• Asnani, G. C., 2005: Tropical Meteorology. Vol. 3, Indian Institute of Tropical Meteorology, 735 pp.

  • Atlas, R., and Coauthors, 2005: Hurricane forecasting with the high-resolution NASA finite volume general circulation model. Geophys. Res. Lett., 32 , L03807. doi:10.1029/2004GL021513.

    • Search Google Scholar
    • Export Citation
  • Bosilovich, M. G., 2008: NASA’s modern era retrospective-analysis for research and applications: Integrating earth observations. [Available online at http://www.earthzine.org/2008/09/26/nasas-modern-era-retrospective-analysis].

    • Search Google Scholar
    • Export Citation
  • Bosilovich, M. G., and Coauthors, 2006: NASA’s modern era retrospective-analysis for research and applications. U.S. CLIVAR Variations, U.S. CLIVAR Project Office, Washington, DC, No. 2, 5–8.

    • Search Google Scholar
    • Export Citation
  • Brown, D. P., 2006: Tropical cyclone report: Hurricane Helene (AL082006) 12–24 September 2006. [Available online at http://www.nhc.noaa.gov/pdf/TCR-AL082006_Helene.pdf].

    • Search Google Scholar
    • Export Citation
  • Burpee, R. W., 1974: Characteristics of North African easterly waves during the summers of 1968 and 1969. J. Atmos. Sci., 31 , 15561570.

    • Search Google Scholar
    • Export Citation
  • Carlson, T. N., , and J. M. Prospero, 1972: The large-scale movements of Saharan air outbreaks over the northern equatorial Atlantic. J. Appl. Meteor., 11 , 283297.

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

    • Search Google Scholar
    • Export Citation
  • Hendricks, A. E., , M. T. Montgomery, , and C. A. Davis, 2004: The role of “vortical” hot towers in the formation of Tropical Cyclone Diana (1984). J. Atmos. Sci., 61 , 12091232.

    • Search Google Scholar
    • Export Citation
  • Hsieh, J-S., , and K. Cook, 2005: Generation of African easterly wave disturbances: Relationship to the African easterly jet. Mon. Wea. Rev., 133 , 13111327.

    • Search Google Scholar
    • Export Citation
  • Jenkins, G. S., , and A. Pratt, 2008: Saharan dust, lightning and tropical cyclones in the eastern tropical Atlantic during NAMMA-06. Geophys. Res. Lett., 35 , L12804. doi:10.1029/2008GL033979.

    • Search Google Scholar
    • Export Citation
  • Jenkins, G. S., , A. Pratt, , and A. Heymsfield, 2008: Possible linkages between Saharan dust and tropical cyclone rain band invigoration in the eastern Atlantic during NAMMA-06. Geophys. Res. Lett., 35 , L08815. doi:10.1029/2008GL034072.

    • Search Google Scholar
    • Export Citation
  • Kiladis, G. N., , C. D. Thorncroft, , and N. M. J. Hall, 2006: Three-dimensional structure and dynamics of African easterly waves. Part I: Observations. J. Atmos. Sci., 63 , 22122230.

    • Search Google Scholar
    • Export Citation
  • Lau, K. M., , and K-M. Kim, 2007: Cooling of the Atlantic by Saharan dust. Geophys. Res. Lett., 34 , L23811. doi:10.1029/2007GL031538.

  • Lin, S-J., 2004: A “vertically Lagrangian” finite-volume dynamical core for global models. Mon. Wea. Rev., 132 , 22932307.

  • Moorthi, S., , and M. J. Suarez, 1992: Relaxed Arakawa-Schubert. A parameterization of moist convection for general circulation models. Mon. Wea. Rev., 120 , 9781002.

    • Search Google Scholar
    • Export Citation
  • Reale, O., , J. Terry, , M. Masutani, , E. Andersson, , L. P. Riishojgaard, , and J. C. Jusem, 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
  • Reale, O., , J. Susskind, , R. Rosenberg, , E. Brin, , E. Liu, , L. P. Riishojgaard, , J. Terry, , and J. C. Jusem, 2008: Improving forecast skill by assimilation of quality-controlled AIRS temperature retrievals under partially cloudy conditions. Geophys. Res. Lett., 35 , L08809. doi:10.1029/2007GL033002.

    • Search Google Scholar
    • Export Citation
  • Reale, O., , W. K. Lau, , J. Susskind, , E. Brin, , E. Liu, , L. P. Riishojgaard, , M. Fuentes, , and R. Rosenberg, 2009: AIRS impact on the analysis and forecast track of Tropical Cyclone Nargis in a global data assimilation and forecasting system. Geophys. Res. Lett., 36 , L06812. doi:10.1029/2008GL037122.

    • Search Google Scholar
    • Export Citation
  • Rienecker, M. M., and Coauthors, 2008: The GEOS-5 data assimilation system. Documentation Versions 5.0.1, 5.1.0, and 5.20. NASA Tech. Rep. Series on Global Modeling and Data Assimilation 27, NASA/TM-2008-104606, 1–118. [Available online at http://gmao.gsfc.nasa.gov/pubs/docs/Rienecker369.pdf].

    • Search Google Scholar
    • Export Citation
  • Shen, B-W., , R. Atlas, , O. Reale, , S-J. Lin, , J-D. Chern, , J. Chang, , C. Henze, , and J-L. Li, 2006: Hurricane forecasts with a global mesoscale-resolving model: Preliminary results with Hurricane Katrina (2005). Geophys. Res. Lett., 33 , L13813. doi:10.1029/2006GL026143.

    • Search Google Scholar
    • Export Citation
  • Susskind, J., 2007: Improved atmospheric soundings and error estimates from analysis of AIRS/AMSU data. Atmospheric and Environmental Remote Sensing Data Processing and Utilization III: Readiness for GEOSS, International Society for Optical Engineering (SPIE Proceedings, Vol. 6684), doi:10.1117/12.734336.

    • Search Google Scholar
    • Export Citation
  • Susskind, J., , C. Barnet, , J. Blaisdell, , L. Iredell, , F. Keita, , L. Kouvaris, , G. Molnar, , and M. Chahine, 2006: Accuracy of geophysical parameters derived from atmospheric infrared sounder/advanced microwave sounding unit as a function of fractional cloud cover. J. Geophys. Res., 111 , D09S17. doi:10.1029/2005JD006272.

    • Search Google Scholar
    • Export Citation
  • Tao, W-K., , S. Long, , J. Simpson, , C-H. Sui, , B. Ferrier, , and M-D. Chou, 1996: Cloud-radiation mechanisms associated with a tropical and a mid-latitude squall line. J. Atmos. Sci., 53 , 26242651.

    • Search Google Scholar
    • Export Citation
  • Wu, W-S., , R. J. Purser, , and D. F. Parrish, 2002: Three-dimensional variational analysis with spatially inhomogeneous covariances. Mon. Wea. Rev., 130 , 29052916.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    (left) Hovmöller of 700-hPa relative vorticity (s−l) and (right) relative humidity from the GEOS-5 analyses, latitudinally averaged (12°–20°N) and covering the SOP-3 period from 15 August to 15 September. The waves, which are the focus of this study, are W1 (0°−40°W longitude, 22–29 August) and W2 (15°–40°W, 12–16 September).

  • View in gallery

    As in Fig. 1, but for the NCEP operational analyses.

  • View in gallery

    Sequence of the observed profiles of the (top) zonal and (bottom) meridional components of the wind, taken daily at 1200 UTC at the Cape Verde station (latitude 14.92°61′11″N, longitude 23.49°46′11″W, and altitude 85 m) during the SOP-3 campaign.

  • View in gallery

    As in Fig. 3, but extracted from the GEOS-5 analyses. Sequence of the observed profiles (at each synoptic time) of the (top) zonal and (bottom) meridional components of the wind, taken at the same latitude and longitude of the station.

  • View in gallery

    (top left) GEOS-5 (at 0.25°) 700-hPa specific humidity (g kg−1) and 900-hPa wind (streamlines) in the GEOS-5 analyses for 0000 UTC 26 Aug, and relative to the 24-, 48-, and 72-h forecasts for 0000 UTC 27, 28, and 29 Aug.

  • View in gallery

    SAL analysis (image courtesy of the CIMSS Tropical Cyclone Group, University of Wisconsin) for 0000 UTC 26 Aug.

  • View in gallery

    As in Fig. 6, but for 0000 UTC 27 Aug.

  • View in gallery

    GEOS-5 (at 0.25°) with a zonal vertical cross section of specific humidity (g kg−1, shaded) and temperature anomaly [°C; red (blue) contour, positive (negative), subtracting the zonal mean between 90° and 15°W] at 20°N for 0000 UTC 27 Aug, 24-h forecast initialized at 0000 UTC 26 Aug. W1 is located at about 25°–30°W.

  • View in gallery

    As in Fig. 8, but extracted from NCEP full-resolution analyses in model levels for validation purposes. The vertical dimension is only approximately comparable with Fig. 8, because the spacing between model levels and pressure levels is different.

  • View in gallery

    As in Fig. 5, but for 0000 UTC 13 Sep and relative to the 24-, 48- and 72-h forecasts for 0000 UTC 14, 15, and 16 Sep.

  • View in gallery

    As in Fig. 8, but showing the cyclonic vorticity (contours at l, 2, 3 × 10−4 s−1 at 20°N for 0000 UTC 14 Sep; darker red) and for the 24-h forecast initialized at 0000 UTC 13 Sep. The developing TS Helene, identifiable by the vertical column of strong cyclonic vorticity, is located at about 32°W.

  • View in gallery

    As in Fig. 11, but extracted from NCEP full-resolution analyses in model levels for validation purposes.

  • View in gallery

    Four-day mean MODIS aerosol optical depth for (a) 26–29 Aug 2006 and (b) 13–16 Sep 2006. The MODIS data show much larger amount of dust for the first wave, which is the nondeveloping one. The 900-hPa flow at (a) 0000 UTC 26 Aug, to indicate W1 position, and the track of Helene, (b) every 6 h from 0000 UTC 13 Sep to 1800 UTC 18 Sep, are superimposed. Solid dots indicate hurricane strength.

  • View in gallery

    As in Fig. 8, but produced from GEOS-5 at the lower resolution of 0.5°.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 29 29 6
PDF Downloads 15 15 3

Atlantic Tropical Cyclogenetic Processes during SOP-3 NAMMA in the GEOS-5 Global Data Assimilation and Forecast System

View More View Less
  • 1 Laboratory for Atmospheres, NASA Goddard Space Flight Center, Greenbelt, Maryland
© Get Permissions
Full access

Abstract

This article investigates the role of the Saharan air layer (SAL) in tropical cyclogenetic processes associated with a nondeveloping and a developing African easterly wave observed during the Special Observation Period (SOP-3) phase of the 2006 NASA African Monsoon Multidisciplinary Analyses (NAMMA). The two waves are chosen because they both interact heavily with Saharan air. A global data assimilation and forecast system, the NASA Goddard Earth Observing System, version 5 (GEOS-5), is being run to produce a set of high-quality global analyses, inclusive of all observations used operationally but with additional satellite information. In particular, following previous works by the same authors, the quality-controlled data from the Atmospheric Infrared Sounder (AIRS) used to produce these analyses have a better coverage than the one adopted by operational centers. From these improved analyses, two sets of 31 five-day high-resolution forecasts, at horizontal resolutions of both half and quarter degrees, are produced. Results indicate that very steep moisture gradients are associated with the SAL in forecasts and analyses, even at great distances from their source over the Sahara. In addition, a thermal dipole in the vertical (warm above, cool below) is present in the nondeveloping case. The Moderate Resolution Imaging Spectroradiometer (MODIS) aboard NASA’s Terra and Aqua satellites shows that aerosol optical thickness, indicative of more dust as opposed to other factors, is higher in the nondeveloping case. Altogether, results suggest that the radiative effect of dust may play some role in producing a thermal structure less favorable to cyclogenesis. Results also indicate that only global horizontal resolutions on the order of 20–30 km can capture the large-scale transport and the fine thermal structure of the SAL, inclusive of the sharp moisture gradients, reproducing the effect of tropical cyclone suppression that has been hypothesized by previous authors from observational and regional modeling perspectives. These effects cannot be fully represented at lower resolutions, therefore global resolution of a quarter of a degree is a minimum critical threshold necessary to investigate Atlantic tropical cyclogenesis from a global modeling perspective.

* Additional affiliation: University of Maryland, Baltimore County, Baltimore, Maryland.

+ Additional affiliation: Science Applications International Corporation, Beltsville, Maryland.

Corresponding author address: Oreste Reale, Laboratory for Atmospheres, NASA GSFC, Code 613, Greenbelt, MD 20771. Email: oreste.reale-1@nasa.gov

This article included in the TCSP NAMMA special collection.

Abstract

This article investigates the role of the Saharan air layer (SAL) in tropical cyclogenetic processes associated with a nondeveloping and a developing African easterly wave observed during the Special Observation Period (SOP-3) phase of the 2006 NASA African Monsoon Multidisciplinary Analyses (NAMMA). The two waves are chosen because they both interact heavily with Saharan air. A global data assimilation and forecast system, the NASA Goddard Earth Observing System, version 5 (GEOS-5), is being run to produce a set of high-quality global analyses, inclusive of all observations used operationally but with additional satellite information. In particular, following previous works by the same authors, the quality-controlled data from the Atmospheric Infrared Sounder (AIRS) used to produce these analyses have a better coverage than the one adopted by operational centers. From these improved analyses, two sets of 31 five-day high-resolution forecasts, at horizontal resolutions of both half and quarter degrees, are produced. Results indicate that very steep moisture gradients are associated with the SAL in forecasts and analyses, even at great distances from their source over the Sahara. In addition, a thermal dipole in the vertical (warm above, cool below) is present in the nondeveloping case. The Moderate Resolution Imaging Spectroradiometer (MODIS) aboard NASA’s Terra and Aqua satellites shows that aerosol optical thickness, indicative of more dust as opposed to other factors, is higher in the nondeveloping case. Altogether, results suggest that the radiative effect of dust may play some role in producing a thermal structure less favorable to cyclogenesis. Results also indicate that only global horizontal resolutions on the order of 20–30 km can capture the large-scale transport and the fine thermal structure of the SAL, inclusive of the sharp moisture gradients, reproducing the effect of tropical cyclone suppression that has been hypothesized by previous authors from observational and regional modeling perspectives. These effects cannot be fully represented at lower resolutions, therefore global resolution of a quarter of a degree is a minimum critical threshold necessary to investigate Atlantic tropical cyclogenesis from a global modeling perspective.

* Additional affiliation: University of Maryland, Baltimore County, Baltimore, Maryland.

+ Additional affiliation: Science Applications International Corporation, Beltsville, Maryland.

Corresponding author address: Oreste Reale, Laboratory for Atmospheres, NASA GSFC, Code 613, Greenbelt, MD 20771. Email: oreste.reale-1@nasa.gov

This article included in the TCSP NAMMA special collection.

1. Introduction

African easterly waves (AEWs) have been recognized as prominent weather-producing events of northern tropical Africa (e.g., Burpee 1974; Asnani 2005) and have been extensively studied from observational and modeling perspectives (e.g., Hsieh and Cook 2005; Kiladis et al. 2006). However, the development of AEWs into tropical depressions remains one of the most challenging problems in the prediction and modeling of Atlantic tropical cyclones.

Investigations of tropical cyclogenesis have been generally carried out from either strictly observational or high-resolution mesoscale points of view (e.g., Hendricks et al. 2004), because the lower resolution used in most global models is deemed to be inadequate to trigger spontaneous cyclogenesis with sufficiently realistic scales. However, in recent years a number of global models have reached the resolution of 10–40 km and have started to display some tropical cyclogenesis capability. Atlas et al. (2005) showed that a semioperational version of the model being run at that time by the National Aeronautics and Space Administration (NASA) was capable of producing well-defined tropical cyclone structures, even if initialized with low-resolution analyses in which cyclones were either not present or not properly analyzed. Shen et al. (2006) used a higher-resolution research version of the same model producing center pressures on the order of 920 hPa for Hurricane Katrina. Reale et al. (2007) analyzed the performance of the European Centre for Medium Range Weather Forecasts (ECMWF) so-called nature run (a 13-month-long simulation in free running mode performed with the then-operational ECMWF model at the horizontal resolution of T511, corresponding to approximately 40 km) and verified that the simulation contained quasi-realistic AEW activity and a realistic number of Atlantic tropical cyclones.

Strictly speaking, the resolutions adopted by the previously referenced studies cannot yet be considered adequate to resolve the fine structure of tropical cyclogenetic processes, because the organization of several convective centers into a rotating system requires cloud-resolving resolutions, which are not yet possible in real-time global numerical weather prediction models. Despite these limitations, it is nonetheless remarkable that the aforementioned high-resolution global models, with different experimental or operational configurations, display distinct cyclogenetic capabilities in the tropics. Moreover, it is important to emphasize that the processes of cyclogenesis and cyclogenesis suppression should be studied from different perspectives, and global models do have two unquestionable advantages: 1) they are better at capturing the large-scale forcings involved, and 2) do not rely upon the somewhat questionable boundary conditions imposed on the domain’s boundaries of limited-area models.

Among the large-scale problems connected with Atlantic tropical cyclogenesis, one of the most debated is the role of the Saharan air layer (SAL). The SAL, a layer of hot dry air rich in dust and produced over the Sahara Desert (Carlson and Prospero 1972) and its transport over the subtropical Atlantic Ocean, have been subject to intense scrutiny in order to understand their possible connection with tropical cyclone development or suppression.

Dunion and Velden (2004) investigated the SAL with the aid of Geostationary Operational Environmental Satellite (GOES) imagery and recognized it as a possible mechanism linked to tropical cyclone suppression, emphasizing three possible cyclogenesis inhibiting actions: increased static stability, increased vertical shear, and evaporatively driven downdrafts that result from dry air intrusions. On a longer temporal scale, Lau and Kim (2007) presented evidence that the mean activity of the SAL is likely to impact the large-scale environment, including SST and mean vertical shear, during the entire hurricane season.

However, Jenkins and Pratt (2008) have investigated the connection between dust, observed lightning, and tropical cyclones during the NASA African Monsoon Multidisciplinary Analyses (NAMMA-06), with emphasis on microphysical processes, confirming that the SAL increases static stability and may reduce SST, but also suggesting that the SAL may invigorate rainbands connected with cyclogenesis and may increase the cyclonic vorticity of AEWs. In the companion paper by the same first author, Jenkins et al. (2008) present more evidence of a possible tropical cyclone rainband invigoration due to Saharan dust.

At this time, it appears that there is no conclusive evidence of a clear, unambiguous role for the SAL with respect to cyclogenesis. The only certain fact is the extreme complexity of the interaction between the SAL and tropical weather systems, which requires much more research.

In this paper, it is shown that a high-resolution global data assimilation system (DAS) coupled with a global model of a comparable or even higher resolution are suitable tools to investigate the role of the SAL not only on the scale of cyclone formation but also from the point of view of the large-scale transport from its source region, and can therefore represent the possible modifications of the SAL, in terms of thermodynamical properties and shear, as the waves propagate over thousands of kilometers. At the same time, it is shown that a model horizontal resolution, not lower than a quarter of a degree, is needed to unveil some of the SAL’s thermodynamic and kinematic features, such as the increasingly narrow structure of the dry air filaments being intruded in a tropical circulation and the sharpness of the boundaries between Saharan and non-Saharan air.

In this work, we explore processes associated with two waves that were observed during the Special Observing Phase (SOP-3) of NAMMA, with the aid of the global data assimilation and forecasting system called the NASA Goddard Earth Observing System, version 5 (GEOS-5). Of particular concern for this article is the representation of the thermal structure during cyclogenesis and specifically the possible influence of the SAL on tropical cyclogenesis. Section 2 describes the data assimilation, computational aspects, and simulations performed, and section 3 analyzes the results. Section 4 presents a general discussion and section 5 states the conclusions of this work.

2. The NASA GEOS-5 data assimilation and forecasting system

The effect of Saharan air on Atlantic tropical cyclogenesis is investigated using a high-resolution global DAS based upon the gridpoint statistical interpolation (GSI) analysis algorithm developed by the National Centers for Environmental Prediction (NCEP) Environmental Modeling Center, documented in Wu et al. (2002). The NASA GEOS-5 combines a modified version of the previously referenced DAS with the NASA atmospheric global forecast model, as documented in Bosilovich et al. (2006) and more extensively in Rienecker et al. (2008). One important aspect of the GEOS-5 development is the optimization of satellite data usage, and in particular of the Earth Observing System instrumentation, having among its goals the production of the Modern Era Retrospective Analysis for Research and Applications (MERRA; Bosilovich 2008). The GEOS-5 has been used by Reale et al. (2008) and Reale et al. (2009), among others. The forecast model shares the same dynamical core (Lin 2004) with the so-called NASA finite-volume general circulation model (fvGCM), which in the version known as GEOS-4 demonstrated remarkable capabilities in hurricane forecasting (Atlas et al. 2005; Shen et al. 2006).

The GEOS-5, however, contains a new physics (convective and boundary layer parameterizations) developed predominantly by the Global Modeling and Assimilation Office (GMAO). The changes in physics made the system quite different from its predecessor. In particular, the convective scheme is a GMAO-modified version of Moorthi and Suarez (1992), and the radiative transfer model is the one developed by the Goddard Climate and Radiation Branch and implemented in various models, among which are the GEOS-5 and a cloud ensemble model (Tao et al. 1996). A full documentation of the system, with detailed description of the various components, is in Rienecker et al. (2008).

The global forecasting skill of this model has increased substantially, with the model’s performance (assessed globally with various metrics such as anomaly correlation plots) becoming comparable to major operational centers (e.g., Reale et al. 2008). Moreover, the heavy reliance on satellite data has made the performance quite symmetric between the two hemispheres, unlike earlier operational models that relied more on conventional observations, and consequently displayed a substantial difference in performance between the Northern and Southern Hemispheres.

However, the capability of representing very deep tropical cyclones, which was remarkable in the previous versions (Atlas et al. 2005; Shen et al. 2006), seems to have disappeared in the current configuration, possibly because of the use of a different convection parameterization.

The most important difference between the GEOS-5 and the system used in these two previous studies is in the initialization. Both Atlas et al. (2005) and Shen et al. (2006), despite running the model at horizontal resolutions of 0.25° and 0.125°, respectively, used initial conditions at about 1° resolution obtained from the NCEP operational analyses. This study makes full use of the data assimilation system included in the GEOS-5 and produces its own set of global, high-resolution 0.5° high-quality analyses. The analyses contain all the operational data used by NCEP at that time but make more extensive use of additional satellite information. In particular, much improved coverage from the Atmospheric Infrared Sounder (AIRS) is utilized, as compared to the one used by operational forecasts. Instead of clear-sky AIRS radiances, which employ AIRS data only from channels completely unaffected by clouds, we use quality-controlled temperature retrievals, also obtained under partly cloudy conditions, following Susskind et al. (2006) and Susskind (2007). Despite the widespread general assumption that clear-sky radiances are the best way to assimilate AIRS data, Reale et al. (2009) have shown retrievals under partly cloudy conditions to provide better analyses of a tropical cyclone in the Indian Ocean than the clear-sky radiance methodology. Finally, it should be stressed that the DAS for this experiment is configured without any bogus vortex or vortex relocation technique. As a consequence, tropical development, if any, is spontaneously produced by the model.

One 35-day global data assimilation, starting at 0000 UTC 13 August 2006, is performed with the GEOS-5 DAS to cover the entire period of the SOP-3 phase of the NAMMA campaign. From the analyses, two sets of thirty-one 5-day forecasts, at two different horizontal resolutions, are performed. The DAS and forecast model configuration is the same used by Reale et al. (2008) but at a higher resolution, as the DAS is being run at a horizontal resolution of 0.5° × 0.67° and the forecast model is run at 0.25° × 0.33° and 0.5° × 0.67°, both with 72 vertical levels.

Analyses and forecasts are validated and verified against operational NCEP analyses. The GEOS-5 analyses correspond remarkably well to the operational NCEP analyses, which will be shown later, but with more information available because of higher resolution and additional satellite information assimilated. Global and hemispheric forecast skills, investigated through the anomaly correlation for 500-hPa geopotential as a function of forecast time, range from approximately 0.76 to 0.80 at day 5 (not shown), indicating that the forecast skill for boreal summer conditions is reasonable and comparable to the slightly superior forecast skill obtained for boreal winter conditions (Reale et al. 2008).

3. Analysis and simulation

a. Validation

The focus of this work is on two events of the SOP-3: one nondeveloping and one developing AEW. These AEWs appeared quite similar in terms of the intensity of the low-level circulation, vertical shear, and other dynamical forcings. Their different evolutions and their interaction with the SAL are investigated. Figure 1 shows a Hovmöller diagram of 700-hPa relative vorticity and relative humidity obtained from the GEOS-5 analyses, to emphasize AEWs observed during the SOP-3 covering the period from 15 August to 15 September 2006. From Fig. 1 (left panel), the two strongest waves of the period, which are the subject of this study, can be seen: 1) the wave located at about 5°W on 23 August, undergoing transition from a continental to an oceanic environment on the following day, and appearing at about 20°W on 26 August (W1); and 2) the wave located at about 15°–40°W on 12–15 September (W2). The former is a nondeveloping wave, to be discussed in section 3b, the latter is the precursor of Helene and will be discussed in section 3c.

In Fig. 2 the same Hovmöller diagram is extracted from operational NCEP analyses (interpolated on pressure levels at 1° × 1°) for validation purposes. There is a very good correlation in the vorticity fields, notwithstanding that slightly stronger vorticity values can be seen in the GEOS-5 analyses. As for the moisture, it is intrinsically a very noisy field but the correspondence between the basic features is also very good.

To further confirm the legitimacy of our analyses, we also use NAMMA data for validation. In Fig. 3, the profiles of the zonal and meridional wind components, obtained from the vertical ground-based 1200 UTC soundings from the Cape Verde Islands, at 14.92°61′11″N, 23.49°46′11″W and an altitude of 85 m, are reported (radiosonde data are available online at http://namma.msfc.nasa.gov). The zonal wind time series clearly shows the African easterly jet (AEJ) located between 800 and 500 hPa with its amplitude being modulated at a time scale of 2–7 days. The interruption of the AEJ is caused by the periodic passage of waves, shown as sign inversions in the meridional wind time series. The northerly to southerly wind direction change that occurred on 26 August coincided with the passage of the positive vorticity over Cape Verde Islands, shown in Figs. 1, 2.

In Fig. 4, zonal and meridional wind profiles extracted at the same coordinates from the GEOS-5 analyses are provided. Despite the fact that the two datasets are not strictly comparable (the GEOS-5 sequence of profiles is computed continuously every 6 h, whereas the plot in Fig. 3 is an interpolation of daily profiles), the GEOS-5 analyses match very well with the observed data, as the four episodes of maximum zonal wind on 19–22 August, 1–3, 5–7, and 11–13 September correspond perfectly to the observations. The height of the maximum wind is at about 600 hPa, exactly as in the observed profiles. Most importantly, even the maximum values of zonal wind are comparable, being on the order of 20 m s−1. Also, a comparison of the meridional component reveals that the major wave passages are captured. Here, the positive values of υ, corresponding to the northerly component (23 and 27 August, 4 and 13 September) on the order of 5–15 m s−1, indicating a wave passage that is shown in Fig. 3, are clearly reproduced in the GEOS-5 high-resolution analyses.

b. Nondeveloping wave

With the strong confidence provided by the good quality of the GEOS-5 analyses, we investigate the structure and kinematics of the major nondeveloping wave observed in the first part of the campaign. This was the third wave observed during the SOP-3 campaign and was the partial object of flights 4 and 5 (on 25 and 26 August, respectively) during its earlier stages. In this work, we investigate the evolution of this wave in the subsequent days.

From Fig. 1 (right panel) an evident sharp strip of dry air shows the same propagation speed and amplitude of the wave W1 (as identified in section 3a) and is clearly associated with a Saharan air outbreak, which shortly precedes the wave. There is a hint of another dry air outbreak appearing on Fig. 1 at about 20°W and lagging of about one day, but this is not fully captured here because it occurs at a higher latitude.

In Fig. 5, 700-hPa specific humidity and 900-hPa flow are shown together to emphasize the relationship between two different levels in the analysis (0000 UTC 26 August) and across 3 forecast times (24-, 48-, and 72-h forecasts, corresponding to verification times of 0000 UTC 27, 28, and 29 August, respectively). The 700- and 900-hPa levels correspond approximately to the vertical center and the base of the SAL in the eastern Atlantic, with the latter emphasizing the low-level circulation. Despite the strength of the wave and the short-lasting apparent formation of a vertically aligned circulation evident in the analyses from lower levels up to almost 500 hPa (not shown), W1 becomes a nondeveloping wave. Figure 5 shows that dry air at 700 hPa, present at 0000 UTC 26 August to the west of the 900-hPa circulation center (at about 17°N, 21°W), is being advected in 24 h on the top of it (about 30°W). After that, the rotating system becomes elongated and then rapidly evolves into an open wave (48- and 72-h forecasts). This is even more evident while analyzing intermediate time steps (not shown).

For further validation, two SAL analysis images (Figs. 6, 7), courtesy of the Cooperative Institute for Meteorological Satellite Studies (CIMSS, Tropical Cyclone Group, University of Wisconsin) are shown for 0000 UTC 26 August and 0000 UTC 27 August 2006, respectively. These images are obtained by combining GOES West and Meteosat imagery (merging them at about 65°W) and indicate dry and/or dusty air. Although dry air of midlatitude origin cannot be separated from SAL, through the sequence of images, the respective origin can be inferred (more information on the data is available online at http://cimss.ssec.wisc.edu/tropic/real-time/wavetrak/info.sal.m8g10split.html). A very good qualitative agreement can be inferred between these two images and the top two panels of Fig. 5, which refer to the same times (0000 UTC 26 and 27 August, respectively). The approximate center of the wave, W1, appears in Figs. 7, 8 to move from about 22° to 30°W. At 0000 UTC 27 August, moister air appears confined to the north of the rotation center, approximately between 20° and 25°N, in both the CIMSS imagery and in the GEOS-5 forecast.

In Fig. 8, a zonal vertical cross section of specific humidity at 20°N is extracted from the GEOS-5 24-h forecast (displayed in Fig. 5) intersecting the SAL outbreak between 40° and 35°W. The signature of the SAL, as a sharply defined “corridor” of extremely dry air centered at about 38°W (with values less 2 g kg−1, surrounded by values larger than 8 g kg−1), can be seen down to 800 hPa. Remarkable moisture gradients are present on both sides. A zonal temperature anomaly is obtained by subtracting the zonal mean between 90° and 15°W. A well-defined thermal dipole, stronger than any other anomaly in the range of longitudes selected (represented by a warm anomaly between 800 and 500 hPa and a cool anomaly between 850 hPa and the surface, centered at about 35°–40°W), can be seen in correspondence to the dry tongue. Since the cross section cuts across the SAL intrusion, it appears that temperature, in the core of the SAL, is up to 3°C warmer than the surroundings at the same latitude. Most important, in the Fig. 8 cross section there is also evidence of a thin dry air intrusion at about 48°–51°W. This intrusion is not associated to a thermal dipole similar to the one seen in the SAL intrusion, as there is indeed an opposite (slightly warmer in the lower levels) dipole. A step-by-step streamline analysis (not shown) compared with infrared geostationary imagery suggests this is a fine intrusion of midlatitude air, produced by northerly midlevel flow on the eastern side of a large anticyclone in the Atlantic (at 40°N, 30°W on 25 August). The flow then recurves westward and is subject to subsidence. It becomes a dry meridionally elongated filament advected by a predominantly easterly flow, well ahead of the wave. The GEOS-5 0.25° forecasts keep these two air masses well separated.

In Fig. 7, the CIMSS imagery, which condenses GOES-E and Meteosat satellite imagery, also shows this dry air filament at about 45°, which is slightly more to the east than the GEOS-5 but well apart from the dry air associated with SAL, at 35°W. The animation of the CIMSS imagery (not shown) shows that later on, the northern side of this filament gets entangled in the westerly flow, which rapidly shreds it apart.

In Fig. 9, for verification purposes, the same cross section seen in Fig. 8 is extracted from the full-resolution NCEP operational analyses in model sigma levels (0.5σ corresponding roughly to 500 hPa). Despite small-scale differences, the NCEP analysis confirms the thermal structure depicted in the GEOS-5 24-h forecast and the presence of a very well-defined dry intrusion at about 35°–40°W. The presence of a similar thermal dipole seen in Fig. 8 is most remarkable. A positive anomaly of up to about 3.5°C, even stronger than the GEOS-5 forecast, is present in the midtroposphere and a corresponding cool anomaly in the low moist layer. Both anomalies are obtained, as in Fig. 8, by subtracting the 90°–15°W mean. The heating in the lower midtroposphere, indicative of a thermal inversion at the base of the SAL, is in partial agreement with Dunion and Velden (2004), although the values seen in the analyses are not as high as the observed ones (5°–10°C). The relatively new aspect of this analysis is that a negative shallow anomaly also appears in the moist low-level layer at the base of the column, mostly confined below 875 hPa in the GEOS-5 (a little higher in the NCEP analyses). It is possible that this surface cooling is caused by downward shortwave reduction (in agreement with Lau and Kim 2007) but is unrealistically subject to some vertical diffusion by the boundary layer scheme in both the GEOS-5 and the NCEP analyses (more so in the NCEP analyses). Finally, it is important to notice that the thin midlatitude dry air intrusion, clearly seen in the GEOS-5 analyses, maintained in the 24-h GEOS-5 forecasts as a separate air mass (Fig. 8), and confirmed in the CIMSS imagery, is not represented at all in the NCEP analyses (Fig. 9), possibly being merged with the surrounding air mass.

c. Developing wave

The Hovmöller computed in Fig. 1 shows a strong wave moving from about 15° to 40°W between 1200 UTC 11 and 15 September (W2). On the right panel of Fig. 1, it can be seen that Saharan air precedes the wave. However, in contrast with W1, which is preceded and followed by dry air (Fig. 1), values of low humidity before and after the wave are less extended in the W2 case, suggesting that the SAL intrusion was not as strong. This is the wave associated with the system which became Tropical Depression 8 at 1200 UTC 12 September, Tropical Storm Helene at 0000 UTC 14 September, and Hurricane Helene at 1200 UTC 16 September (Brown 2006).

In Fig. 10, the GEOS-5 24-, 48-, and 72-h forecasts (for 0000 UTC 14, 15, and 16 September, respectively), together with the initializing analysis of 0000 UTC 13 September, are shown to illustrate the early development stages of Helene. The 900-hPa circulation shows a clearly defined vortex (at about 12°N, 25°W in the analyses) progressing westward and then recurving northwestward, entangled in corresponding high levels of 700-hPa moisture. The displacement error of the predicted center, with respect to the observed position, is smaller than 2° at all times (not shown).

In Fig. 11, a zonal cross section at 13°N (similar to what was done to investigate W1 in Fig. 8) is produced intersecting both the center of the system W2 (which is a precursor of Hurricane Helene) and the SAL intrusion. The vertically aligned vorticity column at about 30°–33°W, stretching from the surface to the the tropopause, is the signature of Tropical Storm Helene. A positive temperature anomaly of more than 2° at about 35°–45°W in the lower midtroposphere is associated with the same Saharan air outbreak that in the Hovmöller in Fig. 1 reaches 40°W on 14 September. The temperature anomalies within the vorticity column from 30°–33°W, associated with Helene, are not related to the SAL but are instead a result of the tropical storm development (a hint of a warm core above 500 hPa and some cooling possibly due to rainfall evaporation in the lower levels). Comparing this cross section with the one relative to the nondeveloping W1 (Fig. 8), two prominent differences in the SAL can be seen: there is minimal cool anomaly in the lowest levels, and the dry air appears more diluted with less sharp horizontal gradients.

In Fig. 12, for validation purposes we plot the same vertical zonal cross section seen in Fig. 11 but this time from the full-resolution NCEP analyses in sigma levels. The vertical column of vorticity present at about 30°–33°W is the analyzed signature of Tropical Storm Helene. A warm anomaly of up to 3°C at about 40°W, corresponding to the SAL intrusion, is also present in the mid–lower troposphere. Aside from the representation of Helene being weaker in the NCEP than in the GEOS-5 (the latter being closer to observations, since Helene at that time was already a named system), the NCEP and GEOS-5 systems agree on the overall scenario of Helene being preceded by a SAL intrusion, which is however less strong than the intrusion associated with W1.

4. Discussion

a. Possible role of the SAL

W1 and W2 do present some similarities. Among these, the large-scale environmental vertical shear, computed from our analyses as the difference between the vertically integrated 100–200- and 800–900-hPa wind speeds, does not reach 10 m s−1. Other environmental conditions, such as sea surface temperature and low-level vorticity, appear very favorable in both W1 and W2 case (not shown); however, only the latter underwent development becoming Helene. Within our global DAS and modeling framework (in which we do not have the capability to resolve smaller-scale SAL-induced modification of the shear at the 600–700-hPa level), the main difference between W1 and W2 appears to be the intensity of the Saharan air intrusion depicted in analyses and GEOS-5 forecasts. In the model, the temperature dipole associated with the SAL affecting W1 can be followed at each time step and can be considered a possible cause of suppression (increasing static stability in the wave) together with the SAL dryness. In the precursor of Helene, while dry air is present especially on its northern and western quadrants, the low-level negative anomaly is minimal or absent and therefore the effect on static stability is smaller.

The positive anomaly can be simply attributed to the signature of warm air that originated over the Sahara, but the cool anomaly in the lower levels does not have any plausible explanation relying only on transport. There is no source of localized cooler temperatures at that latitude, away from landmass, and in a very homogeneous marine tropical environment. At this time, a possible explanation could be that the low-level cool temperatures are at least partly an indirect evidence of dust amounts. The thermal effect of Saharan mineral dust is a net reduction of downwave shortwave radiation in the near-surface levels and a heating in the lower midtroposphere, corresponding to the core of the SAL.

To support this possibility we present, in Fig. 13, the aerosol optical depth measured from the Moderate Resolution Imaging Spectroradiometer (MODIS), as compared between the periods of 26–29 August and 13–16 September 2006. These periods encompass the snapshots of the circulations in Figs. 5, 10. Notwithstanding some contamination of MODIS data by clouds, the air advected by the large-scale flow in W1 circulation, especially on its northern side, is likely to contain a larger amount of dust than the flow in Fig. 13a, because it has a predominantly northerly component to the north of the approximate wave center. The different amount of dust must have a different impact on the thermal structure of the atmosphere. It appears that the high-resolution analyses and forecasts can represent these different thermal structures and that the GEOS-5 model initialized by such analyses can retain it for 24–72 h, advecting it into the circulation and producing a realistic cyclone dissipation. In the case of the GEOS-5 forecast, the nature of the finite-volume dynamics (Lin 2004) is such that it is a particularly suitable tool to generally maintain sharp gradients by minimizing unrealistic diffusion processes. The finite-volume dynamics have been shown to be very efficient in the midlatitudes, where localized temperature gradients associated with sharp fronts can be very realistically simulated and maintained. This work documents that the same skill can be very useful in the tropics, as well as when dealing with Saharan air.

b. Critical role of data coverage and model resolution

In this work, as stated in the introduction, two sets of forecasts were produced from the GEOS-5 high-quality analyses at a resolution of half and a quarter of a degree. The most prominent result is that the forecasts issued at half-degree resolution, when compared to those at quarter-degree resolution, are not as suitable to retain all the information presented in the analyses and, especially with respect to the SAL, tend to irreparably blur the sharpness of its boundaries. Also, the thermal information is sometimes diluted and spread unrealistically. To provide evidence related to this problem, Fig. 14 shows the same section of Fig. 8 (which was obtained from the 24-h forecast at 0.25°) but extracted from the corresponding forecast at 0.5°. The most prominent difference is that the SAL filament is much broader and has less sharply defined borders in Fig. 14. Moreover, there is no clear separation between the midlatitude dry air at about 50°W and the SAL (at about 35°–42°), as seen in Fig. 8. In addition, the cool anomaly in the lower level is substantially less confined and spread in the lower-resolution forecasts. Finally, Fig. 8 shows a moist column at 25°–30°W, which is associated with W1, as this feature is basically absent from the half-degree forecast in Fig. 14.

It is important to stress that not only inadequate resolution of the model or the data assimilation system can suppress the fine structure of the SAL and its thermal properties. Over data-sparse areas, such as the Sahel and the tropical eastern North Atlantic, the optimal use of satellite information, far from being implemented in any operational system, is probably equally important.

One of the purposes of the GEOS-5 development was to contribute to the creation of a new set of analyses, the MERRA (Bosilovich 2008), by using all the satellite information much more extensively. It is important to stress that the GEOS-5 used in this work is almost the same version adopted for MERRA. The differences between the NCEP operational analyses and the GEOS-5 analyses (which are of comparable resolution) evident, for example, between Figs. 11 and 12 in terms of sharpness of moisture boundaries and in the intensity of Helene, are probably a result of the different assimilation of satellite data.

5. Conclusions

This work investigates AEW development during the SOP-3 phase of the NAMMA campaign with the aid of the NASA GEOS-5 data assimilation and forecasting system. A data assimilation run is performed to produce a high-quality set of global analyses that make heavy use of satellite information. From these improved analyses, two sets of thirty-one 5-day forecasts at two different resolutions are produced. Both sets of GEOS-5 forecasts are validated against operational NCEP analyses, calculating anomaly correlations on 500-hPa geopotential height up to day 5, and evaluating forecast track skill for tropical cyclones up to day 3. The forecast skill of the system is found reasonable.

A very strong nondeveloping wave (W1) is selected and compared with the wave W2 that eventually developed into Hurricane Helene. Analyses and forecasts document the presence of a strong temperature dipole associated with a Saharan air intrusion, which initially precedes the core of the wave W1 of about 10°. This dipole is gradually advected into the circulation of the wave and may have contributed to the inhibition of further development by increasing static stability. No such dipole is found for the Saharan air that intruded in Helene’s precursor. MODIS aerosol optical thickness shows that the SAL outbreak associated with the nondeveloping case contained a higher amount of dust. The lower-tropospheric cooling associated with the strong Saharan air outbreak suggests that the high-resolution global simulations and analyses can capture not only the SAL intrusion, but also possibly part of the thermal effect that can result from the reduction of downward shortwave caused by large amounts of Saharan dust. When this thermal structure is well represented in the initial conditions, a high-resolution global model can have more skill in predicting the impact of SAL on a tropical cyclone.

At this time, the role of the SAL on tropical cyclogenesis is still subject to a vigorous debate. Apparently, contrasting evidence has been presented by Dunion and Velden (2004), in which a cyclogenesis suppression is suggested and Jenkins et al. (2008), which provides evidence of rainband invigoration attributed to dust. However, the subjects of these two studies are quite different and their results do not necessarily contradict each other but may indeed both prove valid at different levels. In fact, Dunion and Velden (2004) explained that SAL’s dry air and enhanced shear tend to suppress convection, and that SAL’s dust-induced warming can strengthen the temperature inversion at the SAL base. However, microphysical effects were not clearly known at the time of their study. The more recent study by Jenkins et al. (2008) is focused on microphysical effects and shows that the microphysical effects of SAL’s dust can enhance convection by acting on condensation nuclei. The two studies are not incompatible but they do serve to highlight the great complexity of the problem.

This work, with the aid of high-quality global analyses and modeling capabilities, provides some further evidence that the thermodynamic properties associated with the SAL are detrimental to tropical cyclogenesis, in agreement with Dunion and Velden (2004). However, it is a limited sample and more experiments are needed. It is clear that the details of possible radiative effects by dust on cyclogenesis will have to be investigated using a high-resolution GCM with interactive dust aerosol. This capability is not yet available anywhere within the scientific community, but this work aims to provide a set of requirements to plan such research. In particular, we have emphasized how the resolution is a critical parameter in being able to represent with some degree of confidence the fine sharpness of SAL intrusions in tropical circulations. No simulation can reasonably hope to investigate the complexity of the mechanisms involved with the interaction of the SAL and tropical cyclones at a resolution lower than a quarter of a degree. This is a demanding standard for global modeling but the evidence provided in this paper suggests that it is an inescapable necessity.

Finally, in this work it is shown that even at the respectable resolution of the current operational analyses, the inadequate use of satellite data, especially over data-sparse areas, may be a serious limitation. On the other hand, optimizing the satellite information adopted and most importantly improving the quality of the coverage, may bring substantial improvement to our understanding of processes associated with tropical cyclogenesis.

Acknowledgments

The authors acknowledge support from Dr. Ramesh Kakar, NASA Headquarters, through the NAMMA Project, and the use of NASA high-end computing resources. Thanks are due to three anonymous reviewers for helpful comments.

REFERENCES

  • Asnani, G. C., 2005: Tropical Meteorology. Vol. 3, Indian Institute of Tropical Meteorology, 735 pp.

  • Atlas, R., and Coauthors, 2005: Hurricane forecasting with the high-resolution NASA finite volume general circulation model. Geophys. Res. Lett., 32 , L03807. doi:10.1029/2004GL021513.

    • Search Google Scholar
    • Export Citation
  • Bosilovich, M. G., 2008: NASA’s modern era retrospective-analysis for research and applications: Integrating earth observations. [Available online at http://www.earthzine.org/2008/09/26/nasas-modern-era-retrospective-analysis].

    • Search Google Scholar
    • Export Citation
  • Bosilovich, M. G., and Coauthors, 2006: NASA’s modern era retrospective-analysis for research and applications. U.S. CLIVAR Variations, U.S. CLIVAR Project Office, Washington, DC, No. 2, 5–8.

    • Search Google Scholar
    • Export Citation
  • Brown, D. P., 2006: Tropical cyclone report: Hurricane Helene (AL082006) 12–24 September 2006. [Available online at http://www.nhc.noaa.gov/pdf/TCR-AL082006_Helene.pdf].

    • Search Google Scholar
    • Export Citation
  • Burpee, R. W., 1974: Characteristics of North African easterly waves during the summers of 1968 and 1969. J. Atmos. Sci., 31 , 15561570.

    • Search Google Scholar
    • Export Citation
  • Carlson, T. N., , and J. M. Prospero, 1972: The large-scale movements of Saharan air outbreaks over the northern equatorial Atlantic. J. Appl. Meteor., 11 , 283297.

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

    • Search Google Scholar
    • Export Citation
  • Hendricks, A. E., , M. T. Montgomery, , and C. A. Davis, 2004: The role of “vortical” hot towers in the formation of Tropical Cyclone Diana (1984). J. Atmos. Sci., 61 , 12091232.

    • Search Google Scholar
    • Export Citation
  • Hsieh, J-S., , and K. Cook, 2005: Generation of African easterly wave disturbances: Relationship to the African easterly jet. Mon. Wea. Rev., 133 , 13111327.

    • Search Google Scholar
    • Export Citation
  • Jenkins, G. S., , and A. Pratt, 2008: Saharan dust, lightning and tropical cyclones in the eastern tropical Atlantic during NAMMA-06. Geophys. Res. Lett., 35 , L12804. doi:10.1029/2008GL033979.

    • Search Google Scholar
    • Export Citation
  • Jenkins, G. S., , A. Pratt, , and A. Heymsfield, 2008: Possible linkages between Saharan dust and tropical cyclone rain band invigoration in the eastern Atlantic during NAMMA-06. Geophys. Res. Lett., 35 , L08815. doi:10.1029/2008GL034072.

    • Search Google Scholar
    • Export Citation
  • Kiladis, G. N., , C. D. Thorncroft, , and N. M. J. Hall, 2006: Three-dimensional structure and dynamics of African easterly waves. Part I: Observations. J. Atmos. Sci., 63 , 22122230.

    • Search Google Scholar
    • Export Citation
  • Lau, K. M., , and K-M. Kim, 2007: Cooling of the Atlantic by Saharan dust. Geophys. Res. Lett., 34 , L23811. doi:10.1029/2007GL031538.

  • Lin, S-J., 2004: A “vertically Lagrangian” finite-volume dynamical core for global models. Mon. Wea. Rev., 132 , 22932307.

  • Moorthi, S., , and M. J. Suarez, 1992: Relaxed Arakawa-Schubert. A parameterization of moist convection for general circulation models. Mon. Wea. Rev., 120 , 9781002.

    • Search Google Scholar
    • Export Citation
  • Reale, O., , J. Terry, , M. Masutani, , E. Andersson, , L. P. Riishojgaard, , and J. C. Jusem, 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
  • Reale, O., , J. Susskind, , R. Rosenberg, , E. Brin, , E. Liu, , L. P. Riishojgaard, , J. Terry, , and J. C. Jusem, 2008: Improving forecast skill by assimilation of quality-controlled AIRS temperature retrievals under partially cloudy conditions. Geophys. Res. Lett., 35 , L08809. doi:10.1029/2007GL033002.

    • Search Google Scholar
    • Export Citation
  • Reale, O., , W. K. Lau, , J. Susskind, , E. Brin, , E. Liu, , L. P. Riishojgaard, , M. Fuentes, , and R. Rosenberg, 2009: AIRS impact on the analysis and forecast track of Tropical Cyclone Nargis in a global data assimilation and forecasting system. Geophys. Res. Lett., 36 , L06812. doi:10.1029/2008GL037122.

    • Search Google Scholar
    • Export Citation
  • Rienecker, M. M., and Coauthors, 2008: The GEOS-5 data assimilation system. Documentation Versions 5.0.1, 5.1.0, and 5.20. NASA Tech. Rep. Series on Global Modeling and Data Assimilation 27, NASA/TM-2008-104606, 1–118. [Available online at http://gmao.gsfc.nasa.gov/pubs/docs/Rienecker369.pdf].

    • Search Google Scholar
    • Export Citation
  • Shen, B-W., , R. Atlas, , O. Reale, , S-J. Lin, , J-D. Chern, , J. Chang, , C. Henze, , and J-L. Li, 2006: Hurricane forecasts with a global mesoscale-resolving model: Preliminary results with Hurricane Katrina (2005). Geophys. Res. Lett., 33 , L13813. doi:10.1029/2006GL026143.

    • Search Google Scholar
    • Export Citation
  • Susskind, J., 2007: Improved atmospheric soundings and error estimates from analysis of AIRS/AMSU data. Atmospheric and Environmental Remote Sensing Data Processing and Utilization III: Readiness for GEOSS, International Society for Optical Engineering (SPIE Proceedings, Vol. 6684), doi:10.1117/12.734336.

    • Search Google Scholar
    • Export Citation
  • Susskind, J., , C. Barnet, , J. Blaisdell, , L. Iredell, , F. Keita, , L. Kouvaris, , G. Molnar, , and M. Chahine, 2006: Accuracy of geophysical parameters derived from atmospheric infrared sounder/advanced microwave sounding unit as a function of fractional cloud cover. J. Geophys. Res., 111 , D09S17. doi:10.1029/2005JD006272.

    • Search Google Scholar
    • Export Citation
  • Tao, W-K., , S. Long, , J. Simpson, , C-H. Sui, , B. Ferrier, , and M-D. Chou, 1996: Cloud-radiation mechanisms associated with a tropical and a mid-latitude squall line. J. Atmos. Sci., 53 , 26242651.

    • Search Google Scholar
    • Export Citation
  • Wu, W-S., , R. J. Purser, , and D. F. Parrish, 2002: Three-dimensional variational analysis with spatially inhomogeneous covariances. Mon. Wea. Rev., 130 , 29052916.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

(left) Hovmöller of 700-hPa relative vorticity (s−l) and (right) relative humidity from the GEOS-5 analyses, latitudinally averaged (12°–20°N) and covering the SOP-3 period from 15 August to 15 September. The waves, which are the focus of this study, are W1 (0°−40°W longitude, 22–29 August) and W2 (15°–40°W, 12–16 September).

Citation: Journal of the Atmospheric Sciences 66, 12; 10.1175/2009JAS3123.1

Fig. 2.
Fig. 2.

As in Fig. 1, but for the NCEP operational analyses.

Citation: Journal of the Atmospheric Sciences 66, 12; 10.1175/2009JAS3123.1

Fig. 3.
Fig. 3.

Sequence of the observed profiles of the (top) zonal and (bottom) meridional components of the wind, taken daily at 1200 UTC at the Cape Verde station (latitude 14.92°61′11″N, longitude 23.49°46′11″W, and altitude 85 m) during the SOP-3 campaign.

Citation: Journal of the Atmospheric Sciences 66, 12; 10.1175/2009JAS3123.1

Fig. 4.
Fig. 4.

As in Fig. 3, but extracted from the GEOS-5 analyses. Sequence of the observed profiles (at each synoptic time) of the (top) zonal and (bottom) meridional components of the wind, taken at the same latitude and longitude of the station.

Citation: Journal of the Atmospheric Sciences 66, 12; 10.1175/2009JAS3123.1

Fig. 5.
Fig. 5.

(top left) GEOS-5 (at 0.25°) 700-hPa specific humidity (g kg−1) and 900-hPa wind (streamlines) in the GEOS-5 analyses for 0000 UTC 26 Aug, and relative to the 24-, 48-, and 72-h forecasts for 0000 UTC 27, 28, and 29 Aug.

Citation: Journal of the Atmospheric Sciences 66, 12; 10.1175/2009JAS3123.1

Fig. 6.
Fig. 6.

SAL analysis (image courtesy of the CIMSS Tropical Cyclone Group, University of Wisconsin) for 0000 UTC 26 Aug.

Citation: Journal of the Atmospheric Sciences 66, 12; 10.1175/2009JAS3123.1

Fig. 7.
Fig. 7.

As in Fig. 6, but for 0000 UTC 27 Aug.

Citation: Journal of the Atmospheric Sciences 66, 12; 10.1175/2009JAS3123.1

Fig. 8.
Fig. 8.

GEOS-5 (at 0.25°) with a zonal vertical cross section of specific humidity (g kg−1, shaded) and temperature anomaly [°C; red (blue) contour, positive (negative), subtracting the zonal mean between 90° and 15°W] at 20°N for 0000 UTC 27 Aug, 24-h forecast initialized at 0000 UTC 26 Aug. W1 is located at about 25°–30°W.

Citation: Journal of the Atmospheric Sciences 66, 12; 10.1175/2009JAS3123.1

Fig. 9.
Fig. 9.

As in Fig. 8, but extracted from NCEP full-resolution analyses in model levels for validation purposes. The vertical dimension is only approximately comparable with Fig. 8, because the spacing between model levels and pressure levels is different.

Citation: Journal of the Atmospheric Sciences 66, 12; 10.1175/2009JAS3123.1

Fig. 10.
Fig. 10.

As in Fig. 5, but for 0000 UTC 13 Sep and relative to the 24-, 48- and 72-h forecasts for 0000 UTC 14, 15, and 16 Sep.

Citation: Journal of the Atmospheric Sciences 66, 12; 10.1175/2009JAS3123.1

Fig. 11.
Fig. 11.

As in Fig. 8, but showing the cyclonic vorticity (contours at l, 2, 3 × 10−4 s−1 at 20°N for 0000 UTC 14 Sep; darker red) and for the 24-h forecast initialized at 0000 UTC 13 Sep. The developing TS Helene, identifiable by the vertical column of strong cyclonic vorticity, is located at about 32°W.

Citation: Journal of the Atmospheric Sciences 66, 12; 10.1175/2009JAS3123.1

Fig. 12.
Fig. 12.

As in Fig. 11, but extracted from NCEP full-resolution analyses in model levels for validation purposes.

Citation: Journal of the Atmospheric Sciences 66, 12; 10.1175/2009JAS3123.1

Fig. 13.
Fig. 13.

Four-day mean MODIS aerosol optical depth for (a) 26–29 Aug 2006 and (b) 13–16 Sep 2006. The MODIS data show much larger amount of dust for the first wave, which is the nondeveloping one. The 900-hPa flow at (a) 0000 UTC 26 Aug, to indicate W1 position, and the track of Helene, (b) every 6 h from 0000 UTC 13 Sep to 1800 UTC 18 Sep, are superimposed. Solid dots indicate hurricane strength.

Citation: Journal of the Atmospheric Sciences 66, 12; 10.1175/2009JAS3123.1

Fig. 14.
Fig. 14.

As in Fig. 8, but produced from GEOS-5 at the lower resolution of 0.5°.

Citation: Journal of the Atmospheric Sciences 66, 12; 10.1175/2009JAS3123.1

Save