Article Type:
Research Article

Mesoscale Variations of Water Vapor Inferred from the Millimeter-Wave Imaging Radiometer during TOGA COARE

Merritt N. Deeter Program in Atmospheric and Oceanic Sciences, University of Colorado, Boulder, Colorado

Search for other papers by Merritt N. Deeter in
Current site
Google Scholar
PubMed Close
and
K. Franklin Evans Program in Atmospheric and Oceanic Sciences, University of Colorado, Boulder, Colorado

Search for other papers by K. Franklin Evans in
Current site
Google Scholar
PubMed Close
Print Publication:
01 Feb 1997
DOI:
https://doi.org/10.1175/1520-0450(1997)036<0183:MVOWVI>2.0.CO;2
Page(s):
183–188
Full access

Abstract

Autocorrelation analysis was applied to data acquired by NASA’s Millimeter-wave Imaging Radiometer during the 1992–93 Tropical Ocean Global Atmosphere Coupled Ocean–Atmosphere Response Experiment to quantify the spatial variability of tropical mid- and upper-tropospheric water vapor. Scenes including thin cirrus (as confirmed by lidar measurements) are associated with greater spatial variability of relative humidity (RH) than clear-sky scenes. Typical clear-sky RH spatial variabilities range between 2% and 3%, whereas values for some cirrus scenes are well above 5%. For both cirrus and clear-sky scenes, most of the RH spatial variability is on spatial scales of less than 10 km. Both the amplitude and spatial scale of observed brightness temperature variabilities are in fair quantitative agreement with a model based on the effects of adiabatic vertical displacements generated by boundary layer–forced gravity waves.

Corresponding author address: K. Franklin Evans, Program in Atmospheric and Oceanic Sciences, University of Colorado, Campus Box 311, Boulder CO 80309.

Abstract

Autocorrelation analysis was applied to data acquired by NASA’s Millimeter-wave Imaging Radiometer during the 1992–93 Tropical Ocean Global Atmosphere Coupled Ocean–Atmosphere Response Experiment to quantify the spatial variability of tropical mid- and upper-tropospheric water vapor. Scenes including thin cirrus (as confirmed by lidar measurements) are associated with greater spatial variability of relative humidity (RH) than clear-sky scenes. Typical clear-sky RH spatial variabilities range between 2% and 3%, whereas values for some cirrus scenes are well above 5%. For both cirrus and clear-sky scenes, most of the RH spatial variability is on spatial scales of less than 10 km. Both the amplitude and spatial scale of observed brightness temperature variabilities are in fair quantitative agreement with a model based on the effects of adiabatic vertical displacements generated by boundary layer–forced gravity waves.

Corresponding author address: K. Franklin Evans, Program in Atmospheric and Oceanic Sciences, University of Colorado, Campus Box 311, Boulder CO 80309.

Introduction

Water vapor exerts a powerful positive feedback to global warming. The magnitude of this feedback, however, depends to a large extent on the distribution of the water vapor generated in response to climate change. More subtly, the mesoscale variability of water vapor plays an important role in cloud formation and, potentially, in nonlinear processes such as radiative transfer. Observations of mesoscale variability of water vapor at different levels in the atmosphere may contain useful information about dynamic processes such as gravity waves. For microwave remote sensing of clouds, water vapor variability generates a source of measurement uncertainty that must be quantified. Thus, mean water vapor statistics are often insufficient to completely specify water vapor’s various effects.

In this note, Millimeter-wave Imaging Radiometer (MIR) (Racette et al. 1996; Wang et al. 1995) data are used to quantify the horizontal variability of tropical water vapor in the middle and upper troposphere. While the 6.5-μm channels of geostationary satellites have been used to observe upper-tropospheric humidity globally, the analysis has been in terms of means (Soden and Bretherton 1993; Schmetz et al. 1995) or large-scale variations (Udelhofen and Hartmann 1995). Autocorrelation analyses of MIR brightness temperatures observed during the 1992–93 Tropical Ocean Global Atmosphere Coupled Ocean–Atmosphere Response Experiment (TOGA COARE) (Webster and Lukas 1992) in the western Pacific Ocean are compared for scenes free of clouds and scenes including thin cirrus. A radiative transfer model is used to determine the relationship between variations in brightness temperature and variations in relative humidity.

MIR radiative modeling

During TOGA COARE the MIR instrument was onboard the National Aeronautics and Space Administration (NASA) ER-2 research aircraft. At that time, the MIR recorded brightness temperature values at the frequencies 89, 150, 183.3 ± 1, 3, 7, and 220 GHz. Only the three MIR channels with the greatest sensitivity to water vapor (those centered on the 183.3-GHz absorption line) are analyzed in this work. The MIR scans across the flight track while sampling the microwave signals at 57 beam positions (“tracks”) every 3 s. The beamwidth is 3.5°, giving an upper-troposphere resolution of approximately 0.8 km from the 20-km ER-2 aircraft level.

The MIR measures the upwelling microwave thermal emission from the atmosphere, which can be interpreted as an equivalent blackbody brightness temperature Tb. Since water vapor is the principal absorber at the frequencies analyzed here, Tb values are a measure of both relative humidity (RH) and temperature.

The upwelling brightness temperatures from tropical atmospheric profiles are modeled here to establish a relationship between Tb’s and humidity. Given an atmospheric profile (i.e., temperature, pressure, and water vapor density versus altitude) and the absorption coefficients of the primary atmospheric absorbers as a function of frequency, temperature, and pressure (Liebe 1989), it is straightforward to numerically integrate the radiative transfer equation and calculate a theoretical brightness temperature for each of the MIR frequencies.

No accounting for the effects of liquid or ice clouds was included in this model since we chose to consider only scenes characterized by clear sky or thin cirrus. For cirrus clouds, our calculations show that the ratio of scattering-induced brightness temperature depression at 183 GHz to visible optical depth is less than 1 K. Therefore, scattering effects can reasonably be neglected for optically thin cirrus clouds. Properties of the surface are not relevant for the MIR channels used here because the transmission from the surface in the Tropics is negligible.

The relative humidity profiles used as the input to the radiative transfer model are shown in Fig. 1. Temperature and RH profiles were generated by averaging 2400 soundings recorded throughout the TOGA COARE experiment (Mapes and Zuidema 1996). The sounding data also generated profiles of standard deviations of temperature and RH. Temperature standard deviation values are small (on the order of 1 K or less) throughout the troposphere. On the other hand, as shown in the figure, the soundings data reflect large temporal variability in RH. In addition to the mean RH profile, the figure shows RH profiles formed by computing, at each altitude, the RH standard deviation added to and subtracted from the mean RH. Subsequently, these will be referred to as the “moist” and “dry” RH profiles. For comparison, Fig. 1 also includes the standard McClatchey tropical RH profile (Ellingson et al. 1991).

The forward radiative transfer model vertically integrates contributions to the observed microwave radiance. The model is, therefore, capable of quantifying the differential contributions to the measured signal in the form of the weighting functions shown in Fig. 2. The figure also includes the TOGA COARE mean temperature profile, which, along with the weighting functions, determines the observed brightness temperature. Since the absorptivity of water vapor decreases with increasing detuning Δf (Δf = f − f0) around the water vapor absorption line center (f0 = 183.3 GHz), the 183.3±1-GHz channel senses a higher region of the troposphere than the 183.3±3-GHz channel, whereas the 183.3±7-GHz channel is primarily sensitive to a still lower portion of the atmosphere. A comparison of the weighting functions for the mean and dry RH profiles in the figure shows how decreasing the RH pushes the weighting functions to lower altitudes, which results in increased brightness temperatures.

In principle, it is possible to obtain the water vapor profile from the MIR data (Al-Khalaf et al. 1994; Wang et al. 1995). Such inversion problems usually require one to model the atmosphere as a layered structure, where the number of layers is less than or equal to the number of microwave channels. This problem is complicated by the nonlinearity and potential instability of the inversion. Rather than solving this general problem, a simpler RH retrieval scheme involves merely assuming some nominal profile and then calculating the first-order relationship between brightness temperature variations and RH variations.

Scene selection

Only scenes either apparently lacking clouds or including only optically thin cirrus clouds were considered for analysis. Precipitation and cloud particles can both absorb and scatter microwave radiation and could thus mask the water vapor signal. For this study, nadir-track lidar backscatter data recorded simultaneously with the MIR images were the primary source for identifying both cloudless and cirrus scenes. The Cloud Lidar System (CLS), which accompanied the MIR on board the NASA ER-2 aircraft, recorded backscatter data at 1064-nm wavelength (Spinhirne and Hart 1990). Scenes exhibiting no significant lidar backscatter signal above 2 km were considered to be free of cirrus. Cirrus scenes were identified by the presence of small (but nonzero) backscatter signals between 10 and 15 km, the general lack of any signal between 2 and 10 km, and a strong surface signal, indicating small cirrus optical depth. As indicated by the weighting functions plotted in Fig. 2, boundary layer clouds below 2 km should not significantly contribute to the MIR signals. Convective systems with cloud tops of up to 4 km (thus penetrating the weighting function for the 183.3±7-GHz channel) were observed in a few of the scenes; however, these systems represented only a very small fraction of the entire dataset. There were no indications of cirrus clouds below 10 km.

Visible and infrared imagery recorded by the MODIS (Moderate Resolution Imaging Spectrometer) Airborne Simulator (MAS) were also considered when available. MAS images were valuable as a two-dimensional confirmation of the scene characteristics, as opposed to the more sensitive lidar data, which only provided information along the nadir track.

From a total of 12 ER-2 sorties flown as part of the TOGA COARE campaign between 11 January and 21 February 1993, four clear-sky scenes and six cirrus scenes were selected. The majority of rejected MIR scenes were excluded either because of cloudiness (much of the time the ER-2 was over deep convective systems) or because the lidar was not operational. The relevant temporal and geographic information for these scenes is summarized in Table 1. The scenes typically vary in length from about 50 up to a few hundred kilometers.

MIR data analysis

Mean statistics

For each scene and water vapor channel, brightness temperature values along each of the 57 tracks were averaged. The resulting mean brightness temperatures recorded by the 183.3±1-GHz channel for both the clear-sky and cirrus scenes are compared with model predictions for the mean, dry, and moist RH profiles in Fig. 3. The dotted curves in the figure show the model predictions, with the lowest curve corresponding to the moist RH profile, the middle curve representing the mean RH profile, and the top curve representing the dry RH profile. As expected for clear-sky scenes, the mean brightness temperature curves all fall well above (i.e., on the dry side of) the theoretical curve for the mean RH profile. In contrast, the mean brightness temperature curves for the cirrus scenes are more widely distributed around the dry and mean RH profiles.

Autocorrelation analysis

Water vapor spatial variability was deduced from the MIR brightness temperature data by autocorrelation analysis. We begin by defining X2(Δx) as the mean-square error between the brightness temperature values of two points in a scene separated by Δx along a particular track:
X2xTbxTbxx2
where the angle brackets indicate an average taken first over each value of x in each track and then over all 57 tracks. Assuming that X2(Δx) is formed by the addition of uncorrelated components related to the instrumental noise (which should not be dependent on x or Δx) and the “geophysical variability” GTB(Δx), we have
X2xσ2IδxG2TBx
where σ2I is the variance of the instrumental noise, and δ(Δx) = 1 for Δx = 0 and is 0 otherwise; GTB(Δx) thus represents the rms brightness temperature deviation due to geophysical (i.e., noninstrumental) effects for either of two points separated by a distance Δx (measured along track—i.e., parallel to the flight path). Both terms on the right-hand side of (2) include factors of 2 since σ2I and GTB(Δx) each refer only to deviations associated with a single measurement at a single point. Expanding (1) and defining the scene autocorrelation function R(Δx) as
RxTbxTbxx
it can be shown that
G2TBxRRxσ2Iδx
This equation also forms the basis for determination of σ2I. Brightness temperature variability in scenes that lack significant geophysical variability [GTB(Δx) ≈ 0] arises primarily because of instrumental noise. Thus, comparison of R(0) with R(δx) (where δx is the minimum lag) for scenes exhibiting low geophysical variability directly yields estimates of σ2I. Values of σI estimated by this technique were 1.3 K for the 183.3±1-GHz channel and 1.0 K for both the 183.3±3-GHz and the 183.3±7-GHz channels.
Brightness temperature spatial variability, characterized by GTB(Δx), is caused by humidity and temperature spatial variability. Based on the TOGA COARE radiosondes, we ignore the temperature variability, so that to first order the relationship between Tb and RH variability can be expressed by
GRHxCRHGTBx
where GRH(Δx) is termed the RH geophysical variability (%) and CRH is the conversion factor (% K−1). Because of the nonlinear dependence of brightness temperature on RH, accurate determination of CRH depends on knowledge of the mean (“baseline”) RH characteristic of a particular scene. The conversion factor CRH is calculated separately for each scene and channel. It is determined from a quadratic fit of the modeled brightness temperatures versus relative humidity generated from the appropriate TOGA COARE profile. The relative humidity profiles are generated by adding or subtracting a fixed RH at every level from the mean TOGA COARE profile. The scene-averaged brightness temperature (appropriately adjusted for the nadir track) is thus translated to a mean scene relative humidity. If this scene-averaged RH departs more than 20% from the mean TOGA COARE profile, then either the moist or dry profile is used instead. The conversion factor CRH is the slope of the quadratic fit at the scene-averaged brightness temperature.

The RH variability for a MIR channel represents the layer around where the weighting function peaks. Plots of GRH(Δx) for both the clear-sky and cirrus scenes for the three water vapor channels are shown in Fig. 4. The variability is shown out to half the scene length, so that the autocorrelation contains a sufficient number of independent samples. Generally, GRH grows rapidly from zero lag up to lag values of about 5 km (the smallest lag, one pixel, is about 0.6 km). After about 5 km the variability is either constant or increases only slowly. The approximately linearly increasing GRH for some of the longer scenes is caused by a synoptic-scale water vapor gradient, rather than more complex variations. For both the 183.3±1-GHz and 183.3±3-GHz channels, most of the cirrus scenes exhibit significantly greater RH variability values than the clear-sky scenes. The cirrus scenes have higher brightness temperature variability, and their higher RH causes the conversion factor CRH to be larger. For the 183.3±7-GHz channel, the cirrus and clear-sky scenes exhibit much more similar statistics, indicating that cirrus clouds are not associated with increased RH variability in the lower to middle troposphere. Tabulated values of mean autocorrelation statistics for brightness temperature and relative humidity at lag values of 10, 20, and 40 km are listed in Table 2. The values listed are derived from a five-bin boxcar filter applied to the original autocorrelations, which are then averaged over scenes weighted by the length of the scene. Such filtering, along with the averaging that occurs both within the individual autocorrelations and then within each scene, generates variability values that are much more accurate than the instrumental noise values corresponding to single-point measurements.

The spatial variability of RH might be related to the spatial scale of adiabatic vertical displacements, such as those generated by gravity waves. Here we develop a simple model for the effect of such displacements by considering the effects of vertically shifting an atmospheric column. For simplicity, it is assumed that the vertical displacement Δz is uniform within the nominal weighting function profile of a particular channel. Unsaturated parcels subjected to such a displacement will cool according to the dry-adiabatic lapse rate Γd. Since water vapor (the primary absorber for all of the 183-GHz channels) is displaced along with the entire column, the weighting functions move by the same distance Δz (ignoring the small change in absorption coefficient with temperature and pressure). Relative to neighboring unshifted atmospheric columns, then, thermal emissions from this column will radiate according to a temperature shifted from the nominal temperature profile by the product of Δz and the dry-adiabatic lapse rate Γd. Thus, a brightness temperature variation can be translated to a vertical displacement by dividing by 0.0098 K m−1.

With this model, the brightness temperature variations caused by gravity wave displacements can be estimated. A numerical modeling study (Hauf and Clark 1989) of gravity waves generated by boundary layer convection had rms potential temperature variations at 10.5 km on the order of 0.2 K. For a value of potential temperature lapse rate (i.e., the difference between the environmental and dry-adiabatic lapse rates) in the upper troposphere of 2.5 K km−1, the amplitude of these vertical displacements would be on the order of 80 m rms. The corresponding expected brightness temperature shift ΔTb for displacements of this amplitude would be 0.8 K rms. This value is quite close to the observed GTB for the upper-troposphere channel at 10 km. The horizontal spatial scale of autocorrelations (Fig. 4) is consistent with the scale expected from gravity waves (Hauf and Clark 1989). The transient nature of convective boundary layer–forced gravity waves results in waves characterized by long vertical wavelengths, such that wave displacements should be uniform within the layer defined by the MIR weighting function. The match in the amplitude of brightness temperature variations and the relatively constant autocorrelation beyond the 5-km scale suggest that much of the mesoscale variability in upper-tropospheric water vapor in the Tropics, outside of deep convection, may be due to gravity wave activity.

Conclusions

Tropospheric water vapor exhibits large temporal and spatial variability. Accurate measurements of this variability are required to improve our understanding of cloud physics, monitor the global hydrologic cycle, and establish the accuracy of microwave remote sensing retrieval algorithms. Autocorrelation analysis of brightness temperature images at water vapor line (183 GHz) frequencies was used to study tropical water vapor spatial variability. Lidar data were used to identify both clear-sky scenes and scenes including thin cirrus clouds. Autocorrelations of brightness temperature data were scaled in terms of RH by a simple radiative transfer model, based on mean temperature and RH profiles derived from soundings recorded during the TOGA COARE experiment. While the number of scenes is small, the total area covered by the MIR data used here is not insignificant, and the results should be indicative of tropical water vapor variability outside of deep convection.

The upper-tropospheric relative humidity variations at 10–50-km spatial scale are in the range of 1%–5% rms. Cirrus clouds are associated with increased RH variability above about 7 km. The findings suggest that most of the mid- to upper-tropospheric water vapor variability in the Tropics occurs at a small spatial scale (less than 10 km). Since the brightness temperature variations analyzed here are consistent with those modeled from reported characteristics of convective boundary layer–forced gravity waves, such waves might be a primary source of RH variability. If this hypothesis is verified by further research, analyses of brightness temperature spatial variability might provide a new means for detecting and studying gravity waves.

Acknowledgments

The authors thank Jim Spinhirne of NASA/Goddard for help with the CLS data, Jim Wang of NASA/Goddard for help in obtaining and reading the MIR data, and Brian Mapes for his mean TOGA COARE atmospheric profiles. We thank Terry Clark of the National Center for Atmospheric Research for fruitful discussions about the effects of gravity waves. The Process Studies Program Office and the Distributed Active Archive Center at Goddard Space Flight Center provided support (sponsored by NASA’s Mission to Planet Earth Program) for the production and distribution of the NASA TOGA COARE data used in this paper. Financial support for this research was provided by NASA FIRE-III Grant NAG1-1702.

REFERENCES

  • Al-Khalaf, A. K., D. T. Conlee, and T. T. Wilheit, 1994: Retrieval of atmospheric water vapor profiles from SSM/T-2 microwave radiances. Preprints, Seventh Conf. on Satellite Meteorology and Oceanography, Monterey, CA, Amer. Meteor. Soc., 106–109.

  • Ellingson, R. G., J. Ellis, and S. Fels, 1991: The intercomparison of radiation codes used in climate models: Long wave results. J. Geophys. Res.,96, 8929–8953.

  • Hauf, T., and T. L. Clark, 1989: Three-dimensional numerical experiments on convectively forced internal gravity waves. Quart. J. Roy. Meteor. Soc.,115, 309–333.

  • Liebe, H. J., 1989: MPM—An atmospheric millimeter wave propagation model. Int. J. Infrared Millimeter Waves,10, 631–650.

  • Mapes, B. E., and P. Zuidema, 1996: Radiative–dynamical consequences of dry tongues in the tropical troposphere. J. Atmos. Sci.,53, 620–638.

  • Racette, P., R. F. Adler, J. R. Wang, A. J. Gasiewski, D. M. Jackson, and D. S. Zacharias, 1996: An airborne Millimeter-wave Imaging Radiometer for cloud, precipitation, and atmospheric water vapor studies. J. Atmos. Oceanic Technol.,13, 610–619.

  • Schmetz, J., and Coauthors, 1995: Monthly mean large-scale analyses of upper-tropospheric humidity and wind field divergence derived from three geostationary satellites. Bull. Amer. Meteor. Soc.,76, 1578–1584.

  • Soden, B. J., and F. P. Bretherton, 1993: Upper tropospheric relative humidity from the GOES 6.7 μm channel: Method and climatology for July 1987. J. Geophys. Res.,98, 16 669–16 688.

  • Spinhirne, J. D., and W. D. Hart, 1990: Cirrus structure and radiative parameters from airborne lidar and spectral radiometer observations: The 28 October 1986 FIRE study. Mon. Wea. Rev.,118, 2329.

  • Udelhofen, P. M., and D. L. Hartmann, 1995: Influence of tropical cloud systems on the relative humidity in the upper troposphere. J. Geophys. Res.,100, 7423–7440.

  • Wang, J. R., S. H. Melfi, P. Racette, D. N. Whitemen, L. A. Chang, R. A. Ferrare, K. D. Evans, and F. J. Schmidlin, 1995: Simultaneous measurements of atmospheric water vapor with MIR, Raman lidar, and rawinsondes. J. Appl. Meteor.,34, 1595–1607.

  • Webster, P. J., and R. Lukas, 1992: TOGA COARE: The Coupled Ocean–Atmosphere Response Experiment. Bull. Amer. Meteor. Soc.,73, 1377–1416.

Fig. 1.
Fig. 1.

Relative humidity profiles derived from TOGA COARE soundings compared with the McClatchey tropical standard atmosphere RH profile.

Citation: Journal of Applied Meteorology 36, 2; 10.1175/1520-0450(1997)036<0183:MVOWVI>2.0.CO;2

Fig. 2.
Fig. 2.

Weighting functions for the MIR water vapor channels derived for both the mean and dry TOGA COARE RH profiles.

Citation: Journal of Applied Meteorology 36, 2; 10.1175/1520-0450(1997)036<0183:MVOWVI>2.0.CO;2

Fig. 3.
Fig. 3.

Mean brightness temperatures at 183.3 ± 1 GHz (averaged along track) as functions of the scan angle compared with model predictions for moist RH (lowest dotted line), mean RH (middle dotted line), and dry RH (highest dotted line) profiles.

Citation: Journal of Applied Meteorology 36, 2; 10.1175/1520-0450(1997)036<0183:MVOWVI>2.0.CO;2

Fig. 4.
Fig. 4.

Relative humidity variability GRH(Δx) for the three MIR water vapor channels.

Citation: Journal of Applied Meteorology 36, 2; 10.1175/1520-0450(1997)036<0183:MVOWVI>2.0.CO;2

Table 1.

Temporal and geographical coordinates of MIR scenes.

Table 1.
Table 2.

Variability statistics from autocorrelation analysis for brightness temperature and relative humidity at lags of 10, 20, and 40 km.

Table 2.
Save
Cover Journal of Applied Meteorology and Climatology
Journal of Applied Meteorology and Climatology

  • Al-Khalaf, A. K., D. T. Conlee, and T. T. Wilheit, 1994: Retrieval of atmospheric water vapor profiles from SSM/T-2 microwave radiances. Preprints, Seventh Conf. on Satellite Meteorology and Oceanography, Monterey, CA, Amer. Meteor. Soc., 106–109.

  • Ellingson, R. G., J. Ellis, and S. Fels, 1991: The intercomparison of radiation codes used in climate models: Long wave results. J. Geophys. Res.,96, 8929–8953.

  • Hauf, T., and T. L. Clark, 1989: Three-dimensional numerical experiments on convectively forced internal gravity waves. Quart. J. Roy. Meteor. Soc.,115, 309–333.

  • Liebe, H. J., 1989: MPM—An atmospheric millimeter wave propagation model. Int. J. Infrared Millimeter Waves,10, 631–650.

  • Mapes, B. E., and P. Zuidema, 1996: Radiative–dynamical consequences of dry tongues in the tropical troposphere. J. Atmos. Sci.,53, 620–638.

  • Racette, P., R. F. Adler, J. R. Wang, A. J. Gasiewski, D. M. Jackson, and D. S. Zacharias, 1996: An airborne Millimeter-wave Imaging Radiometer for cloud, precipitation, and atmospheric water vapor studies. J. Atmos. Oceanic Technol.,13, 610–619.

  • Schmetz, J., and Coauthors, 1995: Monthly mean large-scale analyses of upper-tropospheric humidity and wind field divergence derived from three geostationary satellites. Bull. Amer. Meteor. Soc.,76, 1578–1584.

  • Soden, B. J., and F. P. Bretherton, 1993: Upper tropospheric relative humidity from the GOES 6.7 μm channel: Method and climatology for July 1987. J. Geophys. Res.,98, 16 669–16 688.

  • Spinhirne, J. D., and W. D. Hart, 1990: Cirrus structure and radiative parameters from airborne lidar and spectral radiometer observations: The 28 October 1986 FIRE study. Mon. Wea. Rev.,118, 2329.

  • Udelhofen, P. M., and D. L. Hartmann, 1995: Influence of tropical cloud systems on the relative humidity in the upper troposphere. J. Geophys. Res.,100, 7423–7440.

  • Wang, J. R., S. H. Melfi, P. Racette, D. N. Whitemen, L. A. Chang, R. A. Ferrare, K. D. Evans, and F. J. Schmidlin, 1995: Simultaneous measurements of atmospheric water vapor with MIR, Raman lidar, and rawinsondes. J. Appl. Meteor.,34, 1595–1607.

  • Webster, P. J., and R. Lukas, 1992: TOGA COARE: The Coupled Ocean–Atmosphere Response Experiment. Bull. Amer. Meteor. Soc.,73, 1377–1416.

  • Fig. 1.
    View in gallery
    Fig. 1.

    Relative humidity profiles derived from TOGA COARE soundings compared with the McClatchey tropical standard atmosphere RH profile.

  • Fig. 2.
    View in gallery
    Fig. 2.

    Weighting functions for the MIR water vapor channels derived for both the mean and dry TOGA COARE RH profiles.

  • Fig. 3.
    View in gallery
    Fig. 3.

    Mean brightness temperatures at 183.3 ± 1 GHz (averaged along track) as functions of the scan angle compared with model predictions for moist RH (lowest dotted line), mean RH (middle dotted line), and dry RH (highest dotted line) profiles.

  • Fig. 4.
    View in gallery
    Fig. 4.

    Relative humidity variability GRH(Δx) for the three MIR water vapor channels.

Fig. 1.

Relative humidity profiles derived from TOGA COARE soundings compared with the McClatchey tropical standard atmosphere RH profile.
Fig. 2.

Weighting functions for the MIR water vapor channels derived for both the mean and dry TOGA COARE RH profiles.
Fig. 3.

Mean brightness temperatures at 183.3 ± 1 GHz (averaged along track) as functions of the scan angle compared with model predictions for moist RH (lowest dotted line), mean RH (middle dotted line), and dry RH (highest dotted line) profiles.
Fig. 4.

Relative humidity variability GRH(Δx) for the three MIR water vapor channels.
All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 521 346 22
PDF Downloads 64 25 1