• Allan, R. P., , and B. J. Soden, 2007: Large discrepancy between observed and simulated precipitation trends in the ascending and descending branches of the tropical circulation. Geophys. Res. Lett., 34, L18705, doi:10.1029/2007GL031460.

    • Search Google Scholar
    • Export Citation
  • Bhat, G. S., , J. Srinivasan, , and S. Gadgil, 1996: Tropical deep convection, convective available potential energy and sea surface temperature. J. Meteor. Soc. Japan, 74, 155166.

    • Search Google Scholar
    • Export Citation
  • Bony, S., , K.-M. Lau, , and Y. C. Sud, 1997: Sea surface temperature and large-scale circulation influences on tropical greenhouse effect and cloud radiative forcing. J. Climate,10, 2055–2077.

  • Bretherton, C. S., , M. E. Peters, , and L. E. Back, 2004: Relationships between water vapor path and precipitation over the tropical oceans. J. Climate, 17, 15171528.

    • Search Google Scholar
    • Export Citation
  • Cess, R. D., and Coauthors, 1996: Cloud feedback in atmospheric general circulation models: An update. J. Geophys. Res., 101 (D8), 12 79112 794.

    • Search Google Scholar
    • Export Citation
  • Cho, H., , C.-H. Ho, , and Y.-S. Choi, 2012: The observed variation in cloud-induced longwave radiation in response to sea surface temperature over the Pacific warm pool from MTSAT-1R imagery. Geophys. Res. Lett., 39, L18802, doi:10.1029/2012GL052700.

    • Search Google Scholar
    • Export Citation
  • Del Genio, A. D., 2012: Representing the sensitivity of convective cloud systems to tropospheric humidity in general circulation models. Surv. Geophys., 33, 637656, doi:10.1007/s10712-011-9148-9.

    • Search Google Scholar
    • Export Citation
  • Derbyshire, S. H., , I. Beau, , P. Bechtold, , J.-Y. Grandpeix, , J.-M. Piriou, , J.-L. Redelsperger, , and P. M. M. Soares, 2004: Sensitivity of moist convection to environmental humidity. Quart. J. Roy. Meteor. Soc., 130, 30553079.

    • Search Google Scholar
    • Export Citation
  • Fu, Q., , M. Baker, , and D. L. Hartmann, 2002: Tropical cirrus and water vapor: An effective Earth infrared iris feedback? Atmos. Chem. Phys., 2, 3137, doi:10.5194/acp-2-31-2002.

    • Search Google Scholar
    • Export Citation
  • Fu, R., , A. D. Del Genio, , and W. B. Rossow, 1994: Influence of ocean surface conditions on atmospheric thermodynamic structure and deep convection. J. Climate, 7, 10921108.

    • Search Google Scholar
    • Export Citation
  • Gadgil, S., , P. V. Joseph, , and N. V. Joshi, 1984: Ocean–atmosphere coupling over monsoon regions. Nature, 312, 141143.

  • Gettelman, A., , and P. M. Forster, 2002: A climatology of the tropical tropopause layer. J. Meteor. Soc. Japan, 80, 911924.

  • Graham, N. E., , and T. P. Barnett, 1987: Sea surface temperature, surface wind divergence, and convection over tropical oceans. Science,238, 657–659, doi:10.1126/science.238.4827.657.

  • Hartmann, D. L., 2002: An important constraint on tropical cloud–climate feedback. Geophys. Res. Lett., 29, 1951, doi:10.1029/2002GL015835.

    • Search Google Scholar
    • Export Citation
  • Haynes, J. M., , and G. L. Stephens, 2007: Tropical oceanic cloudiness and the incidence of precipitation: Early results from CloudSat. Geophys. Res. Lett., 34, L09811, doi:10.1029/2007GL029335.

    • Search Google Scholar
    • Export Citation
  • Holloway, C. E., , and J. D. Neelin, 2009: Moisture vertical structure, column water vapor, and tropical deep convection. J. Atmos. Sci., 66, 16651683.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc.,77, 437–471.

  • Lau, K.-M., , H.-T. Wu, , and S. Bony, 1997: The role of large-scale atmospheric circulation in the relationship between tropical convection and sea surface temperature. J. Climate,10, 381–392.

  • Lindzen, R. S., , and S. Nigam, 1987: On the role of sea surface temperature gradients in forcing low-level winds and convergence in the tropics. J. Atmos. Sci.,44, 2418–2436.

  • Mace, G. G., , Q. Q. Zhang, , M. Vaughan, , R. Marchand, , G. Stephens, , C. Trepte, , and D. Winker, 2009: A description of hydrometeor layer occurrence statistics derived from the first year of merged CloudSat and CALIPSO data. J. Geophys. Res.,114, D00A26, doi:10.1029/2007jd009755.

  • Martin, E. R., , and C. Schumacher, 2012: The relationship between tropical warm pool precipitation, sea surface temperature, and large-scale vertical motion in IPCC AR4 models. J. Atmos. Sci., 69, 185194.

    • Search Google Scholar
    • Export Citation
  • Meenu, S., , K. Rajeev, , K. Parameswaran, , and A. K. M. Nair, 2010: Regional distribution of deep clouds and cloud top altitudes over the Indian subcontinent and the surrounding oceans. J. Geophys. Res.,115, D05205, doi:10.1029/2009JD011802.

  • Meenu, S., , K. Rajeev, , and K. Parameswaran, 2011: Regional and vertical distribution of semitransparent cirrus clouds over the tropical Indian region derived from CALIPSO data. J. Atmos. Sol. Terr. Phys., 73, 19671979, doi:10.1016/j.jastp.2011.06.007.

    • Search Google Scholar
    • Export Citation
  • Meenu, S., , K. Parameswaran, , and K. Rajeev, 2012: Role of sea surface temperature and wind convergence in regulating convection over the tropical Indian Ocean. J. Geophys. Res., 117, D14102, doi:10.1029/2011JD016947.

    • Search Google Scholar
    • Export Citation
  • Mehta, S. K., , B. V. K. Murthy, , D. N. Rao, , M. V. Ratnam, , K. Parameswaran, , K. Rajeev, , C. S. Raju, , and K. G. Rao, 2008: Identification of tropical convective tropopause and its association with cold point tropopause. J. Geophys. Res., 113, D00B04, doi:10.1029/2007JD009625.

    • Search Google Scholar
    • Export Citation
  • Nair, A. K. M., , K. Rajeev, , S. Sijikumar, , and S. Meenu, 2011: Characteristics of a persistent “pool of inhibited cloudiness” and its genesis over the Bay of Bengal associated with the Asian summer monsoon. Ann. Geophys., 29, 12471252, doi:10.5194/angeo-29-1247-2011.

    • Search Google Scholar
    • Export Citation
  • Pierrehumbert, R. T., 1996: Some remarks on mechanisms for the regulation of tropical sea surface temperature. Clouds, Chemistry and Climate, P. J. Crutzen and V. Ramanathan, Eds., NATO ASI Series I, Vol. 35, Springer, 153–172.

  • Rajeev, K., and Coauthors, 2008: Observational assessment of the potential of satellite-based water vapor and thermal IR brightness temperatures in detecting semitransparent cirrus. Geophys. Res. Lett., 35, L08808, doi:10.1029/2008GL033393.

    • Search Google Scholar
    • Export Citation
  • Ramanathan, V., , and W. Collins, 1991: Thermodynamic regulation of ocean warming by cirrus clouds deduced from observations of the 1987 El Niño. Nature, 351, 2732, doi:10.1038/351027a0.

    • Search Google Scholar
    • Export Citation
  • Reynolds, R. W., , C. L. Gentemann, , and G. K. Corlett, 2010: Evaluation of AATSR and TMI satellite SST data. J. Climate, 23, 152165.

  • Rondanelli, R., , and R. S. Lindzen, 2008: Observed variations in convective precipitation fraction and stratiform area with sea surface temperature. J. Geophys. Res., 113, D16119, doi:10.1029/2008JD010064.

    • Search Google Scholar
    • Export Citation
  • Sabin, T. P., , C. A. Babu, , and P. V. Joseph, 2013: SST–convection relation over tropical oceans. Int. J. Climatol., 33, 14241435, doi:10.1002/joc.3522.

    • Search Google Scholar
    • Export Citation
  • Sijikumar, S., , and K. Rajeev, 2012: Role of the Arabian Sea warm pool on the precipitation characteristics during the monsoon onset period. J. Climate, 25, 18901899.

    • Search Google Scholar
    • Export Citation
  • Stephens, G. L., 2005: Cloud feedbacks in the climate system: A critical review. J. Climate, 18, 237273.

  • Su, H., , J. H. Jiang, , J. Teixeira, , A. Gettelman, , X. Huang, , G. Stephens, , D. Vane, , and V. S. Perun, 2011: Comparison of regime-sorted tropical cloud profiles observed by CloudSat with GEOS5 analyses and two general circulation model simulations. J. Geophys. Res., 116, D09104, doi:10.1029/2010JD014971.

    • Search Google Scholar
    • Export Citation
  • Su, H., and Coauthors, 2013: Diagnosis of regime-dependent cloud simulation errors in CMIP5 models using “A-Train” satellite observations and reanalysis data. J. Geophys. Res., 118, 27622780, doi:10.1029/2012JD018575.

    • Search Google Scholar
    • Export Citation
  • Sud, Y. C., , G. K. Walker, , and K.-M. Lau, 1999: Mechanisms regulating sea-surface temperatures and deep convection in the tropics. Geophys. Res. Lett., 26, 1019–1022, doi:10.1029/1999GL900197.

    • Search Google Scholar
    • Export Citation
  • Sud, Y. C., , G. K. Walker, , Y. P. Zhou, , G. A. Schmidt, , K.-M. Lau, , and R. F. Cahalan, 2008: Effects of doubled CO2 on tropical sea surface temperatures (SSTs) for onset of deep convection and maximum SST: Simulations based inferences. Geophys. Res. Lett.,35, L12707, doi:10.1029/2008GL033872.

  • Tompkins, A. M., 2001: On the relationship between tropical convection and sea surface temperature. J. Climate,14, 633–637.

  • Waliser, D. E., 1996: Formation and limiting mechanisms for very high sea surface temperature: Linking the dynamics and the thermodynamics. J. Climate,9, 161–188.

  • Waliser, D. E., , and N. E. Graham, 1993: Convective cloud systems and warm-pool sea surface temperatures: Coupled interactions and self-regulation. J. Geophys. Res., 98 (D7), 12 88112 893.

    • Search Google Scholar
    • Export Citation
  • Webb, M. J., , F. H. Lambert, , and J. M. Gregory, 2013: Origins of differences in climate sensitivity, forcing and feedback in climate models. Climate Dyn., 40, 677707, doi:10.1007/s00382-012-1336-x.

    • Search Google Scholar
    • Export Citation
  • Williams, K. D., , M. A. Ringer, , and C. A. Senior, 2003: Evaluating the cloud response to climate change and current climate variability. Climate Dyn., 20, 705721.

    • Search Google Scholar
    • Export Citation
  • Winker, D. M., , W. H. Hunt, , and M. J. McGill, 2007: Initial performance assessment of CALIOP. Geophys. Res. Lett.,34, L19803, doi:10.1029/2007GL030135.

  • Wood, R., , and C. S. Bretherton, 2006: On the relationship between stratiform low cloud cover and lower-tropospheric stability. J. Climate, 19, 64256432.

    • Search Google Scholar
    • Export Citation
  • Yuan, J., , D. L. Hartmann, , and R. Wood, 2008: Dynamic effects on the tropical cloud radiative forcing and radiation budget. J. Climate, 21, 23372351.

    • Search Google Scholar
    • Export Citation
  • Zelinka, M. D., , and D. L. Hartmann, 2011: The observed sensitivity of high clouds to mean surface temperature anomalies in the tropics. J. Geophys. Res., 116, D23103, doi:10.1029/2011JD016459.

    • Search Google Scholar
    • Export Citation
  • Zhang, C., 1993: Large-scale variability of atmospheric deep convection in relation to sea surface temperature in the tropics. J. Climate,6, 1898–1913.

  • View in gallery

    Number of collocated observations of 2B-GEOPROF-lidar data as a fraction of TMI SST at an SST interval of 0.3°C for the total period of June 2006–February 2011 and for different seasons during the period: winter [December–February (DJF)], premonsoon [March–May (MAM)], summer monsoon [June–September (JJAS)], and postmonsoon [October–November (ON)].

  • View in gallery

    Altitude distribution of the FOC (%) as a function of SST during 2006–11.

  • View in gallery

    As in Fig. 2, but for different seasons (a) DJF, (b) MAM, (c) JJAS, and (d) ON.

  • View in gallery

    (a) Normalized probability distribution function (PDF) of cloud thickness as a function of SST for clouds having a base altitude of <4 km. (b) Normalized probability of the FOC having a thickness of >4, >6, and >10 km, as a function of SST. (c) FOC as a function of cloud thickness (for clouds with a base altitude of <4 km) for four regimes of SST: 24° < SST < 25°, 25° < SST < 27.5°, 27.5° < SST < 29°, and 29° < SST < 30.5°C.

  • View in gallery

    Average FOC for different altitude bands: (a) 16–17, (b) 8–16, (c) 2–8, and (d) 0–2 km, as a function of SST. Standard deviations are represented by vertical bars. (e) SST dependence of the ratio of the average FOC in (i) 4–8-km-altitude band to that 0–4-km-altitude band and (ii) 8–16-km-altitude band to that at 4–8-km-altitude band.

  • View in gallery

    (a) SST dependence of the FOC derived from 2B-GEOPROF-lidar data for the period of study (2006–11) and for different seasons. Standard deviations are represented by vertical bars. Standard deviations for all individual seasons are comparable to that for the whole period. (b) As in (a), but for the total FOC derived from MODIS data (average for all seasons) collocated with the 2B-GEOPROF-lidar data.

  • View in gallery

    PDF of (a) SST gradient, (b) surface wind divergence, and (c) wind divergence at 150 hPa as a function of SST.

  • View in gallery

    PDF of (a) LTS and (b) CAPE as a function of SST.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 42 42 8
PDF Downloads 25 25 7

Multiyear CloudSat and CALIPSO Observations of the Dependence of Cloud Vertical Distribution on Sea Surface Temperature and Tropospheric Dynamics

View More View Less
  • 1 Space Physics Laboratory, Vikram Sarabhai Space Centre, Thiruvananthapuram, India
© Get Permissions
Full access

Abstract

Utilizing the synergy of the capabilities of CloudSat and Cloud–Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) and ~4.5 years of their observations, this paper investigates the dependence of the altitude distribution and thickness of tropical clouds on sea surface temperature (SST). Variations in the altitude distribution of clouds with SST show three distinct regimes: SST < 27.5°C, 27.5° < SST < 29°C, and SST > 29°C. At an SST < 27.5°C, the convection is rather weak, so that most of the clouds are limited to <2-km altitude with peak occurrence at 1–1.5 km. The frequency of occurrence of the low-altitude clouds as well as the prominence of the peak at ~1.5 km consistently decreases for SST > 24°C. Vertical development of clouds through the 3–12-km-altitude region increases for SST > 27.5°C to achieve maximum cloud occurrence and thickness in the SST range of 29°–30.5°C. Penetration of the deep convective clouds to altitudes >15 km and their frequency of occurrence increase with SST until ~30°C. These observations reveal two differences with the SST dependence of total cloudiness observed using passive imager data: (i) the increase in cloudiness at an SST > 26°–27°C observed using the imager data is found to be influenced by the increase in cirrus clouds generated by deep convective outflows and is not directly driven by the local SST, and (ii) the total cloudiness does not decrease for SST > 29.5°C as observed using imagers, but weakly increases until an SST of ~30.5°C. The role of the spatial gradient of SST and atmospheric dynamical parameters in modulating the observed SST dependence of cloudiness at different SST regimes is investigated.

Corresponding author address: K. Rajeev, Space Physics Laboratory, Vikram Sarabhai Space Centre, Thiruvananthapuram 695 022, India. E-mail: k_rajeev@vssc.gov.in

Abstract

Utilizing the synergy of the capabilities of CloudSat and Cloud–Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) and ~4.5 years of their observations, this paper investigates the dependence of the altitude distribution and thickness of tropical clouds on sea surface temperature (SST). Variations in the altitude distribution of clouds with SST show three distinct regimes: SST < 27.5°C, 27.5° < SST < 29°C, and SST > 29°C. At an SST < 27.5°C, the convection is rather weak, so that most of the clouds are limited to <2-km altitude with peak occurrence at 1–1.5 km. The frequency of occurrence of the low-altitude clouds as well as the prominence of the peak at ~1.5 km consistently decreases for SST > 24°C. Vertical development of clouds through the 3–12-km-altitude region increases for SST > 27.5°C to achieve maximum cloud occurrence and thickness in the SST range of 29°–30.5°C. Penetration of the deep convective clouds to altitudes >15 km and their frequency of occurrence increase with SST until ~30°C. These observations reveal two differences with the SST dependence of total cloudiness observed using passive imager data: (i) the increase in cloudiness at an SST > 26°–27°C observed using the imager data is found to be influenced by the increase in cirrus clouds generated by deep convective outflows and is not directly driven by the local SST, and (ii) the total cloudiness does not decrease for SST > 29.5°C as observed using imagers, but weakly increases until an SST of ~30.5°C. The role of the spatial gradient of SST and atmospheric dynamical parameters in modulating the observed SST dependence of cloudiness at different SST regimes is investigated.

Corresponding author address: K. Rajeev, Space Physics Laboratory, Vikram Sarabhai Space Centre, Thiruvananthapuram 695 022, India. E-mail: k_rajeev@vssc.gov.in

1. Introduction

Clouds play a paramount role in regulating the radiation budget, energy and moisture transport, hydrological cycle, and climate of the earth–atmosphere system (e.g., Stephens 2005). Processes and parameters that control the generation and properties of clouds and cloud-feedback processes (Cess et al. 1996; Stephens 2005) are among the long-standing poorly resolved issues and are major hindrances in the prediction of future climate. Investigations into the variations of cloudiness with SST and large-scale atmospheric circulation and the underlying physics have been carried out extensively during the past few decades, which showed that the development of clouds are governed by a host of processes and parameters, the most prominent among them being the moist static energy near the surface, water vapor content in the troposphere, atmospheric thermodynamical instability, and circulation systems controlling the atmospheric convergence/divergence and vertical winds (Gadgil et al. 1984; Graham and Barnett 1987; Zhang 1993; Waliser 1996; Bony et al. 1997; Tompkins 2001; Rondanelli and Lindzen 2008; Su et al. 2011, 2013; Zelinka and Hartmann 2011; Meenu et al. 2012; Sabin et al. 2013). Influence of the tropospheric water vapor content as a regulator of convection is also well recognized (Bretherton et al. 2004; Derbyshire et al. 2004; Holloway and Neelin 2009; Del Genio 2012; Su et al. 2013). Over the oceanic regions, most of the above-mentioned controlling parameters are substantially influenced by the sea surface temperature (SST) (Lindzen and Nigam 1987; Gadgil et al. 1984; Cho et al. 2012; Su et al. 2013), while the SST itself is regulated through cloud-feedback processes, especially through modification of net radiative fluxes at the surface (Ramanathan and Collins 1991; Sud et al. 1999). Larger values of SST generally yield higher equivalent potential temperature (and hence higher moist static energy) in the atmospheric boundary layer, and can trigger local convection through thermodynamical changes within the atmospheric column (Bony et al. 1997). On the contrary, the spatial gradient of SST, rather than the local SST, is important in driving large-scale circulation, which transports heat and moisture and affects the thermodynamical stability of the atmosphere over large spatial scales (Lindzen and Nigam 1987; Waliser 1996; Williams et al. 2003) and in turn regulates the convection, cloud occurrence, and the associated radiative forcing (e.g., Yuan et al. 2008).

Earlier studies using the total cloudiness observed using imager data showed a general pattern of consistent increase in cloudiness with SST in the range of ~26°–28°C and peak cloud occurrence at an SST of ~28.5°–29.5°C. Most of them showed a decrease in cloudiness with further increase in SST above ~29.5°C (Gadgil et al. 1984; Graham and Barnett 1987; Waliser 1996), while such a reduction was not clearly discernible in all regions or during all seasons (Lau et al. 1997; Meenu et al. 2012). The SST threshold for convection corresponds to the minimum SST that can generate large convective available potential energy (CAPE) through a deep tropospheric layer (Bhat et al. 1996). Lau et al. (1997) argued that the critical SST threshold (e.g., >27°C) required for convection to occur does not have any fundamental microphysical or thermodynamical importance, but arises from the fact that this regime represents a transition between stable and convective atmospheric conditions. Investigations into the vertical distribution of cloud water content and cloud faction observed using CloudSat/Cloud–Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) data and its comparison with the circulation models and reanalysis data revealed that the high-altitude clouds in observations as well as models are concentrated in the regions with large midtropospheric vertical velocity, warm SST, and low lower-tropospheric stability (LTS) (Su et al. 2011, 2013). The CloudSat observations also revealed that the preferential distribution of low-level clouds occurs in the regions of large-scale subsidence, relatively cold SST, and high LTS (Su et al. 2011). Significant seasonal and spatial variations occur in the SST dependence of cloudiness, and the regions of the warmest SST, maximum surface wind convergence, and the largest cloudiness are generally not collocated (Fu et al. 1994; Meenu et al. 2012). The increase in SST associated with global warming (and the potential changes in convection associated with it) as well as the role of clouds in decreasing SST through feedback processes (e.g., Ramanathan and Collins 1991; Pierrehumbert 1996; Fu et al. 2002) have further enhanced the importance of the investigations into the SST–cloudiness relationship (e.g., Sud et al. 2008).

Observations of the SST–cloudiness relationship are influenced by the methods used to determine cloudiness: the passive radiometer observations, on which most of the previous studies were based, have two important limitations. (i) They cannot discriminate between deep convective clouds and optically thick cirrus clouds. While the former could be generated through convection driven by local warm SST, the latter might originate from the outflows of deep convection and are not affected by the local SST. In principle, this can cause deviations in the observed SST–convection relationship compared to the actual. (ii) They cannot provide the actual altitudinal cross section (and distribution) of clouds (except for inferring the cloud-top altitude based on the observed cloud-top brightness temperature). The vertical cross sections of clouds observed using the spaceborne radar CloudSat and spaceborne lidar CALIPSO provide a unique solution to this enigma. The main objectives of this paper are (i) to examine the variations in the vertical distribution and physical thickness of clouds with SST, and (ii) to investigate the potential influence of spatial gradient of SST and parameters that govern atmospheric stability and circulation in modifying the observed SST–cloudiness relationship, utilizing about 4.5 years (June 2006–February 2011) of CloudSat and CALIPSO observations of the vertical distribution of clouds. This study would contribute to improvements in the understanding of the SST–convection–cloudiness relationship, which is especially important for reducing the sensitivity of cloudiness/precipitation to SST in atmospheric models (e.g., Allan and Soden 2007; Martin and Schumacher 2012).

2. Data and method of analysis

The vertical distribution of clouds is obtained from a combined analysis of the data obtained from the CloudSat and CALIPSO satellites, which are part of the A-Train constellation that follows the same satellite orbits with an equatorial crossing time around 1330 local time (separated by few minutes). The CloudSat has a Cloud Profiling Radar (CPR) operating at 94 GHz and provides the altitude profiles of backscattered radar signal from hydrometeors with a vertical resolution of 240 m (Haynes and Stephens 2007). The Cloud–Aerosol Lidar with Orthogonal Polarization (CALIOP) on board CALIPSO provides altitude profiles of lidar backscatter signals (at 532 and 1064 nm) that can provide the vertical profiles of clouds and aerosols with a vertical resolution of 30–60 m (Winker et al. 2007). The CPR can detect optically thick cloud layers and is less sensitive to optically thin clouds such as semitransparent cirrus. In contrast, CALIPSO can detect even very thin cirrus clouds with visible band cloud optical depth as low as 0.03 (e.g., Meenu et al. 2011), but it cannot probe the cloud layers located beneath optically thick clouds.

This study utilizes the CloudSat geometric profile product (2B GEOPROF)-lidar data (version 003) during the period of June 2006–February 2011, provided by the CloudSat data processing center, which utilizes the synergy of the observational capabilities of CloudSat and CALIPSO (Mace et al. 2009). The 2B-GEOPROF-lidar data provide optimally merged analysis of the data from CloudSat and CALIPSO, and give the best picture of the vertical distribution of cloud occurrence as has been compiled so far. Details of the 2B-GEOPROF-lidar dataset, methodology for the combined analysis to derive the cloud mask, the complexities involved while combining the information derived from instruments with different observational capabilities and spatial resolutions, and the uncertainties are extensively discussed in Mace et al. (2009) and the algorithm theoretical basis document (available online at http://www.cloudsat.cira.colostate.edu). The 2B-GEOPROF-lidar dataset provides information for up to five cloud layers (their altitudes of occurrence) with a vertical resolution of 240 m and along-track resolution of 1 km.

The daily-mean SST data are obtained from the Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI), which utilizes the observations at 10.7 GHz, which are nearly transparent to clouds (Reynolds et al. 2010). The TMI SST provides daily observations covering the global region extending from 40°S to 40°N at a pixel resolution of 0.25°, with an expected accuracy of better than 0.5°C (Reynolds et al. 2010). The atmospheric circulation and thermodynamical data (including the potential temperature and CAPE) are obtained from the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis, which integrates the data from a variety of observing systems with numerical models to produce a temporally and spatially consistent synthesis of observations and analyses of variables with a spatial resolution of 2.5° (Kalnay et al. 1996). The surface wind divergence (SWD) during this period is obtained from the spaceborne scatterometer observations from the Quick Scatterometer (QuikSCAT) satellite of the National Aeronautics and Space Administration, which has a spatial resolution of 0.5°.

All the above-mentioned datasets have different spatial resolutions. Collocation of the above-mentioned data is done on the spatial resolution of 2B-GEOPROF-lidar data (1 km), which is better than that of the TMI SST (0.25°). Thus, there will be ~25 2B-GEOPROF-lidar data points within a geographical grid of the SST data, and the spatial variations of SST at scales smaller than 0.25° are neglected. Though the study is made using the high-resolution profiles obtained from 2B-GEOPROF-lidar data on an orbit-by-orbit basis, we focus here only on the seasonal and long-term-mean SST–cloudiness relationship. The above-mentioned daily data over the tropical Indian Ocean, the Bay of Bengal, and the Arabian Sea encompassed between 25°S and 25°N and 30° and 110°E are used in this study. This region covers a wide range of SST (~24°–32°C) and convective conditions required for the study within a limited geographical box of 80° in longitude and 50° in latitude in the tropics.

3. Results

Depending on the season and location, SST values over the study region (25°S–25°N, 30°–110°E) vary in the range of about 24°–32°C [e.g., Fig. 2 of Meenu et al. (2012)]. On average, the SSTs are higher (>28°C) at the equatorial Indian Ocean and lower (24°–28°C) in the Southern Hemisphere Indian Ocean throughout the year. The warmest temperatures are observed at in southern Arabian Sea and the equatorial Indian Ocean from April to May with values exceeding 30°C over a wide region, including the areas covered by the Arabian Sea warm pool (Sijikumar and Rajeev 2012). The lowest SSTs (24°–26°C) are observed in the northern Arabian Sea (during December–March) and the southern Indian Ocean (south of ~10°S during July–September). Figure 1 depicts histograms of the number of collocated 2B-GEOPROF-lidar data points as a function of SST during different seasons and the entire period (2006–11). Maxima in all of the histograms occur in the range of 28°–29°C, where the total number of observations over the entire period exceeds 1 million (per 0.3°C SST interval). Though the number of collocated points decreases considerably for SST > 29°C, their number for the long-term analysis is more than 0.25 million (per 0.3°C SST interval) for all SSTs in range of 24°–30.5°C. We mainly focus here on the variations in cloudiness in this SST range, where the total number of observations is quite large.

Fig. 1.
Fig. 1.

Number of collocated observations of 2B-GEOPROF-lidar data as a fraction of TMI SST at an SST interval of 0.3°C for the total period of June 2006–February 2011 and for different seasons during the period: winter [December–February (DJF)], premonsoon [March–May (MAM)], summer monsoon [June–September (JJAS)], and postmonsoon [October–November (ON)].

Citation: Journal of Climate 27, 2; 10.1175/JCLI-D-13-00062.1

a. Variation of cloud vertical distribution with SST

Figure 2 shows the contour map of the long-term-mean frequency of occurrence of clouds (FOC) as a function of altitude and SST, obtained by combining the whole data (2006–11) over the study region. Based on the altitude variation of cloud occurrence, three distinct regimes of SSTs are discernible in Fig. 2—namely, SST < 27.5°, 27.5° < SST < 29°, and SST > 29°C. (i) At an SST < 27.5°C, the convection is rather weak that most of the total cloudiness is limited to <2-km altitude with a peak occurrence at 1–1.5 km. This feature is akin to the cloud distribution in the high pressure areas because of trapping of air mass in the lower troposphere by subsidence from above (e.g., Su et al. 2011). In this regime, the frequency of occurrence of the low-altitude marine clouds consistently decreases at larger SSTs. At the altitude of their peak occurrence (~1–1.5 km), the cloudiness decreases typically from ~40% at 24°C to ~25% at 27.5°C. Occurrence of clouds above these low-level clouds is rare, though a building up of upper-level clouds in the altitude band of ~10–15 km is observed with an increase in SST. (ii) In the SST range of 27.5° < SST < 29°C, the FOC in the altitude band of ~3–10 km increases with SST, presumably indicating an increase in the convective activity. However, a significant increase in these clouds is seen only above an SST of ~28°C. Though the occurrence of low-altitude clouds decreases with SST in this regime as well, it is less significant compared to that for SST < 27.5°C. (iii) FOC in the altitude region of about 3–12 km attains its peak value in the SST range of 29°–30.5°C, where the SST dependence of the FOC is negligible at all altitudes.

Fig. 2.
Fig. 2.

Altitude distribution of the FOC (%) as a function of SST during 2006–11.

Citation: Journal of Climate 27, 2; 10.1175/JCLI-D-13-00062.1

Figure 2 shows that the FOC rapidly increases above the altitude of ~8 km and maximizes in the altitude band of 13–16 km. The occurrence of high-altitude clouds increases with SST right from ~25°C and the largest occurrence is observed at 14–15-km altitude in the SST range of 29.5°–30.5°C. In addition to the deep convective clouds that originate right from the lower troposphere, the increase in cloudiness at higher altitudes might have been contributed by an increase in cirrus clouds also, as they can be produced from the anvils or remnants of deep convective clouds and have longer lifetime compared to the deep convective clouds. The cirrus production also may be expected to be more at higher SSTs (SST > 29°C), where deep convection is more prominent. In the altitude band of 16–17 km, the frequency of occurrence of clouds is ~25%–40% at an SST of ~30.5°C. The seasonal-mean altitude variations in the FOC with SST during winter (December–February), premonsoon (March–May), summer monsoon (June–September), and postmonsoon (October–November) seasons are shown in Fig. 3. Overall, the SST–cloudiness relationships observed during different seasons are consistent and similar to that observed using the long-term-mean variations illustrated in Fig. 2. However, significant variations are observed in the absolute values of the FOC, which indicate the influence of factors other than SST (such as large-scale atmospheric circulation) in regulating the vertical distribution of clouds.

Fig. 3.
Fig. 3.

As in Fig. 2, but for different seasons (a) DJF, (b) MAM, (c) JJAS, and (d) ON.

Citation: Journal of Climate 27, 2; 10.1175/JCLI-D-13-00062.1

b. Variation of cloud thickness with SST

The SST dependence of cloud thickness is shown in Fig. 4a, which depicts the long-term-mean (2006–11: all seasons) probability distribution of cloud thickness normalized with respect to the total frequency of occurrence of clouds for each SST. The objective here is to investigate the vertical development of clouds that are directly influenced by the local SST and to avoid any influence of cirrus clouds originating from deep convective outflows. Hence, only those clouds having their base altitude occurring below 4 km are included in this analysis. Figure 4a shows that, at an SST < 26°C, most of the clouds have a thickness of <4 km and that the probability of occurrence of clouds with a thickness of >5 km is quite small. The probability of occurrence of thicker clouds significantly increases at an SST > 27.5°C, which is also associated with a systematic reduction in the normalized probability of occurrence of thin clouds, especially those having a thickness of <3 km. The average frequency of occurrence of clouds having a thickness of >4, >6, and >10 km is depicted in Fig. 4b. Only <5% of the total clouds that occur at an SST < 26°C have a thickness of >4 km. However, the probability of occurrence of clouds with a thickness of >4 km increases considerably in the SST range of 27.5°–29°C. Similar is the case with the clouds having a thickness of >6 and >10 km. At an SST > 29°C, the probabilities of occurrence of clouds having a thickness of >4, >6, and >10 km are ~30%, ~20%, and ~10%, respectively. Figure 4c shows the average frequency of occurrence of clouds as a function of cloud thickness for different SST ranges: <27.5°, 27.5°–29°, and >29°C. Figures 4a and 4c indicate that, in all SSTs, the probability of occurrence is largest for thin clouds and the occurrence frequency decreases with the increase in cloud thickness up to ~10–11 km. The rate of this decrease is rapid for lower SST (SST < 27.5°C) and smaller for warmer SST. One of the most striking features seen in Figs. 4a and 4c is the increase in the probability of occurrence of clouds with thickness greater than ~10–11 km. The magnitude of this secondary peak increases significantly with SST and is prominent even up to an SST of ~30.5°C. However, this secondary peak is very weak at an SST < 25°C. The average frequency of occurrence of clouds with thickness >12 km is ~1%, 4%, and 7% for SST < 27.5°, 27.5° < SST < 29°, and SST > 29°C, respectively. Furthermore, the value of cloud thickness at which this secondary maximum occurs increases with SST from ~15 km for SST < 27.5°C to ~16 km for SST > 29°C.

Fig. 4.
Fig. 4.

(a) Normalized probability distribution function (PDF) of cloud thickness as a function of SST for clouds having a base altitude of <4 km. (b) Normalized probability of the FOC having a thickness of >4, >6, and >10 km, as a function of SST. (c) FOC as a function of cloud thickness (for clouds with a base altitude of <4 km) for four regimes of SST: 24° < SST < 25°, 25° < SST < 27.5°, 27.5° < SST < 29°, and 29° < SST < 30.5°C.

Citation: Journal of Climate 27, 2; 10.1175/JCLI-D-13-00062.1

c. Variations in the layer-averaged cloud occurrence with SST

The SST dependences of the average FOC in the altitude bands of <2, 2–8, 8–16, and 16–17 km are depicted in Fig. 5. The average FOC in the 2–8-km-altitude band increases significantly only above ~27.5°C and maximizes in the SST range of 29°–30.5°C, where it remains steady with an average occurrence of ~20%. The occurrence of high-altitude clouds (8–16- and 16–17-km-altitude bands) consistently increases for SST > 25°C (Figs. 5a,b). The magnitude of this increase is also larger than that of clouds in the 2–8-km-altitude band. Figure 5e depicts the ratio of the average FOC at different altitude bands (0–4, 4–8, and 8–16 km). Note that the ratio of the average FOC at 4–8 km to that at 0–4 km is <0.5 at an SST < 27°C, indicating a considerable decrease in cloudiness with altitude in this SST range. This ratio (which can be considered as a proxy to the vertical development of clouds) increases with SST only above ~27°C. In contrast, the ratio of the average FOC at 8–16-km altitude to that at 4–8 km shows a significant increase from a value of ~1.1 at an SST of ~24°C to ~2.2 at an SST of ~26.5°C, followed by a decrease to ~1.7 at an SST > 28°C. As seen in Fig. 2, at an SST < 27.5°C, the occurrence of high-altitude clouds does not appear to be closely associated with the clouds originating in the lower altitudes, while the continuity in their occurrence at the lower, middle, and upper levels are quite evident for SST > 28°C. As seen from Fig. 4, more than 90% of the clouds occurring at an SST < 27°C have a thickness of <4 km. These observations indicate that most of the upper-tropospheric clouds occurring at an SST < 27°C do not arise from the local vertical development from the lower/middle troposphere, but might have originated from the outflows of deep convective systems occurring in the nearby warmer oceanic regions.

Fig. 5.
Fig. 5.

Average FOC for different altitude bands: (a) 16–17, (b) 8–16, (c) 2–8, and (d) 0–2 km, as a function of SST. Standard deviations are represented by vertical bars. (e) SST dependence of the ratio of the average FOC in (i) 4–8-km-altitude band to that 0–4-km-altitude band and (ii) 8–16-km-altitude band to that at 4–8-km-altitude band.

Citation: Journal of Climate 27, 2; 10.1175/JCLI-D-13-00062.1

The SST dependences of the seasonal- and long-term-mean total frequency of occurrence of clouds derived from the 2B-GEOPROF-lidar data at all altitudes are shown in Fig. 6a. In this analysis, a pixel is considered to be cloudy if a cloud appears at any altitude. In principle, this is similar to the frequency of occurrence of clouds that would be observed using the imager data [e.g., Moderate Resolution Imaging Spectroradiometer (MODIS) and National Oceanic and Atmospheric Administration (NOAA) Advanced Very High Resolution Radiometer (AVHRR)], which are reported in the literature (e.g., Meenu et al. 2012). Figure 6a shows that the long-term-mean occurrence of total cloudiness has a consistent increase with SST for SST > 26°C. During the summer monsoon season, this increase in cloudiness is more rapid and tends to saturate at an SST > 28.5°C. In contrast, during the winter and premonsoon seasons, the increase in cloudiness occurs only over warmer regions with SST > 27°C. The threshold SST of ~26°–27°C, above which the total cloudiness increases with SST, is distinctly smaller than the threshold SST of ~27.5°–28°C observed in Fig. 4 for the vertically developing clouds. This difference might be mainly associated with the increase in thick cirrus clouds generated from deep convective outflows, which significantly contribute to the total cloudiness at an SST < 27.5°C.

Fig. 6.
Fig. 6.

(a) SST dependence of the FOC derived from 2B-GEOPROF-lidar data for the period of study (2006–11) and for different seasons. Standard deviations are represented by vertical bars. Standard deviations for all individual seasons are comparable to that for the whole period. (b) As in (a), but for the total FOC derived from MODIS data (average for all seasons) collocated with the 2B-GEOPROF-lidar data.

Citation: Journal of Climate 27, 2; 10.1175/JCLI-D-13-00062.1

Figure 6b shows the corresponding variations of FOC derived from MODIS data (provided along with 2B-GEOPROF-lidar data) that are spatially and temporally collocated with the CloudSat and CALIPSO observations. Overall, the absolute magnitudes of FOC derived from MODIS data are ~12%–17% smaller than that from 2B-GEOPROF-lidar data (Fig. 6a). Part of this difference might be because CALIPSO can observe even semitransparent clouds, while such clouds may not be observable using MODIS (differences in pixel shape/spatial resolutions of the two observations would also contribute to this difference). Notwithstanding the differences in absolute magnitude, the overall variations in the SST dependences of the FOC observed using MODIS and 2B-GEOPROF-lidar data are somewhat similar, with the largest increase in the FOC observed in the SST range of 26°–28.5°C. However, MODIS data show a plateau in the SST dependence of the FOC at 28.5° < SST < 30°C, which is followed by a decrease in the FOC at an SST > 30°C. The above-mentioned variations in the FOC observed using MODIS data are similar to the SST dependence of total cloudiness reported in the literature using other imager data as well (e.g., Meenu et al. 2012). In contrast, the total FOC derived from the 2B-GEOPROF-lidar data shows a weak but systematic increase with SST in the range of 28.5°–30.5°C. This result differs from those obtained from the imager data, where the total cloudiness is generally found to decrease with an increase in SST above ~29.5°C (e.g., Gadgil et al. 1984; Graham and Barnett 1987; Lau et al. 1997; Bony et al. 1997). This difference might be due to the underestimation of semitransparent clouds by MODIS and other imagers (which work based on the detection of thermal-IR radiation emitted from the clouds) (e.g., Rajeev et al. 2008). On the other hand, these thin clouds will be detectable using CALIPSO, and a combination of CloudSat and CALIPSO would detect clouds of all opacity. Note that, in contrast to the nearly steady values of the FOC in the 2–8- and 8–16-km-altitude bands at an SST > 28.5°C (Figs. 5b,c), the FOC in the 16–17-km-altitude region rapidly increases with SST in the SST range from ~28.5° to ~30.5°C. This indicates that the occurrence of thin cirrus clouds in the 16–17-km-altitude band significantly increases with SST even up to ~30.5°C. CALIPSO observations of the large occurrence of semitransparent cirrus below the cold point tropical tropopause were also reported by Meenu et al. (2011). This further supports the above-mentioned inferences on the potential cause for the difference between the SST dependence of cloud occurrence at an SST > 29°C observed using 2B-GEOPROF-lidar and MODIS (or other imager) data.

d. Variations in spatial gradient of SST and atmospheric dynamical parameters with SST

In addition to SST, the convection and cloudiness over a given region are significantly regulated by the spatial gradient of SST (which has a significant influence on the large-scale atmospheric circulation) and atmospheric dynamical factors such as surface wind convergence, lower-tropospheric stability, upper-tropospheric divergence, and CAPE—all of which are crucial in determining the vertical wind velocity and formation of convective clouds (e.g., Wood and Bretherton 2006; Su et al. 2011, 2013; Meenu et al. 2012; Webb et al. 2013). Figure 7a shows the probability distribution function for the magnitude of the spatial gradient of SST as a function of SST, derived from the TMI SST dataset over the study region during 2006–11. The mean SST gradient is largest in the SST range of 26.5°–27.5°C, where the gradient values are generally in the range of 0.1°–0.5°C per degree of latitude/longitude. On the contrary, the SST gradients are remarkably smaller for higher values of SST, especially at an SST > 28.5°C, where the gradients are mostly <0.1°C per degree of latitude/longitude. As seen in Figs. 2 and 4, the occurrence of convective clouds significantly increases with SST above ~27.5°C, where the SST gradient is largest. However, maximum cloudiness occurs at an SST of ~29°–30.5°C, where the SST gradients are substantially smaller. Figure 7a together with Figs. 2 and 4 clearly shows that both SST and its spatial gradient are important in determining the vertical development of clouds. However, their prominence differs at different ranges of SST. The decrease in the magnitude of SST gradient might have slowed down the rate of increase in convection with SST, especially for SST > 29°C.

Fig. 7.
Fig. 7.

PDF of (a) SST gradient, (b) surface wind divergence, and (c) wind divergence at 150 hPa as a function of SST.

Citation: Journal of Climate 27, 2; 10.1175/JCLI-D-13-00062.1

The probability distribution function of SWD (observed using scatterometer data) as a function of TMI SST is depicted in Fig. 7b. The surface winds are mostly diverging (positive values of SWD) at an SST < 26°C, which will inhibit large-scale convection in the lower troposphere and vertical development of clouds. On average, the surface winds have a high probability for convergence at an SST of ~28°–29°C, while at still higher SSTs the probability for convergence is again comparable to that for divergence. Thus, the occurrence of convective clouds caused by the surface wind convergence will be largest at an SST of ~28°–29°C, which is also the SST range at which the occurrence of convective clouds shows a rapid increase with SST (Figs. 2, 4).

The probability distribution function of the upper-tropospheric wind divergence at the 150-hPa level (derived from NCEP reanalysis data) as a function of TMI SST is depicted in Fig. 7c, which shows that the upper-level winds are strongly converging at an SST < 27.5°C. This might cause subsidence of upper-tropospheric air mass and inhibit convection at the lower levels. The 150-hPa-level winds are mostly diverging at an SST > 28°C, and the average wind divergence increases up to an SST ~29°C, above which the average magnitude of the wind divergence does not significantly vary with SST. The variations in the upper-level divergence with SST are in tandem with the corresponding variations in the vertical distribution of clouds in the middle and upper troposphere. However, the occurrence of convective clouds cannot further increase through this process at an SST > 29°C and might have contributed to the observed weakening of the rate of increase in the frequency of occurrence of convective clouds for SST > 29°C.

The potential temperature at the 700- minus that at the 1000-hPa level provides an index of the stability of the lower troposphere. Figure 8a shows the probability distribution function of the LTS as a function of SST—the larger the value of LTS, the more likely the lower troposphere is stable. Figure 8a shows that the LTS is quite high and rather steady (mean value of ~15 K) at an SST < 26.5°C and systematically decreases with a further increase in SST. The LTS is distinctly smaller and remains almost steady (mean value ~12 K) for 29° < SST < 30.5°C. The decrease in LTS would trigger convection in the lower troposphere and this effect systematically increases with SST. Thus, the convection arising from the reduction in lower-tropospheric stability might contribute to the observed rapid increase in convective clouds in the SST range of 27.5°–29°C and weaken further an increase in their occurrence at an SST > 29°C. Figure 8b shows the probability distribution function for CAPE as a function of SST, the pattern of which is strikingly similar to the SST dependence of the vertical distribution of cloud occurrence observed in Figs. 25. Figure 8b is derived from the monthly-mean values of CAPE and SST, which had resulted in the relatively smaller magnitude of CAPE. The average CAPE remains almost steady for SST < 27°C and thereafter steadily increases with SST until ~29.5°C to attain the peak value. The further increase of CAPE with SST is negligible. The increase in CAPE would increase convection and frequency of occurrence of clouds in the middle and upper troposphere. As the CAPE does not significantly increase with SST at an SST > 29.5°C, the occurrence of deep convective cloudiness would be frequent but almost steady at an SST > 29.5°C.

Fig. 8.
Fig. 8.

PDF of (a) LTS and (b) CAPE as a function of SST.

Citation: Journal of Climate 27, 2; 10.1175/JCLI-D-13-00062.1

4. Discussion and conclusions

The main focus of this study is to investigate (i) the SST dependence of the vertical distribution and thickness of tropical clouds and (ii) the potential influence of the SST gradient and other atmospheric dynamical parameters in governing the observed SST dependence of cloudiness. It is clearly shown that the frequency of occurrence of clouds in the middle and upper troposphere, apparently produced by deep convection, increases significantly only for SST > 27.5°C and attains the maximum cloudiness at an SST of ~29°–30.5°C. The frequency of occurrence of clouds does not decrease with an increase in SST, at least until ~30.5°C. Some of these results differ from the SST dependence of total cloudiness derived using imager data that are reported in the literature, in which the total cloudiness is found to increase with SST > ~26°–27°C to attain a peak value at ~29.5°C and to decrease with a further increase in SST. One of the potential causes for these differences between the results based on imager data and derived from the vertical distribution of clouds observed using CloudSat and CALIPSO might stem from the fact that the imager observations may not fully discriminate the thick cirrus clouds generated by the outflows from deep convection. Cirrus clouds generated by convective outflows can be transported even hundreds of kilometers away from their deep convective source regions (e.g., Nair et al. 2011) and are not generated through convection caused by the local SST. This causes biases in the SST–cloudiness relationship derived from the imager data, especially when the deep convection is generated in a region having large spatial gradient of SST. Furthermore, the semitransparent clouds may be underestimated in the imager observations (e.g., Rajeev et al. 2008). A combination of the vertical cross sections of clouds observed using CloudSat and CALIPSO can determine both thick and thin clouds, and provide an excellent opportunity to investigate the SST dependence of the vertical distribution of clouds. This also provides an opportunity to relook at the well-studied SST dependence of total cloudiness reported in the literature.

The present study also shows that the frequency of occurrence of lower-tropospheric clouds decreases with an increase in SST, especially in the SST range of 24°–27°C. This is in agreement with the observations of Bony et al. (1997) showing a decrease in cloud fraction and optical depth with an increase in SST for SST < 27°C. These regions of smaller SST in the tropics are generally characterized by large-scale subsidence driven by the remotely generated convection at warmer oceanic regions (e.g., Waliser and Graham 1993) and favor the occurrence of low-level clouds (e.g., Su et al. 2011). In the study region, the cold SSTs generally occur over the southwestern Indian Ocean during the Asian summer monsoon season and the northern Arabian Sea during winter and are manifested by the descending limb of Hadley or Walker circulation cells. Both of these conditions are associated with the divergence of the lower-tropospheric winds, which causes the trapping of air (and moisture) in the atmospheric boundary layer or lower troposphere and would prevent vertical development of clouds. The observed decrease in the lower-tropospheric cloudiness with an increase in SST might be because of the increase in the upward transport of moisture from the lower troposphere by convection to the middle and upper troposphere, where they cause cloud formation.

While SST controls the moist static energy available at the surface, which is crucial for local convection, the updraft in the atmosphere is also influenced by remote forcing of tropospheric circulation driven by the large-scale spatial gradient of SST and atmospheric thermodynamical variables. The observed variations in the SST dependence of the vertical distribution of cloudiness are explainable based on variations in the above-mentioned parameters. Among them, the variations of CAPE and upper-level divergence with SST are strikingly similar to the corresponding variations in the vertical distribution of clouds with SST, especially the rapid development of convection above 28°C and little variations in the vertical distribution of clouds in the SST band of 29°–30.5°C. One of the striking features observed in Figs. 25 is the increase in the cloud-top altitude up to which significant convection takes place, even for SST > 29.5°C, where the spatial gradient of SST as well as the increase in upper-level divergence is negligible. This might have been primarily caused by (i) an increase in CAPE, though weak, in the SST range of 29°–30°C; and (ii) an increase in the moist static energy available at the surface through an increase in SST.

The frequency of occurrence of upper-level clouds, especially above ~10-km altitude, is found to increase substantially with altitude (Figs. 2, 3). This increase is most prominent in the SST range of 29.5°–30.5°C. It is primarily due to the increase in cirrus clouds, which have longer lifetime compared to the deep convective clouds in the middle and upper troposphere. Another important feature observed in the present study is the increase in the probability of occurrence of deep convective clouds having a thickness of >12 km. Prominence of this secondary peak increases with SST. The probability of occurrence of clouds as a function of cloud thickness in this secondary peak can be explained based on the potential temperature lapse rate. Altitude variation of potential temperature shows that, on average, the lapse rate of potential temperature (LRPT = /dz) is minimum near the convective tropopause altitude, which is the base of the tropical tropopause layer (TTL; Gettelman and Forster 2002). The magnitude of LRPT is generally largest in the lower and middle troposphere and decreases above ~10-km altitude to attain a minimum value in the convective tropopause around 12–15 km (e.g., Meenu et al. 2010). Under highly convective conditions, the LRPT can be very close to zero or sometimes even negative (e.g., Mehta et al. 2008). This means that, if an updraft has sufficient energy to reach in the vicinity of the convective tropopause, then it requires only a little more energy to be pushed up by another couple of kilometers or to be naturally pushed up because of the instability (LRPT < 0°C km−1) in the vicinity of the convective tropopause. Hartmann (2002) proposed that the most active convection will be limited to the altitude range where radiative cooling is efficient: this occurs only below the ~200-hPa level, as the water vapor amount is considerably smaller in the colder regions above. Based on this, they argued that the tropical convective clouds might detrain preferentially near the 200-hPa level. The above-mentioned mechanisms might be mainly responsible for the secondary peak observed in the probability of occurrence of clouds as a function of cloud thickness.

This study refines the extensively investigated SST dependence of cloudiness reported in the literature and identifies the potential causes for the observed differences. However, the variations in cloudiness with SST for SST > 30.5°C and the factors that limit the maximum SST to less than ~32°C are not addressed here. The CALIPSO and CloudSat data have the potential to address this fundamental science question. The tropospheric humidity might be responsible for some of the observed variability in the convective cloudiness (and vice versa), which needs to be investigated.

Acknowledgments

The 2B-GEOPROF-lidar data were obtained from the CloudSat Data Processing Center through its website (www.cloudsat.cira.colostate.edu). The TMI data were produced by Remote Sensing Systems and sponsored by the National Aeronautics and Space Administration Earth Science MEaSUREs DISCOVER Project and are available at online (www.remss.com). The ERS-1/2 and QuikSCAT scatterometer data were taken from CERSAT, at IFREMER, Plouzané, France. NCEP reanalysis data were provided by the NOAA/OAR/ESRL PSD, Boulder, Colorado, from their website (www.esrl.noaa.gov/psd). A. K. M. Nair is supported by ISRO through a research fellowship. Detailed constructive suggestions by two anonymous reviewers considerably contributed to improving the science content of this paper. This study was carried out as part of the NOBLE Project of ISRO-GBP.

REFERENCES

  • Allan, R. P., , and B. J. Soden, 2007: Large discrepancy between observed and simulated precipitation trends in the ascending and descending branches of the tropical circulation. Geophys. Res. Lett., 34, L18705, doi:10.1029/2007GL031460.

    • Search Google Scholar
    • Export Citation
  • Bhat, G. S., , J. Srinivasan, , and S. Gadgil, 1996: Tropical deep convection, convective available potential energy and sea surface temperature. J. Meteor. Soc. Japan, 74, 155166.

    • Search Google Scholar
    • Export Citation
  • Bony, S., , K.-M. Lau, , and Y. C. Sud, 1997: Sea surface temperature and large-scale circulation influences on tropical greenhouse effect and cloud radiative forcing. J. Climate,10, 2055–2077.

  • Bretherton, C. S., , M. E. Peters, , and L. E. Back, 2004: Relationships between water vapor path and precipitation over the tropical oceans. J. Climate, 17, 15171528.

    • Search Google Scholar
    • Export Citation
  • Cess, R. D., and Coauthors, 1996: Cloud feedback in atmospheric general circulation models: An update. J. Geophys. Res., 101 (D8), 12 79112 794.

    • Search Google Scholar
    • Export Citation
  • Cho, H., , C.-H. Ho, , and Y.-S. Choi, 2012: The observed variation in cloud-induced longwave radiation in response to sea surface temperature over the Pacific warm pool from MTSAT-1R imagery. Geophys. Res. Lett., 39, L18802, doi:10.1029/2012GL052700.

    • Search Google Scholar
    • Export Citation
  • Del Genio, A. D., 2012: Representing the sensitivity of convective cloud systems to tropospheric humidity in general circulation models. Surv. Geophys., 33, 637656, doi:10.1007/s10712-011-9148-9.

    • Search Google Scholar
    • Export Citation
  • Derbyshire, S. H., , I. Beau, , P. Bechtold, , J.-Y. Grandpeix, , J.-M. Piriou, , J.-L. Redelsperger, , and P. M. M. Soares, 2004: Sensitivity of moist convection to environmental humidity. Quart. J. Roy. Meteor. Soc., 130, 30553079.

    • Search Google Scholar
    • Export Citation
  • Fu, Q., , M. Baker, , and D. L. Hartmann, 2002: Tropical cirrus and water vapor: An effective Earth infrared iris feedback? Atmos. Chem. Phys., 2, 3137, doi:10.5194/acp-2-31-2002.

    • Search Google Scholar
    • Export Citation
  • Fu, R., , A. D. Del Genio, , and W. B. Rossow, 1994: Influence of ocean surface conditions on atmospheric thermodynamic structure and deep convection. J. Climate, 7, 10921108.

    • Search Google Scholar
    • Export Citation
  • Gadgil, S., , P. V. Joseph, , and N. V. Joshi, 1984: Ocean–atmosphere coupling over monsoon regions. Nature, 312, 141143.

  • Gettelman, A., , and P. M. Forster, 2002: A climatology of the tropical tropopause layer. J. Meteor. Soc. Japan, 80, 911924.

  • Graham, N. E., , and T. P. Barnett, 1987: Sea surface temperature, surface wind divergence, and convection over tropical oceans. Science,238, 657–659, doi:10.1126/science.238.4827.657.

  • Hartmann, D. L., 2002: An important constraint on tropical cloud–climate feedback. Geophys. Res. Lett., 29, 1951, doi:10.1029/2002GL015835.

    • Search Google Scholar
    • Export Citation
  • Haynes, J. M., , and G. L. Stephens, 2007: Tropical oceanic cloudiness and the incidence of precipitation: Early results from CloudSat. Geophys. Res. Lett., 34, L09811, doi:10.1029/2007GL029335.

    • Search Google Scholar
    • Export Citation
  • Holloway, C. E., , and J. D. Neelin, 2009: Moisture vertical structure, column water vapor, and tropical deep convection. J. Atmos. Sci., 66, 16651683.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc.,77, 437–471.

  • Lau, K.-M., , H.-T. Wu, , and S. Bony, 1997: The role of large-scale atmospheric circulation in the relationship between tropical convection and sea surface temperature. J. Climate,10, 381–392.

  • Lindzen, R. S., , and S. Nigam, 1987: On the role of sea surface temperature gradients in forcing low-level winds and convergence in the tropics. J. Atmos. Sci.,44, 2418–2436.

  • Mace, G. G., , Q. Q. Zhang, , M. Vaughan, , R. Marchand, , G. Stephens, , C. Trepte, , and D. Winker, 2009: A description of hydrometeor layer occurrence statistics derived from the first year of merged CloudSat and CALIPSO data. J. Geophys. Res.,114, D00A26, doi:10.1029/2007jd009755.

  • Martin, E. R., , and C. Schumacher, 2012: The relationship between tropical warm pool precipitation, sea surface temperature, and large-scale vertical motion in IPCC AR4 models. J. Atmos. Sci., 69, 185194.

    • Search Google Scholar
    • Export Citation
  • Meenu, S., , K. Rajeev, , K. Parameswaran, , and A. K. M. Nair, 2010: Regional distribution of deep clouds and cloud top altitudes over the Indian subcontinent and the surrounding oceans. J. Geophys. Res.,115, D05205, doi:10.1029/2009JD011802.

  • Meenu, S., , K. Rajeev, , and K. Parameswaran, 2011: Regional and vertical distribution of semitransparent cirrus clouds over the tropical Indian region derived from CALIPSO data. J. Atmos. Sol. Terr. Phys., 73, 19671979, doi:10.1016/j.jastp.2011.06.007.

    • Search Google Scholar
    • Export Citation
  • Meenu, S., , K. Parameswaran, , and K. Rajeev, 2012: Role of sea surface temperature and wind convergence in regulating convection over the tropical Indian Ocean. J. Geophys. Res., 117, D14102, doi:10.1029/2011JD016947.

    • Search Google Scholar
    • Export Citation
  • Mehta, S. K., , B. V. K. Murthy, , D. N. Rao, , M. V. Ratnam, , K. Parameswaran, , K. Rajeev, , C. S. Raju, , and K. G. Rao, 2008: Identification of tropical convective tropopause and its association with cold point tropopause. J. Geophys. Res., 113, D00B04, doi:10.1029/2007JD009625.

    • Search Google Scholar
    • Export Citation
  • Nair, A. K. M., , K. Rajeev, , S. Sijikumar, , and S. Meenu, 2011: Characteristics of a persistent “pool of inhibited cloudiness” and its genesis over the Bay of Bengal associated with the Asian summer monsoon. Ann. Geophys., 29, 12471252, doi:10.5194/angeo-29-1247-2011.

    • Search Google Scholar
    • Export Citation
  • Pierrehumbert, R. T., 1996: Some remarks on mechanisms for the regulation of tropical sea surface temperature. Clouds, Chemistry and Climate, P. J. Crutzen and V. Ramanathan, Eds., NATO ASI Series I, Vol. 35, Springer, 153–172.

  • Rajeev, K., and Coauthors, 2008: Observational assessment of the potential of satellite-based water vapor and thermal IR brightness temperatures in detecting semitransparent cirrus. Geophys. Res. Lett., 35, L08808, doi:10.1029/2008GL033393.

    • Search Google Scholar
    • Export Citation
  • Ramanathan, V., , and W. Collins, 1991: Thermodynamic regulation of ocean warming by cirrus clouds deduced from observations of the 1987 El Niño. Nature, 351, 2732, doi:10.1038/351027a0.

    • Search Google Scholar
    • Export Citation
  • Reynolds, R. W., , C. L. Gentemann, , and G. K. Corlett, 2010: Evaluation of AATSR and TMI satellite SST data. J. Climate, 23, 152165.

  • Rondanelli, R., , and R. S. Lindzen, 2008: Observed variations in convective precipitation fraction and stratiform area with sea surface temperature. J. Geophys. Res., 113, D16119, doi:10.1029/2008JD010064.

    • Search Google Scholar
    • Export Citation
  • Sabin, T. P., , C. A. Babu, , and P. V. Joseph, 2013: SST–convection relation over tropical oceans. Int. J. Climatol., 33, 14241435, doi:10.1002/joc.3522.

    • Search Google Scholar
    • Export Citation
  • Sijikumar, S., , and K. Rajeev, 2012: Role of the Arabian Sea warm pool on the precipitation characteristics during the monsoon onset period. J. Climate, 25, 18901899.

    • Search Google Scholar
    • Export Citation
  • Stephens, G. L., 2005: Cloud feedbacks in the climate system: A critical review. J. Climate, 18, 237273.

  • Su, H., , J. H. Jiang, , J. Teixeira, , A. Gettelman, , X. Huang, , G. Stephens, , D. Vane, , and V. S. Perun, 2011: Comparison of regime-sorted tropical cloud profiles observed by CloudSat with GEOS5 analyses and two general circulation model simulations. J. Geophys. Res., 116, D09104, doi:10.1029/2010JD014971.

    • Search Google Scholar
    • Export Citation
  • Su, H., and Coauthors, 2013: Diagnosis of regime-dependent cloud simulation errors in CMIP5 models using “A-Train” satellite observations and reanalysis data. J. Geophys. Res., 118, 27622780, doi:10.1029/2012JD018575.

    • Search Google Scholar
    • Export Citation
  • Sud, Y. C., , G. K. Walker, , and K.-M. Lau, 1999: Mechanisms regulating sea-surface temperatures and deep convection in the tropics. Geophys. Res. Lett., 26, 1019–1022, doi:10.1029/1999GL900197.

    • Search Google Scholar
    • Export Citation
  • Sud, Y. C., , G. K. Walker, , Y. P. Zhou, , G. A. Schmidt, , K.-M. Lau, , and R. F. Cahalan, 2008: Effects of doubled CO2 on tropical sea surface temperatures (SSTs) for onset of deep convection and maximum SST: Simulations based inferences. Geophys. Res. Lett.,35, L12707, doi:10.1029/2008GL033872.

  • Tompkins, A. M., 2001: On the relationship between tropical convection and sea surface temperature. J. Climate,14, 633–637.

  • Waliser, D. E., 1996: Formation and limiting mechanisms for very high sea surface temperature: Linking the dynamics and the thermodynamics. J. Climate,9, 161–188.

  • Waliser, D. E., , and N. E. Graham, 1993: Convective cloud systems and warm-pool sea surface temperatures: Coupled interactions and self-regulation. J. Geophys. Res., 98 (D7), 12 88112 893.

    • Search Google Scholar
    • Export Citation
  • Webb, M. J., , F. H. Lambert, , and J. M. Gregory, 2013: Origins of differences in climate sensitivity, forcing and feedback in climate models. Climate Dyn., 40, 677707, doi:10.1007/s00382-012-1336-x.

    • Search Google Scholar
    • Export Citation
  • Williams, K. D., , M. A. Ringer, , and C. A. Senior, 2003: Evaluating the cloud response to climate change and current climate variability. Climate Dyn., 20, 705721.

    • Search Google Scholar
    • Export Citation
  • Winker, D. M., , W. H. Hunt, , and M. J. McGill, 2007: Initial performance assessment of CALIOP. Geophys. Res. Lett.,34, L19803, doi:10.1029/2007GL030135.

  • Wood, R., , and C. S. Bretherton, 2006: On the relationship between stratiform low cloud cover and lower-tropospheric stability. J. Climate, 19, 64256432.

    • Search Google Scholar
    • Export Citation
  • Yuan, J., , D. L. Hartmann, , and R. Wood, 2008: Dynamic effects on the tropical cloud radiative forcing and radiation budget. J. Climate, 21, 23372351.

    • Search Google Scholar
    • Export Citation
  • Zelinka, M. D., , and D. L. Hartmann, 2011: The observed sensitivity of high clouds to mean surface temperature anomalies in the tropics. J. Geophys. Res., 116, D23103, doi:10.1029/2011JD016459.

    • Search Google Scholar
    • Export Citation
  • Zhang, C., 1993: Large-scale variability of atmospheric deep convection in relation to sea surface temperature in the tropics. J. Climate,6, 1898–1913.

Save