082.

    • Search Google Scholar
    • Export Citation
  • Whiteman, D. N., 1999: Application of statistical methods to the determination of slope in lidar data. Appl. Opt, 38 , 33603369.

  • Whiteman, D. N., , and Melfi S. H. , 1999: Cloud liquid water, mean droplet radius, and number density measurements using a Raman lidar. J. Geophys. Res, 104 , 3141131419.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Whiteman, D. N., , Melfi S. H. , , and Ferrare R. A. , 1992: Raman lidar system for the measurement of water vapor and aerosols in the Earth's atmosphere. Appl. Opt, 31 , 30683082.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • View in gallery

    Data flow through the Raman lidar algorithms (in gray). Inputs are TWR (tower level in situ mixing ratio data), MWR (microwave radiometer), AERI retrievals, and Raman lidar (RL) photon count/log data. The algorithms are aerosol scattering ratio and backscatter coefficient (ASR), aerosol extinction (EXT), water vapor mixing ratio and relative humidity (MR), linear depolarization ratio (DEP), and the best estimate (BE). The last compiles the primary scientifically relevant fields in a single daily output file, interpolating as necessary to have all fields on the same time–height grid

  • View in gallery

    Contribution of the error in relative humidity (ΔRH/RH, in %) due to uncertainties in the temperature profile (solid lines), fractional change (dashed), or offset (dot–dot–dash) in water vapor mixing ratio for a typical profile on 8 Oct 1999 at 1000 UTC. The total precipitable water amount for this profile is 3.2 cm

  • View in gallery

    Line plot showing an average wide FOV ASR profile and the mean uncorrected and corrected narrow FOV ASR profiles for an alignment tweak period on 20 Aug 1998 between 0000 and 0300 UTC

  • View in gallery

    Aerosol profiles for 1 Sep 1998 at 0530 UTC: (upper left) aerosol scattering ratio and backscatter profiles, with error bars included on the backscatter profiles; (upper right) aerosol extinction profiles derived from nitrogen data only and derived from aerosol backscatter data using the smoothed extinction-to-backscatter ratio (Sa) profile; (lower left) the smoothed Sa profile (see text for details about the smoothing), along with the raw Sa profile; (lower right) the vertical resolution used for the backscatter and extinction measurements. These profiles, with the exception of the smoothed Sa profile, are 10-min averages

  • View in gallery

    Comparison of aerosol optical thickness (AOT) derived from the Raman lidar (at 355 nm) with the CIMEL sun photometer (340 nm) for 2304 coincident, cloud-screened points between Apr 1998 and Jan 2000. The slope of the regression line is 0.900, and the correlation coefficient is 0.884. The rms difference is 30%. The right-hand panel shows the airmass distribution of the CIMEL data (the lidar measures vertically) and shows that approximately 35% of the data used in this comparison are at air masses 2.75 or greater (which corresponds to solar elevation angles of less than 22°)

  • View in gallery

    (left) Comparisons of water vapor mixing ratio, (center) aerosol scattering ratio, and (right) aerosol extinction with NASA scanning Raman lidar on 23 Sep 1996 from 0230 to 0500 UTC

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 261 261 12
PDF Downloads 103 103 9

Automated Retrievals of Water Vapor and Aerosol Profiles from an Operational Raman Lidar

View More View Less
  • 1 Pacific Northwest National Laboratory, Richland, Washington
  • | 2 NASA Langley Research Center, Hampton, Virginia
  • | 3 SAIC, NASA Langley Research Center, Hampton, Virginia
  • | 4 Cooperative Institute for Meteorological Satellite Studies, University of Wisconsin—Madison, Madison, Wisconsin
  • | 5 Sandia National Laboratories, Livermore, California
© Get Permissions
Full access

Abstract

Automated routines have been developed to derive water vapor mixing ratio, relative humidity, aerosol extinction and backscatter coefficient, and linear depolarization profiles, as well as total precipitable water vapor and aerosol optical thickness, from the operational Raman lidar at the Atmospheric Radiation Measurement (ARM) program's site in north-central Oklahoma. These routines have been devised to maintain the calibration of these data products, which have proven sensitive to the automatic alignment adjustments that are made periodically by the instrument. Since this Raman lidar does not scan, aerosol extinction cannot be directly computed below approximately 800 m due to the incomplete overlap of the outgoing laser beam with the detector's field of view. Therefore, the extinction-to-backscatter ratio at 1 km is used with the aerosol backscatter coefficient profile to compute aerosol extinction from 60 m to the level of complete overlap. Comparisons of aerosol optical depth derived using these algorithms with a collocated CIMEL sun photometer for clear-sky days over an approximate 2-yr period show a slope of 0.90 with a correlation coefficient of 0.884. Furthermore, comparing the aerosol extinction profile retrieved from this system with that from the National Aeronautics and Space Administration (NASA) Goddard Space Flight Center's scanning Raman lidar agrees within 10% for the single available case.

Corresponding author address: David D. Turner, Climate Dynamics Group, MS K9-24, Pacific Northwest National Laboratory, P.O. Box 999, Richland, WA 99352. Email: dave.turner@pnl.gov

Current affiliation: University of Wisconsin—Madison, Madison, WI.

Abstract

Automated routines have been developed to derive water vapor mixing ratio, relative humidity, aerosol extinction and backscatter coefficient, and linear depolarization profiles, as well as total precipitable water vapor and aerosol optical thickness, from the operational Raman lidar at the Atmospheric Radiation Measurement (ARM) program's site in north-central Oklahoma. These routines have been devised to maintain the calibration of these data products, which have proven sensitive to the automatic alignment adjustments that are made periodically by the instrument. Since this Raman lidar does not scan, aerosol extinction cannot be directly computed below approximately 800 m due to the incomplete overlap of the outgoing laser beam with the detector's field of view. Therefore, the extinction-to-backscatter ratio at 1 km is used with the aerosol backscatter coefficient profile to compute aerosol extinction from 60 m to the level of complete overlap. Comparisons of aerosol optical depth derived using these algorithms with a collocated CIMEL sun photometer for clear-sky days over an approximate 2-yr period show a slope of 0.90 with a correlation coefficient of 0.884. Furthermore, comparing the aerosol extinction profile retrieved from this system with that from the National Aeronautics and Space Administration (NASA) Goddard Space Flight Center's scanning Raman lidar agrees within 10% for the single available case.

Corresponding author address: David D. Turner, Climate Dynamics Group, MS K9-24, Pacific Northwest National Laboratory, P.O. Box 999, Richland, WA 99352. Email: dave.turner@pnl.gov

Current affiliation: University of Wisconsin—Madison, Madison, WI.

1. Introduction

The remote determination of the optical properties and vertical extent of aerosols is important for several facets of ongoing research, including climate research and atmospheric chemistry. Since meteorological conditions such as relative humidity can greatly affect aerosol physical and optical properties (Pilinis et al. 1995), it is critical to study aerosols in their natural state with remote sensors. Sun photometers and other passive measurements of solar transmission are able to provide some information on the physical and optical properties of aerosols (e.g., Harrison et al. 1994; Dubovik and King 2000), but these are integrated measurements throughout the atmosphere and do not provide any height or layer information. Lidars have been used for many years to provide information on the vertical distribution of aerosols. While single wavelength backscatter lidars have proven to be an effective way to characterize the vertical distribution of aerosols, decomposing the backscatter profile into quantitative aerosol properties requires an accurate lidar calibration and assumptions regarding the aerosol optical properties (Klett 1981). For example, scanning the lidar beam circumvents some of these assumptions by using a multiangle solution to solve for both backscatter and extinction; however, this requires horizontally homogeneous aerosol conditions (Spinhirne et al. 1980). Lidars that are able to measure the aerosol and molecular contributions to the backscatter independently and simultaneously, such as Raman lidars (e.g., Ansmann et al. 1990; Ferrare et al. 1992) and high-spectral resolution lidars (e.g., Shipley et al. 1983; Grund and Eloranta 1991), are able to retrieve aerosol backscatter coefficient and aerosol extinction coefficient profiles independently. However, Raman lidars provide the advantage of also being able to profile water vapor in the same parcels as the aerosol measurements. There are many examples of Raman lidar systems that have measured water vapor and aerosol profiles simultaneously (e.g., Ansmann et al. 1992a; Whiteman et al. 1992; Shibata et al. 1996), but these systems were research systems used to advance not only the knowledge about the atmosphere but about the science of making these measurements and as such have obtained relatively limited datasets.

The U.S. Department of Energy's Atmospheric Radiation Measurement (ARM) Program has fielded a Raman lidar at its Cloud and Radiation Testbed (CART) site in north-central Oklahoma near the township of Lamont (36.61°N, 97.49°W). One of the ARM program goals is to collect a 10-yr dataset that can be used to study, and hence improve, the treatment of radiative transfer in the atmosphere, especially with respect to water vapor, aerosols, and clouds (Stokes and Schwartz 1994). Therefore, this Raman lidar was designed to remotely profile water vapor, aerosols, and clouds autonomously in the lower troposphere to midtroposphere throughout the diurnal cycle for extended time periods (Goldsmith et al. 1998). While other Raman lidars are operated routinely (Bösenberg et al. 2001; Murayama et al. 2001), the ARM system is the only truly operational Raman lidar that we are aware of that operates unattended and autonomously with a goal of collecting data for 24 hours a day, seven days a week. From April 1998 to January 2000, this instrument collected more than 7500 h of data. (The principal causes of down time were system shutdowns due to aperiodic power interruptions at the rural site, which have since been eliminated through the installation of a large uninterruptable power supply, and occasional laser failures.) To analyze these data, we have developed algorithms to retrieve profiles of water vapor mixing ratio, relative humidity, aerosol scattering ratio, aerosol backscatter coefficient, aerosol extinction coefficient, and linear depolarization ratio, as well as integrated values of total precipitable water vapor and aerosol optical thickness. To facilitate the processing, these algorithms were developed to run autonomously. The aerosol algorithms are an extension of the work done by Ferrare et al. (1998), whereas the water vapor algorithms used here were discussed in Turner and Goldsmith (1999) and Turner et al. (2000). The details of these routines, and in particular how the calibration of the various data products are autonomously maintained, are discussed in section 3.

An important geophysical parameter not currently measured by the CART Raman lidar is the atmospheric temperature profile. Temperature is needed to convert water vapor mixing ratio into relative humidity and to compute molecular number density for the aerosol retrieval algorithms. While Raman lidars can be used to retrieve profiles of atmospheric temperature (e.g., Cooney 1972; Nedeljkovic et al. 1993; Arshinov and Bobrovnikov 1999; Behrendt and Reichardt 2000), the Raman lidar at the CART site was not designed to make these measurements. Therefore, we require a routine source of temperature profiles. Another algorithm, which has been developed and automated as part of the ARM program, retrieves profiles of water vapor and temperature from the Atmospheric Emitted Radiance Interferometer (AERI) radiance data (Feltz et al. 1998; Smith et al. 1999). The AERI measures downwelling radiance from 550 to 3000 wavenumbers (18–3.3 μm) at approximately one wavenumber resolution every 8 min. Combined with first-guess profiles of water vapor and temperature retrieved from the Geostationary Operational Environmental Satellite and Eta model (Menzel et al. 1998; Feltz et al. 2000), full tropospheric profiles of temperature can be retrieved automatically. This AERI product provides a critical piece of information for the Raman lidar algorithms. The temperature data retrieved from the AERI, the Raman lidar's water vapor and aerosol data, and the wind data from a radar wind profiler (the CART site has both 915- and 50-MHz wind profilers) provide a complete specification of the state of the atmosphere above the CART site in noncloudy conditions from remote sensors (Turner et al. 2000).

This paper concentrates primarily on the water vapor and aerosol algorithms that have been implemented, especially with regard to the automatic calibration techniques, and some initial results. A short description of the Raman lidar is given in section 2, and details on the computation of water vapor mixing ratio, aerosol scattering ratio, aerosol backscatter, and aerosol extinction are presented in section 3. Comparisons of the aerosol products derived from the Raman lidar with those observed from other sensors are discussed in section 4. The water vapor observations from this lidar have been extensively compared in Turner and Goldsmith (1999); thus, these comparisons are not repeated here.

2. Instrument

A Raman lidar operates by transmitting a pulse of laser energy into the atmosphere and records the time of flight of the backscattered light as a function of wavelength for a few discrete wavelengths. The CART Raman lidar uses a frequency tripled Nd:YAG laser, transmitting nominally 350 mJ pulses of 355 nm light into the atmosphere at 30 Hz. The outgoing laser beam is expanded 15 times to achieve eye safety (an important consideration for an automated system, which is why the fundamental and second harmonic are not transmitted). This also serves to reduce the divergence of the outgoing beam to less than 0.1 mrad, enabling the system to utilize a narrow field of view for detection. The backscattered light is collected with a 61-cm telescope. The system measures backscattered light at the laser wavelength (aerosol return), as well as 408 and 387 nm (water vapor and nitrogen Raman shifted returns, respectively) in 0.26-μs bins, which results in 39-m vertical resolution. Photon counting is used and data are typically averaged for approximately 1740 laser shots, which corresponds to 1-min datasets (a small amount of time is needed to clear the multichannel scalar cards). Since Raman scattering is a relatively weak process compared to elastic scattering, an important attribute of any successful Raman lidar is the ability to block the backscatter at the laser wavelength in the Raman channels. In the CART system, the Raman channels have 1012 blocking; thus, even strong elastic backscatter signals (such as from a low cloud) do not affect the Raman signals.

Raman lidars have been used to profile water vapor (e.g., Melfi et al. 1989) and aerosols (e.g., Ferrare et al. 1992) for some time. However, since Raman scattering is a weak process, Raman lidar measurements during the daytime are hampered due to the high solar background. The CART Raman lidar incorporates a dual field of view (FOV; 2 and 0.3 mrad) detection system and narrowband interference filters (0.4 nm full width, half maximum) to reduce the contribution due to the solar background by only admitting light from the desired direction and at the desired wavelength. This permits this lidar to profile throughout the diurnal cycle. The aerosol, water vapor, and nitrogen returns are recorded for both fields of view simultaneously; therefore, there are two complete, nearly independent, detection systems operating that share only a common telescope and a beam splitter that separates light into the two fields of view. This feature is important for both the aerosol scattering ratio and aerosol extinction computations.

Since the Nd:YAG laser's output is polarized, information about cloud phase can be determined if both the copolarized and cross-polarized signal (with respect to the laser beam's polarization) are measured, as aspherical particles (like ice) have a very large depolarization capability (Sassen 1991). Depolarization measurements have also been used in aerosol analyses (e.g., Gobbi 1998; Murayama et al. 1999). Therefore, an additional feature of this system is that the narrow field of view aerosol return is split into copolarized and cross-polarized channels with respect to the laser's output. The output from these two channels can then be used to compute the linear depolarization ratio but also must be linearly recombined in order to retrieve the total elastic return (which is required for the computation of the aerosol scattering ratio).

The primary design goal of this instrument was to provide continuous atmospheric water vapor profiles above the CART site throughout the diurnal cycle to meet the needs of the ARM program. This has led to design decisions, for example, choices of the fields of view, layout of detection electronics, et cetera, that lead to complications in deriving the aerosol profiles from this system. Some of these complications will be discussed in the algorithm section below. Additional details on the configuration of the Raman lidar, including a schematic of the detection optics, are given in Goldsmith et al. (1998) and Turner and Goldsmith (1999).

3. Algorithms

A suite of four separate algorithms is used to calculate the water vapor profiles, the aerosol scattering ratio and backscatter coefficient profiles, the aerosol extinction data, and the depolarization ratio profile products. In addition to the raw photon count and instrument log data from the Raman lidar that are required by each of the algorithms, additional data streams are also required. The additional data streams required are water vapor mixing ratio data at the surface, 25-, and 60-m tower levels and microwave radiometer total precipitable water data for the water vapor algorithm, and AERI temperature retrievals for the water vapor and aerosol codes. The algorithm that computes the aerosol extinction profiles also requires that the aerosol backscatter coefficient profiles be previously derived in order to compute the extinction-to-backscatter ratio (see section 3c). To account for the differential transmission due to aerosols, which can be as large as 9% and 22% in the water vapor mixing ratio and aerosol scattering ratio, respectively (Whiteman et al. 1992), these routines must be run in an iterative manner. Estimates of the aerosol profiles generated during the first iteration of the aerosol scattering ratio and extinction codes serve as inputs to the second iteration of the aerosol scattering ratio and extinction codes serve as inputs to the second iteration of the aerosol scattering ratio algorithm and as input to the water vapor algorithm. Finally, the “best-estimate” code collects the relevant fields for a general scientific user, interpolates them to a common altitude grid, and stores them in a single output file. Table 1 lists the data products included in this daily file. All of the codes automatically generate “quick-look” images of the various geophysical parameters, which enables the user to quickly identify interesting periods and facilitates quality control efforts. These output netCDF files are stored in the ARM data archive and are available to the general scientific community,1 and the images can be accessed via the ARM World Wide Web pages.2 A flowchart depicting the data flow through these algorithms is given in Fig. 1. Note that these algorithms are not run on the computer that controls the operation of the Raman lidar but on a separate centralized data system. Furthermore, they are usually run with a lag time of 2 days to allow ample time for the ancillary data to be transferred from the other instruments to the processing computer and for the intermediary data products (AERI retrievals, water vapor mixing ratio at the tower levels from the in situ sensors) to be generated.

All of the processing algorithms share a common two-step procedure to correct the signals before calculating the desired products. First, the signals must be corrected for system “dead time” or scalar loss. This has been discussed in Goldsmith et al. (1998) for this instrument. Second, the background in each channel, which is especially appreciable in the water vapor channel (408 nm) during the daytime, is then estimated from data collected 15 μs before the laser fires and is subtracted from the profile.

a. Water vapor mixing ratio

Water vapor mixing ratio, which is defined as the mass of water vapor to the mass of dry air, is proportional to the ratio of the water vapor Raman signal to the nitrogen Raman signal. The technique has been covered extensively in the literature (e.g., Ansmann et al. 1992a; Whiteman et al. 1992; Turner and Goldsmith 1999); thus, we refer the reader to these articles and the references therein for details on this calculation.

After the initial processing steps account for the system dead time and background subtraction, overlap correction must be applied, as full overlap is not achieved in the wide FOV until approximately 800 m and in the narrow FOV until approximately 5 km. Whiteman et al. (1992) provide an excellent example of how this can be directly determined for a Raman lidar. We follow this example, which extends the minimum height where water vapor mixing ratio can be measured by the Raman lidar to 60 m. Since there is a wavelength dependence of the Rayleigh and particle scattering, this must be accounted for in this ratio. Whiteman et al. (1992) provides an illustration of the possible size of these errors if not corrected for. The differential attenuation due to molecular scattering is computed from a standard atmospheric model, whereas the differential attenuation due to particle scattering requires the aerosol extinction profile (section 3c) in order to be determined.

The final step in the calibration of this ratio is to determine a height-independent calibration factor. While in principle a Raman lidar can be calibrated in absolute terms, uncertainties in the ratio of the Raman lidar cross sections of water vapor and nitrogen currently limit this calibration to about 10% (Penney and Lapp 1976). Other techniques have also been proposed to calibrate Raman lidars (e.g., Vaughan et al. 1988; Sherlock et al. 1999) with some success, but the common approach is to calibrate the water vapor mixing ratio to another independent observation of water vapor. Therefore, we have chosen to make use of the coincident measurements of water vapor by other instruments at the ARM site to derive the calibration factor. Radiosondes, while used in many previous experiments to calibrate Raman lidars (e.g., Ansmann et al. 1992a; Ferrare et al. 1995), have been shown to have significant sonde-to-sonde variability as well as a dry bias (Turner et al. 1998; Miller et al. 1999); thus, we decided to not utilize the radiosonde in the calibration of the CART Raman lidar. Instead, we calibrate the Raman lidar's water vapor mixing ratio by adjusting the calibration value to achieve agreement in total precipitable water vapor with that retrieved from a dual-channel microwave radiometer (Turner and Goldsmith 1999). For a given day, clear-sky data from the nighttime periods from that day as well as the two adjacent days are used to determine a single calibration factor, which is then used to calibrate all of the profiles for that day (both nighttime and daytime). If the microwave radiometer data are not available, or overcast conditions prevent the lidar from profiling to at least 8 km (and thus unable to sense essentially all of the precipitable water vapor in the column), then the calibration algorithm utilizes the in situ measurement on a collocated 60-m tower to calibrate the ratio. Similar to the calibration to precipitable water vapor, nighttime data from the current day and the two adjacent days are used to determine this single calibration factor, which is applied to all of the data for that day. Comparing the two calibration techniques (microwave radiometer to tower) over a 21-month period (Apr 1998–Dec 1999) revealed a 3% difference between the two, where the tower-derived calibration value is slightly lower than the former. This bias could be due to differences in the absolute calibration of either the microwave radiometer (Radiometrics WVR1100) or the salt-bath calibrated in situ probe (Vaisala HMP35D). Recent experiments at the CART site were conducted to specifically investigate how different ways to measure water vapor in an absolute sense compared to each other (Revercomb 2000), but analysis of data collected from these experiments is ongoing and is beyond the scope of this paper. Another possible explanation is that the overlap correction applied to the Raman lidar data was incorrectly specified at 60 m, causing this bias. Fortunately, the above experiment also included a small remote-controlled aircraft that made repeated vertical profiling flights, as well as a scanning Raman lidar, both of which will be able to address the validity of this overlap correction after this data is reduced and calibrated. Both the microwave and tower calibration techniques had a 4% standard deviation about a mean calibration value during this 21-month period, which is a combination of uncertainties in both the Raman lidar and the calibration instrument. The uncertainty in the calibration value of roughly 5% is the largest source of uncertainty in the water vapor mixing ratio measurement made by the Raman lidar. A more detailed evaluation of the water vapor mixing ratio profiles from this instrument, including comparisons with radiosondes and aircraft in situ profiles, is given in Turner and Goldsmith (1999).

As indicated above, the Raman lidar does not measure temperature profiles directly but uses temperature profiles retrieved by the AERI to convert water vapor mixing ratio to relative humidity. Feltz et al. (2000) have shown that the AERI's retrieved temperature profile agrees within 1 K rms to temperatures measured by coincident radiosondes in clear-sky conditions. Figure 2 illustrates that, for a typical profile, the uncertainty in the relative humidity profile in the lower troposphere and midtroposphere due to this temperature uncertainty is approximately the same size as the uncertainty in the absolute calibration of the water vapor mixing ratio.

b. Aerosol scattering ratio

Aerosol scattering ratio (ASR) is defined as the ratio of the molecular plus aerosol return to the molecular return, (βm(λ, z) + βa(λ, z))/βm(λ, z). Since the elastic return at the laser wavelength has contributions both from Rayleigh and Mie scattering, and the nitrogen return is a function of molecular scattering only, the ratio of these two signals is proportional to the ASR. Note that all channels are corrected for system dead time and have the background subtracted before the ratio is computed. This ratio is then corrected for overlap and the differential transmission of aerosols and particles in a manner similar to the water vapor mixing ratio. After these corrections are applied, the ratio is calibrated by normalizing it such that this ratio is unity in aerosol-free air. Other investigators (e.g., Ferrare et al. 1992) have normalized this ratio over altitude ranges between 6 and 10 km, which is usually aerosol free in nonvolcanic conditions (Russell et al. 1979). Our calibration procedure is described below.

While the method to compute ASR profiles is somewhat similar to the procedure used to compute water vapor mixing ratio profiles, there are additional issues that complicate the computation of ASR profiles for this system. Given the desire to measure depolarization, the high (narrow FOV) channel elastic signal is broken into its beam-parallel (called the high aerosol channel Sλ,s) and the beam-perpendicular (called the depolarization channel Sλ,p) components. These two channels must be linearly recombined to generate the elastic signal, which is needed in this ratio. The weight (w) to apply is not unity because there is an extra fixed neutral density filter in the high elastic channel that is not present in the depolarization channel. This extra neutral density filter is used to prevent the beam-parallel signal from saturating the detector. To compute ASR, the two channels must be recombined to create the high channel elastic signal as Sλ(z) = wSλ,s(z) + Sλ,p(z) (Cairo et al. 1999). The inverse of the weight w is also the calibration factor needed for the linear depolarization ratio, δ(z), which is defined as
i1520-0426-19-1-37-eq1
This weight was determined by collecting background light (i.e., the laser was not operating) during a low heavy overcast condition during the day. It was assumed the collected light was randomly polarized and thus the calibration value w was determined such that δ(z) was unity. From this procedure, w was derived to be 66.0. Initial attempts to validate this using artificial depolarized light (such as from an arc lamp) have shown that this value is accurate within approximately 10%.

A second issue that affects the high channel (narrow FOV) ASR data is the periodic alignment “tweaks” that are performed by the lidar system automatically every few hours. These tweaks are small adjustments to the final steering mirror that attempt to optimize the position of the outgoing laser beam in the detector's field of view. The narrow FOV design requires these adjustments, as small changes in thermal equilibrium, mechanical vibrations, or other events can change the alignment slightly over time and cause the eventual misalignment of the system. The technique is as follows. First, the lidar control computer determines that the sky is cloud free by ensuring that the wide FOV elastic signal has no sharp increases in range and the nitrogen signal no large decreases (extinction) with range over the last 10 min. If the sky is determined to be “cloud free” from the above test, the system quickly scans the beam along the north–south axis, mapping out the strength of the return in the high nitrogen channel between (currently) 1.9 and 2.4 km as a function of alignment position. This altitude range was chosen as a compromise between the mitigating effects of the overlap in the narrow FOV channels and excellent signal to noise required for an expedient alignment process (J. E. M. Goldsmith 1999, personal communication). After making the complete scan, it returns to the position corresponding to the highest return. This same procedure is repeated for the east–west axis. Ideally, after both axes are scanned, the beam's position is now optimized. This entire process takes about 1 min. However, due to hysteresis and the fact that the adjustment motors currently do not have absolute position encoding, or the fact that clouds may enter the FOV during the scan process, it is not guaranteed that the final position is precisely the optimal position. Occasionally, the final position actually has worsened the alignment compared to the previous alignment. However, there is no easy way for the autonomous system to determine whether or not an alignment adjustment is necessary; thus, these tweaks are performed every 3 h given cloud-free conditions.

These occasional slight “misalignments” are more readily seen in the narrow FOV aerosol (beam parallel) channel than in any of the other six channels. The three wide FOV channels are insensitive to these small changes in the final alignment of the laser beam because of their wider FOV. The high aerosol channel is more sensitive than the other three narrow FOV (water vapor, nitrogen, and depolarization) channels because of its physical position with respect to these channels on the detection optics table. If all of the narrow FOV channels were in perfect alignment with each other, any small changes in the final alignment of the laser beam would affect each of these channels the same; thus, in the ratio this effect would be largely removed. However, due to a lack of space on the optics table and the fact that the depolarization beam splitter reflects the beam-parallel light 90° while passing the beam-perpendicular light through, the high aerosol channel is set normal (vertical) to the optics table. The other six channels are arranged parallel to each other on the surface of the table. This arrangement makes the high aerosol channel more susceptible to small mechanical vibrations (such as the door to the lidar trailer closing); thus, it is easier for this channel to become misaligned with respect to the other channels.

The overlap function for the high elastic channel is not stable from one alignment period to another (compared to the other channels), so that using a static overlap correction can result in large (typically 20%–30%, but occasionally larger) errors in this channel. To overcome this, the wide FOV ASR data, which are insensitive to these small changes in the final alignment of the laser beam, are used to determine the overlap correction to be applied to the narrow FOV ASR data, similar to the technique described by Keckhut et al. (1993). By averaging together all of the clear-sky ASR data between two alignment tweaks, where the cloudy cases are identified from a simple threshold-based mask applied to the uncorrected ASR data, a single profile from each of the two fields of view is created for the alignment period. These two mean profiles are smoothed with a simple five-point boxcar filter before computing the ratio to reduce any sharp transitions in either of the two profiles, since the overlap correction should be relatively smooth. The ratio of the narrow to wide FOV profiles then determines the correction to be applied to the narrow FOV data for this period. After applying this correction, a variety of simple tests (e.g., threshold checks) are performed to verify that no anomalies have been introduced (e.g., values significantly below 1). Figure 3 illustrates how the corrected narrow FOV ASR profile is in much better agreement with the wide FOV profile than the original profile. After this correction has been applied to each of the narrow FOV profiles in this alignment tweak period, the wide and narrow FOV ASR profiles are linearly merged together between 1.5 and 1.9 km to provide a composite ASR profile. This procedure is repeated for each alignment tweak period.

To finish the calibration of the ASR profile, a height-independent calibration factor must be determined for the composite profile. The automatic calibration software searches for a 2-km-thick region of clear air between 5 and 9 km to compute this calibration (normalization) value. Starting at 9 km and working downward, the standard deviation of the ASR and the slope of the regression line of the ASR as a function of height is computed in a 2-km-thick layer. If the standard deviation is below 0.03 and the slope is less than 0.025, then the region is determined to be cloud free and the normalization factor is computed. If either threshold is exceeded, then the 2-km calibration region is lowed by 100 m successively until a calibration value can be determined or the calibration region extends below 5 km. If the latter occurs, then the calibration value is unable to be determined automatically; thus, the original static calibration value is maintained for this tweak interval.

In order to determine the uncertainty in the height-independent calibration value for the ASR, nighttime data from the low channel were analyzed. Using data from August 1998 through January 2000, nightly calibration values for the low channel ASR were determined using the above technique. Over this period, the standard deviation of the calibration value was found to be 4%. Therefore, the calibration uncertainty in the lowest 1.5 km where the majority of the tropospheric aerosols reside, which comes entirely from the low channel, is on the order of 4%. Similarly, the standard deviation of the high channel normalization factor in clear-sky conditions, which represents the calibration uncertainty in the high channel and thus the uncertainty in the ASR data above 1.5 km, is about 7%.

After the ASR profiles have been calibrated, the aerosol backscatter coefficient profiles are derived. Using temperature data taken from the AERI retrievals (although radiosonde temperature profiles could be used, if desired), profiles of atmospheric molecular number density are calculated. The backscatter profiles are then computed from the ASR profiles in the manner of Ferrare et al. 1998. Figure 4 (upper-left panel) shows a typical profile of aerosol scattering ratio and aerosol backscatter coefficient, along with the error bars for the backscatter coefficient measurement, for 1 September 1998.

c. Aerosol extinction coefficient

The computation of aerosol extinction coefficient is different than the computation of aerosol scattering ratio or water vapor mixing ratio because it is not the ratio of two signals. Ansmann et al. (1990) demonstrated how aerosol extinction can be derived from the Raman scattered signal from a well-mixed gas, such as nitrogen in the troposphere. The total extinction coefficient is equal to the derivative with respect to range of the logarithm of the nitrogen signal, from which the contribution due to molecular extinction is subtracted to yield the aerosol extinction coefficient. This molecular extinction contribution is calculated from density profiles derived from AERI temperature profiles.

Because the derivative is a measure of the change of slope in the signal with range, effects due to the transmitter–receiver overlap must be avoided. Consequently, aerosol extinction is typically only derived at altitudes where the overlap correction is unity; that is, the laser beam is entirely within the detector's field of view. For the CART Raman lidar, Goldsmith et al. (1998) have shown that for the wide FOV channels, full overlap does not occur until at least 500 m, and the narrow FOV does not achieve full overlap until almost 5 km. Our analysis has shown that full overlap for the wide FOV channels is not achieved until approximately 800 m. This presents two problems. First, since the minimum altitude for the direct calculation of aerosol extinction is 800 m, an alternate strategy is needed to extend the extinction profile to the surface. Second, because the narrow FOV channel does not achieve full overlap until 5 km, the wide FOV data must be used to reach this level. However, there is too much background noise in this channel during the daytime to permit rapid (i.e., 10–15 min) retrievals of aerosol extinction; thus, the derived extinction profile from this channel is extremely noisy well below 5 km. Therefore, we have developed an automated technique that corrects for the overlap in the narrow FOV nitrogen channel by using the data from the wide FOV nitrogen channel in a manner analogous to the way in which the narrow FOV ASR data is corrected. This technique allows us to derive aerosol extinction data directly from the narrow FOV nitrogen data down to approximately 1 km. This greatly improves the quality of the merged extinction profile, as the narrow FOV nitrogen data is significantly less noisy than the wide FOV data.

A few more details are necessary to understand how the extinction profiles are derived from the nitrogen data. Ansmann et al. (1990) assumed that the wavelength dependence of the aerosol extinction was 1, that is, that υ = 1 in
i1520-0426-19-1-37-eq2
While this wavelength dependence is a function of aerosol size and composition, Ferrare et al. (1998) discuss the range of errors in the aerosol extinction assuming a constant wavelength dependence and indicate that they are within approximately ±10% for aerosols typically seen at the Southern Great Plains (SGP) site. Using one year's worth of collocated CIMEL sun photometer data (Holben et al. 1998) from April 1998 to April 1999, we have determined the mean value of the wavelength dependence (υ) to be 0.95 between 340 and 380 nm. The other detail is the determination of the number of points to use in finding the slope of the log of the nitrogen data (which is equivalent to finding the derivative). In order to maintain approximately constant random error with altitude, the vertical resolution must change with altitude. For our calculations, the vertical resolution R(z) is given by
i1520-0426-19-1-37-eq3
from which the number of points Np used to fit the line (which is used to find the derivative) is found from
i1520-0426-19-1-37-eq4
where Δz is 39 m for the CART Raman lidar. Using the zmin and zmax values given in Table 2, the resulting vertical resolution for the aerosol extinction profiles is 312 m at 1 km, 500 m at 3 km, and 1500 m at 7 km. The vertical resolution of the aerosol extinction data is shown in the lower right-hand panel of Fig. 4. By varying the height resolution, we are able to maintain relatively constant fractional error of about 10% in the extinction measurement through layers where the aerosol extinction is greater than approximately 0.03 km−1.

Godin et al. (1999) have demonstrated that modifying the number of points as a function of altitude can introduce biases when taking a numerical derivative, but that these biases are generally less than 10% where the number of points used in the derivative is about 25. This corresponds to a vertical resolution of about 1 km for the aerosol extinction profiles derived from this system, which implies that the bias errors below approximately 6 km are small. Furthermore, ARM Raman lidar aerosol extinction profiles are predominately negligible above 6 km; thus, these bias errors should not be a factor.

At this stage, we have extinction profiles that extend from 0.8 to 10 km. However, due to the overlap in the wide FOV channel, we need to use an alternate strategy to extend these profiles to the surface. To accomplish this, we use the aerosol backscatter coefficient profiles together with the aerosol extinction profiles to derive the extinction-to-backscatter ratio Sa above 0.8 km. In a 10-min average, the typical random error in the Sa measurements is less than 15%, if the extinction coefficient is greater than 0.03 km−1. For the calculations that follow, Sa is smoothed using a median filter over a 1-h period and 500 m in the vertical. Points where the relative error in the extinction or backscatter are above 50% are excluded from the averaging, as are Sa values outside of the range of 5–130 sr. We then use triangular interpolation and a weighted extrapolation routine to fill in any gaps in the Sa dataset that were created by the quality-control thresholds above; these gaps are caused by noisy extinction data or very small values of backscatter in relatively clean conditions. The smoothed Sa dataset is then extrapolated to the surface. A simple smoothing filter is applied to smooth any sharp transitions that may have resulted from the extrapolation. Then, aerosol extinction profiles are calculated directly from the backscatter profile using this interpolated–extrapolated–smoothed Sa dataset. This provides an aerosol extinction profile from the surface to 10 km. Figure 4 (upper-right panel) shows typical profiles derived via this method: an extinction profile derived directly from the nitrogen data, and the extinction profile derived from the backscatter data using the smoothed Sa data. The raw and smoothed extinction-to-backscatter ratio data for this case are shown in the lower-left panel of Fig. 4. After extending the extinction profiles to the surface using the above technique, these profiles are then integrated from the surface to 7 km, or below cloud-base height, to provide aerosol optical thickness at 355 nm in the lower troposphere.

Random error bars are shown on the 10-min averages of backscatter, extinction, and raw extinction-to-backscatter ratio measurements in Fig. 4. Ansmann et al. (1992b) and Ferrare et al. (1998) have undertaken a rigorous treatment of errors in these calculations. Whiteman (1999) has discussed various sources of errors in aerosol extinction measurements, including the assumption of the statistical distribution of the data and any errors in the molecular number density calculation, and the impact on both the extinction coefficient and its error. While the error in the molecular number density is small, we have accounted for it in our calculations. However, we have used the least squares regression technique to determine the slope of the log of the nitrogen data (i.e., extinction). The least squares technique provides a true measure of aerosol extinction, but due to possible differences in the statistical distribution of the data, our estimates of the error in the extinction could be significantly different than the true error (Whiteman 1999). The random error in the smoothed Sa data is approximately 10%; however, this error does not account for changes in aerosol type, hydroscopic growth, et cetera, which may cause Sa to vary within the lowest 800 m.

d. Linear depolarization ratio

Lidar depolarization measurements have been used for years by many groups to ascertain the presence of nonspherical particles in the atmosphere (e.g., Sassen 1991; Cairo et al. 1999; Gobbi 1998). A basic measurement is the linear depolarization measurement δ, which is defined as the ratio of the backscattered signals that are polarized orthogonal and parallel to the linearly polarized outgoing laser beam. The CART Raman lidar does record these two backscatter signals at 355 nm. From these, the total linear depolarization ratio is calculated and, as indicated in section 3b, this product is calibrated to approximately 10%. The vertical resolution of this data product is 39 m near the surface and is slowly degraded to 312 m at 14 km (to maintain a relative random error of 10% or less at all altitudes in the 10-min profile).

4. Results

There are two ways to evaluate retrievals of aerosol extinction: compare with profiles from another instrument capable of measuring aerosol profiles (such as another Raman lidar or in situ measurements on an aircraft), or compare integrated quantities to another instrument. The latter approach is easier to accomplish, because there are several instruments at the SGP CART site capable of measuring aerosol optical thickness routinely, such as a CIMEL sun photometer (Holben et al. 1998) and multifilter rotating shadowband radiometers (Harrison et al. 1994; Harrison and Michalsky 1994). However, this technique provides little information as to the root cause behind any differences.

Figure 5 shows the results of comparing the aerosol optical thickness derived from cloud-screened CIMEL data at 340 nm (Smirnov et al. 2000) to the Raman lidar's retrievals of aerosol optical thickness (AOT) at 355 nm when the Raman lidar (RL) did not detect clouds. There are 2304 coincident samples in this dataset from April–May 1998 and August 1998 to January 2000 (the aft detection optics were out of alignment during Jun and Jul 1998). A linear regression on aerosol optical thickness from the lidar to that from the 340-nm channel from the CIMEL results in a slope of 0.900 and an intercept of −0.029. The slope indicates that there is a difference of approximately 10% between the two measurements, but given the wavelength dependence of 0.95 between 340 and 380 nm one would expect a difference of approximately 4%. The mean difference (CIMEL − RL) is 0.052, and the rms is 0.068 (approximately 30%). The correlation coefficient between the two datasets is 0.884. A histogram of the differences in AOT between the two instruments is shown in the middle panel of Fig. 5. However, since the Raman lidar profiles aerosols vertically and the CIMEL derives optical thickness from direct normal signal, the two instruments are measuring AOT along different paths through the atmosphere. Therefore, horizontal inhomogeniety in the aerosol field will add scatter in the Raman lidar–CIMEL comparisons. The right-hand panel of Fig. 5 shows the airmass values for the CIMEL data used in this analysis. Approximately 35% of the CIMEL measurements used here had airmass values above 2.75 (which corresponds to solar elevation angles of 22° or less). However, utilizing only CIMEL data with airmass values less than 2.75 explains less than 10% of the differences between the two instruments, indicating that the majority of the scatter in the differences between the two instruments is due to differences in their measurement technique rather than horizontal inhomogeniety.

There have been a few opportunities to compare the aerosol profiles derived from the CART Raman lidar to other profile measurements. Kato et al. (2000) derived aerosol extinction profiles from aircraft nephelometer and absorption measurements and compared them with Raman lidar measurements during several days in September 1997 and August 1998. In order to do this, they computed the humidification factor using data collected at the surface level by ARM's aerosol observing system (a series of in situ devices that sample the air at 10 m, including three-wavelength and integrating nephelometers, optical particle counter, and light absorption photometer), and assumed that this relationship held throughout the lower troposphere. This study has shown reasonable agreement (generally within 25%), given the uncertainties in the derivation of the extinction profiles from the aircraft data, although the profiles derived from the in situ measurements appear biased low with respect to the Raman lidar. Such biases, although smaller, have been observed in other experiments where lidar and nephelometer measurements have been compared (Ferrare et al. 2000a).

Because the largest uncertainty in the aerosol extinction retrievals from the CART Raman lidar is in the lowest 800 m due to the assumption of a constant extinction-to-backscatter ratio in this regime, validation of the retrievals in this region is critical. The National Aeronautics and Space Administration Goddard Space Flight Center (NASA/GSFC) scanning Raman lidar (SRL; Whiteman and Melfi 1999; Ferrare et al. 1995; Whiteman et al. 1992) was located at the SGP site alongside the CART lidar from 10 to 30 September 1996 during a special campaign focused on water vapor. This system has very similar nighttime abilities as the CART Raman lidar, except it has the ability to scan in a single vertical plane while the CART lidar does not. This scanning capability allows the SRL to compute aerosol extinction profiles directly from the nitrogen data to very near the surface using scan data (Ferrare et al. 1998). Unfortunately, this time period had low aerosol loading amounts, and there is only one night (23 Sep) where there were significant aerosols aloft to evaluate this comparison. Figure 6 demonstrates the differences between the two lidar systems for both mixing ratio, aerosol scattering ratio, and aerosol extinction for data collected between 0230 and 0500 UTC on 23 September 1996. The water vapor profiles are in excellent agreement, differing by less than 3% from the surface to 7 km. The aerosol scattering ratio data show larger differences, showing the CART system to be approximately 5% lower than the SRL in the first kilometer. The aerosol extinction from the CART system is about 8%–10% lower than that derived from the SRL. In this case, the constant extinction-to-backscatter ratio (Sa) profile utilized by the CART Raman lidar algorithm between the surface and 1000 m (59.4 sr) is almost exactly the mean of the Sa profile observed by the scanning Raman lidar (59.0 sr). The standard deviation of the scanning lidar's Sa profile over this altitude range is 4.5 sr. Although this single comparison indicates that the extinction retrievals are in relatively good agreement, the possible low bias requires further validation of these profiles against those from other systems.

The aerosol extinction profiles from the CART Raman lidar have also been compared to those derived from the collocated micropulse lidar (MPL; Campbell et al. 2000). However, wavelength differences between the two lidars (the MPL's laser operates at 523 nm), calibration uncertainties in the MPL, and the fact that the MPL only measures backscatter at the laser wavelength and thus is unable to derive aerosol extinction independently of aerosol backscatter, make the comparison of aerosol extinction profiles from two lidars more qualitative than quantitative. Therefore, to determine if the apparent bias shown in the comparison of the CART and NASA/GSFC Raman lidars is significant, more comparisons of aerosol extinction profiles with other sensors not at the CART site will need to be performed. These comparisons will use additional data from the NASA/GSFC Raman lidar, profiles of aerosol optical thickness from an airborne sun photometer, and profiles of in situ measurements collected during future intensive observation periods.

5. Conclusions

There are two steps in the development of any operational observational system: the development of the instrument and the development of the analysis software to derive useful data from the raw data. The Raman lidar at the ARM CART site in north-central Oklahoma was designed to be a turn-key instrument to profile atmospheric water vapor and aerosols. The addition of a large uninterruptible power supply has greatly improved the fraction of time in which the lidar is operational by allowing the system to run through the frequent power glitches observed at the rural site. This has resulted in the lidar being operational more than 95% of the time for several months during the end of 1999 and early in 2000, making this instrument the first truly operational Raman lidar. From April 1998 to January 2000, this system has collected over 7500 h of aerosol and water vapor data. This paper has concentrated on the automated algorithms that we have developed to retrieve various aerosol and humidity parameters from this operational lidar, and the uncertainties in the calibration of the various products. These uncertainties, as well as the ranges and resolutions of each, are summarized in Table 1.

Aerosol optical thickness has been derived from these Raman lidar aerosol extinction measurements and compared to a collocated CIMEL sun photometer. The comparison between the lidar's AOT and the 340-nm AOT from CIMEL shows a slope of 0.900, with a correlation coefficient of 0.884, and a rms difference of 30%. This rather large rms difference is due to a number of factors, including 1) failure of the periodic alignment tweaks, 2) the inability of directly measure Sa below 1 km, and 3) horizontal differences in AOT, as the lidar is a vertical measurement while the CIMEL is a slant path measurement. However, the Raman lidar is able to retrieve AOT below cloud base, which cannot be done with a sun photometer. Comparisons of the Raman lidar profiles with profiles derived from in situ measurements by Kato et al. (2000) show agreement within 25%, which is consistent with other similar comparisons. Comparisons with the NASA/GSFC scanning Raman lidar, which is able to measure aerosol extinction directly below 1 km, demonstrates reasonable agreement (approximately 10%) between the two lidars for the single available case. To properly evaluate the Raman lidar's retrievals of aerosol extinction, profiles of AOT measured by an airborne sun photometer, such as those collected during the Tropospheric Aerosol Radiative Forcing Observational Experiment, are required (Ferrare et al. 2000a). Extensive comparisons of the water vapor products from this lidar have been presented in Turner and Goldsmith (1999).

The high-resolution Raman water vapor and aerosol measurements, together with the temperature measurements retrieved from the AERI radiance data, provide a much more detailed representation of the atmospheric state than can be achieved using radiosondes or other in situ measurements alone. These high-resolution data are being used to investigate synoptic features such as drylines and cold fronts (Turner et al. 2000) and radiative transfer applications, to develop an aerosol climatology for the SGP region (Turner et al. 2001), to generate statistics on the extinction-to-backscatter ratio (Ferrare et al. 2001), and to investigate the interaction between water vapor and aerosols (Ferrare et al. 1998, 2000b). Raman lidar data such as these are critical for the study of the last two, as Raman lidars can directly measure both aerosol extinction and backscatter independently in the ambient environment. This operational lidar, together with its automated algorithms, is providing an enormous wealth of high-resolution data that can be and are being used for these analyses.

Acknowledgments

The SGP CART Raman lidar, CIMEL sun photometer, and AERI data were obtained from the 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 sunphotometers managed by B. N. Holben, NASA/GSFC. We would like to thank Dave Whiteman for the NASA/GSFC scanning Raman lidar data used in this analysis. We would also like to thank the anonymous reviewers whose comments greatly improved this document. The DOE ARM and NASA EOS Validation programs provided funding for this work. Pacific Northwest National Laboratory is operated for the U.S. Department of Energy by Battelle under Contract DE-AC06-76RLO-1830.

REFERENCES

  • Ansmann, A., , Riebesell M. , , and Weitkamp C. , 1990: Measurement of atmospheric aerosol extinction profiles with a Raman lidar. Opt. Lett, 15 , 746748.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ansmann, A., , Riebesell M. , , Wandinger U. , , Weitkamp C. , , Voss E. , , Lahmann W. , , and Michaelis W. , 1992a: Combined Raman elastic-backscatter lidar for vertical profiling of moisture, aerosol extinction, and lidar ratio. Appl. Phys, B55 , 1828.

    • Search Google Scholar
    • Export Citation
  • Ansmann, A., , Wandinger U. , , Riebesell M. , , Weitkamp C. , , and Michaelis W. , 1992b: Independent measurement of extinction and backscatter profiles in cirrus clouds by using a combined Raman elastic-backscatter lidar. Appl. Opt, 31 , 71137131.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Arshinov, Y., , and Bobrovnikov S. , 1999: Use of a Farby-Perot interferometer to isolate pure rotational Raman spectra of diatomic molecules. Appl. Opt, 38 , 46354638.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Behrendt, A., , and Reichardt J. , 2000: Atmospheric temperature profiling in the presence of clouds with a pure rotational Raman lidar using an interference-filter-based polychromator. Appl. Opt, 39 , 13721378.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bösenberg, J., and and Coauthors, 2001: EARLINET: A European aerosol research lidar network. Advances in Laser Remote Sensing: Selected Papers Presented at the 20th International Laser Radar Conference, A. Dabas, C. Loth, and J. Pelon, Eds., Ecole Polytechnique, 155–158.

    • Search Google Scholar
    • Export Citation
  • Cairo, F., , Donfrancesco G. Di , , Adriani A. , , Pulvirenti L. , , and Fierli F. , 1999: Comparisons of various linear depolarization parameters measured by lidar. Appl. Opt, 38 , 44254432.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Campbell, J. R., , Hlavka D. L. , , Spinhirne J. D. , , Ferrare R. A. , , and Turner D. D. , 2000: Automated aerosol retrieval algorithms for ARM micro pulse lidars. Preprints, Symp. on Lidar Atmosphere Monitoring, Long Beach, CA, Amer. Meteor. Soc., 71–74.

    • Search Google Scholar
    • Export Citation
  • Cooney, J. A., 1972: Measurement of atmospheric temperature profiles by Raman backscatter. J. Appl. Meteor, 11 , 108112.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dubovik, O., , and King M. D. , 2000: A flexible inversion algorithm for retrieval of aerosol optical properties from Sun and sky radiance measurements. J. Geophys. Res, 105 , 2067320696.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Feltz, W. F., , Smith W. L. , , Knuteson R. O. , , Revercomb H. E. , , Woolf H. M. , , and Howell H. B. , 1998: Meteorological applications of temperature and water vapor retrievals from the ground-based atmospheric emitted radiance interferometer (AERI). J. Appl. Meteor, 37 , 857875.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Feltz, W. F., and and Coauthors, 2000: Retrieval and visualization of AERIPLUS temperature and moisture profiles for assimilation into ARM single-column models. Proc. 10th Atmospheric Radiation Measurement (ARM) Science Team Meeting, San Antonio, TX. [Available online at http://www.arm.gov/docs/documents/technical/conf_0003/feltz_wf.pdf.].

    • Search Google Scholar
    • Export Citation
  • Ferrare, R. A., , Melfi S. H. , , Whiteman D. N. , , and Evans K. D. , 1992: Raman lidar measurements of Pinatubo aerosols over southeastern Kansas during November-December 1991. Geophys. Res. Lett, 19 , 15991602.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ferrare, R. A., , Melfi S. H. , , Whiteman D. N. , , Evans K. D. , , Schmidlin F. J. , , and O'C. Starr D. , 1995: A comparison of water vapor measurements made by Raman lidar and radiosondes. J. Atmos. Oceanic Technol, 12 , 11771195.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ferrare, R. A., , Melfi S. H. , , Whiteman D. N. , , Evans K. D. , , and Leifer R. , 1998: Raman lidar measurements of aerosol extinction and backscattering: Methods and comparisons. J. Geophys. Res, 103 , 1966319672.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ferrare, R. A., and and Coauthors, 2000a: Comparisons of LASE, aircraft, and satellite measurements of aerosol optical properties and water vapor during TARFOX. J. Geophys. Res, 105 , 99359948.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ferrare, R. A., , Turner D. D. , , Heilman L. A. , , Feltz W. F. , , Tooman T. P. , , Dubovik O. , , and Halthore R. , 2000b: Characterization of the atmospheric state over the SGP using Raman lidar and AERI/GOES measurements. Proc. 10th Atmospheric Radiation Measurement (ARM) Science Team Meeting, San Antonio, TX. [Available online at http://www.arm.gov/docs/documents/technical/conf_0003/ferrare-ra.pdf.].

    • Search Google Scholar
    • Export Citation
  • Ferrare, R. A., , Turner D. D. , , Heilman L. A. , , Feltz W. F. , , Dubovik O. , , and Tooman T. P. , 2001: Raman lidar measurements of aerosol extinction-to-backscatter ratio over the Southern Great Plains. J. Geophy. Res.,106, 20 333–20 348.

    • Search Google Scholar
    • Export Citation
  • Gobbi, G. P., 1998: Polarization lidar returns from aerosols and thin clouds: A framework for the analysis. Appl. Opt, 37 , 55055508.

  • Godin, S., and and Coauthors, 1999: Ozone differential absorption lidar algorithm intercomparison. Appl. Opt, 38 , 62256236.

  • Goldsmith, J. E. M., , Blair F. H. , , Bisson S. E. , , and Turner D. D. , 1998: Turn-key Raman lidar for profiling atmospheric water vapor, clouds, and aerosols. Appl. Opt, 37 , 49794990.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Grund, C. J., , and Eloranta E. W. , 1991: University of Wisconsin high spectral resolution lidar. Opt. Eng, 30 , 612.

  • Harrison, L., , and Michalsky J. , 1994: Objective algorithms for the retrieval of optical depths from ground-based measurements. Appl. Opt, 33 , 51265132.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Harrison, L., , Michalsky J. , , and Berndt J. , 1994: Automated multifilter rotating shadow-band radiometer: An instrument for optical depth and radiation measurements. Appl. Opt, 33 , 51185125.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Holben, B. N., and and Coauthors, 1998: AERONET—A federated instrument network and data archive for aerosol characterization. Remote Sens. Environ, 66 , 116.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kato, S., and and Coauthors, 2000: A comparison of aerosol optical thickness derived from ground-based and airborne measurements. J. Geophys. Res, 105 , 97919806.

    • Search Google Scholar
    • Export Citation
  • Keckhut, P., , Hauchecorne A. , , and Chanin M. L. , 1993: A critical review of the database acquired for the long-term surveillance of the middle atmosphere by the French Rayleigh lidars. J. Atmos. Oceanic Technol, 10 , 850867.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Klett, J. D., 1981: Stable analytical inversion solution for processing lidar returns. Appl. Opt, 20 , 211220.

  • Melfi, S. H., , Whiteman D. N. , , and Ferrare R. A. , 1989: Observation of atmospheric fronts using Raman lidar moisture measurements. J. Appl. Meteor, 28 , 789806.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Menzel, W. P., , Holt F. C. , , Schmidt T. J. , , Aune R. M. , , Schreiner A. J. , , Wade G. S. , , and Gray D. G. , 1998: Application of GOES-8/9 soundings to weather forecasting and nowcasting. Bull. Amer. Meteor. Soc, 79 , 20592077.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Miller, E. R., , Wang J. , , and Cole H. L. , 1999: Correction for dry bias in Vaisala radiosonde RH data. Proc. Ninth Atmospheric Radiation Measurement (ARM) Science Team Meeting, Tucson, AZ. [Available online at http://www.arm.gov/pub/docs/documents/technical/conf_9903/miller-er-99.pdf.].

    • Search Google Scholar
    • Export Citation
  • Murayama, T., , Okamoto H. , , Kaneyasu N. , , Kamataki H. , , and Miura K. , 1999: Application of lidar depolarization measurement in the atmospheric boundary layer: Effects of dust and sea-salt particles. J. Geophys. Res, 104 , 3178131792.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Murayama, T., and and Coauthors, 2001: Lidar network observation of Asian dust. Advances in Laser Remote Sensing: Selected Papers Presented at the 20th International Laser Radar Conference, A. Dabas, C. Loth, and J. Pelon, Eds., Ecole Polytechnique, 169–172.

    • Search Google Scholar
    • Export Citation
  • Nedeljkovic, D., , Hauchecorne A. , , and Chanin M. L. , 1993: Rotational Raman lidar to measure the atmospheric temperature from the ground to 30 km. IEEE Trans. Geosci. Remote Sens, 31 , 90101.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Penney, G. M., , and Lapp M. , 1976: Raman-scattering cross-section for water vapor. J. Opt. Soc. Amer, 66 , 422425.

  • Pilinis, C., , Sandis S. N. , , and Seinfeld J. H. , 1995: Sensitivity of direct climate forcing by atmospheric aerosols to aerosol size and composition. J. Geophys. Res, 100 , 1873918754.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Revercomb, H. E., 2000: Science plan for the 3d ARM water vapor IOP and the ARM-FIRE water vapor experiment (AFWEX) at the ARM SGP CART site, Lamont, OK. [Available online at http://www.arm.gov/docs/iops/2000/sgp2000afwex/afwex.html.].

    • Search Google Scholar
    • Export Citation
  • Russell, P. B., , Swissler T. J. , , and McCormick M. P. , 1979: Methodology for error analysis and simulation of lidar measurements. Appl. Opt, 18 , 37833797.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sassen, K., 1991: The polarization lidar technique for cloud research: A review and current assessment. Bull. Amer. Meteor. Soc, 72 , 18481866.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sherlock, V., , Hauchecorne A. , , and Lenoble J. , 1999: Methodology for the independent calibration of Raman backscatter water-vapor lidar systems. Appl. Opt, 38 , 58165837.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shibata, T., , Sakai T. , , Hayashi M. , , Ojio T. , , Kwon S-A. , , and Iwasaka Y. , 1996: Raman lidar observations: Simultaneous measurements of water vapor, temperature, and aerosol vertical profiles, part II. J. Geomag. Geoelectr, 48 , 11371144.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Shipley, S. T., , Tracy D. H. , , Eloranta E. W. , , Trauger J. T. , , Sroga J. T. , , Roesler F. L. , , and Weinman J. A. , 1983: High spectral resolution lidar to measure optical scattering properties of atmospheric aerosols: 1. Theory and implementation. Appl. Opt, 22 , 37163724.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Smirnov, A., , Holben B. N. , , Eck T. F. , , Dubovik O. , , and Slutsker I. , 2000: Cloud screening and quality control algorithms for the AERONET database. Remote Sens. Environ.,73, 337–349.

    • Search Google Scholar
    • Export Citation
  • Smith, W. L., , Feltz W. F. , , Knuteson R. O. , , Revercomb H. E. , , Howell H. B. , , and Woolf H. M. , 1999: The retrieval of planetary boundary layer structure using ground-based infrared spectral radiance measurements. J. Atmos. Oceanic Technol, 16 , 323333.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Spinhirne, J. D., , Reagan J. A. , , and Hermann B. M. , 1980: Vertical distribution of aerosol extinction cross section and inference of aerosol imaginary index in the troposphere by lidar technique. J. Appl. Meteor, 19 , 426438.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stokes, G. M., , and Schwartz S. E. , 1994: The Atmospheric Radiation Measurement (ARM) Program: Programmatic background and design of the cloud and radiation testbed. Bull. Amer. Meteor. Soc, 75 , 12011221.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Turner, D. D., , and Goldsmith J. E. M. , 1999: Twenty-four-hour Raman lidar measurements during the Atmospheric Radiation Measurement program's 1996 and 1997 water vapor intensive observation periods. J. Atmos. Oceanic Technol, 16 , 10621076.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Turner, D. D., , Shippert T. R. , , Brown P. D. , , Clough S. A. , , Knuteson R. O. , , Revercomb H. E. , , and Smith W. L. , 1998: Long-term analysis of observed and line-by-line calculations of longwave surface spectral radiance and the effect of scaling the water vapor profiles. Proc. 8th Atmospheric Radiation Measurement (ARM) Science Team Meeting, Tucson, AZ. [Available online at http://www.arm.gov/pub/docs/documents/technical/conf_9803/turner-98.pdf.].

    • Search Google Scholar
    • Export Citation
  • Turner, D. D., , Feltz W. F. , , and Ferrare R. A. , 2000: Continuous water vapor profiles from ground-based operational active and passive remote sensors. Bull. Amer. Meteor. Soc, 81 , 13011317.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Turner, D. D., , Ferrare R. A. , , and Brasseur L. A. , 2001: Average aerosol extinction and water vapor profiles over the Southern Great Plains. Geophys. Res. Lett., in press.

    • Search Google Scholar
    • Export Citation
  • Vaughan, G., , Waring D. P. , , Thomas L. , , and Metev V. , 1988: Humidity measurements in the free troposphere using Raman backscatter. Quart. J. Roy. Meteor. Soc, 114 , 30683082.

    • Search Google Scholar
    • Export Citation
  • Whiteman, D. N., 1999: Application of statistical methods to the determination of slope in lidar data. Appl. Opt, 38 , 33603369.

  • Whiteman, D. N., , and Melfi S. H. , 1999: Cloud liquid water, mean droplet radius, and number density measurements using a Raman lidar. J. Geophys. Res, 104 , 3141131419.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Whiteman, D. N., , Melfi S. H. , , and Ferrare R. A. , 1992: Raman lidar system for the measurement of water vapor and aerosols in the Earth's atmosphere. Appl. Opt, 31 , 30683082.

    • Crossref
    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Data flow through the Raman lidar algorithms (in gray). Inputs are TWR (tower level in situ mixing ratio data), MWR (microwave radiometer), AERI retrievals, and Raman lidar (RL) photon count/log data. The algorithms are aerosol scattering ratio and backscatter coefficient (ASR), aerosol extinction (EXT), water vapor mixing ratio and relative humidity (MR), linear depolarization ratio (DEP), and the best estimate (BE). The last compiles the primary scientifically relevant fields in a single daily output file, interpolating as necessary to have all fields on the same time–height grid

Citation: Journal of Atmospheric and Oceanic Technology 19, 1; 10.1175/1520-0426(2002)019<0037:AROWVA>2.0.CO;2

Fig. 2.
Fig. 2.

Contribution of the error in relative humidity (ΔRH/RH, in %) due to uncertainties in the temperature profile (solid lines), fractional change (dashed), or offset (dot–dot–dash) in water vapor mixing ratio for a typical profile on 8 Oct 1999 at 1000 UTC. The total precipitable water amount for this profile is 3.2 cm

Citation: Journal of Atmospheric and Oceanic Technology 19, 1; 10.1175/1520-0426(2002)019<0037:AROWVA>2.0.CO;2

Fig. 3.
Fig. 3.

Line plot showing an average wide FOV ASR profile and the mean uncorrected and corrected narrow FOV ASR profiles for an alignment tweak period on 20 Aug 1998 between 0000 and 0300 UTC

Citation: Journal of Atmospheric and Oceanic Technology 19, 1; 10.1175/1520-0426(2002)019<0037:AROWVA>2.0.CO;2

Fig. 4.
Fig. 4.

Aerosol profiles for 1 Sep 1998 at 0530 UTC: (upper left) aerosol scattering ratio and backscatter profiles, with error bars included on the backscatter profiles; (upper right) aerosol extinction profiles derived from nitrogen data only and derived from aerosol backscatter data using the smoothed extinction-to-backscatter ratio (Sa) profile; (lower left) the smoothed Sa profile (see text for details about the smoothing), along with the raw Sa profile; (lower right) the vertical resolution used for the backscatter and extinction measurements. These profiles, with the exception of the smoothed Sa profile, are 10-min averages

Citation: Journal of Atmospheric and Oceanic Technology 19, 1; 10.1175/1520-0426(2002)019<0037:AROWVA>2.0.CO;2

Fig. 5.
Fig. 5.

Comparison of aerosol optical thickness (AOT) derived from the Raman lidar (at 355 nm) with the CIMEL sun photometer (340 nm) for 2304 coincident, cloud-screened points between Apr 1998 and Jan 2000. The slope of the regression line is 0.900, and the correlation coefficient is 0.884. The rms difference is 30%. The right-hand panel shows the airmass distribution of the CIMEL data (the lidar measures vertically) and shows that approximately 35% of the data used in this comparison are at air masses 2.75 or greater (which corresponds to solar elevation angles of less than 22°)

Citation: Journal of Atmospheric and Oceanic Technology 19, 1; 10.1175/1520-0426(2002)019<0037:AROWVA>2.0.CO;2

Fig. 6.
Fig. 6.

(left) Comparisons of water vapor mixing ratio, (center) aerosol scattering ratio, and (right) aerosol extinction with NASA scanning Raman lidar on 23 Sep 1996 from 0230 to 0500 UTC

Citation: Journal of Atmospheric and Oceanic Technology 19, 1; 10.1175/1520-0426(2002)019<0037:AROWVA>2.0.CO;2

Table 1.

List of data products included in the final best-estimate output

Table 1.
Table 2.

The various constants used in the determination of the vertical resolution of the extinction data from both the wide and narrow FOV channels

Table 2.
Save