• Ansmann, A., , Fruntke J. , , and Engelmann R. , 2010: Updraft and downdraft characterization with Doppler lidar: Cloud-free versus cumuli-topped mixed layer. Atmos. Chem. Phys., 10, 78457858.

    • Search Google Scholar
    • Export Citation
  • Baker, W. E., and Coauthors, 1995: Lidar-measured winds from space: An essential component for weather and climate prediction. Bull. Amer. Meteor. Soc., 76, 869888.

    • Search Google Scholar
    • Export Citation
  • Banta, R. M., , Pichugina Y. L. , , and Brewer W. A. , 2006: Turbulent velocity variance profiles in the stable boundary layer generated by a nocturnal low-level jet. J. Atmos. Sci., 63, 27002719.

    • Search Google Scholar
    • Export Citation
  • Barlow, J. F., , Dunbar T. M. , , Nemitz E. G. , , Wood C. R. , , Gallagher M. W. , , Davies F. , , O’Connor E. , , and Harrison R. M. , 2011: Boundary layer dynamics over London, UK, as observed using Doppler lidar during REPARTEE-II. Atmos. Chem. Phys., 11, 21112125.

    • Search Google Scholar
    • Export Citation
  • Blackadar, A. K., 1957: Boundary layer wind maximum and their significance for the growth of nocturnal inversions. Bull. Amer. Meteor. Soc., 38, 283290.

    • Search Google Scholar
    • Export Citation
  • Browning, R. A., , and Wexler R. , 1968: The determination of kinematic properties of a wind field using Doppler radar. J. Appl. Meteor., 7, 105113.

    • Search Google Scholar
    • Export Citation
  • Calhoun, R., , Heap R. , , Princevac M. , , Newsom R. , , Fernando H. , , and Ligon D. , 2006: Virtual towers using coherent Doppler lidar during the Joint Urban 2003 dispersion experiment. J. Appl. Meteor. Climatol., 45, 11161126.

    • Search Google Scholar
    • Export Citation
  • Chan, P. W., , and Lee Y. F. , 2012: Application of short-range lidar in wind shear alerting. J. Atmos. Oceanic Technol., 29, 207220.

  • Collier, C. G., and Coauthors, 2005: Dual-Doppler lidar measurements for improving dispersion models. Bull. Amer. Meteor. Soc., 86, 825838.

    • Search Google Scholar
    • Export Citation
  • Doyle, J. D., , Grubišić V. , , Brown W. O. J. , , De Wekker S. F. J. , , Dörnbrack A. , , Jiang Q. , , Mayor S. D. , , and Weissmann M. , 2009: Observations and numerical simulations of subrotor vortices during T-REX. J. Atmos. Sci., 66, 12291249.

    • Search Google Scholar
    • Export Citation
  • Drechsel, S., , Mayr G. J. , , Chong M. , , Weissmann M. , , Dörnbrack A. , , and Calhoun R. , 2009: Three-dimensional wind retrieval: Application of MUSCAT to dual-Doppler lidar. J. Atmos. Oceanic Technol., 26, 635646.

    • Search Google Scholar
    • Export Citation
  • Drechsel, S., , Mayr G. J. , , Chong M. , , and Chow F. K. , 2010: Volume scanning strategies for 3D wind retrieval from dual-Doppler lidar measurements. J. Atmos. Oceanic Technol., 27, 18811892.

    • Search Google Scholar
    • Export Citation
  • Drobinski, P., , and Foster R. C. , 2003: On the origin of near-surface streaks in the neutrally-stratified planetary boundary layer. Bound.-Layer Meteor., 108, 247256.

    • Search Google Scholar
    • Export Citation
  • Drobinski, P., , Carlotti P. , , Newsom R. K. , , Banta R. M. , , Foster R. C. , , and Redelsperger J. L. , 2004: The structure of the near-neutral surface layer. J. Atmos. Sci., 61, 699714.

    • Search Google Scholar
    • Export Citation
  • Drobinski, P., , Carlotti P. , , Redelsperger J. L. , , Banta R. M. , , Masson V. , , and Newsom R. K. , 2007: Numerical and experimental investigation of the neutral atmospheric surface layer. J. Atmos. Sci., 64, 137156.

    • Search Google Scholar
    • Export Citation
  • Etling, D., , and Brown R. A. , 1993: A review of large-eddy dynamics in the planetary boundary layer. Bound.-Layer Meteor., 65, 215248.

    • Search Google Scholar
    • Export Citation
  • Frehlich, R., 1999: Performance of maximum likelihood estimators of mean power and Doppler velocity with a priori knowledge of spectral width. J. Atmos. Oceanic Technol., 16, 17021709.

    • Search Google Scholar
    • Export Citation
  • Frehlich, R., 2001: Estimation of velocity error for Doppler lidar measurements. J. Atmos. Oceanic Technol., 18, 16281639.

  • Frehlich, R., , Hannon S. M. , , and Henderson S. W. , 1994: Performance of a 2-μm coherent Doppler lidar for wind measurements. J. Atmos. Oceanic Technol., 11, 15171528.

    • Search Google Scholar
    • Export Citation
  • Frehlich, R., , Hannon S. M. , , and Henderson S. W. , 1997: Coherent Doppler lidar measurements of winds in the weak signal regime. Appl. Opt., 36, 34913499.

    • Search Google Scholar
    • Export Citation
  • Fujiwara, C., , Yamashita K. , , Nakanishi M. , , and Fujiyoshi Y. , 2011: Dust devil–like vortices in an urban area detected by a 3D scanning Doppler lidar. J. Appl. Meteor. Climatol., 50, 534547.

    • Search Google Scholar
    • Export Citation
  • Goodall, C., 1990: A survey of smoothing techniques. Modern Methods of Data Analysis, J. Fox and J. S. Long, Eds., Sage Publications, 126–176.

  • Grund, C. J., , Banta R. M. , , George J. L. , , Howell J. N. , , Post M. J. , , Richter R. A. , , and Weickmann A. M. , 2001: High-resolution Doppler lidar for boundary layer and cloud research. J. Atmos. Oceanic Technol., 18, 376393.

    • Search Google Scholar
    • Export Citation
  • Henderson, S. W., , Yuen E. Y. , , and Fry E. S. , 1986: Fast resonance-detection technique for single-frequency operation of injection-seeded Nd:YAG lasers. Opt. Lett., 11, 715717.

    • Search Google Scholar
    • Export Citation
  • Henderson, S. W., , Hale C. P. , , Magee J. R. , , Kavaya M. J. , , and Huffaker A. V. , 1991: Eye-safe coherent laser radar system at 2.1 μm using Tm,Ho:YAG lasers. Opt. Lett., 16, 773775.

    • Search Google Scholar
    • Export Citation
  • Henderson, S. W., , Suni P. J. M. , , Hale C. P. , , Hannon S. M. , , Magee J. R. , , Bruns D. L. , , and Yuen E. H. , 1993: Coherent laser radar at 2 μm using solid-state lasers. IEEE Trans. Geosci. Remote Sens., 31, 415.

    • Search Google Scholar
    • Export Citation
  • Henderson, S. W., , Gatt P. , , Rees D. , , and Huffaker R. M. , 2005: Wind lidar. Laser Remote Sensing, T. Fujii and T. Fukuchi, Eds., CRC Press, 469–472.

  • Hogan, R. J., , Grant A. L. M. , , Illingworth A. J. , , Pearson G. N. , , and O’Connor E. J. , 2009: Vertical velocity variance and skewness in clear and cloud-topped boundary layers as revealed by Doppler lidar. Quart. J. Roy. Meteor. Soc., 135, 635643.

    • Search Google Scholar
    • Export Citation
  • Huffaker, R. M., , and Hardesty R. M. , 1996: Remote sensing of atmospheric wind velocities using solid-state and CO2 coherent laser systems. Proc. IEEE, 84, 181204.

    • Search Google Scholar
    • Export Citation
  • Huffaker, R. M., , Lawrence T. R. , , Post M. J. , , Priestley J. T. , , Hall F. F. Jr., , Richter R. A. , , and Keeler R. J. , 1984: Feasibility studies for a global wind measuring satellite system (Windsat): Analysis of simulated performance. Appl. Opt., 23, 25232536.

    • Search Google Scholar
    • Export Citation
  • Hunt, J. C. R., , and Morrison J. F. , 2000: Eddy structure in turbulent boundary layers. Eur. J. Mech., 19B, 673694.

  • Ishii, S., and Coauthors, 2010: Coherent 2 μm differential absorption and wind lidar with conductively cooled laser and two-axis scanning device. Appl. Opt., 49, 18091817.

    • Search Google Scholar
    • Export Citation
  • Ishii, S., and Coauthors, 2012: Partial CO2 column-averaged dry-air mixing ratio from measurements by coherent 2-μm differential absorption and wind lidar with laser frequency offset locking. J. Atmos. Oceanic Technol., 29, 11691181.

    • Search Google Scholar
    • Export Citation
  • Iwai, H., , Murayama Y. , , Ishii S. , , Mizutani K. , , Ohno Y. , , and Hashiguchi T. , 2011: Strong updraft at a sea-breeze front and associated vertical transport of near-surface dense aerosol observed by Doppler lidar and ceilometer. Bound.-Layer Meteor., 141, 117142.

    • Search Google Scholar
    • Export Citation
  • Kaimal, J. C., , Wyngaard J. C. , , Haugen D. A. , , Coté O. R. , , Izumi Y. , , Caughey S. J. , , and Readings C. J. , 1976: Turbulence structure in the convective boundary layer. J. Atmos. Sci., 33, 21522169.

    • Search Google Scholar
    • Export Citation
  • Käsler, Y., , Rahm S. , , Simmet R. , , and Kühn M. , 2010: Wake measurements of a multi-MW wind turbine with coherent long-range pulsed Doppler wind lidar. J. Atmos. Oceanic Technol., 27, 15291532.

    • Search Google Scholar
    • Export Citation
  • Kelley, N., , Jonkman B. J. , , Scott G. N. , , and Pichugina Y. L. , 2007: Comparing pulsed Doppler lidar with sodar and direct measurements for wind assessment. Proc. American Wind Energy Association 2007 Conf. and Exhibition, Los Angeles, CA, National Renewable Energy Laboratory, NREL/CP-50041792, 21 pp.

  • Khanna, S., , and Brasseur J. G. , 1998: Three-dimensional buoyancy- and shear-induced local structure of the atmospheric boundary layer. J. Atmos. Sci., 55, 710743.

    • Search Google Scholar
    • Export Citation
  • Koch, G. J., , Beyon J. Y. , , Barnes B. W. , , Petros M. , , Yu J. , , Amzajerdian F. , , Kavaya M. J. , , and Singh U. N. , 2007: High-energy 2μm Doppler lidar for wind measurements. Opt. Eng., 46, 116201, doi:10.1117/1.2802584.

    • Search Google Scholar
    • Export Citation
  • Konrad, T. G., 1970: The dynamics of the convective process in clear air as seen by radar. J. Atmos. Sci., 27, 11381147.

  • Köpp, F., , Schwiesow R. L. , , and Werner Ch. , 1984: Remote measurements of boundary-layer wind profiles using a CW Doppler lidar. J. Climate Appl. Meteor., 23, 148154.

    • Search Google Scholar
    • Export Citation
  • Köpp, F., , Rahm S. , , and Smalikho I. , 2004: Characterization of aircraft wake vortices by 2-μm pulsed Doppler lidar. J. Atmos. Oceanic Technol., 21, 194206.

    • Search Google Scholar
    • Export Citation
  • Lenschow, D. H., , Lothon M. , , Mayor S. D. , , Sullivan P. P. , , and Canut G. , 2012: A comparison of higher-order vertical velocity moments in the convective boundary layer from lidar with in situ measurements and large-eddy simulation. Bound.-Layer Meteor., 143, 107123.

    • Search Google Scholar
    • Export Citation
  • Levin, M. J., 1965: Power spectrum parameter estimation. IEEE Trans. Inf. Theory, 11, 100107.

  • Lhermitte, R. M., , and Miller L. J. , 1970: Doppler radar methodology for the observation of convective storms. Preprints, 14th Conf. Radar Meteorology, Tucson, AZ, Amer. Meteor. Soc., 133–138.

  • Lin, C.-L., , Moeng C.-H. , , Sullivan P. P. , , and McWilliams J. C. , 1997: The effect of surface roughness on flow structures in a neutrally stratified planetary boundary layer. Phys. Fluids, 9, 32353249.

    • Search Google Scholar
    • Export Citation
  • Lothon, M., , Lenschow D. H. , , and Mayor S. D. , 2006: Coherence and scale of vertical velocity in the convective boundary layer from a Doppler lidar. Bound.-Layer Meteor., 121, 521536.

    • Search Google Scholar
    • Export Citation
  • Mann, J., , Peña A. , , Bingöl F. , , Wagner R. , , and Courtney M. S. , 2010: Lidar scanning of momentum flux in and above the atmospheric surface layer. J. Atmos. Oceanic Technol., 27, 959976.

    • Search Google Scholar
    • Export Citation
  • Mayor, S. D., , Lenschow D. H. , , Schwiesow R. L. , , Mann J. , , Frush C. L. , , and Simon M. K. , 1997: Validation of NCAR 10.6-μm CO2 Doppler lidar radial velocity measurements and comparison with a 915-MHz profiler. J. Atmos. Oceanic Technol., 14, 11101126.

    • Search Google Scholar
    • Export Citation
  • Menzies, R. T., 1986: Doppler lidar atmospheric wind sensors: A comparative performance evaluation for global measurement applications from earth orbit. Appl. Opt., 25, 25462553.

    • Search Google Scholar
    • Export Citation
  • Menzies, R. T., , and Hardesty R. M. , 1989: Coherent Doppler lidar for measurements of wind fields. Proc. IEEE, 77, 449462.

  • Mizutani, K., , Itabe T. , , Ishii S. , , Aoki T. , , Asai K. , , Sato A. , , Fukuoka H. , , and Ishikawa T. , 2008: Conductive-cooled 2 micron laser for CO2and wind observations. Lidar Remote Sensing for Environmental Monitoring IX, U. N. Singh, K. Asai, and A. Jayaraman, Eds., International Society for Optical Engineering (SPIE Proceedings, Vol. 7153), doi:10.1117/12.804807.

  • Newsom, R. K., , Ligon D. , , Calhoun R. , , Heap R. , , Cregan E. , , and Princevac M. , 2005: Retrieval of microscale wind and temperature fields from single- and dual-Doppler lidar data. J. Appl. Meteor., 44, 13241345.

    • Search Google Scholar
    • Export Citation
  • Newsom, R. K., , Calhoun R. , , Ligon D. , , and Allwine J. , 2008: Linearly organized turbulence structures observed over a suburban area by dual-Doppler lidar. Bound.-Layer Meteor., 127, 111130.

    • Search Google Scholar
    • Export Citation
  • Oda, R., , Iwai H. , , Ishii S. , , Sekizawa S. , , Mizutani K. , , and Murayama Y. , 2011: Structure of turbulence in the urban atmospheric boundary layer detected in the Doppler lidar observation. J. Japan Soc. Civ. Eng., 67B, I_313I_318.

    • Search Google Scholar
    • Export Citation
  • Pichugina, Y. L., , Banta R. M. , , Brewer W. A. , , Sandberg S. P. , , and Hardesty R. M. , 2012: Doppler lidar–based wind-profile measurement system for offshore wind-energy and other marine boundary layer applications. J. Appl. Meteor. Climatol., 51, 327349.

    • Search Google Scholar
    • Export Citation
  • Post, M. J., , and Cupp R. E. , 1990: Optimizing a pulsed Doppler lidar. Appl. Opt., 29, 41454158.

  • Post, M. J., , Schwiesow R. L. , , Cupp R. E. , , Haugen D. A. , , and Newman J. T. , 1978: A comparison of anemometer- and lidar-sensed wind velocity data. J. Appl. Meteor., 17, 11791181.

    • Search Google Scholar
    • Export Citation
  • Rothermel, J., , Kessinger C. , , and Davis D. L. , 1985: Dual-Doppler lidar measurement of winds in the JAWS experiment. J. Atmos. Oceanic Technol., 2, 138147.

    • Search Google Scholar
    • Export Citation
  • Rye, B. J., , and Hardesty R. M. , 1993: Discrete spectral peak estimation in incoherent backscatter heterodyne lidar. I: Spectral accumulation and the Cramer-Rao lower bound. IEEE Trans. Geosci. Remote Sens., 31, 1627.

    • Search Google Scholar
    • Export Citation
  • Sawada, M., and Coauthors, 2012: Transient downslope winds under the influence of stationary lee waves from the Zao mountain range. J. Meteor. Soc. Japan, 90, 79100.

    • Search Google Scholar
    • Export Citation
  • Shun, C. M., , and Chan P. W. , 2008: Applications of an infrared Doppler lidar in detection of wind shear. J. Atmos. Oceanic Technol., 25, 637655.

    • Search Google Scholar
    • Export Citation
  • Stull, R. B., 1988: An Introduction to Boundary Layer Meteorology. Kluwer Academic, 666 pp.

  • Sugimoto, N., and Coauthors, 2008: Lidar network observations of tropospheric aerosols. Lidar Remote Sensing for Environmental Monitoring IX, U. N. Singh, K. Asai, and A. Jayaraman, Eds., International Society for Optical Engineering (SPIE Proceedings, Vol. 7153), doi:10.1117/12.806540.

  • Tsunematsu, N., , Iwai H. , , Ishii S. , , Murayama Y. , , Yasui M. , , and Mizutani K. , 2009: The formation of sharp multi-layered wind structure over Tokyo associated with sea-breeze circulation. SOLA, 5, 14.

    • Search Google Scholar
    • Export Citation
  • Werner, C., 2005: Doppler wind lidar. Lidar: Range-Resolved Optical Remote Sensing of the Atmosphere, C. Weitkamp, Ed., Series in Optical Sciences, Vol. 102, Springer, 339–342.

  • Westbrook, C. D., , and Illingworth A. J. , 2009: Testing the influence of small crystals on ice size spectra using Doppler lidar observations. Geophys. Res. Lett., 36, L12810, doi:10.1029/2009GL038186.

    • Search Google Scholar
    • Export Citation
  • Westbrook, C. D., , Illingworth A. J. , , O’Connor E. J. , , and Hogan R. J. , 2010: Doppler lidar measurements of oriented planar ice crystals falling from supercooled and glaciated layer clouds. Quart. J. Roy. Meteor. Soc., 136, 260276.

    • Search Google Scholar
    • Export Citation
  • Wilczak, J. M., , and Tillman J. E. , 1980: The three-dimensional structure of convection in the atmospheric surface layer. J. Atmos. Sci., 37, 24242443.

    • Search Google Scholar
    • Export Citation
  • Xia, Q., , Lin C.-L. , , Calhoun R. , , and Newsom R. K. , 2008: Retrieval of urban boundary layer structures from Doppler lidar data. Part I: Accuracy assessment. J. Atmos. Sci., 65, 320.

    • Search Google Scholar
    • Export Citation
  • Young, G. S., , Kristovich D. S. , , Hjelmfelt M. R. , , and Foster R. C. , 2002: Rolls, streets, waves and more: A review of quasi-two-dimensional structures in the atmospheric boundary layer. Bull. Amer. Meteor. Soc., 83, 9971001.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Map and topography of Kanto District. Grayscale bar indicates topography height in meters above mean sea level. Black square indicates location of NICT headquarters. Black triangle is location of target building 25 km from Co2DiaWiL. Arrow delineates path from Co2DiaWiL to target.

  • View in gallery

    Concept underlying comparison between mean vertical wind profile from Co2DiaWiL and that from radiosondes. Gray line is radiosonde trajectory. Thee and are height resolutions of 20°- and 70°-elevation conical scans, which are sampling volumes of Co2DiaWiL at each height bin. Radiosonde data were averaged over or . Since observation heights of radiosondes changed with time, closest mean profile in time from Co2DiaWiL was used for comparison.

  • View in gallery

    Sketch of experimental setup for pseudo-dual-Doppler technique, as seen from north. Locations of Co2DiaWiL, steering mirror, and sonic anemometer are indicated by black circles. Large black arrow is pointing north. Coordinate system is defined with Co2DiaWiL at origin. Between-beam angle = 50.76°.

  • View in gallery

    (a) One-second-averaged radial wind velocities of hard target return for sequence of 7200 samples. (b) Pdf of radial wind velocities of hard target return. Average radial wind velocity is , and standard deviation of radial wind velocity is .

  • View in gallery

    Height–time display of (a) radial wind velocity (colored bar in m s−1) and (b) wideband SNR (colored bar in dB) observed by vertical staring mode of Co2DiaWiL, between 0200 and 0400 JST 23 Feb 2010. Positive velocities are updrafts and negative velocities are downdrafts. Data resolution is 1 s temporally and 76 m vertically. (c) Profile of horizontal wind speed (solid line) and direction (dashed line) from radiosonde launched at 0545 JST 23 Feb 2010. (d) Profile of temperature (red line) and relative humidity (blue line) from radiosonde launched at 0545 JST 23 Feb 2010.

  • View in gallery

    Comparison of precision of radial wind velocity of Co2DiaWiL (dots) with theoretical Cramer–Rao lower bound calculated with Eq. (4).

  • View in gallery

    Range–time display of quasi-horizontal radial wind velocity observed by Co2DiaWiL between 0902 and 1102 JST 29 Sep 2010. Negative velocities, represented in blue, indicate flow toward Co2DiaWiL. Positive velocities, in red and yellow, indicate flow away from Co2DiaWiL.

  • View in gallery

    (a) Wideband SNR vs range and (b) radial wind velocity vs range from Co2DiaWiL looking quasi-horizontally at 0947 (red dots) and 1050 JST (black dots) 29 Sep 2010. (c) Ten-minute-averaged power spectra of radial wind velocity fluctuations between 0942 and 0952 JST (red line) and between 1045 and 1055 JST (black line) on 29 Sep 2010. Spectra are smoothed using the Hanning moving average filter. Error bars show the 95% confidence intervals; slope line (dotted line) is shown for reference.

  • View in gallery

    (a) Height–time display of 1-min-averaged radial wind velocity (colored bar in m s−1) with −30-dB wideband SNR threshold observed by vertical staring mode of Co2DiaWiL, between 0000 and 2400 JST 23 Feb 2010. Negative velocities, represented in blue, indicate downdrafts. Positive velocities, in red and yellow, indicate updrafts. From 1000 to 1500 JST, pulse repetition frequency was 15 Hz due to CO2 concentration being measured. (b) Height–time display of 1-s-averaged radial wind velocity with −20-dB wideband SNR threshold observed by vertical staring mode of Co2DiaWiL, between 0330 and 0530 JST 23 Feb 2010. (c) Vertical profiles of temperature from radiosonde launched at 0745 (solid line) and 1331 JST (dashed line) 23 Feb 2010. Vertical profile of (d) u- and (e) υ-component measurements of radiosonde and Co2DiaWiL. Squares indicate data from radiosondes, and black circles indicate data from Co2DiaWiL. Radiosonde launch time was 0745 JST 23 Feb 2010. Components u and υ were from 70°-elevation Co2DiaWiL conical scan.

  • View in gallery

    Time series of (a) 20-s and (b) 1-min-averaged wind data from Co2DiaWiL and sonic anemometer from 1357 to 1539 JST 20 Aug 2010. Squares on thin line represent sonic anemometer data projected onto direction of Co2DiaWiL beam. Data from Co2DiaWiL are indicated by black circles on thick line.

  • View in gallery

    Scatterplots of (a) 1-s, (b) 20-s, and (c) 1-min-averaged wind data, 1100–1141, 1207–1302, 1357–1539, and 1610–1810 JST 20 Aug 2010. Sonic anemometer measurements are given on the y axis, and Co2DiaWiL measurements are given by the x axis. In the legend, N is number of points included, R is correlation coefficient, RMSD is root-mean-square difference between Co2DiaWiL and sonic anemometer measurements, MAE is mean absolute error, and slope and intercept are for the least squares fit line.

  • View in gallery

    Vertical profile of (a) u- and (b) υ-component measurements of radiosonde and Co2DiaWiL. Squares indicate data from radiosondes, and black circles indicate data from Co2DiaWiL. Radiosonde launch time was 0545 JST 21 Feb 2010. Components u and υ were from a 20°-elevation Co2DiaWiL conical scan.

  • View in gallery

    Comparison of (a) u and (b) υ components from 0245 JST 19 Feb 2010 through 1345 JST 25 Feb 2010. Co2DiaWiL values are from 20°-elevation VAD data of Co2DiaWiL and are given on the x axis. The radiosonde measurements are given on the y axis. Legend as in Fig. 11.

  • View in gallery

    As in Fig. 13, but Co2DiaWiL values are from 70°-elevation VAD data of Co2DiaWiL. Squares represent data whose observation altitude was higher than that of 20°-elevation VAD data.

  • View in gallery

    Time series of 1-min-averaged (a) u and (b) υ components retrieved from pseudo-dual-Doppler lidar (black circles on thick line) and corresponding components of sonic anemometer (squares on thin line) from 1914 to 2021 JST 22 Jun 2010.

  • View in gallery

    Time series of wideband SNRs with laser beam oriented directly toward sonic anemometer (thick line), and laser beam reflected by steering mirror (thin line) from 1914 to 2021 JST 22 Jun 2010.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 139 139 36
PDF Downloads 89 89 22

Performance and Technique of Coherent 2-μm Differential Absorption and Wind Lidar for Wind Measurement

View More View Less
  • 1 National Institute of Information and Communications Technology, Koganei, Tokyo, Japan
© Get Permissions
Full access

Abstract

A coherent 2-μm differential absorption and wind lidar (Co2DiaWiL) has been built with a high-power Q-switched Tm,Hm:YLF laser to measure CO2 concentration and radial wind speed. The performance of the Co2DiaWiL is described and analyzed, with a view to demonstrating system capabilities for remote measurements of wind velocities in the atmospheric boundary layer and free troposphere. Bias in the velocity measurements was estimated at −0.0069 m s−1 using measurements from a stationary hard target. The Co2DiaWiL achieved a velocity precision of 0.12 m s−1, derived from the magnitude of random error in radial wind velocity measurements. These measurements were made for ranges out to 20–25 km by using a horizontally fixed beam mode for average times of 1 min. Quantitative intercomparisons of 1-min averages between the Co2DiaWiL and a sonic anemometer revealed a correlation coefficient of 0.99. This study demonstrated measurements of horizontal wind profiles, by making radial wind velocity measurements with the Co2DiaWiL using conical scanning. Profile differences at higher levels could be attributed to probable large horizontal separations of the radiosondes and the low signal-to-noise ratio of the Co2DiaWiL. A pseudo-dual-Doppler technique was developed to retrieve horizontal wind components with a single-Doppler lidar and a steering mirror. Intercomparisons of the 1-min-averaged u and υ components from the pseudo-dual-Doppler lidar measurements with those from the sonic anemometer revealed correlation coefficients of 0.84 and 0.83, respectively.

Current affiliation: Department of Life and Environmental Sciences, Chiba Institute of Technology, Narashino, Chiba, Japan.

Corresponding author address: Hironori Iwai, Okinawa Electromagnetic Technology Center, National Institute of Information and Communications Technology, 4484 Aza-Onna, Onna, Kunigami, Okinawa 904-0411, Japan. E-mail: iwai@nict.go.jp

Abstract

A coherent 2-μm differential absorption and wind lidar (Co2DiaWiL) has been built with a high-power Q-switched Tm,Hm:YLF laser to measure CO2 concentration and radial wind speed. The performance of the Co2DiaWiL is described and analyzed, with a view to demonstrating system capabilities for remote measurements of wind velocities in the atmospheric boundary layer and free troposphere. Bias in the velocity measurements was estimated at −0.0069 m s−1 using measurements from a stationary hard target. The Co2DiaWiL achieved a velocity precision of 0.12 m s−1, derived from the magnitude of random error in radial wind velocity measurements. These measurements were made for ranges out to 20–25 km by using a horizontally fixed beam mode for average times of 1 min. Quantitative intercomparisons of 1-min averages between the Co2DiaWiL and a sonic anemometer revealed a correlation coefficient of 0.99. This study demonstrated measurements of horizontal wind profiles, by making radial wind velocity measurements with the Co2DiaWiL using conical scanning. Profile differences at higher levels could be attributed to probable large horizontal separations of the radiosondes and the low signal-to-noise ratio of the Co2DiaWiL. A pseudo-dual-Doppler technique was developed to retrieve horizontal wind components with a single-Doppler lidar and a steering mirror. Intercomparisons of the 1-min-averaged u and υ components from the pseudo-dual-Doppler lidar measurements with those from the sonic anemometer revealed correlation coefficients of 0.84 and 0.83, respectively.

Current affiliation: Department of Life and Environmental Sciences, Chiba Institute of Technology, Narashino, Chiba, Japan.

Corresponding author address: Hironori Iwai, Okinawa Electromagnetic Technology Center, National Institute of Information and Communications Technology, 4484 Aza-Onna, Onna, Kunigami, Okinawa 904-0411, Japan. E-mail: iwai@nict.go.jp

1. Introduction

Coherent Doppler lidar is one of the most useful active remote sensing instruments, and has the ability to accurately retrieve atmospheric winds at high spatial and temporal resolutions (e.g., Menzies and Hardesty 1989; Henderson et al. 1991, 1993; Huffaker and Hardesty 1996; Grund et al. 2001). Because of recent advances in eye-safe solid-state laser technology, coherent Doppler lidar is becoming more accessible to the meteorological research community (Henderson et al. 2005; Werner 2005). Many important insights into various atmospheric flow patterns, such as the sea breeze (Iwai et al. 2011; Tsunematsu et al. 2009), vortices associated with lee waves (Doyle et al. 2009; Sawada et al. 2012), aircraft wake vortices (Köpp et al. 2004), and dust devil–like vortices (Fujiwara et al. 2011), have been obtained from coherent Doppler lidar. The applications of coherent Doppler lidar have recently been studied to improve wind shear alerts at airports (Chan and Lee 2012; Shun and Chan 2008) and the energy outputs of wind power plants (Käsler et al. 2010; Mann et al. 2010; Pichugina et al. 2012).

Global measurements of wind are essential for numerical weather predictions, climate change studies, and many other meteorological studies. Coherent Doppler lidar has been proposed for global wind measurements using space-based platforms (Baker et al. 1995; Huffaker et al. 1984; Menzies 1986). The National Institute of Information and Communications Technology (NICT) is developing coherent Doppler lidar technology for satellite-based observations of wind on a global scale. Further development of high output energy solid-state laser technology might be required for long-range wind sensing (Koch et al. 2007). Therefore, a 2-μm conductively cooled laser-diode-pumped single-frequency Q-switched solid-state laser with 2.4-W power (80 mJ and 30 Hz) has been developed at NICT for basic studies on a future spaceborne Doppler lidar (Mizutani et al. 2008). Conductively cooling is a key technology to operate the high-output energy solid-state lasers on board satellites. We built a coherent 2-μm differential absorption and wind lidar (Co2DiaWiL) with this laser to measure CO2 concentrations and radial wind speeds (Ishii et al. 2010).

We conducted experiments to test the capabilities of the Co2DiaWiL to take wind measurements. First, bias and random error in the Co2DiaWiL velocity measurements were investigated by using horizontally and vertically fixed beam modes, respectively. Bias is the total systematic error in contrast to random error. Precision is the standard deviation of random error. We compared the radial component measured by a fixed beam Co2DiaWiL with the same component measured by a nearby sonic anemometer to check the fundamental accuracy of the Co2DiaWiL. The term accuracy involves a combination of random components and a bias component. Horizontal wind (both u and υ components) and the wind profile under various conditions were determined by comparing the Co2DiaWiL results with conventional radiosonde profiles.

Doppler lidar has the ability to observe the structure and evolution of boundary layer flow at high spatial and temporal resolution. However, direct measurements are limited to the radial wind component when using a single-Doppler lidar. So-called dual-Doppler analysis techniques have existed to retrieve the vector wind fields. However, dual (or multiple)-Doppler lidar measurements have been rare (Rothermel et al. 1985; Collier et al. 2005; Newsom et al. 2005; Calhoun et al. 2006; Xia et al. 2008; Newsom et al. 2008; Drechsel et al. 2009, 2010) due to the high cost and complexity of operating Doppler lidar. There has been limited research on comparisons between wind fields retrieved from dual-Doppler lidar measurements and other instruments (Drechsel et al. 2009). We developed a pseudo-dual-Doppler technique in this study to retrieve the horizontal wind components (u and υ) with a single-Doppler lidar and steering mirror, and conducted a field experiment to test the performance of the technique. We compared horizontal wind components from the pseudo-dual-Doppler technique with those from a sonic anemometer.

2. Description and data analysis of Co2DiaWiL

a. Laser transmitter and receiver

The Co2DiaWiL was developed to measure CO2 concentration and radial wind speed (Ishii et al. 2010). It was housed in a container, which was stationed on the rooftop of a building at NICT headquarters (35.71°N, 139.49°E; height is 75 m above mean sea level; denoted by the closed square in Fig. 1), 20 m above ground level. This may serve as a test bed for future airborne and spaceborne lidar missions. A 2-μm conductively cooled laser-diode-pumped single-frequency Q-switched Tm,Ho:YLF laser (Mizutani et al. 2008) was used in the Co2DiaWiL to make long-range CO2 and wind measurements. This laser has an operating wavelength of 2.05 μm, an output energy of 80 mJ, a pulse width of 150 ns [full width at half maximum (FWHM)], and a pulse repetition frequency of 30 Hz. While the previous system described in Ishii et al. (2010) had two master oscillators (MOs) that were single-frequency continuous-wave (CW) Tm,Ho:YLF lasers for injection seeding, the present system had three MOs to take long-range measurements of CO2 concentration (Ishii et al. 2012). A block diagram of the present system was presented in Ishii et al. (2012). The system parameters for wind measurements are listed in Table 1.

Fig. 1.
Fig. 1.

Map and topography of Kanto District. Grayscale bar indicates topography height in meters above mean sea level. Black square indicates location of NICT headquarters. Black triangle is location of target building 25 km from Co2DiaWiL. Arrow delineates path from Co2DiaWiL to target.

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

Table 1.

Co2DiaWiL specifications for wind measurements.

Table 1.

The MO at a wavelength of 2051.250 nm, corresponding to the far wing of the R30 absorption line of CO2, is used for the wind measurements. The laser is controlled only by adjusting the resonator temperature and piezoelectric movement of the output coupler element. Part of the laser beam transmitted from the MO is diffracted by a single-crystal, germanium acousto-optic modulator. The acousto-optic modulator upshifts the frequency of the MO laser beam by 105 MHz to create an intermediate frequency, and the diffracted beam is injected into the acousto-optic Q-switched (AO Q-sw) slave oscillator (SO) in a ring-configuration resonator. A single-frequency Q-switched laser pulse is obtained by injection seeding with the upshifted MO laser beam matched to the SO with the ramp-and-fire technique (Henderson et al. 1986). The SO ring cavity is swept until a resonance is detected that triggers the AO Q-sw.

The pulsed laser beam is transmitted to the atmosphere by a 0.1-m-diameter off-axis telescope. After being expanded by the telescope, the pulsed laser beam is pointed and scanned with a waterproof two-axis scanner mounted on the container roof. The scanner is capable of full hemispherical coverage, 0.01° precision, and scanning speeds of up to 60° s−1. A scan sequence is downloaded in operation to a motion controller card (NOVA Electronics, MC8043P) housed in the personal computer (PC) that acquired data for the Co2DiaWiL, and the scanner independently executes the scan sequence until it is completed. The elevation and azimuth angles of the scanner are acquired from the PC every single shot, by reading the outputs of an encoder built into a scanner controller.

The signal backscattered by moving aerosol particles is photomixed with a portion of the laser beam transmitted from the MO on an InGaAs-PIN photodiode (DET1). A small portion of the pulsed laser beam is also photomixed with offline laser output on a balanced InGaAs-PIN photodiode (DET2) to monitor the frequency of the outgoing laser pulse. The two photoreceivers, DET1 and DET2, convert the frequencies of the laser pulse and backscattered signals into intermediate frequencies for heterodyne detection. The outputs of the photoreceivers are passed through a preamplifier and bandpass filter.

b. Data acquisition and signal processing

The outputs of DET1 and DET2 are digitized at a 500-MHz sampling frequency , using 8-bit analog-to-digital (A/D) converters (Gage Applied Technologies, CompuScope 82G) starting from a 30-Hz trigger. The current system digitized 131 072 samples for the DET1 output (backscattered signals) and 4096 samples for the DET2 output (monitor pulse). The 131 072 samples of backscattered signal at a sampling frequency of 500 MHz are corresponding to a data system limited range of 39.2 km. The 131 072 samples of backscattered signals are divided into 512 segments of 256 samples (corresponding to 76.75-m lengths), and the segments are called range gates. The monitor pulse is available as a reference for subsequent signal processing steps to calculate ranges and correct pulse-by-pulse frequencies. It is necessary to correct frequencies because the pulse-by-pulse frequency jitter of the laser is about 1 MHz (Ishii et al. 2012), corresponding to 1 m s−1 at 2 μm, and the goal of the velocity performance requires about 0.1 m s−1 precision in measurements. The digital signals of backscattered signals and monitor pulses are stored in the PC of the Co2DiaWiL to acquire data, and they are later processed by software developed in C–C++ programming language. We have been developing a real-time processing system using graphics processing units, or, more commonly, video cards.

The radial wind velocity is calculated by
e1
with laser wavelength (2051.250 nm) and Doppler-shifted frequency , which is the difference between the frequencies of backscattered signals and monitor pulses. It is essential to accurately estimate the frequencies of backscattered signals and monitor pulses to achieve highly precise wind measurements. A Levin’s maximum likelihood discrete spectral peak estimator [Rye and Hardesty 1993; originally presented by Levin (1965) and extended to include spectral accumulation] is used for estimating these frequencies. An algorithm proposed by Frehlich et al. (1997) is used to produce the spectrum of noise-corrected and frequency-corrected backscattered signals at each range gate. The calculations involve the following four steps:
  1. A noise spectrum is produced with the discrete Fourier transform (DFT) from data at the tail of each backscattered signal, where the aerosol signal is negligible. The noise spectrum is accumulated for accurate estimates with pulses.
  2. A noise-corrected backscattered signal is produced from the original data sequence of backscattered signals by the whitening algorithm for each lidar pulse. A noise-corrected backscattered signal spectrum is produced by dividing the DFT of the backscattered signal by the noise spectrum. A noise-corrected backscattered signal is calculated by using the inverse DFT of the spectrum for the noise-corrected backscattered signal.
  3. The power spectrum of the monitor pulse is calculated by using a 4096-point DFT. The frequency of monitor pulse is that corresponding to the peak of the spectrum, estimated with the maximum likelihood estimator that will be described later. Frequency is corrected pulse by pulse for each lidar pulse using the noise-corrected backscattered signal and frequency of the monitor pulse. This process shifts the zero Doppler-shifted frequency to .
  4. The spectrum of the noise-corrected and frequency-corrected backscattered signal is produced with the DFT at each range gate. The spectrum is accumulated for accurate estimates with pulses.
We assume that the spectrum of the monitor pulse and backscattered signal at each range gate has a Gaussian profile, and the spectrum of noise is equal to unity. The spectral model given by Rye and Hardesty (1993) is
e2
where represents an expected value for the spectrum; refers to a frequency normalized to sampling frequency (500 MHz); ; and and are the spectral peak and spectral width normalized to the sampling frequency, respectively. Here, is the wideband signal-to-noise ratio (wideband SNR: ratio of total signal power to noise power over the entire spectral bandwidth) and subscript indicates the ith spectral component. The maximum likelihood estimator is a promising algorithm to estimate parameter by maximizing the likelihood function of the spectral data (Rye and Hardesty 1993). The maximum likelihood estimates of parameters are those values that maximize the following log-likelihood function, (Rye and Hardesty 1993):
e3
where is a constant, is the total number of spectral estimates, and is the spectrum of the monitor pulse or backscattered signal at each range gate. Here, the spectrum is calculated from the spectrum model described in Eq. (2). Maximum likelihood estimate is obtained by adjusting until is maximized. Since the maximum likelihood estimator of signal power (i.e., wideband SNR) cannot be applied to cases of large SNR because of large errors in estimation and numerical instability (Frehlich 1999), wideband SNR is calculated with Eq. (7) of Frehlich et al. (1997).

3. Experimental procedures

a. Bias

We investigated the presence of bias in the radial wind velocity measurements of Co2DiaWiL to evaluate systematic error in its velocity measurements. Co2DiaWiL was pointed at the hard target of a stationary building at about a range of 25 km (Fig. 1) to determine this bias. The bias was calculated from 7200 records of 1-s-averaged radial wind velocities (1-s average corresponding to 30 laser pulses) starting from 0802 Japan standard time (JST) 12 September 2010.

b. Random error

We evaluated the experimentally observed standard deviation of vertical velocity as a function of wideband SNR to quantify random error in the velocity measurements. Vertical velocity was observed between 0200 and 0400 JST 23 February 2010. Standard deviation was calculated from 10-min records of 1-s-averaged vertical velocity by using the velocity-difference method of Frehlich (2001). The theoretical Cramer–Rao lower bound (CRLB) on the standard deviation of radial wind velocity estimate is defined as the following equation from Rye and Hardesty (1993):
e4
where is the number of accumulated pulses and is the number of spectral channels.

c. Long-range measurements

Because the Co2DiaWiL has high laser output power (2.4 W), it can measure radial wind velocities at long ranges. Ishii et al. (2010) showed that it is capable of measuring radial wind velocities over horizontal ranges up to 20 km, for accumulation times of 3 min. The radial wind velocities were estimated with the first spectral moment of backscattered signals in Ishii et al. (2010). We collected data on radial wind velocities estimated with the maximum likelihood estimator for accumulation times of 1 min (1-min average corresponding to 1800 laser pulses) on 29 September 2010. The Co2DiaWiL was pointed at the same building used to estimate velocity measurement bias (section 3a).

The Co2DiaWiL was pointed toward the zenith on 23 February 2010 to measure the profile of vertical wind and CO2 concentration. The pulse repetition frequency was 15 Hz from 1000 to 1500 JST, when the Co2DiaWiL measured CO2 concentration. The vertical velocities were calculated for accumulation times of 1 min.

d. Sonic anemometer comparisons

The radial wind velocities measured with Co2DiaWiL were compared with wind speeds measured with a three-axis ultrasonic anemometer (SONIC SAT-540) mounted on a tower at a height of 59 m above ground level (AGL). The location for the anemometer was about 120 m south of Co2DiaWiL. The sonic anemometer measured the three components of wind at a rate of 10 Hz. We assumed that the sonic anemometer sampling volume is a spherical volume with a diameter of 0.1 m based on the physical separation of the three measurement paths within the spatial geometry of the sensing head and estimated the sampling volume to be about 0.0005 m3. Wind speed was measured within a range of 0–60 m s−1, and with a precision of ±(4% of reading +0.05 m s−1) and a resolution of 0.01 m s−1. The laser beam of Co2DiaWiL was oriented along an azimuth of 191.57° and an elevation angle of 17.94°. The sampling volume was roughly a slant cylinder, 76 m long and 0.08 m in diameter, centered at 153 m from the Co2DiaWiL. Because the outgoing laser pulse scattered by system optics made very near-field signal returns unusable, the minimum range of the Co2DiaWiL was restricted to 153 m. The center of the sampling volume and sonic anemometer were about 25 m apart, and the sampling volume center was about 7 m higher than the anemometer. The three wind components (uSAT, υSAT, wSAT) observed by the sonic anemometer were projected onto the direction of the radial wind velocity measured by the Co2DiaWiL with the following equation:
e5
where VrSAT is the projected wind speed and and are the azimuth and elevation angles of the Co2DiaWiL laser beam, respectively.

e. Radiosonde comparisons

A total of 36 Vaisala RS92-SGP radiosondes were launched from NICT headquarters from 14 through 25 February 2010. The radiosonde launch site was about 300 m north-northwest of the Co2DiaWiL. The radiosondes transmitted observed data every 2 s to a Vaisala MW15 ground receiver unit, where the data were processed using Vaisala proprietary software (DigiCORA version). The height resolution is about 10 m, given the typical ascent rate of 5 m s−1. The DigiCORA software used changes in the GPS location to compute the u and υ components of horizontal wind with an accuracy of 0.2 m s−1. Of the 36 radiosondes listed in Table 2, 25 were used for comparison here. Comparisons were done during wintertime, when strong westerly winds prevailed over the observation site.

Table 2.

Launch times of 25 Vaisala RS92-SGP radiosondes used for comparison, and maximum height and number of samples corresponding to 20°- and 70°-elevation Co2DiaWiL conical scans.

Table 2.

The Co2DiaWiL executed 20°-elevation and 70°-elevation conical scans to enable a comparison with the radiosondes, and vertical profiles of horizontal wind velocity were retrieved with the velocity–azimuth display (VAD) technique described by Browning and Wexler (1968). The VAD technique could retrieve the u and υ components of horizontal wind from radial wind velocity data around horizontal circles centered on the vertical of the lidar scanner. The 20°- and 70°-elevation conical scans took 1 min each to complete, and included about 120 radial wind velocities with 15 pulse accumulations. During a pulse accumulation time of 0.5 s, the azimuth angle moved through about 3°. The 20°- and 70°-elevation conical scans were carried out alternately. The height resolutions of the conical scans were about 26 m for the former and 72 m for the latter. Since the height resolution of 70°-elevation conical scans was much coarser than that of the radiosondes, the wind measurements with the 20°-elevation conical scans were suitable for comparisons of wind data between the Co2DiaWiL and radiosonde under the strong vertical shear conditions of horizontal wind at low altitudes. A high-elevation angle was required for wind measurements at higher altitudes. Table 2 summarizes launch times of radiosondes used to compare, and the maximum height and sample number of the corresponding 20°- and 70°-elevation conical scans of the Co2DiaWiL. The lidar signal was weaker at upper levels (i.e., longer ranges), and the number of “bad” velocity estimates increased significantly, producing inaccurate estimates of the vertical profiles of horizontal wind velocity.

The two instruments generally sampled velocities at different locations. The Co2DiaWiL provided simultaneous measurements over all heights at a fixed location (Eulerian sense), whereas the radiosonde measured velocities at different heights, times, and locations. Figure 2 outlines the concept underlying the procedure to compare the mean vertical wind profile obtained from the Co2DiaWiL and that obtained from the radiosondes. Since the height resolution of the radiosondes was greater than that of the Co2DiaWiL, radiosonde data were averaged over the sampling volume of the Co2DiaWiL at each height bin. The Co2DiaWiL data were used to obtain time-averaged profiles of u and υ components, and these mean profiles were compared with instantaneous data from the radiosondes at the same height. Since radiosonde observation heights changed with time, the closest mean profile in time from the Co2DiaWiL was used for comparison.

Fig. 2.
Fig. 2.

Concept underlying comparison between mean vertical wind profile from Co2DiaWiL and that from radiosondes. Gray line is radiosonde trajectory. Thee and are height resolutions of 20°- and 70°-elevation conical scans, which are sampling volumes of Co2DiaWiL at each height bin. Radiosonde data were averaged over or . Since observation heights of radiosondes changed with time, closest mean profile in time from Co2DiaWiL was used for comparison.

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

f. Pseudo-dual-Doppler measurements

Figure 3 shows a sketch of the experimental setup for the pseudo-dual-Doppler technique that was viewed from the north. The locations of the Co2DiaWiL, steering mirror, and sonic anemometer are indicated by the closed circles. The steering mirror was about 160 m from the Co2DiaWiL, whose laser beam was redirected toward the sonic anemometer by the mirror. The angle of the steering mirror was adjusted to reflect the laser beam from the Co2DiaWiL toward the sonic anemometer. The steering mirror was the same as that used by Ishii et al. (2010). Since the diameter of the laser beam was 8 cm and that of steering mirror was 15 cm, the elevation and azimuth angles of the scanner had to be adjusted with a precision of 0.01°. The pseudo-dual-Doppler measurements were made using the following four steps:

  1. The laser beam of the Co2DiaWiL was oriented toward the sonic anemometer, and radial wind velocity was observed at 1-s temporal resolution.
  2. The laser beam of the Co2DiaWiL was oriented toward the steering mirror, which took about 2.5 s.
  3. Radial wind velocity was observed at 1-s temporal resolution.
  4. The Co2DiaWiL laser beam was oriented toward the sonic anemometer, which took about 2.5 s.
Fig. 3.
Fig. 3.

Sketch of experimental setup for pseudo-dual-Doppler technique, as seen from north. Locations of Co2DiaWiL, steering mirror, and sonic anemometer are indicated by black circles. Large black arrow is pointing north. Coordinate system is defined with Co2DiaWiL at origin. Between-beam angle = 50.76°.

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

Radial wind velocities and were observed every 6 s and averaged every minute. Horizontal wind components (u and υ) were calculated with the following two equations:
e6
e7
where (, , ) and (, , ) are coordinates of the sonic anemometer with respect to the Co2DiaWiL and steering mirror, and and correspond to distances from the Co2DiaWiL and steering mirror to the sonic anemometer (Fig. 3). The is vertical velocity at the sonic anemometer, and it could not directly be observed by the Co2DiaWiL. Since the absolute values of 1-min- averaged measured by the sonic anemometer were lower than 0.5 m s−1 over the observation period from 1914 to 2021 JST 22 June 2010, the second terms in Eqs. (6) and (7) were small relative to the first terms and can be ignored.

4. Results

a. Bias

Figure 4a plots the 1-s-averaged radial wind velocities of the hard target return for a sequence of 7200 samples, whereas Fig. 4b shows the probability density function (pdf) of the radial wind velocities. We found that the average radial wind velocity for these 7200 samples was −0.0069 m s−1 (−0.69 cm s−1), indicating that there was minimal bias in the velocity measurements. The standard deviation of radial wind velocity was 0.081 m s−1 (8.1 cm s−1). Accurate radial wind velocity measurements could be made, and these error values could be compared with those from other Doppler lidars. For example, Henderson et al. (1991) reported a mean value from 96 single-shot radial wind velocity estimates from a flashlamp-pumped 2.1-μm Doppler lidar of −3.3 cm s−1 and a standard deviation of 11 cm s−1. Frehlich et al. (1994) used the same Doppler lidar and reported a mean value from 410 single-shot radial wind velocity estimates of −2.24 cm s−1 and a standard deviation of 7.18 cm s−1. Post and Cupp (1990) used a transverse-excited atmospheric CO2 Doppler lidar at a wavelength of 10.59 μm and reported a mean value from 1000 single-shot radial wind velocity estimates of −0.15 m s−1 and a standard deviation of 0.64 m s−1.

Fig. 4.
Fig. 4.

(a) One-second-averaged radial wind velocities of hard target return for sequence of 7200 samples. (b) Pdf of radial wind velocities of hard target return. Average radial wind velocity is , and standard deviation of radial wind velocity is .

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

b. Random error

Figure 5a shows a height–time display of 1-s-averaged vertical velocities observed by the Co2DiaWiL between 0200 and 0400 JST 23 February 2010. Wind velocity transverse to the lidar beam axis () was estimated over the observation period by using radiosonde data from NICT headquarters (Fig. 1) from 0545 JST 23 February. Figure 5c plots a profile of the horizontal wind speed from the radiosonde with = 2–8 m s−1. The horizontal distance was = 2–8 m during the dwell time, = 1 s, of vertical velocity measurements. Since the range gate length is about 76 m, = 0.026–0.105. The velocity-difference method outperformed the covariance-based or spectral-based methods in this regime (Frehlich 2001). Figure 5b shows a height–time display of wideband SNR for the same period as that in Fig. 5a. A strong temperature inversion formed at about 350 m AGL that was obtained from the radiosonde data; and a weak inversion formed at about 1500 m AGL (Fig. 5d). There was higher relative humidity under the low-level inversion, which was related to the greater wideband SNR observed by the Co2DiaWiL at lower levels. The relative humidity was approximately constant between the two temperature inversions, and gradually decreased above the high-level inversion. Figure 6 plots standard deviations () of vertical velocity evaluated with the velocity-difference method for 296 individual 10-min-long vertical velocity records. Here, the values are plotted versus wideband SNR and compared with values from theoretical CRLB calculated from Eq. (4). How closely a Doppler lidar achieves CRLB is a function of 1) hardware performance (such as frequency drift by the laser, nonlinear amplifiers, digitization errors, and spectral noise content of the electronics); 2) the signal processing algorithm, including an incorrect model for the spectrum; and 3) inhomogeneities in the wind field due to turbulence over the range gate (Frehlich et al. 1994; Grund et al. 2001). As seen from Fig. 6, the Co2DiaWiL operates near the theoretical limit under both high and low wideband SNR conditions. Errors are constant (about 0.12 m s−1) in the high-SNR region because of the effect of speckle noise. This noise is derived from the interference of randomly phased backscattered fields from individual aerosol particles, and is independent of wideband SNR. Therefore, speckle noise is the dominant noise source in the high-SNR region (Rye and Hardesty 1993), and it limits the precision of velocity. Errors increase with reduced values of wideband SNR to approximately 0.4 m s−1 at a wideband SNR of −15 dB.

Fig. 5.
Fig. 5.

Height–time display of (a) radial wind velocity (colored bar in m s−1) and (b) wideband SNR (colored bar in dB) observed by vertical staring mode of Co2DiaWiL, between 0200 and 0400 JST 23 Feb 2010. Positive velocities are updrafts and negative velocities are downdrafts. Data resolution is 1 s temporally and 76 m vertically. (c) Profile of horizontal wind speed (solid line) and direction (dashed line) from radiosonde launched at 0545 JST 23 Feb 2010. (d) Profile of temperature (red line) and relative humidity (blue line) from radiosonde launched at 0545 JST 23 Feb 2010.

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

Fig. 6.
Fig. 6.

Comparison of precision of radial wind velocity of Co2DiaWiL (dots) with theoretical Cramer–Rao lower bound calculated with Eq. (4).

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

c. Long-range measurements

Figure 7 shows 1-min-averaged radial wind velocity in range–time format, with a −30 dB wideband SNR threshold. Positive values (warm colors) for these data indicate motion away from the Co2DiaWiL. During a 2-h period in the morning (0902–1102 JST 29 September 2010), the maximum horizontal measurement range fluctuated from 20 to 25 km. Since the Co2DiaWiL was pointed nearly east and the northerly surface winds ranging from 2 to 5 m s−1 observed by ambient air pollution monitoring stations (not shown) prevailed across the Co2DiaWiL observation range, the Co2DiaWiL observed a cross-stream component at an observation height of m. Figure 8a plots wideband SNR as a function of range at 0947 and 1050 JST, and Fig. 8b plots radial wind velocity as a function of range over the same period. Wideband SNR gradually decreased with range, and it was approximately −30 dB beyond 20 km. Since high wideband SNRs (6–10 dB) and zero-velocity returns (Fig. 8b) were detected at 25 km, range measurements to the hard target were consistent. The quasi-horizontal radial wind velocity plot in Fig. 7 indicates turbulent motion within the atmospheric boundary layer (ABL). Some of the turbulent structure is organized after about 1030 JST. Radial wind velocity at 1050 JST in Fig. 8b indicates an organized wave structure. Figure 8c plots the power spectra of radial wind velocities in an approximately 17-km domain along the line of sight, after the mean component has been removed. The spectra were averaged over 10-min periods (0942–0952 and 1045–1055 JST) and smoothed using the Hanning moving average filter (Goodall 1990). ABL height at 1000 JST estimated from a Mie lidar of the National Institute for Environmental Studies (Sugimoto et al. 2008; http://www-lidar.nies.go.jp/Tokyo/), which was located about 20 km east of NICT headquarters, was about 1 km. Kaimal et al. (1976) demonstrated that the spectral peak wavelength in the cross-stream component was about without dependence and that the cross-stream component had a spectral slope where the wavelength was longer than the peak wavelength. Consequently, we used the spectral slope as a reference where the wavelength was shorter than 1.5 km (dotted line in Fig. 8c). The small peak around 460 m of the spectrum observed between 0942 and 0952 JST presumably corresponds to near-surface streaks that are spatially organized eddies generally aligned with the mean wind. Large-eddy simulations (LESs: e.g., Lin et al. 1997; Drobinski and Foster 2003) and Doppler lidar observations (e.g., Drobinski et al. 2004; Newsom et al. 2008) revealed near-surface streaks with an average spacing of hundreds of meters in ABL flows with neutral stratification. Using LES, Lin et al. (1997) derived a relationship between the streak spacing () and the distance from ground and ABL height :
e8
where and . At m above ground with m, Eq. (8) gives m, which is in good agreement with the small peak wavelength that was observed. The smaller peaks around 249, 272, and 339 m of the spectrum observed between 0942 and 0952 JST presumably corresponds to cat’s paws observed on water surface by Hunt and Morrison (2000) or near-surface plumes elongated in the mean wind direction by Wilczak and Tillman (1980). This multiscale eddy structure near the surface was observed by the National Oceanic and Atmospheric Administration’s (NOAA) High-Resolution Doppler Lidar (Drobinski et al. 2007). Stability parameter, (Stull 1988), where is the Monin–Obukhov length, is often used to identify the respective roles played by dynamics and thermal instability in the shape of horizontal convective rolls (HCRs: e.g., Etling and Brown 1993; Young et al. 2002) and streaks (e.g., Khanna and Brasseur 1998). We calculated the stability parameter using the tower-mounted sonic anemometer (section 3d) by averaging over a 10-min period. The stability parameter at 0947 JST was 5.34 and at 1050 JST was 9.63. Using LES, Khanna and Brasseur (1998) demonstrated that the streak spacing decreased and HCRs formed with a wavelength of about at higher . The peak around 1.4 km of the spectrum observed between 1045 and 1055 JST is presumably associated with HCRs. Although the spectral peak wavelength of 1.4 km is shorter than m, one possible reason for the disparity is the estimation error of , which was obtained from a Mie lidar located about 20 km from the observation site. The shorter wavelength variations corresponding to the smaller peak around 800 m were present from near Co2DiaWiL to about 7 km (Figs. 7 and 8b). The smaller spectral peak wavelength of 800 m is roughly the same degree of and presumably associated with the thermal streets, where individual thermals are aligned in parallel rows along the mean wind direction (Konrad 1970).
Fig. 7.
Fig. 7.

Range–time display of quasi-horizontal radial wind velocity observed by Co2DiaWiL between 0902 and 1102 JST 29 Sep 2010. Negative velocities, represented in blue, indicate flow toward Co2DiaWiL. Positive velocities, in red and yellow, indicate flow away from Co2DiaWiL.

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

Fig. 8.
Fig. 8.

(a) Wideband SNR vs range and (b) radial wind velocity vs range from Co2DiaWiL looking quasi-horizontally at 0947 (red dots) and 1050 JST (black dots) 29 Sep 2010. (c) Ten-minute-averaged power spectra of radial wind velocity fluctuations between 0942 and 0952 JST (red line) and between 1045 and 1055 JST (black line) on 29 Sep 2010. Spectra are smoothed using the Hanning moving average filter. Error bars show the 95% confidence intervals; slope line (dotted line) is shown for reference.

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

Figure 9a shows the one-day profile for the 1-min averaged vertical velocities with the −30-dB SNR threshold. The data were collected on 23 February 2010 with the laser beam pointed toward the zenith. Weak updrafts and downdrafts were observed during the early morning (0000–0800 JST) up to an altitude of approximately 2.5 km AGL before the signal dropped into noise. Weak aerosol backscatter was observed between 4 and 6 km AGL. High-altitude cirrus clouds were observed between 7 and 11 km AGL. Prevailing westerlies of 40–60 m s−1 blew in the altitude (Figs. 9d and 9e). The tropopause height was around 10.5 km according to the in situ measurements made with a radiosonde launched at 0745 JST from NICT headquarters (Fig. 9c). Figure 9b shows the profile of the 1-s-averaged vertical velocities with a −20-dB SNR threshold between 7.0 and 9.5 km AGL from 0330 to 0530 JST. The vertical velocity measurements indicate a downward motion within the cirrus cloud, presumably ice crystal virga, ranging from 1 to 3 m s−1. The high spatial and velocity resolution of Co2DiaWiL offers new capabilities for studying cloud dynamics in detail (e.g., Westbrook and Illingworth 2009; Westbrook et al. 2010). After 0830 JST, the vertical velocities were observed from near ground to 4–5.5 km AGL. Since the ABL height estimated from a radiosonde launched at 1331 JST (Fig. 9c) was about 1.3 km, the Co2DiaWil seamlessly measured the vertical air motion in the ABL and lower troposphere.

Fig. 9.
Fig. 9.

(a) Height–time display of 1-min-averaged radial wind velocity (colored bar in m s−1) with −30-dB wideband SNR threshold observed by vertical staring mode of Co2DiaWiL, between 0000 and 2400 JST 23 Feb 2010. Negative velocities, represented in blue, indicate downdrafts. Positive velocities, in red and yellow, indicate updrafts. From 1000 to 1500 JST, pulse repetition frequency was 15 Hz due to CO2 concentration being measured. (b) Height–time display of 1-s-averaged radial wind velocity with −20-dB wideband SNR threshold observed by vertical staring mode of Co2DiaWiL, between 0330 and 0530 JST 23 Feb 2010. (c) Vertical profiles of temperature from radiosonde launched at 0745 (solid line) and 1331 JST (dashed line) 23 Feb 2010. Vertical profile of (d) u- and (e) υ-component measurements of radiosonde and Co2DiaWiL. Squares indicate data from radiosondes, and black circles indicate data from Co2DiaWiL. Radiosonde launch time was 0745 JST 23 Feb 2010. Components u and υ were from 70°-elevation Co2DiaWiL conical scan.

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

d. Sonic anemometer comparisons

The Co2DiaWiL observed radial wind velocities with 1-s temporal resolution from 1100 to 1141, 1207 to 1302, 1357 to 1539, and 1610 to 1810 JST 20 August 2010. Figure 10 plots a time series of 20-s and 1-min-averaged wind data from the Co2DiaWiL and sonic anemometer, from 1357 to 1539 JST. The 20-s-averaged wind data (Fig. 10a) are generally in good agreement for wind fluctuations, although there are some greater differences in measured wind speeds for rapid fluctuations. Whereas the sonic anemometer represents point velocity measurements, the Co2DiaWiL-measured velocity at a single range gate represents the mean radial wind velocity of a cylindrical volume 76 m in length and 0.08 m in diameter. The smaller eddies (<0.1 m) measured by the sonic anemometer may not be large enough to influence the entire Co2DiaWiL sampling volume. The 1-min-averaged wind data (Fig. 10b) show how well the two datasets track each other, and there is excellent agreement in wind fluctuations.

Fig. 10.
Fig. 10.

Time series of (a) 20-s and (b) 1-min-averaged wind data from Co2DiaWiL and sonic anemometer from 1357 to 1539 JST 20 Aug 2010. Squares on thin line represent sonic anemometer data projected onto direction of Co2DiaWiL beam. Data from Co2DiaWiL are indicated by black circles on thick line.

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

Figure 11 compares 1-s-, 20-s-, and 1-min-averaged wind speeds from the Co2DiaWiL and sonic anemometer. The correlation coefficient between 1-s-averaged radial wind velocities from the two instruments was 0.96, with a root-mean-square (rms) difference of 0.51 m s−1. The longer averaging time yielded a better correlation coefficient and rms difference, which were the same as in the results found by Köpp et al. (1984). They compared a CW 10.6-μm CO2 Doppler lidar to a sonic anemometer. They found correlation coefficients of 0.97 and 0.98 and rms differences of 0.19 and 0.12 m s−1 for 20- and 60-s averaging periods, respectively. The correlation coefficient for 1-min radial wind velocities from the Co2DiaWiL and sonic anemometer was 0.99 and their rms difference was 0.28 m s−1. Good correlation was also indicated by the linear least squares technique, which gave straight-line slopes of 0.98, and straight-line intercepts on the vertical (i.e., sonic anemometer) axis of 0.11 m s−1. The mean absolute error between Co2DiaWiL and sonic anemometer radial wind velocity measurements was 0.22 m s−1, and differences between their data during this period were mostly within ±0.5 m s−1 (never >1 m s−1). This is good agreement, considering the different sampling volumes and techniques. Since the wind speed was in a range of −4 to 4 m s−1, comparison under stronger wind conditions would be needed. The correlation coefficient of 0.98 for 1-min-averaged data is similar to comparisons of CW 10.6-μm CO2 Doppler lidars (with cylindrical volumes of 20–100 m in length and 0.03–0.3 m in diameter) to the cup anemometer used by Post et al. (1978) (0.96–0.99), the propeller anemometer used by Mayor et al. (1997) (0.97), and the sonic anemometer used by Köpp et al. (1984) (0.98). Kelley et al. (2007) found good agreement between the NOAA’s High-Resolution Doppler Lidar [2-μm wavelength, cylindrical volume 30 m in length, and 0.2 m in diameter: Grund et al. (2001)] and 1-min averaged measurements from a sonic anemometer, with a correlation coefficient of 0.95 and an rms difference of 0.34 m s−1. From the above it is noted that the differences in laser wavelengths and sensing volumes of Doppler lidars had no significant impact on comparisons with sonic anemometers.

Fig. 11.
Fig. 11.

Scatterplots of (a) 1-s, (b) 20-s, and (c) 1-min-averaged wind data, 1100–1141, 1207–1302, 1357–1539, and 1610–1810 JST 20 Aug 2010. Sonic anemometer measurements are given on the y axis, and Co2DiaWiL measurements are given by the x axis. In the legend, N is number of points included, R is correlation coefficient, RMSD is root-mean-square difference between Co2DiaWiL and sonic anemometer measurements, MAE is mean absolute error, and slope and intercept are for the least squares fit line.

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

The results reveal that the two independent wind measurements agree closely, on average, despite the inherent differences between remote, volume-averaged, and point measurements. Note that the agreement largely depends on the accumulation time to make both datasets more representative. Some turbulent eddies sensed by the Co2DiaWiL may not have appeared in the sonic anemometer sampling volume, and other smaller eddies sensed by the sonic anemometer may not have been large enough to affect the entire Co2DiaWiL sampling volume sufficiently to alter its volume-averaged velocity value. Another difference is that the center of the Co2DiaWiL sampling volume was about 25 m from the sonic anemometer, and about 7 m higher.

e. Radiosonde comparisons

Figures 9d, 9e, 12a, and 12b plot the vertical profiles of u- and υ-component measurements from the radiosonde and Co2DiaWiL. The agreement in these figures is generally good, although there are some systematic differences that do not seem to be caused by measurement errors in either instrument. Figures 12a and 12b show a pair of profiles retrieved from a 20°-elevation conical scan of a nocturnal low-level jet (e.g., Blackadar 1957). The density of aerosols under cloud-free conditions was too low for the Co2DiaWiL to obtain reliable returns at heights greater than about 1300 m for the 20°-elevation conical scan. While the horizontal wind speed derived from the Co2DiaWiL was slightly lower than that from the radiosonde below the level of maximum wind speed, the Co2DiaWiL υ components were about 1 m s−1 higher than those from the radiosonde between the maximum wind speed level and 1000 m AGL. The error of horizontal wind speed below the level of maximum wind speed derived from the Co2DiaWiL might be seen due to the influence of shear and turbulence generated by the nocturnal low-level jet (Banta et al. 2006). Since a radius of the 20°-elevation VAD horizontal circle is longer (e.g., 1374 m at 500 m AGL) and atmospheric variances along the circle are larger with height, the error above the maximum wind speed level might be seen due to the influence of atmospheric variances. Figures 9c and 9d show a pair of profiles retrieved from a 70°-elevation conical scan in the morning. The Co2DiaWiL υ components were about 1 m s−1 higher than those of the radiosonde in the wind shear layer between 500 and 1000 m AGL. There were some errors in the Co2DiaWiL horizontal wind speed above 2 km AGL. The Co2DiaWiL u component ranged from 40 to 50 m s−1 at altitudes from 7.5 to 9 km, and there was good agreement between the Co2DiaWiL and radiosonde.

Fig. 12.
Fig. 12.

Vertical profile of (a) u- and (b) υ-component measurements of radiosonde and Co2DiaWiL. Squares indicate data from radiosondes, and black circles indicate data from Co2DiaWiL. Radiosonde launch time was 0545 JST 21 Feb 2010. Components u and υ were from a 20°-elevation Co2DiaWiL conical scan.

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

Figure 13 compares the u and υ components retrieved from a 20°-elevation conical scan of the Co2DiaWiL and those from radiosondes. The correlation coefficient was 0.97 for u components and 0.99 for υ components. These correlations were also found by using the linear least squares technique, which gave straight-line slopes of 1.01 for u components and 0.98 for υ components, and straight-line intercepts on the vertical (i.e., radiosonde) axis of 0.12 m s−1 and 0.09 m s−1 for the respective components. The slope of the linear regression line near one and a small intercept indicate that winds measured with Co2DiaWiL are accurate. The rms differences were 0.75 m s−1 and 0.80 m s−1 for the respective components. The mean absolute error between the Co2DiaWiL and radiosonde measurements was 0.57 m s−1 for u components and 0.61 m s−1 for υ components, and their differences in data during the period were mostly within ±1.5 m s−1 (never >3 m s−1). The differences in wind are not unreasonably large, in light of the great difference between wind measurement techniques. The Co2DiaWiL velocities were determined from the Doppler shift of light scattered by aerosols, whereas radiosonde velocities were determined from changes in the balloon range. The Co2DiaWiL sampling volume was centered over the scanner on average, whereas the balloon drifted with the wind. Thus, the two techniques measured different atmospheric regions that may have had dissimilar wind values.

Fig. 13.
Fig. 13.

Comparison of (a) u and (b) υ components from 0245 JST 19 Feb 2010 through 1345 JST 25 Feb 2010. Co2DiaWiL values are from 20°-elevation VAD data of Co2DiaWiL and are given on the x axis. The radiosonde measurements are given on the y axis. Legend as in Fig. 11.

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

Figure 14 compares between the u and υ components retrieved from a 70°-elevation conical scan of the Co2DiaWiL and those measured by radiosondes. The correlation coefficient was 0.99 for u components and 0.97 for υ components. The rms differences were 1.42 m s−1 and 1.20 m s−1 for the respective components. The larger rms differences compared with the 20°-elevation conical scan comparison can be attributed to the numerous and huge differences between the Co2DiaWiL and radiosonde measurements. The mean absolute error between the Co2DiaWiL and radiosonde measurements was 0.99 m s−1 for u components and 0.86 m s−1 for υ components. There were greater differences (>3 m s−1) where the observation altitude was higher than that for the 20°-elevation VAD data. There was large wind shear (i.e., strong wind) at those heights. As previously mentioned, the lidar signal was weaker under cloud-free conditions at upper levels, and the wind retrieval error increased. We therefore attributed the larger differences in our comparisons to probable large-scale horizontal separations of the radiosonde and low SNR of the Co2DiaWiL.

Fig. 14.
Fig. 14.

As in Fig. 13, but Co2DiaWiL values are from 70°-elevation VAD data of Co2DiaWiL. Squares represent data whose observation altitude was higher than that of 20°-elevation VAD data.

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

f. Pseudo-dual-Doppler measurements

Figure 15 plots the time series of 1-min-averaged u and υ components retrieved from pseudo-dual-Doppler measurements and those from the sonic anemometer, from 1914 to 2021 JST 22 June 2010. The correlation coefficients of these components were as high as 0.84 for the former and 0.83 for the latter ( = 66). A comparison revealed rms differences in the u and υ components of 0.23 and 0.29 m s−1, respectively, with mean absolute errors of 0.19 and 0.19 m s−1, respectively. The time sequence plot shows how well the two datasets tracked each other. The Co2DiaWiL data in Fig. 15 generally have lower values than those of the sonic anemometer under rapidly changing wind conditions from 1914 to 1934 JST. This was expected for the comparison, since the time resolution of the Co2DiaWiL (⅙ = 0.167 Hz) was coarser than that of the sonic anemometer (10 Hz).

Fig. 15.
Fig. 15.

Time series of 1-min-averaged (a) u and (b) υ components retrieved from pseudo-dual-Doppler lidar (black circles on thick line) and corresponding components of sonic anemometer (squares on thin line) from 1914 to 2021 JST 22 Jun 2010.

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

Figure 16 plots the time series of wideband SNRs with the laser beam oriented directly toward the sonic anemometer (SNR1 corresponding to ) and the laser beam reflected by the steering mirror (SNR2 corresponding to ) from 1914 to 2021 JST 22 June 2010. The SNR2 was 3–5 dB lower than the SNR1, because the laser beam was scattered at the steering mirror. From the relationship between the velocity standard deviation and wideband SNR shown in Fig. 6, is estimated at about 0.14 m s−1 and is estimated at about 0.17 m s−1. The standard deviations (, ) of (u, υ) retrieved from the pseudo-dual-Doppler measurements are related to and by Lhermitte and Miller (1970):
e9
where is a between-beam angle defined in Fig. 3, which is equal to 50.76°. Assuming , Eq. (9) gives 0.20 m s−1. Errors in retrieved u and υ components deduced from the configuration of the Co2DiaWiL, steering mirror, and sonic anemometer contribute somewhat to the overall error.
Fig. 16.
Fig. 16.

Time series of wideband SNRs with laser beam oriented directly toward sonic anemometer (thick line), and laser beam reflected by steering mirror (thin line) from 1914 to 2021 JST 22 Jun 2010.

Citation: Journal of Atmospheric and Oceanic Technology 30, 3; 10.1175/JTECH-D-12-00111.1

Drechsel et al. (2009) presented the results of a comparison between extracted wind profiles from dual-Doppler lidar measurements and those from radiosondes and a wind profiler. They retrieved the wind field over a typical scan period of 16 min using a Cressman weighting function with radii of 1000 m horizontally and 500 m vertically. They concluded the differences between individual profile grid points from dual-Doppler lidar measurements and radiosondes stemmed from different spatial and temporal resolutions. The horizontal wind velocities in this study were retrieved with the pseudo-dual-Doppler technique with a spatial resolution of about 76 m × 76 m × 8 cm and a temporal resolution of 1 min. To the best of our knowledge, this is the first time dual-Doppler lidar measurements have been directly compared with point-scale in situ measurements.

5. Summary and conclusions

A dataset acquired at the NICT headquarters site was analyzed to validate the capabilities of the Co2DiaWiL for wind measurements. Hard-target data revealed that bias in radial wind velocity measurements was very small (−0.0069 m s−1), using an average of 1 s that was composed of 30 laser pulses. Random error in radial wind velocity measurements was analyzed by assessing the random contribution to the time series of the vertical velocity. The Co2DiaWiL achieved a precision in velocity of 0.12 m s−1. Radial wind velocity was measured out to 20–25-km ranges, using the horizontally fixed beam mode and averaging times of 1 min. A comparison of radial wind velocity (averaged over 1 s) with those from a tower-mounted sonic anemometer demonstrated good agreement between them. This was despite the inherent difference between remote, volume-averaged, and point measurements. A comparison with radiosonde profiles allowed consistency between Co2DiaWiL and radiosonde measurements to be evaluated. The differences between winds measured by radiosonde and the Co2DiaWiL at a lower elevation angle (20° elevation) were not unreasonably large, in view of the great disparity between these wind measurement techniques. Larger differences (>3 m s−1) between the two at upper levels were likely caused by their large horizontal separations and the low SNR of the Co2DiaWiL.

The pseudo-dual-Doppler technique provided a rare opportunity for us to directly compare the dual-Doppler lidar measurements with point-scale in situ measurements. The main feature of the pseudo-dual-Doppler technique is the low cost of dual-Doppler lidar measurements. Since it is difficult to maintain laser beam alignment for the mirror when the mirror separation has to be so far, an automatic system to align the beam would be needed to introduce the pseudo-dual-Doppler technique into practical use. A potential application of the pseudo-dual-Doppler technique is in multipoint vertical velocity measurements, with a single-Doppler lidar and multiple steering mirrors. Doppler lidar permits the entire mixed layer to be monitored with vertical resolution, continuously over long periods. This makes detailed studies of the atmospheric boundary layer possible (Lothon et al. 2006; Hogan et al. 2009; Ansmann et al. 2010; Barlow et al. 2011; Oda et al. 2011; Lenschow et al. 2012). Further advances in boundary layer studies are expected using multipoint vertical velocity measurements.

Planned improvements to the technology would be the development of higher-pulse-repetition-rate lasers using a high-resolution A/D converter, which would provide better approximation of the CRLB for the precision of velocity. Another would be to develop a real-time processing system using graphics processing units and a visualization system, to detect the initiation of convection associated with torrential rainfall in urban areas.

Finally, to realize a spaceborne Doppler lidar based on our coherent lidar system, it is important to develop a highly reliable and space-qualified laser, a heterodyne detection system, and other electric components.

Acknowledgments

The authors thank the staff of the Hydrospheric Atmospheric Research Center of Nagoya University for providing the GSP radiosonde sounding system, the Vaisala DigiCORA II MW15. The authors also wish to thank Dr. Katsuhiro Nakagawa of NICT for his support in setting up the sonic anemometer.

REFERENCES

  • Ansmann, A., , Fruntke J. , , and Engelmann R. , 2010: Updraft and downdraft characterization with Doppler lidar: Cloud-free versus cumuli-topped mixed layer. Atmos. Chem. Phys., 10, 78457858.

    • Search Google Scholar
    • Export Citation
  • Baker, W. E., and Coauthors, 1995: Lidar-measured winds from space: An essential component for weather and climate prediction. Bull. Amer. Meteor. Soc., 76, 869888.

    • Search Google Scholar
    • Export Citation
  • Banta, R. M., , Pichugina Y. L. , , and Brewer W. A. , 2006: Turbulent velocity variance profiles in the stable boundary layer generated by a nocturnal low-level jet. J. Atmos. Sci., 63, 27002719.

    • Search Google Scholar
    • Export Citation
  • Barlow, J. F., , Dunbar T. M. , , Nemitz E. G. , , Wood C. R. , , Gallagher M. W. , , Davies F. , , O’Connor E. , , and Harrison R. M. , 2011: Boundary layer dynamics over London, UK, as observed using Doppler lidar during REPARTEE-II. Atmos. Chem. Phys., 11, 21112125.

    • Search Google Scholar
    • Export Citation
  • Blackadar, A. K., 1957: Boundary layer wind maximum and their significance for the growth of nocturnal inversions. Bull. Amer. Meteor. Soc., 38, 283290.

    • Search Google Scholar
    • Export Citation
  • Browning, R. A., , and Wexler R. , 1968: The determination of kinematic properties of a wind field using Doppler radar. J. Appl. Meteor., 7, 105113.

    • Search Google Scholar
    • Export Citation
  • Calhoun, R., , Heap R. , , Princevac M. , , Newsom R. , , Fernando H. , , and Ligon D. , 2006: Virtual towers using coherent Doppler lidar during the Joint Urban 2003 dispersion experiment. J. Appl. Meteor. Climatol., 45, 11161126.

    • Search Google Scholar
    • Export Citation
  • Chan, P. W., , and Lee Y. F. , 2012: Application of short-range lidar in wind shear alerting. J. Atmos. Oceanic Technol., 29, 207220.

  • Collier, C. G., and Coauthors, 2005: Dual-Doppler lidar measurements for improving dispersion models. Bull. Amer. Meteor. Soc., 86, 825838.

    • Search Google Scholar
    • Export Citation
  • Doyle, J. D., , Grubišić V. , , Brown W. O. J. , , De Wekker S. F. J. , , Dörnbrack A. , , Jiang Q. , , Mayor S. D. , , and Weissmann M. , 2009: Observations and numerical simulations of subrotor vortices during T-REX. J. Atmos. Sci., 66, 12291249.

    • Search Google Scholar
    • Export Citation
  • Drechsel, S., , Mayr G. J. , , Chong M. , , Weissmann M. , , Dörnbrack A. , , and Calhoun R. , 2009: Three-dimensional wind retrieval: Application of MUSCAT to dual-Doppler lidar. J. Atmos. Oceanic Technol., 26, 635646.

    • Search Google Scholar
    • Export Citation
  • Drechsel, S., , Mayr G. J. , , Chong M. , , and Chow F. K. , 2010: Volume scanning strategies for 3D wind retrieval from dual-Doppler lidar measurements. J. Atmos. Oceanic Technol., 27, 18811892.

    • Search Google Scholar
    • Export Citation
  • Drobinski, P., , and Foster R. C. , 2003: On the origin of near-surface streaks in the neutrally-stratified planetary boundary layer. Bound.-Layer Meteor., 108, 247256.

    • Search Google Scholar
    • Export Citation
  • Drobinski, P., , Carlotti P. , , Newsom R. K. , , Banta R. M. , , Foster R. C. , , and Redelsperger J. L. , 2004: The structure of the near-neutral surface layer. J. Atmos. Sci., 61, 699714.

    • Search Google Scholar
    • Export Citation
  • Drobinski, P., , Carlotti P. , , Redelsperger J. L. , , Banta R. M. , , Masson V. , , and Newsom R. K. , 2007: Numerical and experimental investigation of the neutral atmospheric surface layer. J. Atmos. Sci., 64, 137156.

    • Search Google Scholar
    • Export Citation
  • Etling, D., , and Brown R. A. , 1993: A review of large-eddy dynamics in the planetary boundary layer. Bound.-Layer Meteor., 65, 215248.

    • Search Google Scholar
    • Export Citation
  • Frehlich, R., 1999: Performance of maximum likelihood estimators of mean power and Doppler velocity with a priori knowledge of spectral width. J. Atmos. Oceanic Technol., 16, 17021709.

    • Search Google Scholar
    • Export Citation
  • Frehlich, R., 2001: Estimation of velocity error for Doppler lidar measurements. J. Atmos. Oceanic Technol., 18, 16281639.

  • Frehlich, R., , Hannon S. M. , , and Henderson S. W. , 1994: Performance of a 2-μm coherent Doppler lidar for wind measurements. J. Atmos. Oceanic Technol., 11, 15171528.

    • Search Google Scholar
    • Export Citation
  • Frehlich, R., , Hannon S. M. , , and Henderson S. W. , 1997: Coherent Doppler lidar measurements of winds in the weak signal regime. Appl. Opt., 36, 34913499.

    • Search Google Scholar
    • Export Citation
  • Fujiwara, C., , Yamashita K. , , Nakanishi M. , , and Fujiyoshi Y. , 2011: Dust devil–like vortices in an urban area detected by a 3D scanning Doppler lidar. J. Appl. Meteor. Climatol., 50, 534547.

    • Search Google Scholar
    • Export Citation
  • Goodall, C., 1990: A survey of smoothing techniques. Modern Methods of Data Analysis, J. Fox and J. S. Long, Eds., Sage Publications, 126–176.

  • Grund, C. J., , Banta R. M. , , George J. L. , , Howell J. N. , , Post M. J. , , Richter R. A. , , and Weickmann A. M. , 2001: High-resolution Doppler lidar for boundary layer and cloud research. J. Atmos. Oceanic Technol., 18, 376393.

    • Search Google Scholar
    • Export Citation
  • Henderson, S. W., , Yuen E. Y. , , and Fry E. S. , 1986: Fast resonance-detection technique for single-frequency operation of injection-seeded Nd:YAG lasers. Opt. Lett., 11, 715717.

    • Search Google Scholar
    • Export Citation
  • Henderson, S. W., , Hale C. P. , , Magee J. R. , , Kavaya M. J. , , and Huffaker A. V. , 1991: Eye-safe coherent laser radar system at 2.1 μm using Tm,Ho:YAG lasers. Opt. Lett., 16, 773775.

    • Search Google Scholar
    • Export Citation
  • Henderson, S. W., , Suni P. J. M. , , Hale C. P. , , Hannon S. M. , , Magee J. R. , , Bruns D. L. , , and Yuen E. H. , 1993: Coherent laser radar at 2 μm using solid-state lasers. IEEE Trans. Geosci. Remote Sens., 31, 415.

    • Search Google Scholar
    • Export Citation
  • Henderson, S. W., , Gatt P. , , Rees D. , , and Huffaker R. M. , 2005: Wind lidar. Laser Remote Sensing, T. Fujii and T. Fukuchi, Eds., CRC Press, 469–472.

  • Hogan, R. J., , Grant A. L. M. , , Illingworth A. J. , , Pearson G. N. , , and O’Connor E. J. , 2009: Vertical velocity variance and skewness in clear and cloud-topped boundary layers as revealed by Doppler lidar. Quart. J. Roy. Meteor. Soc., 135, 635643.

    • Search Google Scholar
    • Export Citation
  • Huffaker, R. M., , and Hardesty R. M. , 1996: Remote sensing of atmospheric wind velocities using solid-state and CO2 coherent laser systems. Proc. IEEE, 84, 181204.

    • Search Google Scholar
    • Export Citation
  • Huffaker, R. M., , Lawrence T. R. , , Post M. J. , , Priestley J. T. , , Hall F. F. Jr., , Richter R. A. , , and Keeler R. J. , 1984: Feasibility studies for a global wind measuring satellite system (Windsat): Analysis of simulated performance. Appl. Opt., 23, 25232536.

    • Search Google Scholar
    • Export Citation
  • Hunt, J. C. R., , and Morrison J. F. , 2000: Eddy structure in turbulent boundary layers. Eur. J. Mech., 19B, 673694.

  • Ishii, S., and Coauthors, 2010: Coherent 2 μm differential absorption and wind lidar with conductively cooled laser and two-axis scanning device. Appl. Opt., 49, 18091817.

    • Search Google Scholar
    • Export Citation
  • Ishii, S., and Coauthors, 2012: Partial CO2 column-averaged dry-air mixing ratio from measurements by coherent 2-μm differential absorption and wind lidar with laser frequency offset locking. J. Atmos. Oceanic Technol., 29, 11691181.

    • Search Google Scholar
    • Export Citation
  • Iwai, H., , Murayama Y. , , Ishii S. , , Mizutani K. , , Ohno Y. , , and Hashiguchi T. , 2011: Strong updraft at a sea-breeze front and associated vertical transport of near-surface dense aerosol observed by Doppler lidar and ceilometer. Bound.-Layer Meteor., 141, 117142.

    • Search Google Scholar
    • Export Citation
  • Kaimal, J. C., , Wyngaard J. C. , , Haugen D. A. , , Coté O. R. , , Izumi Y. , , Caughey S. J. , , and Readings C. J. , 1976: Turbulence structure in the convective boundary layer. J. Atmos. Sci., 33, 21522169.

    • Search Google Scholar
    • Export Citation
  • Käsler, Y., , Rahm S. , , Simmet R. , , and Kühn M. , 2010: Wake measurements of a multi-MW wind turbine with coherent long-range pulsed Doppler wind lidar. J. Atmos. Oceanic Technol., 27, 15291532.

    • Search Google Scholar
    • Export Citation
  • Kelley, N., , Jonkman B. J. , , Scott G. N. , , and Pichugina Y. L. , 2007: Comparing pulsed Doppler lidar with sodar and direct measurements for wind assessment. Proc. American Wind Energy Association 2007 Conf. and Exhibition, Los Angeles, CA, National Renewable Energy Laboratory, NREL/CP-50041792, 21 pp.

  • Khanna, S., , and Brasseur J. G. , 1998: Three-dimensional buoyancy- and shear-induced local structure of the atmospheric boundary layer. J. Atmos. Sci., 55, 710743.

    • Search Google Scholar
    • Export Citation
  • Koch, G. J., , Beyon J. Y. , , Barnes B. W. , , Petros M. , , Yu J. , , Amzajerdian F. , , Kavaya M. J. , , and Singh U. N. , 2007: High-energy 2μm Doppler lidar for wind measurements. Opt. Eng., 46, 116201, doi:10.1117/1.2802584.

    • Search Google Scholar
    • Export Citation
  • Konrad, T. G., 1970: The dynamics of the convective process in clear air as seen by radar. J. Atmos. Sci., 27, 11381147.

  • Köpp, F., , Schwiesow R. L. , , and Werner Ch. , 1984: Remote measurements of boundary-layer wind profiles using a CW Doppler lidar. J. Climate Appl. Meteor., 23, 148154.

    • Search Google Scholar
    • Export Citation
  • Köpp, F., , Rahm S. , , and Smalikho I. , 2004: Characterization of aircraft wake vortices by 2-μm pulsed Doppler lidar. J. Atmos. Oceanic Technol., 21, 194206.

    • Search Google Scholar
    • Export Citation
  • Lenschow, D. H., , Lothon M. , , Mayor S. D. , , Sullivan P. P. , , and Canut G. , 2012: A comparison of higher-order vertical velocity moments in the convective boundary layer from lidar with in situ measurements and large-eddy simulation. Bound.-Layer Meteor., 143, 107123.

    • Search Google Scholar
    • Export Citation
  • Levin, M. J., 1965: Power spectrum parameter estimation. IEEE Trans. Inf. Theory, 11, 100107.

  • Lhermitte, R. M., , and Miller L. J. , 1970: Doppler radar methodology for the observation of convective storms. Preprints, 14th Conf. Radar Meteorology, Tucson, AZ, Amer. Meteor. Soc., 133–138.

  • Lin, C.-L., , Moeng C.-H. , , Sullivan P. P. , , and McWilliams J. C. , 1997: The effect of surface roughness on flow structures in a neutrally stratified planetary boundary layer. Phys. Fluids, 9, 32353249.

    • Search Google Scholar
    • Export Citation
  • Lothon, M., , Lenschow D. H. , , and Mayor S. D. , 2006: Coherence and scale of vertical velocity in the convective boundary layer from a Doppler lidar. Bound.-Layer Meteor., 121, 521536.

    • Search Google Scholar
    • Export Citation
  • Mann, J., , Peña A. , , Bingöl F. , , Wagner R. , , and Courtney M. S. , 2010: Lidar scanning of momentum flux in and above the atmospheric surface layer. J. Atmos. Oceanic Technol., 27, 959976.

    • Search Google Scholar
    • Export Citation
  • Mayor, S. D., , Lenschow D. H. , , Schwiesow R. L. , , Mann J. , , Frush C. L. , , and Simon M. K. , 1997: Validation of NCAR 10.6-μm CO2 Doppler lidar radial velocity measurements and comparison with a 915-MHz profiler. J. Atmos. Oceanic Technol., 14, 11101126.

    • Search Google Scholar
    • Export Citation
  • Menzies, R. T., 1986: Doppler lidar atmospheric wind sensors: A comparative performance evaluation for global measurement applications from earth orbit. Appl. Opt., 25, 25462553.

    • Search Google Scholar
    • Export Citation
  • Menzies, R. T., , and Hardesty R. M. , 1989: Coherent Doppler lidar for measurements of wind fields. Proc. IEEE, 77, 449462.

  • Mizutani, K., , Itabe T. , , Ishii S. , , Aoki T. , , Asai K. , , Sato A. , , Fukuoka H. , , and Ishikawa T. , 2008: Conductive-cooled 2 micron laser for CO2and wind observations. Lidar Remote Sensing for Environmental Monitoring IX, U. N. Singh, K. Asai, and A. Jayaraman, Eds., International Society for Optical Engineering (SPIE Proceedings, Vol. 7153), doi:10.1117/12.804807.

  • Newsom, R. K., , Ligon D. , , Calhoun R. , , Heap R. , , Cregan E. , , and Princevac M. , 2005: Retrieval of microscale wind and temperature fields from single- and dual-Doppler lidar data. J. Appl. Meteor., 44, 13241345.

    • Search Google Scholar
    • Export Citation
  • Newsom, R. K., , Calhoun R. , , Ligon D. , , and Allwine J. , 2008: Linearly organized turbulence structures observed over a suburban area by dual-Doppler lidar. Bound.-Layer Meteor., 127, 111130.

    • Search Google Scholar
    • Export Citation
  • Oda, R., , Iwai H. , , Ishii S. , , Sekizawa S. , , Mizutani K. , , and Murayama Y. , 2011: Structure of turbulence in the urban atmospheric boundary layer detected in the Doppler lidar observation. J. Japan Soc. Civ. Eng., 67B, I_313I_318.

    • Search Google Scholar
    • Export Citation
  • Pichugina, Y. L., , Banta R. M. , , Brewer W. A. , , Sandberg S. P. , , and Hardesty R. M. , 2012: Doppler lidar–based wind-profile measurement system for offshore wind-energy and other marine boundary layer applications. J. Appl. Meteor. Climatol., 51, 327349.

    • Search Google Scholar
    • Export Citation
  • Post, M. J., , and Cupp R. E. , 1990: Optimizing a pulsed Doppler lidar. Appl. Opt., 29, 41454158.

  • Post, M. J., , Schwiesow R. L. , , Cupp R. E. , , Haugen D. A. , , and Newman J. T. , 1978: A comparison of anemometer- and lidar-sensed wind velocity data. J. Appl. Meteor., 17, 11791181.

    • Search Google Scholar
    • Export Citation
  • Rothermel, J., , Kessinger C. , , and Davis D. L. , 1985: Dual-Doppler lidar measurement of winds in the JAWS experiment. J. Atmos. Oceanic Technol., 2, 138147.

    • Search Google Scholar
    • Export Citation
  • Rye, B. J., , and Hardesty R. M. , 1993: Discrete spectral peak estimation in incoherent backscatter heterodyne lidar. I: Spectral accumulation and the Cramer-Rao lower bound. IEEE Trans. Geosci. Remote Sens., 31, 1627.

    • Search Google Scholar
    • Export Citation
  • Sawada, M., and Coauthors, 2012: Transient downslope winds under the influence of stationary lee waves from the Zao mountain range. J. Meteor. Soc. Japan, 90, 79100.

    • Search Google Scholar
    • Export Citation
  • Shun, C. M., , and Chan P. W. , 2008: Applications of an infrared Doppler lidar in detection of wind shear. J. Atmos. Oceanic Technol., 25, 637655.

    • Search Google Scholar
    • Export Citation
  • Stull, R. B., 1988: An Introduction to Boundary Layer Meteorology. Kluwer Academic, 666 pp.

  • Sugimoto, N., and Coauthors, 2008: Lidar network observations of tropospheric aerosols. Lidar Remote Sensing for Environmental Monitoring IX, U. N. Singh, K. Asai, and A. Jayaraman, Eds., International Society for Optical Engineering (SPIE Proceedings, Vol. 7153), doi:10.1117/12.806540.

  • Tsunematsu, N., , Iwai H. , , Ishii S. , , Murayama Y. , , Yasui M. , , and Mizutani K. , 2009: The formation of sharp multi-layered wind structure over Tokyo associated with sea-breeze circulation. SOLA, 5, 14.

    • Search Google Scholar
    • Export Citation
  • Werner, C., 2005: Doppler wind lidar. Lidar: Range-Resolved Optical Remote Sensing of the Atmosphere, C. Weitkamp, Ed., Series in Optical Sciences, Vol. 102, Springer, 339–342.

  • Westbrook, C. D., , and Illingworth A. J. , 2009: Testing the influence of small crystals on ice size spectra using Doppler lidar observations. Geophys. Res. Lett., 36, L12810, doi:10.1029/2009GL038186.

    • Search Google Scholar
    • Export Citation
  • Westbrook, C. D., , Illingworth A. J. , , O’Connor E. J. , , and Hogan R. J. , 2010: Doppler lidar measurements of oriented planar ice crystals falling from supercooled and glaciated layer clouds. Quart. J. Roy. Meteor. Soc., 136, 260276.

    • Search Google Scholar
    • Export Citation
  • Wilczak, J. M., , and Tillman J. E. , 1980: The three-dimensional structure of convection in the atmospheric surface layer. J. Atmos. Sci., 37, 24242443.

    • Search Google Scholar
    • Export Citation
  • Xia, Q., , Lin C.-L. , , Calhoun R. , , and Newsom R. K. , 2008: Retrieval of urban boundary layer structures from Doppler lidar data. Part I: Accuracy assessment. J. Atmos. Sci., 65, 320.

    • Search Google Scholar
    • Export Citation
  • Young, G. S., , Kristovich D. S. , , Hjelmfelt M. R. , , and Foster R. C. , 2002: Rolls, streets, waves and more: A review of quasi-two-dimensional structures in the atmospheric boundary layer. Bull. Amer. Meteor. Soc., 83, 9971001.

    • Search Google Scholar
    • Export Citation
Save