• Andersson, E. & , and Masutani M. 2010: Collaboration on Observing System Simulation Experiments (Joint OSSE). ECMWF Newsletter, No. 123, ECMWF, Reading, United Kingdom, 14–16.

  • Arnold, C. P., Jr., , and Dey C. H. , 1986: Observing-systems simulation experiments: Past, present, and future. Bull. Amer. Meteor. Soc., 67, 687695, doi:10.1175/1520-0477(1986)067<0687:OSSEPP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Atlas, R., 1997: Atmospheric observation and experiments to assess their usefulness in data assimilation. J. Meteor. Soc. Japan, 75, 111130.

    • Search Google Scholar
    • Export Citation
  • Baker, W. E., and Coauthors, 1995: Lidar-measured winds from space: A key component for weather and climate prediction. Bull. Amer. Meteor. Soc., 76, 869888, doi:10.1175/1520-0477(1995)076<0869:LMWFSA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Baker, W. E., and Coauthors, 2014: Lidar-measured wind profiles: The missing link in the global observing system. Bull. Amer. Meteor. Soc.,95, 543–564, doi:10.1175/BAMS-D-12-00164.1.

  • Chen, Y., , Han Y. , , and Weng F. , 2012: Comparison of two transmittance algorithms in the Community Radiative Transfer Model: Application to AVHRR. J. Geophys. Res., 117, D06206, doi:10.1029/2011JD016656.

    • Search Google Scholar
    • Export Citation
  • Emmitt, G. D., 1987: Error analysis for total wind vector computations using one component measurements from a space-based Doppler lidar. Proc. Fourth Conf. on Coherent Laser Radar: Technology and Applications, Aspen, CO, Optical Society of America, 217–220.

  • Frehlich, R. G., 2000: Simulation of coherent Doppler lidar performance for space-based platforms. J. Appl. Meteor., 39, 245262, doi:10.1175/1520-0450(2000)039<0245:SOCDLP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kleist, D. T., , Parrish D. F. , , Derber J. C. , , Treadon R. , , Wu W.-S. , , and Lord S. , 2009: Introduction of the GSI into the NCEP Global Data Assimilation System. Wea. Forecasting, 24, 16911705, doi:10.1175/2009WAF2222201.1.

    • Search Google Scholar
    • Export Citation
  • Liu, Q., , and Weng F. , 2006: Advanced doubling–adding method for radiative transfer in planetary atmosphere. J. Atmos. Sci., 63, 34593465, doi:10.1175/JAS3808.1.

    • Search Google Scholar
    • Export Citation
  • Marseille, G.-J., , Stoffelen A. , , and Barkmeijer J. , 2008: Impact assessment of prospective spaceborne Doppler wind lidar observation scenarios. Tellus,60A, 234–248, doi:10.1111/j.1600-0870.2007.00289.x.

    • Search Google Scholar
    • Export Citation
  • Masutani, M., and Coauthors, 2010: Observing system simulation experiments at the National Centers for Environmental Prediction. J. Geophys. Res., 115, D07101, doi:10.1029/2009JD012528.

    • Search Google Scholar
    • Export Citation
  • McKay, J. A., 1998: Modeling of direct detection Doppler wind lidar. I. The edge technique. Appl. Opt., 37, 64806486, doi:10.1364/AO.37.006480.

    • 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, doi:10.1109/5.24130.

    • Search Google Scholar
    • Export Citation
  • NRC, 2007: Earth Science and Applications from Space: National Imperatives for the Next Decade and Beyond. The National Academies Press, 456 pp.

  • Reale, O., , Terry J. , , Masutani M. , , Andersson E. , , Riishøjgaard L. P. , , and Jusem J. C. , 2007: Preliminary evaluation of the European Centre for Medium-Range Weather Forecasts’ (ECMWF) Nature Run over the tropical Atlantic and African monsoon region. Geophys. Res. Lett., 34, L22810, doi:10.1029/2007GL031640.

    • Search Google Scholar
    • Export Citation
  • Riishøjgaard, L. P., , Atlas R. , , and Emmitt G. D. , 2004: The impact of Doppler lidar wind observations on a single-level meteorological analysis. J. Appl. Meteor., 43, 810820, doi:10.1175/2091.1.

    • Search Google Scholar
    • Export Citation
  • Riishøjgaard, L. P., , Ma Z. , , Masutani M. , , Woollen J. S. , , Emmitt G. D. , , Wood S. A. , , and Greco S. , 2012: Observation system simulation experiments for a global wind observing sounder. Geophys. Res. Lett., 39, L17805, doi:10.1029/2012GL051814.

    • Search Google Scholar
    • Export Citation
  • Rohaly, G. D., , and Krishnamurti T. N. , 1993: An observing system simulation experiment for the Laser Atmospheric Wind Sounder (LAWS). J. Appl. Meteor., 32, 14531471, doi:10.1175/1520-0450(1993)032<1453:AOSSEF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Seablom, M. S., , Talabac S. J. , , Higgins G. J. , , and Womack B. T. , 2007: Simulation for the design of next-generation global Earth observing systems. Atmospheric and Environmental Remote Sensing Data Processing and Utilization III: Readiness for GEOSS, M. D. Goldberg et al., Eds., International Society for Optical Engineering (SPIE Proceedings, Vol. 6684), 668413, doi:10.1117/12.735746.

  • Stoffelen, A., and Coauthors, 2005: The atmospheric dynamics mission for global wind field measurement. Bull. Amer. Meteor. Soc., 86, 7387, doi:10.1175/BAMS-86-1-73.

    • Search Google Scholar
    • Export Citation
  • Stoffelen, A., , Marseille G. J. , , Bouttier F. , , Vasiljevic D. , , De Haan S. , , and Cardinali C. , 2006: ADM-Aeolus Doppler wind lidar Observing System Simulation Experiment. Quart. J. Roy. Meteor. Soc., 132, 19271947, doi:10.1256/qj.05.83.

    • Search Google Scholar
    • Export Citation
  • Tan, D. G. H., , and Andersson E. , 2005: Simulation of the yield and accuracy of wind profile measurements from the Atmospheric Dynamics Mission (ADM-Aeolus). Quart. J. Roy. Meteor. Soc., 131, 1737–1757, doi:10.1256/qj.04.02.

    • Search Google Scholar
    • Export Citation
  • Wood, S. A., Jr., , Emmitt G. D. , , and Greco S. , 2000: DLSM: A coherent and direct detection lidar simulation model for simulating space-based and aircraft-based lidar winds. Laser Radar Technology and Applications V, G. W. Kamerman et al., Eds., International Society for Optical Engineering (SPIE Proceedings, Vol. 4035), 2, doi:10.1117/12.397788.

  • Žagar, N., , Gustafsson N. , , and Källén E. , 2004: Dynamical response of equatorial waves in four-dimensional variational data assimilation. Tellus, 56A, 2946, doi:10.1111/j.1600-0870.2004.00036.x.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    The distribution of GWOS DWL measurements in the 6-h assimilation window at the analysis time 0000 UTC 21 Jul 2005 (from Riishøjgaard et al. 2012).

  • View in gallery

    Time series of the wind RMSE at (a)–(d) 200 and (e)–(h) 850 hPa for analyses verifying daily from 7 Jul through 15 Aug 2005 in the (a),(e) global, (b),(f) tropical, (c),(g) NH, and (d),(h) SH regions. Analyses for each run are initialized at 0000 UTC and verified against the nature run. CTRL (black), DWL1 (red), DWL2 (green), and DWL4 (blue).

  • View in gallery

    The RMSE comparison of tropical wind analyses over the complete cycling assimilation period. (a) Reference OSSE analysis from the CTRL: averaged RMSE of tropical wind for the period, as a function of pressure from 20 to 1000 hPa. (b)–(d) The analysis difference between the RMSEs of lidar (DWL1, DWL2, DWL4) and the CTRL runs, respectively. Green (red) shaded areas in (b)–(d) denote positive (negative) lidar DWL impact. The interval is 0.4 m s−1.

  • View in gallery

    As in Fig. 3, but for the Northern Hemisphere.

  • View in gallery

    As in Fig. 3, but for the Southern Hemisphere.

  • View in gallery

    As in Fig. 3, but for the day 3 wind forecast in the tropics.

  • View in gallery

    The impact of DWL wind measurements from various GWOS configurations on 200- and 850-hPa tropical wind forecasts measured by the RMSE, averaged over 40 cases. Error bars represent statistical significance at the 95% level.

  • View in gallery

    The RMSE comparison of tropical wind forecast at 120 h verifying daily at 0000 UTC 7 Jul–15 Aug 2005. (a) Reference OSSE forecast from the CTRL: averaged RMSE of tropical wind for the 120-h forecast period, as a function of pressure from 20 to 1000 hPa. (b)–(d) The forecast difference between the RMSEs of lidar (DWL1, DWL2, DWL4) and the CTRL runs, respectively. Green (red) shaded areas in (b)–(d) denote positive (negative) lidar DWL impact. The interval is 0.4 ms−1.

  • View in gallery

    As in Fig. 8, but for the Northern Hemisphere.

  • View in gallery

    As in Fig. 9, but for the Southern Hemisphere.

  • View in gallery

    A 700-hPa RMS forecast error comparison for temperature averaged over the period from 7 Jul to 15 Aug 2005 in the (a) NH and (b) SH.

  • View in gallery

    As in Fig. 8, but for the RMSE comparison of temperature in the SH. The interval is 0.04°C.

  • View in gallery

    The average 500-hPa geopotential height anomaly correlation scores by forecast time in the (a) NH and (b) SH. The error bars represent the significance of the difference between the lidar (DWL1, DWL2, DWL4) and the CTRL runs at the 95% confidence level.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 1 1 1
PDF Downloads 0 0 0

Impact of Different Satellite Wind Lidar Telescope Configurations on NCEP GFS Forecast Skill in Observing System Simulation Experiments

View More View Less
  • 1 Earth System Science Interdisciplinary Center, University of Maryland, College Park, and Joint Center for Satellite Data Assimilation, College Park, Maryland
  • 2 Joint Center for Satellite Data Assimilation, College Park, Maryland, and Joint Observing and Information Systems Department, World Meteorological Organization, Geneva, Switzerland
  • 3 Earth System Science Interdisciplinary Center, University of Maryland, College Park, and Joint Center for Satellite Data Assimilation, and NOAA/NWS/NCEP/EMC, College Park, Maryland
  • 4 NOAA/NWS/NCEP/EMC, College Park, and I.M. Systems Group, Rockville, Maryland
  • 5 Simpson Weather Associates, Charlottesville, Virginia
© Get Permissions
Full access

Abstract

The Global Wind Observing Sounder (GWOS) concept, which has been developed as a hypothetical space-based hybrid wind lidar system by NASA in response to the 2007 National Research Council (NRC) decadal survey, is expected to provide global wind profile observations with high vertical resolution, precision, and accuracy when realized. The assimilation of Doppler wind lidar (DWL) observations anticipated from the GWOS is being conducted as a series of observing system simulation experiments (OSSEs) at the Joint Center for Satellite Data Assimilation (JCSDA). A companion paper (Riishøjgaard et al.) describes the simulation of this lidar wind data and evaluates the impact on global numerical weather prediction (NWP) of the baseline GWOS using a four-telescope configuration to provide independent line-of-sight wind speeds, while this paper sets out to assess the NWP impact of GWOS equipped with alternative paired configurations of telescopes. The National Centers for Environmental Prediction (NCEP) Gridpoint Statistical Interpolation (GSI) analysis system and the Global Forecast System (GFS) were used, at a resolution of T382 with 64 layers, as the assimilation system and the forecast model, respectively, in these lidar OSSEs. A set of 45-day assimilation and forecast experiments from 2 July to 15 August 2005 was set up and executed.

In this OSSE study, a control simulation utilizing all of the data types assimilated in the operational GSI/GFS system was compared to three OSSE simulations that added lidar wind data from the different configurations of telescopes (one-, two-, and four-look configurations). First, the root-mean-square error (RMSE) of wind analysis is compared against the nature run. A significant reduction of the stratospheric RMSE of wind analyses is found for all latitudes when lidar wind profiles are used in the assimilation system. The forecast impacts of lidar data on the wind and mass forecasts are also presented. In addition, the anomaly correlations (AC) of geopotential height forecasts at 500 hPa were evaluated to compare the control and different GWOS telescope configuration experiments. The results show that the assimilation of lidar data from the GWOS (one, two, or four looks) can improve the NCEP GFS wind and mass field forecasts. The addition of the simulated lidar wind observations leads to a statistically significant increase in AC scores.

Denotes Open Access content.

Current affiliation: Riverside Technology, Inc., College Park, Maryland.

Corresponding author address: Zaizhong Ma, Joint Center for Satellite Data Assimilation, 5830 University Research Court, College Park, MD 20740-3818. E-mail: zaizhong.ma@noaa.gov

Abstract

The Global Wind Observing Sounder (GWOS) concept, which has been developed as a hypothetical space-based hybrid wind lidar system by NASA in response to the 2007 National Research Council (NRC) decadal survey, is expected to provide global wind profile observations with high vertical resolution, precision, and accuracy when realized. The assimilation of Doppler wind lidar (DWL) observations anticipated from the GWOS is being conducted as a series of observing system simulation experiments (OSSEs) at the Joint Center for Satellite Data Assimilation (JCSDA). A companion paper (Riishøjgaard et al.) describes the simulation of this lidar wind data and evaluates the impact on global numerical weather prediction (NWP) of the baseline GWOS using a four-telescope configuration to provide independent line-of-sight wind speeds, while this paper sets out to assess the NWP impact of GWOS equipped with alternative paired configurations of telescopes. The National Centers for Environmental Prediction (NCEP) Gridpoint Statistical Interpolation (GSI) analysis system and the Global Forecast System (GFS) were used, at a resolution of T382 with 64 layers, as the assimilation system and the forecast model, respectively, in these lidar OSSEs. A set of 45-day assimilation and forecast experiments from 2 July to 15 August 2005 was set up and executed.

In this OSSE study, a control simulation utilizing all of the data types assimilated in the operational GSI/GFS system was compared to three OSSE simulations that added lidar wind data from the different configurations of telescopes (one-, two-, and four-look configurations). First, the root-mean-square error (RMSE) of wind analysis is compared against the nature run. A significant reduction of the stratospheric RMSE of wind analyses is found for all latitudes when lidar wind profiles are used in the assimilation system. The forecast impacts of lidar data on the wind and mass forecasts are also presented. In addition, the anomaly correlations (AC) of geopotential height forecasts at 500 hPa were evaluated to compare the control and different GWOS telescope configuration experiments. The results show that the assimilation of lidar data from the GWOS (one, two, or four looks) can improve the NCEP GFS wind and mass field forecasts. The addition of the simulated lidar wind observations leads to a statistically significant increase in AC scores.

Denotes Open Access content.

Current affiliation: Riverside Technology, Inc., College Park, Maryland.

Corresponding author address: Zaizhong Ma, Joint Center for Satellite Data Assimilation, 5830 University Research Court, College Park, MD 20740-3818. E-mail: zaizhong.ma@noaa.gov

1. Introduction

Wind is one of the basic variables describing the state of the atmosphere together with temperature, pressure, and humidity. Global measurements of the wind field are essential for weather prediction, climate change studies, hurricane prediction, and many other meteorological studies. While the current measurements of atmospheric winds include many types of observations—from ground stations, buoys, ships, radiosondes, aircraft, ground-based wind profilers, radiometers, scatterometers, and cloud-tracking satellites—wind measurements are still inadequate in the present global observing system (GOS), and there is still no direct three-dimensional (3D) structure measurement of the wind field throughout the troposphere and lower stratosphere (approximately 0–30 km) on a global scale. A lack of wind observations over data-sparse areas [i.e., oceans, tropics, and the Southern Hemisphere (SH)] can result in rather nonuniform errors in numerical weather prediction (NWP) analyses and subsequent forecast failures (Marseille et al. 2008). Direct observations of wind profiles have been recognized as the most urgently needed observation type for climate studies as well as numerical weather prediction.

Space-based Doppler wind lidar (DWL) has been identified as the key technology necessary to meet the global wind profiling requirement (Emmitt 1987). Two measurement techniques—namely, coherent and direct detection—are available for the measuring of Mie and Rayleigh backscattering effects. The operating principle of the two measurement systems is different. The coherent heterodyne system measures winds based on aerosol backscatter (Frehlich 2000; Menzies and Hardesty 1989). However, the direct-detection system can measure winds from atmospheric molecules as well as aerosols. More detailed information on technology used in Doppler wind lidar is found in, for example, McKay (1998), Tan and Andersson (2005), Stoffelen et al. (2006), and Baker et al. (1995, 2014).

The first space-based DWL, called the Atmospheric Dynamics Mission Aeolus (ADM-Aeolus; Stoffelen et al. 2005), is scheduled to be launched into a polar orbit by the European Space Agency (ESA) in 2015. ADM-Aeolus with the direct-detection UV concept as the baseline is a demonstration mission measuring profiles of single horizontal line-of-sight (LOS) wind components (Marseille et al. 2008; Tan and Andersson 2005) with a priority on the quality rather than quantity of retrieved winds. Improved NWP analysis and forecast skill is expected from ADM-Aeolus.

The Global Wind Observing Sounder (GWOS) utilizes both coherent-detection and direct-detection pulsed DWL in a “hybrid” approach to vertically profile horizontal LOS wind vectors from space. Compared to the European ADM-Aeolus mission with one telescope, the GWOS is composed of four telescopes, two forward pointing and two aft pointing, each oriented nominally ±45° in both azimuth and elevation relative to the spacecraft’s velocity vectors. Therefore, much more lidar wind measurements from the GWOS can be expected for NWP than from ADM-Aeolus.

The benefit from the space-based Doppler wind observations needs to be accurately and fully assessed before the GWOS concept mission is selected to launch. The observing system simulation experiment (OSSE; Arnold and Dey 1986) is a well-established method of providing an objective and quantitative evaluation of future observing systems and instruments. Several OSSEs have investigated the impact of global winds using a space-based DWL (Rohaly and Krishnamurti 1993; Atlas 1997). These OSSEs used idealized representations of lidar data products. A more complete description of the GWOS mission is required to evaluate the impact of this specific concept, to perform comparisons with alternative measurements, and to improve the engineering designs and signal processing for future measurements from space.

The lidar OSSEs for the GWOS concept mission began at the National Centers for Environmental Prediction (NCEP) in 2006. Since the OSSEs mimic the procedures used to analyze global observations for specifying the state of the atmosphere, it is very crucial for validating the realism of the OSSE. The validation has been performed by comparing the statistics of results of assimilating the simulated observations for 6 weeks compared with the corresponding statistics obtained from assimilating real observations (Riishøjgaard et al. 2012). The impact of the four-telescope lidar wind data on NWP has already been investigated (Riishøjgaard et al. 2012), and the results show the lidar wind data from GWOS have a significant impact on the anomaly correlations of 500-hPa geopotential heights [an improvement of about 1.5% in the Northern Hemisphere (NH) and 1.9% in the SH in the day 5 forecasts].

Although the four-telescope lidar wind impact from GWOS on the NCEP OSSE system has been recently assessed, the performance from the separate telescopes on the GWOS satellite has not been evaluated yet. Hence, as the follow-on OSSE study, we first design three lidar experiments—a one-look study with a single telescope (azimuth angle: 45°), a two-look study with a pair of telescopes (azimuth angles: 45° and 315°), and a four-look study with two pairs of telescopes (azimuth angles from one pair is 45° and 315°, the other pair is 135° and 225°)—in this paper and then investigate the potential impact of these three experiments on the analyses and forecasts by comparing them against the control, which assimilates all the same observational data as the NCEP operational datasets in 2005.

The paper is arranged as follows. The model details are presented in section 2, which also gives the overview of the GWOS concept mission and describes the nature run from the European Centre for Medium-Range Weather Forecasts (ECMWF) global model and the lidar data assimilation methodologies in the Gridpoint Statistical Interpolation (GSI) analysis system. Section 3 describes the experimental design for the different pairs of telescopes in the GWOS concept mission, section 4 shows the preliminary comparison results to assess the GWOS lidar data impact on the analysis and forecast, and section 5 gives the conclusions and an outlook for further research related to the GWOS OSSE study.

2. Description of GWOS mission and OSSE system

a. The GWOS concept mission

The GWOS concept is 1 of the 15 priority missions recommended for NASA over the next decade in response to the 2007 National Research Council decadal survey for Earth science (NRC 2007). As the concept mission serving as the user case for our model-driven sensor web operations concept, GWOS would improve our understanding and prediction of atmospheric dynamics and global atmospheric transport. It would achieve these objectives by making space-based direct lidar measurements of vertical profiles of the horizontal wind field to provide a complete global 3D picture of the dynamical state, clouds permitting. A simplified description of the GWOS concept mission will be given here.

The recommended technological approach for the GWOS concept is a hybrid DWL that utilizes the complementary capabilities of two Doppler lidar technologies—a coherent Doppler lidar sensing winds from the aerosol backscattered laser signal at a wavelength of 2 μm and a direct-detection Doppler lidar sensing winds from the molecular backscattered laser signal at 355 nm (Seablom et al. 2007). The measurement principle is based on the fact that the Doppler shift of the return from an emitted pulse of monochromatic electromagnetic radiation can be translated into information about the radial velocity of the air at the origin of the return. Only the velocity component aligned with the lidar beam is measured and the observations are, therefore, often referred to as LOS wind measurements (Riishøjgaard et al. 2004). The GWOS can provide a direct wind measurement with an accurate height assignment with a global coverage.

b. The nature run and observational data simulation

Simulation experiments use a model-generated proxy for the real atmosphere, commonly called the “nature run.” The nature run used in this lidar OSSE study is a 13-month forecast (from 1 May 2005 to 1 June 2006) using the ECMWF Integrated Forecasting System (IFS) with a resolution of T511L91 (Reale et al. 2007; Andersson and Masutani 2010). The horizontal resolution is about 40 km and the output is saved every 3 h. The nature run was evaluated and very realistic hurricanes and midlatitude cyclone statistics were reported (Reale et al. 2007). Its cloud distribution is much more realistic than in the previous nature run (Masutani et al. 2010). Statistics of the midlatitude jet were also studied and were found to be realistic.

Since exactly the same OSSE and NWP system as Riishøjgaard et al. (2012) have been used to provide further investigations of additional telescope configurations, here we briefly summarize the process of simulating observations introduced by Riishøjgaard et al. (2012). The conventional data in this lidar OSSE study, which include wind, temperature, surface pressure, and humidity, were simulated according to a 2005–06 NCEP operational usage at the appropriate time and location. Radiance data were generated by processing ECWMF nature run profiles of temperature and humidity with the Community Radiative Transfer Model (CRTM, version 1.2.2) developed by NOAA’s Joint Center for Satellite Data Assimilation (Liu and Weng 2006; Chen et al. 2012). The simulated satellite radiances (clear sky) from the Atmospheric Infrared Sounder (AIRS; Aqua), AMSU-A (Aqua; NOAA-15, -16, -18), AMSU-B (NOAA-15, -16, -17), HIRS2 (NOAA-14), HIRS3 (NOAA-15, -16, -17), HIRS4 (NOAA-18), MSU (NOAA-14), the Microwave Humidity Sounder (MHS; NOAA-18), and the GOES sounder (GOES-10, -12) were assimilated with the GSI system in the experiments. More detailed of the data simulation also can be found on the Joint OSSE home page (http://www.emc.ncep.noaa.gov/research/JointOSSEs/).

The Doppler Lidar Simulation Model (DLSM) is an evolution of existing Doppler lidar simulation models (Wood et al. 2000) that are currently used to provide spaced-based Doppler lidar wind simulations for OSSEs (Riishøjgaard et al. 2012). The distribution of simulated DWL wind from GWOS is shown in Fig. 1. Here a random 6-h period—that is, ±3 h at the central time 0000 UTC 21 July 2005— is picked up because the number of lidar wind observations per 6-h cycle is almost identical in the whole experiment period.

Fig. 1.
Fig. 1.

The distribution of GWOS DWL measurements in the 6-h assimilation window at the analysis time 0000 UTC 21 Jul 2005 (from Riishøjgaard et al. 2012).

Citation: Journal of Atmospheric and Oceanic Technology 32, 3; 10.1175/JTECH-D-14-00057.1

The synthetic lidar observations and corresponding errors produced by the DLSM are designed to be realistic and faithful to real-world performance. A consequence is that the observations have a range of accuracies as a function of atmospheric conditions, with some observations containing gross errors unsuitable for assimilation. The observation error standard deviation for the GWOS DWL data varies from 0.2 to more than 10 m s−1. But the assumed observation errors’ standard deviation in this OSSE study was constrained to fall within the range from 1.8 to 10 m s−1 to avoid overfitting the wind observations.

Furthermore, it is important to mention that no random or systematic observation errors were added to the simulated data with the exception of the DWL observations from the GWOS. As explained by Riishøjgaard et al. (2012), the fact that the simulated observations were obtained by sampling a free-running forecast produced by a model (ECWMF IFS) that was different from the assimilating model [global data assimilation system (GDAS)] and running at a higher horizontal resolution does add an element of “error” to the observations, since they would be extremely unlikely to perfectly represent an achievable state of the assimilating model. Therefore, the same strategies for the observational errors discussed in Riishøjgaard et al. (2012) are used in this paper.

c. Lidar OSSE system

In these GWOS lidar observing system simulation experiments, the December 2009 [fiscal year 2010 (FY10)] version of the NCEP GSI/GFS global data assimilation system (Kleist et al. 2009) was used. The experiment setting is consistent with the operational GSI/GFS system at NCEP, except that the model resolution of T382L64 has been used rather than the current (April 2012) operational resolution of T574. A short-term forecast (6 h) is run to obtain a first guess for the data assimilation, which uses a ± 3-h data cutoff window, and the analyses and forecasts are centered at 0000, 0600, 1200, and 1800 UTC.

The lidar wind operator has been developed and tested to assimilate the LOS lidar measurements within the GSI data assimilation system. The observation operator for horizontal line-of-sight winds is relatively simple, consisting of an interpolation of the horizontal wind component of the background field to the observation time and location, followed by a projection on the line of sight of the lidar. The analysis is obtained by minimizing the scalar cost function, defined as
e1
where the vector represents the background or prior estimate of the control vector x, and (analysis state) when minimized; is the background error covariance matrix; is the observational and representativeness error covariance matrix; the vector y contains the available observations, for example, the lidar wind data in this OSSE; and is the observational operator that transforms from the model state to the observation space as described above.

A key point about this lidar OSSE system is that the nature run is performed with the ECMWF Integrated Forecasting System, while the lidar impact experiments are carried out with the NCEP GSI/GFS system. So, the problem of “identical twins,” where the same analysis/forecast system is used, has been avoided in our OSSEs.

3. Experimental design

To assess the impact of the simulated lidar wind data from the different configurations of telescopes in the GWOS system, four experiments with/without lidar data are given in Table 1. In the OSSEs, the control experiment (CTRL) is performed to assimilate the simulated data for all of the observations with a real-data counterpart in the NCEP operational data stream during the experimental period from 1200 UTC 1 July to 0000 UTC 15 August 2005. A set of three DWL assessment experiments, in which the simulated GWOS observations were added to the set of observations used for the CTRL, are also carried out in the same period.

Table 1.

Summary of GWOS lidar wind OSSE configurations.

Table 1.

Two subsets for each experiment are conducted: first, a cycling assimilation experiment for the experiment period, in which the GSI assimilation system is used for the analysis component and the GFS forecast model is used for the 6-h forecast component; and second, one forecast per day (only from the corresponding analysis at 0000 UTC with the GFS forecast model) being run out to 120 h due to the limited computer resource. Note that a cycling data assimilation was run, extending over a spin-up period to allow the assimilation system and forecast model to adjust to the new lidar wind data from 1 to 6 July, followed by an experimental period from 7 July through 15 August. Therefore, a total of 40 cases were taken to verify against the nature run in this OSSE, that is, a wind analysis and 120-h forecasts from 7 July to 15 August.

In Table 2, the averaged values for lidar data count in the period from 2 July to 15 August are calculated before/after the quality control (QC) procedure of the GSI data assimilation system. The lidar measurements will be increased with more telescopes—that is, about 8000 lidar measurements are assimilated per cycle in DWL1—while this number in DWL2 is doubled. In addition, note that more than 5% of the lidar measurements are rejected by the quality control procedure in the GSI system.

Table 2.

A comparison of the lidar measurement counts from three DWL data assimilation experiments before/after quality control in the GSI system. The average number of lidar observations per assimilation cycle is calculated in the period from 2 Jul to 15 Aug 2005.

Table 2.

4. Preliminary results

This section presents the results of the lidar wind OSSEs conducted for 40 cases selected from each experiment. The impact of the different sets of lidar wind data in three experiments (DWL1, DWL2, and DWL4) is measured by their capability to resolve the analysis corrections and the subsequent forecast improvement.

The primary fields used for the verification are the tropical winds and the extratropical 500-hPa geopotential heights. Several objective statistical measures to verify forecast quality are commonly used, that is, the root-mean-square error (RMSE) and the anomaly correlation (AC) of forecasts are against verifying analyses. In these assessments, both analyses and forecasts are verified against the nature run for each experiment, using the NCEP operational verification package. All datasets were reduced to a 2.5° × 2.5° horizontal resolution before calculating the RMSE and anomaly correlations with this package. The extratropical verification is done for the latitudinal bands of 20°–80° in each hemisphere, while the tropical verification is done in the band from 20°N to 20°S.

a. Comparison of the analysis impact

The comparison of wind analyses with respect to the nature run is investigated first. A time series of RMSE from wind analysis is displayed in Fig. 2 for different pressure levels and different regions of the globe; that is, Fig. 2a presents a comparison of wind RMSE for four OSSEs (CTRL, DWL1, DWL2, and DWL4) at 200 hPa over the globe. Compared with the CTRL, a positive impact from DWL data on the analysis is clearly seen at 200 hPa (four panels on the left) and 850 hPa (four panels on the right).

Fig. 2.
Fig. 2.

Time series of the wind RMSE at (a)–(d) 200 and (e)–(h) 850 hPa for analyses verifying daily from 7 Jul through 15 Aug 2005 in the (a),(e) global, (b),(f) tropical, (c),(g) NH, and (d),(h) SH regions. Analyses for each run are initialized at 0000 UTC and verified against the nature run. CTRL (black), DWL1 (red), DWL2 (green), and DWL4 (blue).

Citation: Journal of Atmospheric and Oceanic Technology 32, 3; 10.1175/JTECH-D-14-00057.1

Table 3 summarized the averaged analysis impacts from Fig. 2 for four verification areas: globe, tropics, Northern Hemisphere, and Southern Hemisphere. From Table 3 it is concluded that DWL1 improves the wind analysis by 0.37 m s−1 as compared to the CTRL when averaged over the globe. DWL2 achieves a 0.50 m s−1 wind analysis improvement, while DWL4 with 0.68 m s−1 shows the best DWL impact from a four-look GWOS. The performance of DWL2 is better than DWL1 over the globe, but note this scenario does not double the impact of DWL1, indicating some redundancy in the DWL observations. Furthermore, not surprisingly, a larger impact from lidar data on the wind analysis is clearly seen in the tropics (Figs. 2b,f) than in the Northern (or Southern) Hemisphere (Figs. 2c,d,g,h), because of the reduced coverage of other observations over the ocean in the tropics and because the mass data do not provide wind information in the tropics.

Table 3.

The time-averaged RMSE for the wind analysis shown in Fig. 2. The boldface values show that GWOS lidar wind observations improved the analysis mostly in the tropics for both 200 and 850 hPa.

Table 3.

Figure 3 shows the time–height distribution of RMSE from the wind analysis against the nature run for the experiment period, as a function of pressures from 20 to 1000 hPa in the tropics. The averaged tropical wind RMSE from the CTRL is displayed in Fig. 3a. It is noted that the large tropical wind RMSE value appears at the high altitudes between 100 and 200 hPa. The differences between the lidar runs (DWL1, DWL2, and DWL4) and the CTRL are calculated and shown in Figs. 3b–d. It is clear that the most pressure levels are covered by green, which means the wind RMSE from the lidar run is smaller than that from the CTRL. In other words, DWL wind can produce the significant positive impact on the wind analysis for all three lidar scenarios in the tropics.

Fig. 3.
Fig. 3.

The RMSE comparison of tropical wind analyses over the complete cycling assimilation period. (a) Reference OSSE analysis from the CTRL: averaged RMSE of tropical wind for the period, as a function of pressure from 20 to 1000 hPa. (b)–(d) The analysis difference between the RMSEs of lidar (DWL1, DWL2, DWL4) and the CTRL runs, respectively. Green (red) shaded areas in (b)–(d) denote positive (negative) lidar DWL impact. The interval is 0.4 m s−1.

Citation: Journal of Atmospheric and Oceanic Technology 32, 3; 10.1175/JTECH-D-14-00057.1

The wind analysis impact of three DWL wind configurations for the Northern Hemisphere and Southern Hemisphere regions are also investigated in Figs. 4 and 5, respectively. On average, although the impact of DWL observations to the wind analysis is positive for both areas, it is clear that the impact in the Northern or Southern Hemisphere is substantially smaller than in the tropics, especially for the Northern Hemisphere.

Fig. 4.
Fig. 4.

As in Fig. 3, but for the Northern Hemisphere.

Citation: Journal of Atmospheric and Oceanic Technology 32, 3; 10.1175/JTECH-D-14-00057.1

Fig. 5.
Fig. 5.

As in Fig. 3, but for the Southern Hemisphere.

Citation: Journal of Atmospheric and Oceanic Technology 32, 3; 10.1175/JTECH-D-14-00057.1

b. Comparison of the forecast impact

To compare the impact of GWOS lidar data from the different configurations of telescopes on the 120-h forecasts, four experiments (CTRL, DWL1, DWL2, and DWL4) have been performed, initializing with the 0000 UTC analyses from 2 July to 15 August 2005. Differences between the forecasts from the four parallel runs are only due to the impact of the simulated GWOS lidar data. Hence, their respective quality difference is a measure of lidar observation impact. Following the discussion of wind analysis impact in section 4a, the forecast impact on wind and the skill scores by assimilating DWL observations is presented below.

1) Wind

As an example of GWOS lidar impact on the forecast, Fig. 6 presents the time series of wind RMSE at day 3 against the nature run for pressures from 20 to 1000 hPa in the tropics. A zonal belt of the biggest wind RMSE from the CTRL appears between 100 and 200 hPa in Fig. 6a, and the central value of the wind is more than 10 m s−1. The differences between the three lidar runs and the CTRL are shown in Figs. 6b–d. Compared to the control run, it is clearly seen that all three lidar runs can decrease the wind forecast RMSE after the DWL observations are assimilated, and the large benefit from the lidar observations still appears at the higher altitudes. During these three lidar runs, as expected, DWL1 with a one-look configuration produces the smallest positive impact on the day 3 forecast, while the DWL4 with four-look configuration still can produce the larger impact because it has much better analysis than the two other DWL configurations as schematically displayed in Fig. 3.

Fig. 6.
Fig. 6.

As in Fig. 3, but for the day 3 wind forecast in the tropics.

Citation: Journal of Atmospheric and Oceanic Technology 32, 3; 10.1175/JTECH-D-14-00057.1

To illustrate the wind lidar impact at different forecast ranges, tropical wind RMSEs of the forecasts from these four runs versus the nature run at 200 and 850 hPa are shown in Fig. 7. All forecasts are verified using the nature run, and the bottom part of Fig. 7 shows the differences in skill with respect to the CTRL. The positive impact from simulated DWL wind observations is initially large; however, the effect tends to decrease rapidly over time at both levels. This behavior is typical for the tropics, and it illustrates the challenge of using observations in a dynamically consistent way (Žagar et al. 2004). Boxes in the lower portions of Fig. 7 mark the 95% confidence interval; the results outside of these boxes are considered statistically significant. It is noted that the DWL1 results are not statistically significant after day 2 at the 850-hPa level, whereas the DWL4 with a four-look configuration, however, still shows significant improvement on the day 5 forecasts.

Fig. 7.
Fig. 7.

The impact of DWL wind measurements from various GWOS configurations on 200- and 850-hPa tropical wind forecasts measured by the RMSE, averaged over 40 cases. Error bars represent statistical significance at the 95% level.

Citation: Journal of Atmospheric and Oceanic Technology 32, 3; 10.1175/JTECH-D-14-00057.1

Besides 200 and 850 hPa, the forecast performance comparison of lidar impact for all pressure levels is also investigated in Fig. 8. The green shows everywhere within the 120-h forecast and the largest positive impact from lidar is located around 150 hPa in Figs. 8b–d. Again, DWL4 with GWOS four-look configuration provides the best forecast in the tropics as expected from the largest DWL coverage and the resulting best analyses in all three DWL configurations as shown in Fig. 3.

Fig. 8.
Fig. 8.

The RMSE comparison of tropical wind forecast at 120 h verifying daily at 0000 UTC 7 Jul–15 Aug 2005. (a) Reference OSSE forecast from the CTRL: averaged RMSE of tropical wind for the 120-h forecast period, as a function of pressure from 20 to 1000 hPa. (b)–(d) The forecast difference between the RMSEs of lidar (DWL1, DWL2, DWL4) and the CTRL runs, respectively. Green (red) shaded areas in (b)–(d) denote positive (negative) lidar DWL impact. The interval is 0.4 ms−1.

Citation: Journal of Atmospheric and Oceanic Technology 32, 3; 10.1175/JTECH-D-14-00057.1

The impact of DWL data on wind forecasts in the NH and SH are shown in Figs. 9 and 10, respectively. One can see that, in general, the impact of DWL data on wind forecasts in the NH and SH is similar to that in the tropics, that is, including DWL data reduced the wind RMSE in the troposphere and had a neutral impact on the stratosphere.

Fig. 9.
Fig. 9.

As in Fig. 8, but for the Northern Hemisphere.

Citation: Journal of Atmospheric and Oceanic Technology 32, 3; 10.1175/JTECH-D-14-00057.1

Fig. 10.
Fig. 10.

As in Fig. 9, but for the Southern Hemisphere.

Citation: Journal of Atmospheric and Oceanic Technology 32, 3; 10.1175/JTECH-D-14-00057.1

2) Temperature

The lidar wind observations from the GWOS concept do not contain direct information on the mass field in the GSI data assimilation system. The increments for the mass components due to the lidar wind data, if any, should come through the dynamics of the model and the balance constraints in the GSI. In general, we found neutral or slightly positive impacts for the temperature when the GWOS lidar data were included in the GSI assimilation system. This can be seen in Fig. 11, where the 700-hPa RMSE for temperature with forecast time is displayed for these four experiments in both the NH and SH. The verification is done against the nature run discussed above. The forecast impact of the GWOS lidar after 72 h was quite clear for the period being tested.

Fig. 11.
Fig. 11.

A 700-hPa RMS forecast error comparison for temperature averaged over the period from 7 Jul to 15 Aug 2005 in the (a) NH and (b) SH.

Citation: Journal of Atmospheric and Oceanic Technology 32, 3; 10.1175/JTECH-D-14-00057.1

To compare fully with the temperature forecasts, RMSE from the CTRL and quality differences with the other three DWL runs for all pressure levels are presented in Fig. 12 for the SH. It is clear to see that the improvement from DWL1 (in Fig. 12b) is very small and much closer to neutral, while green dominates at the most pressure levels. Furthermore, we also note that the assimilation of more GWOS lidar wind observations in DWL2 and DWL4 (shown in Figs. 12c,d, respectively) can produce a larger positive impact than that in DWL1, but no significant improvement in the accuracy of the temperature forecasts is seen.

Fig. 12.
Fig. 12.

As in Fig. 8, but for the RMSE comparison of temperature in the SH. The interval is 0.04°C.

Citation: Journal of Atmospheric and Oceanic Technology 32, 3; 10.1175/JTECH-D-14-00057.1

3) Geopotential height anomaly correlation

Table 4 summarizes the forecast impacts from the 500-hPa geopotential height anomaly correlation for the averaged day 5 AC scores in the NH and SH based on the selected 40 cases. The sequence in the NH is CTRL (0.849) < DWL1 (0.852) < DWL2 (0.856) < DWL4 (0.864). Similar results can be found in the SH as well. In particular, the addition of the DWL four-look configuration leads to a statistically significant increase in the AC score on day 5 of approximately 1.5% and 1.9% in the Northern and Southern Hemispheres, respectively. In the Southern Hemisphere, for example, the AC score is approximately 0.820 and 0.839 for the CTRL and DWL4 experiments, respectively.

Table 4.

The averaged day 5 AC scores and the percentage of the improved AC scores from three DWL experiments.

Table 4.

Figure 13 compares the skills of the 500-hPa height forecasts measured by the anomaly correlation coefficient (the AC score) for all four experiments in the Northern Hemisphere (Fig. 13a) and the Southern Hemisphere (Fig. 13b). All forecasts are verified using the nature run. The figure shows that the addition of the simulated lidar wind observations leads to a statistically significant increase in AC score within the 120-h forecast period, and a more positive impact can be expected with more DWL lidar wind being assimilated. The bottom part of Fig. 13 shows the differences in skill with respect to the CTRL. Differences that exceed the error bars for the respective color are statistically significant at the 95% level. It is clear that DWL4 results for the 120-h forecasts are all located outside the boxes for both the NH and the SH, meaning the statistically significant improvement has been produced by the GWOS four-look configuration.

Fig. 13.
Fig. 13.

The average 500-hPa geopotential height anomaly correlation scores by forecast time in the (a) NH and (b) SH. The error bars represent the significance of the difference between the lidar (DWL1, DWL2, DWL4) and the CTRL runs at the 95% confidence level.

Citation: Journal of Atmospheric and Oceanic Technology 32, 3; 10.1175/JTECH-D-14-00057.1

It is also noted that the day 5 forecast in the DWL4 with four-look configuration does not show any improvement compared to the DWL2 with two-look in the SH, while the improvement in the NH is significant. It has been known that the GFS has much worse forecast skills in the SH than in the NH for all seasons of the year (F.Yang 2014, personal communication; the review of the GFS forecast skill in 2013 can be found at http://www.emc.ncep.noaa.gov/gmb/wx24fy/doc/GFS.performance.review.2013.pdf). It is likely the GFS has larger systematic model biases in the SH. The impact of DWL wind in the SH is less identifiable than in the NH due to large error growth. More forecast cases are also required to obtain a robust response in the SH.

5. Summary and conclusions

Observing system simulation experiments are important for understanding the impact of new data (i.e., the GWOS space-based satellite lidar wind) on NWP forecasts. The lidar impacts from the different GWOS lidar wind configurations have been investigated through a set of the observing system simulation experiments performed with a 2009 version of the NCEP GSI/GFS operational system. Four separate OSSEs (CTRL, DWL1, DWL2, and DWL4) have been performed to compare forecast skill with and without GWOS wind lidar data at a horizontal resolution of T382. Analyses and forecasts from the four OSSE runs were verified against the nature run using the NCEP operational verification package.

The preliminary results indicate that the assimilation of the DWL data from the different configurations of GWOS (either one-, two-, or four-look configuration) can substantially improve the NCEP GFS analysis and forecasts. More potential benefits for wind can be found in the tropics from DWL data than the other areas in the globe. And larger improvement for a high AC score can be found in the SH, although the positive impact is also found in the NH. Overall, the GWOS lidar observations with the four-look configuration (DWL4) have the largest positive impact on both analyses and forecasts in the three sets of DWL assessment experiments.

Since the DWL data coverage from GWOS would be the same for ADM-Aeolus with a single data track and a single perspective (Baker et al. 2014), DWL1 with the one-look configuration in this OSSE study can be considered to some extent as the European Space Agency (ESA)’s ADM-Aeolus DWL. The addition of the simulated lidar wind observations leads to a statistical increase in AC scores on day 5 of approximately 0.3% (from DWL1) and 1.5% (from DWL4) in the NH, and an increase of approximately 1.0% (from DWL1) and 1.9% (from DWL4) in the SH. Based on the current spaceborne DWL observation scenario, therefore, more potential impact from the GWOS concept will be expected on NWP analysis quality than from ADM-Aeolus.

This OSSE study demonstrates the potential benefits to NWP from DWL wind observations, through comparing three different DWL observation scenarios. Future work in this area is envisaged to evolve as follows: 1) extending the simulations over the remainder of the nature run in the 2005 hurricane season, 2) conducting separate assessments of the impacts of direct detection and coherent detection, and 3) investigating the impact of DWL winds using the NCEP GSI hybrid data assimilation system.

Acknowledgments

Support for this work was provided by NASA (R. Kakar) through ROSES (Grant NNX08AQ44G). Computational resources for the experiments were made available by NOAA/NCEP. The T511NR was produced by Dr. Erik Andersson of ECMWF. The initial simulation of GOES radiance data was conducted by Tong Zhu of NESDIS. We acknowledge Fanglin Yang and James G. Yoe of NECP/EMC for their thoughtful comments.

REFERENCES

  • Andersson, E. & , and Masutani M. 2010: Collaboration on Observing System Simulation Experiments (Joint OSSE). ECMWF Newsletter, No. 123, ECMWF, Reading, United Kingdom, 14–16.

  • Arnold, C. P., Jr., , and Dey C. H. , 1986: Observing-systems simulation experiments: Past, present, and future. Bull. Amer. Meteor. Soc., 67, 687695, doi:10.1175/1520-0477(1986)067<0687:OSSEPP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Atlas, R., 1997: Atmospheric observation and experiments to assess their usefulness in data assimilation. J. Meteor. Soc. Japan, 75, 111130.

    • Search Google Scholar
    • Export Citation
  • Baker, W. E., and Coauthors, 1995: Lidar-measured winds from space: A key component for weather and climate prediction. Bull. Amer. Meteor. Soc., 76, 869888, doi:10.1175/1520-0477(1995)076<0869:LMWFSA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Baker, W. E., and Coauthors, 2014: Lidar-measured wind profiles: The missing link in the global observing system. Bull. Amer. Meteor. Soc.,95, 543–564, doi:10.1175/BAMS-D-12-00164.1.

  • Chen, Y., , Han Y. , , and Weng F. , 2012: Comparison of two transmittance algorithms in the Community Radiative Transfer Model: Application to AVHRR. J. Geophys. Res., 117, D06206, doi:10.1029/2011JD016656.

    • Search Google Scholar
    • Export Citation
  • Emmitt, G. D., 1987: Error analysis for total wind vector computations using one component measurements from a space-based Doppler lidar. Proc. Fourth Conf. on Coherent Laser Radar: Technology and Applications, Aspen, CO, Optical Society of America, 217–220.

  • Frehlich, R. G., 2000: Simulation of coherent Doppler lidar performance for space-based platforms. J. Appl. Meteor., 39, 245262, doi:10.1175/1520-0450(2000)039<0245:SOCDLP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kleist, D. T., , Parrish D. F. , , Derber J. C. , , Treadon R. , , Wu W.-S. , , and Lord S. , 2009: Introduction of the GSI into the NCEP Global Data Assimilation System. Wea. Forecasting, 24, 16911705, doi:10.1175/2009WAF2222201.1.

    • Search Google Scholar
    • Export Citation
  • Liu, Q., , and Weng F. , 2006: Advanced doubling–adding method for radiative transfer in planetary atmosphere. J. Atmos. Sci., 63, 34593465, doi:10.1175/JAS3808.1.

    • Search Google Scholar
    • Export Citation
  • Marseille, G.-J., , Stoffelen A. , , and Barkmeijer J. , 2008: Impact assessment of prospective spaceborne Doppler wind lidar observation scenarios. Tellus,60A, 234–248, doi:10.1111/j.1600-0870.2007.00289.x.

    • Search Google Scholar
    • Export Citation
  • Masutani, M., and Coauthors, 2010: Observing system simulation experiments at the National Centers for Environmental Prediction. J. Geophys. Res., 115, D07101, doi:10.1029/2009JD012528.

    • Search Google Scholar
    • Export Citation
  • McKay, J. A., 1998: Modeling of direct detection Doppler wind lidar. I. The edge technique. Appl. Opt., 37, 64806486, doi:10.1364/AO.37.006480.

    • 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, doi:10.1109/5.24130.

    • Search Google Scholar
    • Export Citation
  • NRC, 2007: Earth Science and Applications from Space: National Imperatives for the Next Decade and Beyond. The National Academies Press, 456 pp.

  • Reale, O., , Terry J. , , Masutani M. , , Andersson E. , , Riishøjgaard L. P. , , and Jusem J. C. , 2007: Preliminary evaluation of the European Centre for Medium-Range Weather Forecasts’ (ECMWF) Nature Run over the tropical Atlantic and African monsoon region. Geophys. Res. Lett., 34, L22810, doi:10.1029/2007GL031640.

    • Search Google Scholar
    • Export Citation
  • Riishøjgaard, L. P., , Atlas R. , , and Emmitt G. D. , 2004: The impact of Doppler lidar wind observations on a single-level meteorological analysis. J. Appl. Meteor., 43, 810820, doi:10.1175/2091.1.

    • Search Google Scholar
    • Export Citation
  • Riishøjgaard, L. P., , Ma Z. , , Masutani M. , , Woollen J. S. , , Emmitt G. D. , , Wood S. A. , , and Greco S. , 2012: Observation system simulation experiments for a global wind observing sounder. Geophys. Res. Lett., 39, L17805, doi:10.1029/2012GL051814.

    • Search Google Scholar
    • Export Citation
  • Rohaly, G. D., , and Krishnamurti T. N. , 1993: An observing system simulation experiment for the Laser Atmospheric Wind Sounder (LAWS). J. Appl. Meteor., 32, 14531471, doi:10.1175/1520-0450(1993)032<1453:AOSSEF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Seablom, M. S., , Talabac S. J. , , Higgins G. J. , , and Womack B. T. , 2007: Simulation for the design of next-generation global Earth observing systems. Atmospheric and Environmental Remote Sensing Data Processing and Utilization III: Readiness for GEOSS, M. D. Goldberg et al., Eds., International Society for Optical Engineering (SPIE Proceedings, Vol. 6684), 668413, doi:10.1117/12.735746.

  • Stoffelen, A., and Coauthors, 2005: The atmospheric dynamics mission for global wind field measurement. Bull. Amer. Meteor. Soc., 86, 7387, doi:10.1175/BAMS-86-1-73.

    • Search Google Scholar
    • Export Citation
  • Stoffelen, A., , Marseille G. J. , , Bouttier F. , , Vasiljevic D. , , De Haan S. , , and Cardinali C. , 2006: ADM-Aeolus Doppler wind lidar Observing System Simulation Experiment. Quart. J. Roy. Meteor. Soc., 132, 19271947, doi:10.1256/qj.05.83.

    • Search Google Scholar
    • Export Citation
  • Tan, D. G. H., , and Andersson E. , 2005: Simulation of the yield and accuracy of wind profile measurements from the Atmospheric Dynamics Mission (ADM-Aeolus). Quart. J. Roy. Meteor. Soc., 131, 1737–1757, doi:10.1256/qj.04.02.

    • Search Google Scholar
    • Export Citation
  • Wood, S. A., Jr., , Emmitt G. D. , , and Greco S. , 2000: DLSM: A coherent and direct detection lidar simulation model for simulating space-based and aircraft-based lidar winds. Laser Radar Technology and Applications V, G. W. Kamerman et al., Eds., International Society for Optical Engineering (SPIE Proceedings, Vol. 4035), 2, doi:10.1117/12.397788.

  • Žagar, N., , Gustafsson N. , , and Källén E. , 2004: Dynamical response of equatorial waves in four-dimensional variational data assimilation. Tellus, 56A, 2946, doi:10.1111/j.1600-0870.2004.00036.x.

    • Search Google Scholar
    • Export Citation
Save