• Ackerman, T. P., and Stokes G. M. , 2003: The Atmospheric Radiation Measurement Program. Phys. Today, 56 , 3844.

  • Chou, S-H., and Atlas D. , 1982: Satellite estimates of ocean–air heat fluxes during cold air outbreaks. Mon. Wea. Rev., 110 , 14341450.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chou, S-H., Atlas R. M. , Shie C-L. , and Ardizzone J. , 1982: Estimates of surface humidity and latent heat fluxes over oceans from SSM/I data. Mon. Wea. Rev., 123 , 24052425.

    • Search Google Scholar
    • Export Citation
  • Clough, S. A., Kneizys F. X. , Rothman L. S. , and Gallery W. O. , 1981: Atmospheric spectra transmittance and radiance: FASCODE 1B. Proc. SPIE, 277 , 152166.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Clough, S. A., Kneizys F. X. , and Davies R. W. , 1989: Line shape and the water vapor continuum. Atmos. Res., 97 , 229241.

  • Courtier, P., and Coauthors, 1998: The ECMWF implementation of three-dimensional variational assimilation (3D-Var). I: Formulation. Quart. J. Roy. Meteor. Soc., 124 , 17831807.

    • Search Google Scholar
    • Export Citation
  • Crook, N. A., 1996: Sensitivity of moist convection forced by boundary layer processes to low-level thermodynamic fields. Mon. Wea. Rev., 124 , 17671785.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dabberdt, W. F., and Schlatter T. W. , 1996: Research opportunities from emerging atmospheric observing and modeling capabilities. Bull. Amer. Meteor. Soc., 77 , 305323.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dunion, J. P., and Veldon C. S. , 2004: The impact of the Saharan air layer on Atlantic tropical cyclone activity. Bull. Amer. Meteor. Soc., 85 , 353365.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Elliott, W. P., and Gaffen D. J. , 1991: On the utility of radiosonde humidity archives for climate studies. Bull. Amer. Meteor. Soc., 72 , 15071520.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Emanuel, K., and Coauthors, 1995: Report of the first prospectus development team of the U.S. Weather Research Program to NOAA and the NSF. Bull. Amer. Meteor. Soc., 76 , 11941208.

    • Search Google Scholar
    • Export Citation
  • Feltz, W. F., Posselt D. , Mecikalski J. R. , Wade G. S. , and Smith T. J. , 2003a: Rapid boundary layer water vapor transitions. Bull. Amer. Meteor. Soc., 84 , 2930.

    • Search Google Scholar
    • Export Citation
  • Feltz, W. F., Smith W. L. , Howell H. B. , Knuteson R. O. , Woolf H. , and Revercomb H. E. , 2003b: Near continuous profiling of temperature, moisture, and atmospheric stability using the Atmospheric Emitted Radiance Interferometer (AERI). J. Appl. Meteor., 42 , 584597.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gutzler, D. S., 1993: Uncertainties in climatological tropical humidity profiles: Some implications for estimating the greenhouse effect. J. Climate, 6 , 978982.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Harries, J. E., Brindley H. E. , Sagoo P. J. , and Bantges R. J. , 2001: Increases in greenhouse forcing inferred from the outgoing longwave radiation spectra of the Earth in 1970 and 1997. Nature, 410 , 355357.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Houghton, J. T., Ding Y. , Griggs D. J. , Noguer M. , van der Linden P. J. , Dai X. , Maskell K. , and Johnson C. A. , 2001: Climate Change 2001: The Scientific Basis. Cambridge University Press, 892 pp.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437471.

  • Kanamitsu, M., Ebisuzaki W. , Woollen J. , Yang S-K. , Hnilo J. J. , Fiorino M. , and Potter G. L. , 2002: NCEP-DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83 , 16311643.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Knuteson, R. O., and Coauthors, 2004: Atmospheric Emitted Radiance Interferometer. Part I: Instrument design. J. Atmos. Oceanic Technol., 21 , 17631776.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, F., Vogelmann A. M. , and Ramanathan V. , 2004: Saharan dust aerosol radiative forcing measured from space. J. Climate, 17 , 25582570.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Miloshevich, L. M., Paukkunen A. , Vömel H. , and Oltmans S. J. , 2004: Development and validation of a time-lag correction for Vaisala radiosonde humidity measurements. J. Atmos. Oceanic Technol., 21 , 13051327.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Minnett, P. J., Knuteson R. O. , Best F. A. , Osborne B. J. , Hanafin J. A. , and Brown O. B. , 2001: The Marine-Atmospheric Emitted Radiance Interferometer (M-AERI), a high-accuracy, sea-going infrared spectroradiometer. J. Atmos. Oceanic Technol., 18 , 9941013.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Nalli, N. R., and Coauthors, 2005: Profile observations of the Saharan air layer during AEROSE 2004. Geophys. Res. Lett., 32 .L05815, doi:10.1029/2004GL022028.

    • Search Google Scholar
    • Export Citation
  • Nalli, N. R., and Coauthors, 2006: Ship-based measurements for infrared sensor validation during Aerosol and Ocean Science Expedition 2004. J. Geophys. Res., 111 .D09S04, doi:10.1029/2005JD006385.

    • Search Google Scholar
    • Export Citation
  • Prospero, J. M., and Carlson T. N. , 1972: Vertical and areal distribution of Saharan dust over the western equatorial North Atlantic Ocean. J. Geophys. Res., 77 , 52555265.

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

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

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

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

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Weckwerth, T. M., Wulfmeyer V. , Wakimoto R. M. , Hardesty R. M. , Wilson J. W. , and Banta R. M. , 1999: NCAR-NOAA lower-tropospheric water vapor workshop. Bull. Amer. Meteor. Soc., 80 , 23392357.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Williams, E., Prager E. , and Wilson D. , 2002: Research combines with public outreach on a cruise ship. Eos, Trans. Amer. Geophys. Union, 83 , 590596.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zipser, E. J., and Johnson R. H. , 1998: Systematic errors in radiosonde humidities a global problem? Preprints, 10th Symp. on Measurements, Observations and Instrumentation, Phoenix, AZ, Amer. Meteor. Soc, 72–73.

  • View in gallery
    Fig. 1.

    The track covered by the NOAA Ronald H. Brown during AEROSE from 29 Feb to 26 Mar 2004. The circles indicate the positions and times of radiosonde launches; the thick black line indicates the locations and times of M-AERI operation.

  • View in gallery
    Fig. 2.

    Downwelling atmospheric radiance spectrum acquired by M-AERI during the AEROSE cruise at 1050 UTC 3 Mar 2004. The spikes in the spectrum between ∼1300 and ∼1800 cm−1, and below ∼550 cm−1 are caused by the atmospheric transmissivity being so low that the detectors do not receive good signals from the internal black bodies. The subsequent radiometric calibration is very noisy. Only the spectral regions marked in red are used in the retrieval of temperature and humidity profiles. Note the change in the vertical scale in the two parts of the graph.

  • View in gallery
    Fig. 3.

    The temperature and water vapor mixing ratio structure from 3 Mar 2004 retrieved by M-AERI with the first-guess profiles constructed from (a) all seven available radiosonde profiles on this day, (b) every other radiosonde profile (four in total), (c) NCEP model analysis, (d) ECMWF model analysis, and (e) ECMWF model forecast. The red vertical lines in all panels indicate the times of radiosonde launches. In (b) the green vertical lines indicate the radiosondes that were not used for the first guess.

  • View in gallery
    Fig. 4.

    A comparison of ABL profiles of air temperature (blue) and the dewpoint temperature (red) measured on 3 Mar 2004 by radiosondes with profiles retrieved from M-AERI data at radiosonde launch times. Radiosonde measurements are compared to M-AERI-retrieved profiles with the first guess constructed using (a) all radiosonde profiles, (b) every other radiosonde profile, (c) NCEP model data, (d) ECMWF analyses, and (e) ECMWF forecasts. Dashed lines show the radiosonde measurements, dots show M-AERI retrievals, and the solid lines in (c)–(e) represent the model profiles.

  • View in gallery
    Fig. 5.

    The vertical profile of the RMS difference between the radiosonde measurements and radiosonde-based retrievals for the entire AEROSE cruise (where M-AERI was operating) for the air temperature, dewpoint temperature, and the water vapor mixing ratio, where (a) all radiosonde data were used to built the first guess, and (b) only every other radiosonde was used.

  • View in gallery
    Fig. 6.

    The profiles of the RMS difference between the radiosonde measurement and the NCEP-based retrieval (solid line), the NCEP model profiles and the NCEP-based retrieval (dashed line), and the radiosonde measurement and the NCEP model (dotted line) for the (a) air temperature and (b) dewpoint temperature.

  • View in gallery
    Fig. 7.

    As in Fig. 6, but for the profiles from the ECMWF model analysis.

  • View in gallery
    Fig. 8.

    As in Fig. 6, but for the profiles from the ECMWF model forecast.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 99 30 4
PDF Downloads 48 30 3

Profiling the Lower Troposphere over the Ocean with Infrared Hyperspectral Measurements of the Marine-Atmosphere Emitted Radiance Interferometer

Malgorzata SzczodrakRosenstiel School of Marine and Atmospheric Science, University of Miami, Miami, Florida

Search for other papers by Malgorzata Szczodrak in
Current site
Google Scholar
PubMed
Close
,
Peter J. MinnettRosenstiel School of Marine and Atmospheric Science, University of Miami, Miami, Florida

Search for other papers by Peter J. Minnett in
Current site
Google Scholar
PubMed
Close
,
Nicholas R. NalliQSS Group, Inc., Lanham, and NOAA/NESDIS/Office of Research and Applications, Camp Springs, Maryland

Search for other papers by Nicholas R. Nalli in
Current site
Google Scholar
PubMed
Close
, and
Wayne F. FeltzCooperative Institute for Meteorological Satellite Studies, University of Wisconsin—Madison, Madison, Wisconsin

Search for other papers by Wayne F. Feltz in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

Measurements of the spectra of infrared emission from the atmosphere were taken by a Marine-Atmospheric Emitted Radiance Interferometer (M-AERI) deployed on the NOAA ship Ronald H. Brown during the Aerosol and Ocean Science Expedition (AEROSE) in the tropical Atlantic Ocean from 29 February to 26 March 2004. The spectra are used to retrieve profiles of temperature and humidity in the lower troposphere up to a height of 3000 m. The M-AERI retrievals of the atmospheric structure require an initial guess profile. In this work, retrievals obtained from four separate initializations are compared, using 1) radiosondes launched from the Ronald H. Brown, 2) NOAA/NWS/NCEP model reanalyses, 3) ECMWF model analyses, and 4) ECMWF model forecasts. The performance of the M-AERI retrievals for all four first-guess sources is then evaluated against the radiosonde measurements. The M-AERI retrievals initialized using radiosondes reproduce the radiosonde profiles quite well and capture much of the observed vertical structure as should be expected. Of the retrievals initialized with model fields, those obtained using the ECMWF data yielded results closest to the radiosonde observations and enabled detection of the Saharan air layer (SAL) evident during AEROSE. However, the NCEP reanalysis, as well as the corresponding retrievals, failed to detect the SAL. These results demonstrate the ability of the M-AERI profile retrievals to identify the anomalous humidity distributions in the lower troposphere, but underscore the need for suitable vertical resolution in the first-guess profile used in the retrievals under such conditions.

Corresponding author address: Dr. M. Szczodrak, Meteorology and Physical Oceanography, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, FL 33149. Email: goshka@rsmas.miami.edu

Abstract

Measurements of the spectra of infrared emission from the atmosphere were taken by a Marine-Atmospheric Emitted Radiance Interferometer (M-AERI) deployed on the NOAA ship Ronald H. Brown during the Aerosol and Ocean Science Expedition (AEROSE) in the tropical Atlantic Ocean from 29 February to 26 March 2004. The spectra are used to retrieve profiles of temperature and humidity in the lower troposphere up to a height of 3000 m. The M-AERI retrievals of the atmospheric structure require an initial guess profile. In this work, retrievals obtained from four separate initializations are compared, using 1) radiosondes launched from the Ronald H. Brown, 2) NOAA/NWS/NCEP model reanalyses, 3) ECMWF model analyses, and 4) ECMWF model forecasts. The performance of the M-AERI retrievals for all four first-guess sources is then evaluated against the radiosonde measurements. The M-AERI retrievals initialized using radiosondes reproduce the radiosonde profiles quite well and capture much of the observed vertical structure as should be expected. Of the retrievals initialized with model fields, those obtained using the ECMWF data yielded results closest to the radiosonde observations and enabled detection of the Saharan air layer (SAL) evident during AEROSE. However, the NCEP reanalysis, as well as the corresponding retrievals, failed to detect the SAL. These results demonstrate the ability of the M-AERI profile retrievals to identify the anomalous humidity distributions in the lower troposphere, but underscore the need for suitable vertical resolution in the first-guess profile used in the retrievals under such conditions.

Corresponding author address: Dr. M. Szczodrak, Meteorology and Physical Oceanography, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, FL 33149. Email: goshka@rsmas.miami.edu

1. Introduction

Atmospheric water vapor is a crucial component of the earth’s atmosphere and has a major impact on the earth’s climate. It is well established that water vapor is the principal greenhouse gas in the atmosphere (e.g., Houghton et al. 2001). Water vapor contributes to the vertical distribution of the radiative heating and cooling, and thus helps to control the radiative fluxes at the surface and the top of the atmosphere. The three-dimensional distribution of atmospheric moisture is intimately linked with the dynamics and thermodynamics of the atmosphere, as well as the vertical and horizontal distribution of clouds. Knowledge of the state of the atmosphere in terms of the vertical and horizontal distribution of the temperature and water vapor is therefore necessary for the prediction of clouds and precipitation as well as in many branches of atmospheric research, such as boundary layer processes, atmospheric chemistry, hydrology, polar meteorology, climate studies, and severe weather event prediction (e.g., Stull 1988; Weckwerth et al. 1999). The improvement of water vapor measurements is argued to be a key factor in improving in the quantitative predictions of convective rainfall (Emanuel et al. 1995; Dabberdt and Schlatter 1996) and convective parameters (Crook 1996; Zipser and Johnson 1998), reducing the uncertainties in radiative budget in the Tropics (Gutzler 1993), and detecting climate change (Harries et al. 2001). As an absorbing gas, water vapor is also a major factor in the satellite remote sensing of surface properties, including infrared measurements of sea surface temperature (SST).

A large fraction of the total atmospheric water vapor is contained in the lower troposphere, and especially in the atmospheric boundary layer (ABL). The ABL is defined as the layer of atmosphere that is directly influenced by the earth’s surface through turbulent exchanges. The top of the ABL has been observed to range anywhere between ∼100 and 4000 m, with 1000 m being typical. The importance of the ABL to atmospheric dynamics is well documented (e.g., Stull 1988). For example, the flux of moisture and heat from the surface to the atmosphere is critical for accurate weather and climate prediction (Chou and Atlas 1982; Chou et al. 1995). Moisture and heat at the surface are transferred into the atmosphere within the surface layer and are then transported to other regions mainly by horizontal and vertical turbulence and convective motion. Accurate characterization of the atmospheric moisture in the ABL is thus especially important.

Radiosondes have been a traditional method for obtaining profiles of temperature and humidity in the atmosphere but their costs, cumbersome operation, and limited temporal resolution led to the development of ground-based passive infrared remote sensing techniques. The University of Wisconsin Space Science and Engineering Center (SSEC) developed the Atmospheric Emitted Radiance Interferometer (AERI; Knuteson et al. 2004) to facilitate the retrieval of the temperature and humidity structure of the lower troposphere up to approximately 3 km above the ground, based on the measurements of downwelling atmospheric infrared radiation (Smith et al. 1999). The AERI is currently deployed at a number of land sites of the Department of Energy’s Atmospheric Radiation Measurement (ARM) Program (Stokes and Schwartz 1994; Ackerman and Stokes 2003), providing real-time measurements with approximately 10-min time resolution of the vertical distribution of the temperature and humidity in the lower troposphere. The Marine-Atmospheric Emitted Radiance Interferometer (M-AERI; Minnett et al. 2001) is a seagoing version of the AERI that has been deployed on a number of research vessels as well as for long-duration operations on a commercial cruise liner, Explorer of the Seas (Williams et al. 2002).

The retrieval of atmospheric profiles of temperature and humidity from AERI instruments relies on an iterative solution of the radiative transfer equation that requires initial “first guess” profiles. For the ARM observing sites, these are derived from statistics of many years of local radiosonde data. Profile retrievals over the oceans from seagoing M-AERIs, however, present several particular challenges—one being that radiosonde databases representative of marine conditions are not available. It is therefore necessary to use a different approach.

Combining M-AERI retrievals with numerical forecast fields provides a possible means to continuously monitor the evolution of the marine atmosphere. This work investigates the utility of using atmospheric temperature and humidity fields produced in the U.S. National Oceanic and Atmospheric Administration (NOAA)/National Centers for Environmental Prediction (NCEP) reanalysis product, as well as the European Centre for Medium-Range Weather Forecasts (ECMWF) model, as the first guess for the retrieval of the structure of the lower marine troposphere from M-AERI. The performance of the NCEP and ECMWF model-based retrievals of the atmospheric profiles is evaluated by comparison with profiles obtained by coincident radiosonde measurements and with M-AERI retrievals obtained using the radiosonde profiles as a first guess. This study is based on data collected during the 2004 Aerosol and Ocean Science Expedition (AEROSE; Nalli et al. 2006).

2. Instruments and data

AEROSE was a multidisciplinary oceanographic field campaign conducted in the tropical North Atlantic Ocean on board the NOAA Research Vessel Ronald H. Brown in March 2004. The main goals of the expedition were 1) to obtain measurements of the microphysical properties and evolution of the Saharan air layer (SAL) and dust aerosol as they traverse the Atlantic Ocean, and 2) to assess the impact on Saharan dust on the atmospheric radiation and satellite remote sensing in visible and infrared wavelengths (e.g., Nalli et al. 2006). Thus, our current study implicitly provides an assessment of the M-AERI sounding ability under conditions of extreme atmospheric aerosol loading.

The Ronald H. Brown set out from Bridgetown, Barbados, on 29 February and traveled eastward toward Africa. Near the African coast, the ship turned north toward the Canary Islands and after a port call in Las Palmas de Gran Canaria, the ship returned to San Juan, Puerto Rico, on 26 March 2004. The cruise track is shown in Fig. 1.

A University of Miami M-AERI was installed on board the Ronald H. Brown and took near-continuous measurements of the atmospheric infrared radiance, delivering a radiance spectrum approximately every 10 min throughout most of the cruise (with the exception of several days of stormy weather and an instrument malfunction). Between 2 and 26 March, radiosondes were launched from the Ronald H. Brown approximately every 3 h (Nalli et al. 2005). Because of the high frequency of radiosonde measurements, it is possible to use a selected subset of radiosonde profiles as the first guess to initialize the M-AERI retrieval and to use the remaining withheld data for validation of the retrievals.

a. M-AERI

The M-AERI is a Fourier transform infrared interferometric spectroradiometer that measures spectra spanning the ∼3–18 μm wavelength interval with a resolution of ∼0.5 cm−1. It uses two infrared detectors cooled to ∼78 K by a Stirling cycle mechanical cooler to reduce the noise-equivalent temperature difference to levels well below 0.1 K. The radiometric calibration of the M-AERI is done frequently using two internal blackbody cavities, each with an effective emissivity of >0.998. The mirror scan sequence includes measurements of the reference cavities before and after taking each set of spectra from the ocean and atmosphere. The control computer integrates interferometric measurements over a preselected time interval, usually a few tens of seconds, to obtain a satisfactory signal-to-noise ratio. A typical cycle of measurements, including two uplooking views of the atmosphere at different zenith angles, one downlooking view of the ocean, and blackbody target calibration measurements, takes about 10 min. The instrument is described more fully by Minnett et al. (2001).

An example of the atmospheric emission spectrum measured by the M-AERI on 3 March 2004 during the AEROSE cruise is shown in Fig. 2 after being apodized. The atmospheric emission spectra contain information about the vertical distribution of the temperature and the water vapor in the lower troposphere. This information is available in the M-AERI data each measurement cycle, approximately every 10 min.

b. Atmospheric profile retrieval

The method of inverting infrared spectra to retrieve the profiles of temperature and water vapor in the lowest 3 km of the atmosphere is described in Smith et al. (1999). Based on this method, the AERIPROF algorithm is routinely used with the land-based AERI at a number of the ARM observing sites (Feltz et al. 2003a, b). In this retrieval approach the downwelling radiance is computed for an assumed (first guess) atmospheric state, defined by profiles of temperature and humidity, and compared with the radiance spectra measured by M-AERI. The difference between the measured and the computed radiance spectra is propagated backward, and the assumed atmospheric profiles are modified. This procedure is repeated until the radiance computed from the modified guess profiles matches the measured radiance within some tolerance. At this point the retrieval is said to have converged and the modified profile is the retrieval result.

The downwelling radiance is computed from the guess profile using a forward radiative transfer model based upon the Fast Atmospheric Signature Code (FASCODE; Clough et al. 1981) line-by-line radiative transfer model with the Clough–Kneizys–Davies (CKD) version 2.1 water vapor continuum absorption model (Clough et al. 1989). The retrieval uses only selected regions of each M-AERI spectrum that correspond to the CO2 and water vapor absorption bands that contain the most accessible information on the vertical distribution of the atmospheric temperature and water vapor. These regions are marked in red in Fig. 2.

The retrieval technique uses the “onion peeling” approach, where profiles are modified first at the surface and then changes are made progressively higher in the atmosphere. The choice of this approach was driven by the fact that the weighting functions for the wavelengths used in the temperature and humidity retrieval peak at or near the surface. Because the information content in M-AERI spectra diminishes with height, the retrieved profiles are essentially limited to approximately 3 km above the sea surface.

As mentioned above, the retrieval process requires an initial guess of the temperature and moisture profiles. This initial guess can be a radiosonde profile or a profile from a numerical forecast model. Satellite-retrieved profiles can also be used.

Because the retrieval is only valid for cloud-free radiances, a simple cloud-detection algorithm is embedded in AERIPROF that compares brightness temperatures computed for a “cloudy atmosphere” (i.e., including in the forward calculation a cloud layer placed between certain levels) against the observed brightness temperature. If the difference is less than a selected threshold, the conditions are marked as possibly cloudy and no retrieval is performed.

The AERIPROF algorithm has been extensively validated in clear skies over the Southern Great Plains (SGP) ARM site by comparisons with the coincident radiosonde observations for temperature retrieval, and by comparisons with humidity profiles from radiosondes, scaled so the integrated water vapor matched that measured by a collocated microwave radiometer, and the Raman lidar data for the water vapor profile retrieval. The retrieved and observed temperature profile root-mean-square (RMS) differences were less than 1 K in the first 3 km of the atmosphere (Feltz et al. 2003a, b). The retrieved water vapor mixing ratio profiles agreed with the microwave radiometer–scaled radiosondes, and the Raman lidar data agreed to within 5% RMS (Turner et al. 2000). These retrieval accuracies can be considered representative for the regions with conditions similar to those at the SGP ARM site, especially in total moisture content quantified as column precipitable water vapor (PWV). The average PWV for the SGP site is about 2 cm. The average PWV encountered during AEROSE was 2.8 cm, but it varied between 0.75 cm when the SAL was present aloft and 5.2 cm when the marine atmosphere was unaffected by the SAL.

c. Vaisala RS80/90 radiosondes

In situ measurements of temperature and water vapor profiles were obtained throughout the cruise using commercial Vaisala RS80-H and RS90 balloon-borne radiosondes. Wind vector profiles were obtained from RS80-H rawinsondes for all but the last 2 days of the operations. Typical accuracies of the radiosonde measurements are ±0.5 K for temperature, ±1 hPa for pressure, and about ±5% for relative humidity (Elliott and Gaffen 1991). The humidity measurement by Vaisala radiosondes is also subject to a measurement bias, for which several sources have been identified (Miloshevich et al. 2004). Some of the biases depend on the temperature of the measurement and these begin to affect the humidity measurements at temperatures below −20°C. The contamination bias that was caused by nonwater molecules from the packaging material clinging to the sensor polymer has been practically eliminated in radiosondes produced after June 2000, when the packaging was changed. Procedures have been developed to correct radiosonde measurements for the biases (Miloshevich et al. 2004). Because the work reported here is limited to the tropical lower troposphere (where the temperature is generally greater than −20°C) and new RS80-H and RS90 radiosondes were used, no correction was applied to the original radiosonde measurement.

d. NCEP data

Atmospheric profiles from the “NCEP Reanalysis 2” were obtained from the NOAA–Cooperative Institute for Research in Environmental Sciences (CIRES) Climate Diagnostics Center, Boulder, Colorado, through their Web site (online at http://www.cdc.noaa.gov/). The NCEP reanalysis product provides global variables describing the state of the atmosphere with spatial resolution of 2.5° × 2.5° four times a day at 0000, 0600, 1200, and 1800 UTC. Atmospheric temperature and humidity are given at 17 pressure levels (1000, 925, 850, 700, 600, 500, 400, 300, 250, 200, 150, 100, 70, 50, 30, 20, and 10 hPa). A detailed description of the NCEP reanalysis project is given by Kalnay et al. (1996) and Kanamitsu et al. (2002). For the purpose of obtaining the first guess to initialize the M-AERI retrieval process, the NCEP dataset was interpolated in space to produce model profiles 4 times per day at the position of the Ronald H. Brown. These model profiles are subsequently interpolated in time to those of the M-AERI spectral measurements.

e. ECMWF data

Global analysis and forecast datasets were obtained from ECMWF for the period of AEROSE. Each ECMWF dataset contains a set of atmospheric variables on a spatial grid of 0.5° × 0.5° and at 60 vertical levels. The ECMWF analysis is produced daily for 0000, 0600, 1200, and 1800 UTC. The 6-hourly analyses are obtained by 2 four-dimensional variational data assimilation (4DVAR) minimization cycles running from 0300 to 1500 and from 1500 to 0300 UTC. The analysis is performed by comparing the observations directly with a very short forecast. The differences between the observed values and the equivalent values predicted by the short-range forecast are used to make a correction to the first-guess field in order to produce the atmospheric analysis (Courtier et al. 1998). The upper-air analysis is combined with the surface analyses of snow, ice, SST, soil wetness, ocean waves, 2-m temperature, and humidity, to produce the initial state for the next short-range forecast. The ECMWF forecast is produced twice daily, at 0000 and 1200 UTC. The 0000 UTC product contains forecasts for 0300 and 0900 UTC, and the 1200 UTC product provides forecasts for 1500 and 2100 UTC. For this study, we estimated the accuracy of the ECMWF profiles by interpolating in time and space to the times and positions of the radiosonde launches from the Ronald H. Brown and comparing the interpolated ECMWF fields with the measured profiles. For the initialization of the profile retrievals from M-AERI, the first-guess profiles were built 4 times daily at the ECMWF analysis or forecast times by interpolating the ECMWF profiles in space to the ship position at the analysis time. These four-times-per-day profiles were then interpolated in time to match the times of the M-AERI atmospheric spectral measurements.

3. Retrieval results

To illustrate the sensitivity of the derivation of the M-AERI profiles to the choice of the first-guess initialization, we conducted a set of retrievals for 3 March 2004 using several different schemes. The five panels of Fig. 3 show the structure of ABL temperature and water vapor mixing ratio, as retrieved by M-AERI, with the first-guess profiles constructed from (a) all available radiosonde profiles on this day (∼3 hourly), and (b) using only alternate radiosonde profiles (∼6 hourly), (c) NCEP analysis, (d) ECMWF analysis, and (e) ECMWF forecast. The red vertical lines in all five panels indicate the times of radiosonde launches; the green vertical lines in Fig. 3b indicate the radiosondes that were not used in generating the first-guess profiles. There is a great similarity between the atmospheric structure retrieved with the first guess based on the full set of radiosondes (Fig. 3a), and a first guess based on only every other radiosonde profile (Fig. 3b). The qualitative similarity between these retrievals implies that improving the first guess by increasing the refresh rate of radiosonde observations from ∼6 to ∼3 h has relatively little effect on the general, gross features of the retrieval.

The lower-tropospheric properties retrieved with the ECMWF analysis profiles as the first guess (Fig. 3d) and with ECMWF forecast profiles (Fig. 3e) are also very similar. This outcome is not surprising because both the ECMWF analysis and short-range forecast fields are themselves very similar. The temperature and moisture retrievals based on the ECMWF first guesses show similar features in the retrieved atmospheric structure with those obtained using radiosondes, in particular the encroaching elevated dry-air layer (the SAL) above 1500 m apparent in the water vapor mixing ratios. Elevated dry-air layers in this part of the northern Atlantic are predominantly of African origin and frequently carry large loads of dust aerosol from the Sahara Desert area (Prospero and Carlson 1972; Dunion and Veldon 2004). The Saharan air outbreaks are very common during late spring, summer, and early fall (Dunion and Veldon 2004; Li et al. 2004). Such an outbreak of the SAL was indeed observed from the high frequency of radiosondes launched during AEROSE (Nalli et al. 2005), and is resolved in both the radiosonde and ECMWF-based M-AERI retrievals (Figs. 3a, 3b, 3d and 3e). However, the dry-air layer is not resolved in the retrievals with the first guess derived from NCEP data, as shown in Fig. 3c.

Atmospheric profiles of air temperature and dewpoint temperature for the seven radiosondes launched on 3 March 2004 are compared in Fig. 4 (dashed lines) with those retrieved from M-AERI data (dots). Radiosonde measurements are compared in Fig. 4a with M-AERI retrievals, with the first guess constructed using all seven radiosonde profiles, and using alternate radiosondes (four total) in Fig. 4b, using NCEP reanalyses in Fig. 4c, using ECMWF analyses in Fig. 4d, and ECMWF forecasts in Fig. 4e. In Figs. 4c–e the solid lines show the model profiles used as first guess in the retrievals. In Fig. 4a, it is clearly apparent that the M-AERI-retrieved profiles with the first guess given by all available radiosondes agree very well with the actual radiosonde measurements. This is also the case when only alternate radiosonde profiles are used for the initialization (0726, 1319, and 1927 UTC launch times). When NCEP profiles are used as a first guess in the retrieval, the agreement between the resulting profiles with the radiosonde measurements is significantly less accurate, especially for the moisture retrieval. It is seen in Fig. 4c that NCEP-based retrieval completely fails to resolve the dry SAL that is clearly visible in the radiosonde measurements. Although the agreement is better for air temperature than humidity, the NCEP-based retrieval fails to resolve several temperature inversions. Nevertheless, the retrieved profiles still represent an improvement over the model profiles used as the first guess (solid lines), indicating that the spectral information leads to a more realistic profile when initialized with one that does not resolve the major atmospheric features present at the time.

The profile retrievals based on the ECMWF analyses and forecasts (Figs. 4d and 4e) are much closer to the radiosonde-measured profiles of temperature and humidity than those produced using the NCEP data. In particular, the M-AERI does not convincingly retrieve the SAL aloft if it is not first resolved in the first guess. The ECMWF profiles resolve the SAL, whereas the NCEP model fields do not. There appears to be no significant difference in agreement with measurements for retrievals based on the ECMWF analysis or forecast profiles.

The profile retrievals obtained with the radiosonde measurements used as the first guess allow for a relatively easy convergence to the measured radiosonde profiles. The difference between the measurement and retrieval driven by the first guess constructed from radiosonde profiles therefore establishes a baseline for the accuracy of the retrievals. Figure 5a shows the vertical profile of the RMS difference between the radiosonde measurement and radiosonde-based retrievals (temperature, dewpoint temperature, and the water vapor mixing ratio) for the entire cruise, where all radiosonde data were used to construct the first guess. All retrieved profiles obtained at radiosonde launch times during AEROSE and the corresponding radiosonde profiles are included in the computation of the RMS error (the number of profiles retrieved at radiosonde times varies from 73 to 110, depending on what constituted the first guess). For the air temperature the RMS uncertainty is less than 0.5 K throughout the entire column under consideration (<3 km). The RMS uncertainty in the retrieved dewpoint temperature is <3 K throughout most of the lower troposphere, with the exception being the levels near the surface, where it reaches ∼5 K. The water vapor mixing ratio is generally below 1.5 g kg−1 throughout, but again with the exception of the levels near the surface where it increases to about 2 g kg−1. As would be expected, the RMS between the retrievals and radiosondes increases as the “quality” of the first guess decreases. For example, in Fig. 5b, the RMS discrepancy in air temperature above ∼500 m increases to as much as 2 K because of the 50% reduced radiosonde refresh for the first guess. The dewpoint temperature RMS in this case increases almost twofold at some levels. The same holds for the water vapor mixing ratio.

The results of using the NCEP analysis as a source of profiles for the first guess of the retrieval are displayed in Fig. 6. Figure 6a shows the profile of the RMS difference between the radiosonde measurement of the air temperature and the NCEP-driven retrieval (solid line). Also shown are the RMS differences of the NCEP model profiles against the NCEP-based retrieval (dashed line) and the radiosonde measurement (dotted line). In all three cases the RMS difference remains <2.5 K, which is comparable to the case when the retrieval is initialized with alternate radiosondes. The difference between the dashed and dotted lines indicates that the information contained in M-AERI spectra brings the profile closer to the measurements than the model profile below ∼1000 m. Figure 6b shows the corresponding profile of RMS discrepancies for the dewpoint temperature. The RMS difference between the retrieval and the radiosonde varies between 4 and 10.5 K (more than 3 times the benchmark case), generally increasing with height except near the surface. Introducing the M-AERI data moves the model profiles closer to the measurements by 0.5–3.5 K in RMS difference above approximately 400 m. Below the 350-m level, the RMS difference (measurement − retrieval) is greater than that for (measurement − model).

Figures 7 and 8 are analogous to Fig. 6 and reveal the performance of the ECMWF analysis (Fig. 7) and forecast (Fig. 8) profiles as the first guess for the M-AERI retrieval of the lower-tropospheric structure. The profiles of the RMS difference between the measurement and the retrievals driven by both the ECMWF analyses and forecasts are similar. In both cases the air temperature RMS difference remains <2 K throughout the boundary layer and <1 K for the lowest 500 m. This behavior is comparable to the cases using the alternate radiosondes and NCEP profiles as the first guess. For the dewpoint temperature, the RMS difference ranges from ∼2 K slightly above the surface to ∼11 K at 3000 m. This is similar to the case of NCEP-based retrievals, but the ECMWF-initialized retrievals are closer to the sondes below ∼1700 m, where the average RMS difference is ∼5 K, as opposed to ∼7 K for those initialized with the NCEP data. A strong increase in the RMS for the ECMWF retrievals occurs at about the 1800-m level. Below this level, the retrievals are up to 2 K closer to the measurements than the model. The results presented in Figs. 6 –8 are summarized in Table 1. For each of the models used as a first guess, Table 1 list the atmospheric layers where adding the radiance information (in the form of M-AERI spectra) improves, degrades, or bears no effect on the agreement with the radiosonde measurement. The level where the difference between the retrieval and the numerical model profiles is the greatest is also given in Table 1. In the case of the NCEP model, the level of maximum effect is at ∼600 m for both the temperature and the dewpoint temperature. For the ECMWF, it is near surface for the analysis and ∼300 m for the forecast for air temperature; and, for dewpoint temperature, near the top of the 3-km layer for the analysis and ∼2200 m for the forecast.

4. Summary and discussion

Spectral measurements of atmospheric radiance measured by the M-AERI during the AEROSE campaign were used to retrieve profiles of temperature and humidity in the lower marine atmosphere. Combining M-AERI retrievals with numerical analysis/forecast profiles offers a possibility to continuously monitor evolution of the marine atmospheric boundary layer from oceangoing vessels without depending on radiosonde measurements. In this work, we present an AEROSE case study to establish the limits of accuracy of this approach. Among other things, AEROSE provided a unique source of data with which to validate the high-frequency M-AERI retrievals, including measurements from relatively high-frequency radiosondes launches (∼8/24 h). The AEROSE M-AERI data also contain several potential geophysical sources of retrieval error, including low-level clouds, an unusually strong BL temperature, and water vapor inversion associated with the SAL, sea spray, and, most significantly, high levels of atmospheric dust aerosols.

Retrievals of lower-atmospheric structure from both AERI and M-AERI require initial profiles (i.e., a first guess). A comprehensive statistical database of marine ABL profiles is not available for M-AERI retrievals (unlike the land-based AERI instruments), and consequently an alternative source of the initial profile is required. Here we explored the following data sources as the first guess: 1) in situ atmospheric profiles from coincident radiosondes, 2) NCEP model reanalysis, 3) ECMWF model analysis, and 4) ECMWF forecasts. Although retrievals initialized with the full set of radiosonde data are not independent when using the same radiosonde data for validation, they nevertheless provide the benchmark accuracy for retrievals initialized with model data.

For the benchmark case of the retrieval initialized by the first guess built from all available radiosonde profiles (typically eight per 24 h), the RMS difference for the air temperature was found to be <0.5 K throughout the atmosphere below 3 km. Corresponding values for the RMS discrepancies in the dewpoint temperature were found to be between 2 and 5 K, with a vertical average of 2.8 K.

When the first guess was constructed from a subset (i.e., alternates) of available radiosonde profiles, the withheld radiosonde profiles were used to validate the retrievals. Compared to the benchmark values, the retrieval deteriorates in terms of RMS to 1 K for the air temperature and 2–6 K, with a vertical average of 3.9 K, for the dewpoint temperature. When the retrieval is driven by the first guess built from the NCEP analysis data, the RMS difference for the temperature remains below 2 K, as in the case of the retrieval driven by alternate radiosonde profiles, while the RMS dewpoint temperature error typically increases to 2–4 times the benchmark case, dependent on height. The retrievals based on ECMWF data, whether analysis or forecast, show similar RMS differences with respect to measurements. For the dewpoint temperature retrieval, those initialized with the ECMWF first guess appear to be more accurate than those driven by NCEP data, especially below 1700 m (approximately 2-K RMS error). In general, the information contained in the M-AERI spectra improves upon the first guess, as can be seen from the lower RMS differences between the measurement and the retrieval versus the measurement and the model.

From the analysis of the RMS differences, the most accurate retrievals are obtained if the first guess is constructed from the radiosonde profiles, even if those measurements are sparse. This is due in part to the relatively high vertical resolution of the radiosondes. The ECMWF-initialized retrievals are more accurate than those obtained using the NCEP analyses. However, the simple analysis of RMS differences does not tell the whole story. Inspection of the individual profiles reveals that both the ECMWF profiles, and the retrieved profiles that used the ECMWF data as the first guess, resemble the shape of the measured profile much closer than either the NCEP analysis or the NCEP-based retrieval. In particular, the ECMWF data both captured the presence of the dry SAL aloft, whereas the NCEP data failed to detect the SAL. The smearing of the SAL boundary in the NCEP data propagated into the M-AERI retrievals.

The height of the layers of dry air in both the ECMWF model and the M-AERI-retrieved profiles were not always a perfect match to the radiosonde measurement. The misalignment of the vertical location of the dry layers in the profiles obtained by different methods can lead to relatively large RMS differences in the vicinity of the inversion height, even though the inversion may have been resolved. Yet, for the ECMWF models, the largest difference between the retrieved and modeled profiles occurs either close to the surface for the temperature, or near the top of the 3-km layer for the water vapor. For the NCEP model, which does not resolve the SAL, the largest difference between the modeled and retrieved profiles occurs near the 600-m level, which roughly corresponds to the bottom of the SAL on the days when it was observed.

In general, M-AERI retrievals improve upon the NCEP and ECMWF profiles below the 600-m level for temperature and around 1900 m for the dewpoint temperature, with the exception of the 0–300-m layer for the NCEP model. Slight degradation occurs also in the 100–400-m layer for the ECMWF forecast, but the observed changes lie within the radiosonde measurement error. Also, the degradation observed for both temperature and water vapor at levels above 1900 m is, in most cases, within the uncertainty of the radiosonde measurements. In fact, the results suggest that M-AERI retrievals might provide a better representation of the atmospheric state of the lower marine troposphere than the models because the differences between the retrievals and the radiosonde measurements are within the measurement error for most levels.

Acknowledgments

This research was funded by NASA (NAG511104). AEROSE 2004 was supported by the NOAA/NESDIS/ORA Satellite Meteorology and Climatology Division (M. Goldberg), the NOAA Educational Partnership Program (Grant NA17AE1625), and NOAA Grant NA17AE1623. We extend our appreciation to P. Clemente-Colón (NOAA/NESDIS) and V. Morris (Howard University/NCAS) for their contributions as co-principal investigators of the AEROSE mission. We also thank W. Wolf (QSS Group, Inc.) for collecting and providing access to the ECMWF model data over the AEROSE domain that was used in support of NASA Aqua Atmospheric Infrared Sounder (AIRS) validation. Finally, the at-sea support during AEROSE by the captain, officers, crew, and colleagues on board the NOAA Ronald H. Brown is gratefully acknowledged. Constructive comments by the anonymous reviewers have helped clarify the presentation.

REFERENCES

  • Ackerman, T. P., and Stokes G. M. , 2003: The Atmospheric Radiation Measurement Program. Phys. Today, 56 , 3844.

  • Chou, S-H., and Atlas D. , 1982: Satellite estimates of ocean–air heat fluxes during cold air outbreaks. Mon. Wea. Rev., 110 , 14341450.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chou, S-H., Atlas R. M. , Shie C-L. , and Ardizzone J. , 1982: Estimates of surface humidity and latent heat fluxes over oceans from SSM/I data. Mon. Wea. Rev., 123 , 24052425.

    • Search Google Scholar
    • Export Citation
  • Clough, S. A., Kneizys F. X. , Rothman L. S. , and Gallery W. O. , 1981: Atmospheric spectra transmittance and radiance: FASCODE 1B. Proc. SPIE, 277 , 152166.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Clough, S. A., Kneizys F. X. , and Davies R. W. , 1989: Line shape and the water vapor continuum. Atmos. Res., 97 , 229241.

  • Courtier, P., and Coauthors, 1998: The ECMWF implementation of three-dimensional variational assimilation (3D-Var). I: Formulation. Quart. J. Roy. Meteor. Soc., 124 , 17831807.

    • Search Google Scholar
    • Export Citation
  • Crook, N. A., 1996: Sensitivity of moist convection forced by boundary layer processes to low-level thermodynamic fields. Mon. Wea. Rev., 124 , 17671785.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dabberdt, W. F., and Schlatter T. W. , 1996: Research opportunities from emerging atmospheric observing and modeling capabilities. Bull. Amer. Meteor. Soc., 77 , 305323.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dunion, J. P., and Veldon C. S. , 2004: The impact of the Saharan air layer on Atlantic tropical cyclone activity. Bull. Amer. Meteor. Soc., 85 , 353365.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Elliott, W. P., and Gaffen D. J. , 1991: On the utility of radiosonde humidity archives for climate studies. Bull. Amer. Meteor. Soc., 72 , 15071520.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Emanuel, K., and Coauthors, 1995: Report of the first prospectus development team of the U.S. Weather Research Program to NOAA and the NSF. Bull. Amer. Meteor. Soc., 76 , 11941208.

    • Search Google Scholar
    • Export Citation
  • Feltz, W. F., Posselt D. , Mecikalski J. R. , Wade G. S. , and Smith T. J. , 2003a: Rapid boundary layer water vapor transitions. Bull. Amer. Meteor. Soc., 84 , 2930.

    • Search Google Scholar
    • Export Citation
  • Feltz, W. F., Smith W. L. , Howell H. B. , Knuteson R. O. , Woolf H. , and Revercomb H. E. , 2003b: Near continuous profiling of temperature, moisture, and atmospheric stability using the Atmospheric Emitted Radiance Interferometer (AERI). J. Appl. Meteor., 42 , 584597.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gutzler, D. S., 1993: Uncertainties in climatological tropical humidity profiles: Some implications for estimating the greenhouse effect. J. Climate, 6 , 978982.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Harries, J. E., Brindley H. E. , Sagoo P. J. , and Bantges R. J. , 2001: Increases in greenhouse forcing inferred from the outgoing longwave radiation spectra of the Earth in 1970 and 1997. Nature, 410 , 355357.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Houghton, J. T., Ding Y. , Griggs D. J. , Noguer M. , van der Linden P. J. , Dai X. , Maskell K. , and Johnson C. A. , 2001: Climate Change 2001: The Scientific Basis. Cambridge University Press, 892 pp.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437471.

  • Kanamitsu, M., Ebisuzaki W. , Woollen J. , Yang S-K. , Hnilo J. J. , Fiorino M. , and Potter G. L. , 2002: NCEP-DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83 , 16311643.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Knuteson, R. O., and Coauthors, 2004: Atmospheric Emitted Radiance Interferometer. Part I: Instrument design. J. Atmos. Oceanic Technol., 21 , 17631776.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, F., Vogelmann A. M. , and Ramanathan V. , 2004: Saharan dust aerosol radiative forcing measured from space. J. Climate, 17 , 25582570.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Miloshevich, L. M., Paukkunen A. , Vömel H. , and Oltmans S. J. , 2004: Development and validation of a time-lag correction for Vaisala radiosonde humidity measurements. J. Atmos. Oceanic Technol., 21 , 13051327.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Minnett, P. J., Knuteson R. O. , Best F. A. , Osborne B. J. , Hanafin J. A. , and Brown O. B. , 2001: The Marine-Atmospheric Emitted Radiance Interferometer (M-AERI), a high-accuracy, sea-going infrared spectroradiometer. J. Atmos. Oceanic Technol., 18 , 9941013.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Nalli, N. R., and Coauthors, 2005: Profile observations of the Saharan air layer during AEROSE 2004. Geophys. Res. Lett., 32 .L05815, doi:10.1029/2004GL022028.

    • Search Google Scholar
    • Export Citation
  • Nalli, N. R., and Coauthors, 2006: Ship-based measurements for infrared sensor validation during Aerosol and Ocean Science Expedition 2004. J. Geophys. Res., 111 .D09S04, doi:10.1029/2005JD006385.

    • Search Google Scholar
    • Export Citation
  • Prospero, J. M., and Carlson T. N. , 1972: Vertical and areal distribution of Saharan dust over the western equatorial North Atlantic Ocean. J. Geophys. Res., 77 , 52555265.

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

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

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

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

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Weckwerth, T. M., Wulfmeyer V. , Wakimoto R. M. , Hardesty R. M. , Wilson J. W. , and Banta R. M. , 1999: NCAR-NOAA lower-tropospheric water vapor workshop. Bull. Amer. Meteor. Soc., 80 , 23392357.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Williams, E., Prager E. , and Wilson D. , 2002: Research combines with public outreach on a cruise ship. Eos, Trans. Amer. Geophys. Union, 83 , 590596.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zipser, E. J., and Johnson R. H. , 1998: Systematic errors in radiosonde humidities a global problem? Preprints, 10th Symp. on Measurements, Observations and Instrumentation, Phoenix, AZ, Amer. Meteor. Soc, 72–73.

Fig. 1.
Fig. 1.

The track covered by the NOAA Ronald H. Brown during AEROSE from 29 Feb to 26 Mar 2004. The circles indicate the positions and times of radiosonde launches; the thick black line indicates the locations and times of M-AERI operation.

Citation: Journal of Atmospheric and Oceanic Technology 24, 3; 10.1175/JTECH1961.1

Fig. 2.
Fig. 2.

Downwelling atmospheric radiance spectrum acquired by M-AERI during the AEROSE cruise at 1050 UTC 3 Mar 2004. The spikes in the spectrum between ∼1300 and ∼1800 cm−1, and below ∼550 cm−1 are caused by the atmospheric transmissivity being so low that the detectors do not receive good signals from the internal black bodies. The subsequent radiometric calibration is very noisy. Only the spectral regions marked in red are used in the retrieval of temperature and humidity profiles. Note the change in the vertical scale in the two parts of the graph.

Citation: Journal of Atmospheric and Oceanic Technology 24, 3; 10.1175/JTECH1961.1

Fig. 3.
Fig. 3.

The temperature and water vapor mixing ratio structure from 3 Mar 2004 retrieved by M-AERI with the first-guess profiles constructed from (a) all seven available radiosonde profiles on this day, (b) every other radiosonde profile (four in total), (c) NCEP model analysis, (d) ECMWF model analysis, and (e) ECMWF model forecast. The red vertical lines in all panels indicate the times of radiosonde launches. In (b) the green vertical lines indicate the radiosondes that were not used for the first guess.

Citation: Journal of Atmospheric and Oceanic Technology 24, 3; 10.1175/JTECH1961.1

Fig. 4.
Fig. 4.

A comparison of ABL profiles of air temperature (blue) and the dewpoint temperature (red) measured on 3 Mar 2004 by radiosondes with profiles retrieved from M-AERI data at radiosonde launch times. Radiosonde measurements are compared to M-AERI-retrieved profiles with the first guess constructed using (a) all radiosonde profiles, (b) every other radiosonde profile, (c) NCEP model data, (d) ECMWF analyses, and (e) ECMWF forecasts. Dashed lines show the radiosonde measurements, dots show M-AERI retrievals, and the solid lines in (c)–(e) represent the model profiles.

Citation: Journal of Atmospheric and Oceanic Technology 24, 3; 10.1175/JTECH1961.1

Fig. 5.
Fig. 5.

The vertical profile of the RMS difference between the radiosonde measurements and radiosonde-based retrievals for the entire AEROSE cruise (where M-AERI was operating) for the air temperature, dewpoint temperature, and the water vapor mixing ratio, where (a) all radiosonde data were used to built the first guess, and (b) only every other radiosonde was used.

Citation: Journal of Atmospheric and Oceanic Technology 24, 3; 10.1175/JTECH1961.1

Fig. 6.
Fig. 6.

The profiles of the RMS difference between the radiosonde measurement and the NCEP-based retrieval (solid line), the NCEP model profiles and the NCEP-based retrieval (dashed line), and the radiosonde measurement and the NCEP model (dotted line) for the (a) air temperature and (b) dewpoint temperature.

Citation: Journal of Atmospheric and Oceanic Technology 24, 3; 10.1175/JTECH1961.1

Fig. 7.
Fig. 7.

As in Fig. 6, but for the profiles from the ECMWF model analysis.

Citation: Journal of Atmospheric and Oceanic Technology 24, 3; 10.1175/JTECH1961.1

Fig. 8.
Fig. 8.

As in Fig. 6, but for the profiles from the ECMWF model forecast.

Citation: Journal of Atmospheric and Oceanic Technology 24, 3; 10.1175/JTECH1961.1

Table 1.

Levels (m) in the 3-km layer over the ocean surface where the M-AERI-retrieved profiles of temperature and the dewpoint temperature represent an improvement, no change, or degradation over the numerical model profiles in terms of RMS difference between the retrieved or modeled profiles and the radiosonde observations. The level where the difference between the model and retrieved profiles is the greatest is listed as the “max effect level.”

Table 1.
Save