• Donner, L. J., 1993: A cumulus parameterization including mass fluxes, vertical momentum dynamics, and mesoscale effects. J. Atmos. Sci., 50 , 889906.

    • Search Google Scholar
    • Export Citation
  • Donner, L. J., , C. J. Seman, , and R. S. Hemler, 1999: Three-dimensional cloud-system modeling of GATE convection. J. Atmos. Sci., 56 , 18851912.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K. A., 1991: A scheme for representing cumulus convection in large-scale models. J. Atmos. Sci., 48 , 23132335.

  • Emanuel, K. A., , and M. Živković-Rothman, 1999: Development and evaluation of a convection scheme for use in climate models. J. Atmos. Sci., 56 , 17661782.

    • Search Google Scholar
    • Export Citation
  • Fowler, L. D., , and D. A. Randall, 1996: Liquid and ice microphysics in the CSU General Circulation Model. Part III: Sensitivity to modeling assumptions. J. Climate, 9 , 561586.

    • Search Google Scholar
    • Export Citation
  • Fowler, L. D., , and D. A. Randall, 1999: Simulation of upper tropospheric clouds with the Colorado State University general circulation model. J. Geophys. Res., 104 , 61016121.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., 1998: Toward cloud resolving modeling of large-scale tropical circulations: A simple cloud microphysics parameterization. J. Atmos. Sci., 55 , 32833298.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., 2000: Cloud microphysics and the tropical climate: Cloud-resolving model perspective. J. Climate, 13 , 23062322.

  • Grabowski, W. W., 2001: Coupling cloud processes with the large-scale dynamics using the Cloud-Resolving Convection Parameterization (CRCP). J. Atmos. Sci., 58 , 978997.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., 2002: Large-scale organization of moist convection in idealized aquaplanet simulations. Int. J. Numer. Methods Fluids, 39 , 843853.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., 2003: MJO-like coherent structures: Sensitivity simulations using the Cloud-Resolving Convection Parameterization (CRCP). J. Atmos. Sci., 60 , 847864.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , and P. K. Smolarkiewicz, 1999: CRCP: A cloud resolving convection parameterization for modeling the tropical convecting atmosphere. Physica D, 133 , 171178.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , and M. W. Moncrieff, 2002: Large-scale organization of tropical convection in two-dimensional explicit numerical simulations: Effects of interactive radiation. Quart. J. Roy. Meteor. Soc., 128 , 23492375.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , and P. K. Smolarkiewicz, 2002: A multiscale anelastic model for meteorological research. Mon. Wea. Rev., 130 , 939956.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , M. W. Moncrieff, , and J. T. Kiehl, 1996a: Long-term behavior of precipitating tropical cloud systems: A numerical study. Quart. J. Roy. Meteor. Soc., 122 , 10191042.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , X. Wu, , and M. W. Moncrieff, 1996b: Cloud resolving modeling of tropical cloud systems during Phase III of GATE. Part I: Two-dimensional experiments. J. Atmos. Sci., 53 , 36843709.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , X. Wu, , M. W. Moncrieff, , and W. D. Hall, 1998: Cloud resolving modeling of tropical cloud systems during Phase III of GATE. Part II: Effects of resolution and the third spatial dimension. J. Atmos. Sci., 55 , 32643282.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , X. Wu, , and M. W. Moncrieff, 1999: Cloud resolving modeling of tropical cloud systems during Phase III of GATE. Part III: Effects of microphysical parameterizations. J. Atmos. Sci., 56 , 23842402.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , J-I. Yano, , and M. W. Moncrieff, 2000: Cloud-resolving modeling of tropical circulations driven by large-scale SST gradients. J. Atmos. Sci., 57 , 20222039.

    • Search Google Scholar
    • Export Citation
  • Hayashi, Y-Y., , and A. Sumi, 1986: The 30–40 day oscillations simulated in an “aqua planet” model. J. Meteor. Soc. Japan, 64 , 451467.

    • Search Google Scholar
    • Export Citation
  • Kelly, M. A., , D. A. Randall, , and G. L. Stephens, 1999: A simple radiative-convective model with a hydrological cycle and interactive clouds. Quart. J. Roy. Meteor. Soc., 125 , 837869.

    • Search Google Scholar
    • Export Citation
  • Khairoutdinov, M. F., , and D. A. Randall, 2001: A cloud resolving model as a cloud parameterization in the NCAR Community Climate System Model: Preliminary results. Geophys. Res. Lett., 28 , 36173620.

    • Search Google Scholar
    • Export Citation
  • Kiehl, J. T., , J. J. Hack, , and B. P. Briegleb, 1994: The simulated Earth radiation budget of the National Center for Atmospheric Research community climate model CCM2 and comparisons with the Earth Radiation Budget Experiment (ERBE). J. Geophys. Res., 99 , 2081520827.

    • Search Google Scholar
    • Export Citation
  • Lipps, F. B., , and R. S. Hemler, 1986: Numerical simulation of deep tropical convection associated with large-scale convergence. J. Atmos. Sci., 43 , 17961816.

    • Search Google Scholar
    • Export Citation
  • McFarquhar, G. M., , and A. J. Heymsfield, 1997: Parameterization of tropical cirrus ice crystal size distributions and implications for radiative transfer: Results from CEPEX. J. Atmos. Sci., 54 , 21872200.

    • Search Google Scholar
    • Export Citation
  • Rennó, N. O., , K. A. Emanuel, , and P. H. Stone, 1994: Radiative-convective model with an explicit hydrologic cycle. 1: Formulation and sensitivity to model parameters. J. Geophys. Res., 99 , 1442914441.

    • Search Google Scholar
    • Export Citation
  • Robe, F. R., , and K. A. Emanuel, 1996: Moist convective scaling: Some inferences from three-dimensional cloud ensemble simulations. J. Atmos. Sci., 53 , 32653275.

    • Search Google Scholar
    • Export Citation
  • Smolarkiewicz, P. K., , L. G. Margolin, , and A. Wyszogrodzki, 2001: A class of nonhydrostatic global models. J. Atmos. Sci., 58 , 349364.

    • Search Google Scholar
    • Export Citation
  • Sui, C-H., , K. M. Lau, , W-K. Tao, , and J. Simpson, 1994: The tropical water and energy cycles in a cumulus ensemble model. Part I: Equilibrium climate. J. Atmos. Sci., 51 , 711728.

    • Search Google Scholar
    • Export Citation
  • Tao, W-K., , and S-T. Soong, 1986: A study of the response of deep tropical clouds to mesoscale processes: Three-dimensional numerical experiments. J. Atmos. Sci., 43 , 26532676.

    • Search Google Scholar
    • Export Citation
  • Tao, W-K., , J. Simpson, , and S-T. Soong, 1987: Statistical properties of a cloud ensemble: A numerical study. J. Atmos. Sci., 44 , 31753187.

    • Search Google Scholar
    • Export Citation
  • Tompkins, A. M., 2000: The impact of dimensionality on long-term cloud-resolving model simulations. Mon. Wea. Rev., 128 , 15211535.

  • Weller, R. A., , and S. P. Anderson, 1996: Surface meteorology and air–sea fluxes in the western equatorial Pacific warm pool during the TOGA Coupled Ocean–Atmosphere Response Experiment. J. Climate, 9 , 19591990.

    • Search Google Scholar
    • Export Citation
  • Wu, X., , W. D. Hall, , W. W. Grabowski, , M. W. Moncrieff, , W. D. Collins, , and J. T. Kiehl, 1999: Long-term behavior of cloud systems in TOGA COARE and their interactions with radiative and surface processes. Part II: Effects of cloud microphysics on cloud–radiation interaction. J. Atmos. Sci., 56 , 31773195.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Results for simulation SP0 at (left) the equator and (right) at 51°S. (top) Hovmöller diagrams of the surface precipitation for the entire simulation. Precipitation intensity larger than 0.2 and 5 mm h−1 are shown using gray and black shading, respectively. (middle and bottom) Zonal distributions of the total surface heat flux (sensible plus latent) and the surface precipitation, respectively, at latitudes of the Hovmöller diagrams and on day 80

  • View in gallery

    As in Fig. 1, but for simulation LP0

  • View in gallery

    Evolution of the globally averaged (top) density-weighted temperature and (bottom) precipitable water for LP0 (dashed lines) and SP0 (solid lines)

  • View in gallery

    As in Fig. 3, but for LP and SP

  • View in gallery

    Quasi-equilibrium globally averaged profiles of the temperature, water vapor mixing ratio, relative humidity, and cloud fraction for SP0 (dashed lines) and LP0 (solid lines)

  • View in gallery

    The difference between globally averaged profiles of the equivalent potential temperature θe for SP0 and LP0. The profiles shown represent deviations from the average of the two θe profiles. Dashed (solid) lines show profiles for SP0 (LP0). The profiles marked “1” (thin lines) are for days 1–5, whereas those marked as “2” (thick lines) show the quasi-equilibrium profiles

  • View in gallery

    Quasi-equilibrium globally averaged profiles of the latent heating in SP0 and LP0 for days 71–80. The prescribed radiative cooling QR is shown for a reference

  • View in gallery

    As in Fig. 5, but for SP and LP

  • View in gallery

    As in Fig. 6, but for SP and LP

  • View in gallery

    Evolution of globally averaged cloud fraction profiles in LP and SP. Thin solid lines show profiles for days 1–5, thin dashed profiles are for days 11–15, and thick lines are the quasi-equilibrium profiles (days 56–60)

  • View in gallery

    Globally averaged radiative cooling profiles in LP and SP for days 1–5 (solid lines) and days 11–15 (dashed lines)

  • View in gallery

    Quasi-equilibrium globally averaged profiles of (left) the radiative cooling for LP and (right) SP. Thin solid, thick solid, and dashed lines represent total, longwave, and shortwave radiative cooling, respectively

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 9 9 2
PDF Downloads 2 2 0

Impact of Cloud Microphysics on Convective–Radiative Quasi Equilibrium Revealed by Cloud-Resolving Convection Parameterization

View More View Less
  • 1 National Center for Atmospheric Research,* Boulder, Colorado
© Get Permissions
Full access

Abstract

This paper investigates how cloud microphysics impact global convective–radiative quasi equilibrium on a constant-SST aquaplanet. A novel approach, the cloud-resolving convection parameterization (CRCP), also known as “superparameterization,” resolves both global dynamics and cloud-scale dynamics, as well as small-scale coupling among convective, radiative, and surface processes, within a single computational framework. As a result, the effects of cloud microphysics on the tropical large-scale dynamics and climate can be investigated with greater confidence than traditional approaches allow. Idealized simulations highlight extreme effects of the microphysical parameterizations to expose the impacts.

The results suggest that cloud microphysics impact quasi-equilibrium temperature and moisture profiles substantially, but the relative humidity is only weakly affected. Small cloud and precipitation particles result in a climate that is warmer and moister. This is explained by the impact of cloud microphysics on the coupling between convection and surface exchange during the approach to quasi equilibrium. With interactive radiation, this effect is supplemented by gradual evolution of mean cloud fraction profiles and the mean radiative cooling. The mean quasi-equilibrium radiative cooling is weaker in the simulation featuring small cloud and precipitation particles. Cloud–radiation interactions, explicitly treated in CRCP, play a significant role in setting up the quasi-equilibrium cloudiness.

The simulations further support the conjecture that the main impact of cloud microphysics in the Tropics is on the net energy budget at the ocean surface. The net energy flux into the ocean in quasi equilibrium is considerably smaller in simulations having small cloud and precipitation particles.

Corresponding author address: Dr. Wojciech W. Grabowski, NCAR, P.O. Box 3000, Boulder, CO 80307-3000. Email: grabow@ncar.ucar.edu

Abstract

This paper investigates how cloud microphysics impact global convective–radiative quasi equilibrium on a constant-SST aquaplanet. A novel approach, the cloud-resolving convection parameterization (CRCP), also known as “superparameterization,” resolves both global dynamics and cloud-scale dynamics, as well as small-scale coupling among convective, radiative, and surface processes, within a single computational framework. As a result, the effects of cloud microphysics on the tropical large-scale dynamics and climate can be investigated with greater confidence than traditional approaches allow. Idealized simulations highlight extreme effects of the microphysical parameterizations to expose the impacts.

The results suggest that cloud microphysics impact quasi-equilibrium temperature and moisture profiles substantially, but the relative humidity is only weakly affected. Small cloud and precipitation particles result in a climate that is warmer and moister. This is explained by the impact of cloud microphysics on the coupling between convection and surface exchange during the approach to quasi equilibrium. With interactive radiation, this effect is supplemented by gradual evolution of mean cloud fraction profiles and the mean radiative cooling. The mean quasi-equilibrium radiative cooling is weaker in the simulation featuring small cloud and precipitation particles. Cloud–radiation interactions, explicitly treated in CRCP, play a significant role in setting up the quasi-equilibrium cloudiness.

The simulations further support the conjecture that the main impact of cloud microphysics in the Tropics is on the net energy budget at the ocean surface. The net energy flux into the ocean in quasi equilibrium is considerably smaller in simulations having small cloud and precipitation particles.

Corresponding author address: Dr. Wojciech W. Grabowski, NCAR, P.O. Box 3000, Boulder, CO 80307-3000. Email: grabow@ncar.ucar.edu

1. Introduction

This study is motivated by the discussion in Grabowski (2000, hereafter G00). In a nutshell, the essence of tropical climate is the interaction between moist convection and large-scale dynamics. This interaction involves dynamical, thermodynamical, radiative, and surface processes acting across a wide range of scales. The structure of the tropical atmosphere should depend on cloud microphysics that are responsible for the release of latent heat, the formation of precipitation, and its evaporation outside clouds [see discussions in Emanuel (1991), his Fig. 6 in particular; and in Emanuel and Živković-Rothman (1999)]. Because of the range of spatial and temporal scales involved, effects of cloud processes on tropical dynamics and climate have been previously investigated using idealized 1D models (e.g., Emanuel 1991; Renno et al. 1994; Kelly et al. 1999) or traditional general circulation models (GCMs; see, e.g., Fowler and Randall 1996, 1999). However, computational limitations dictate that these approaches apply parameterizations of not only cloud microphysics but also cloud dynamics.

These traditional approaches are poorly suited to quantitatively address the relation between cloud physics and tropical climate. First, these approaches include simple representations of cloud-scale and mesoscale dynamics. This is important for cloud microphysics, as well as the representation of convective organization and cloud life cycles. For instance, the organization of moist convection into mesoscale convective systems is seldom considered in traditional parameterizations (a noticeable exception is Donner 1993). Secondly, effects of cloud processes in large-scale models are felt immediately by large-scale dynamics. In reality, small-scale and mesoscale dynamics are the first to respond, and only the residual imbalances are available to drive large-scale circulations. For instance, the effect of the size of cloud particles on radiative fluxes operates at cloud-scale level, not at scales resolved by large-scale models. Consequently, the link between cloud microphysics and the tropical climate is best investigated using a model that resolves cloud-scale and mesoscale dynamics.

Unfortunately, resolving both large-scale and cloud-scale dynamics using traditional computational approaches on timescales relevant to climate is computationally limited. A cloud-resolving model with a horizontal grid spacing of 1 km requires a horizontal domain of the order of 107–108 km2 to explicitly couple cloud-scale, mesoscale, and large-scale dynamics. Such simulations, arguably possible on most powerful modern computers, are impractical because up to a million time steps are required to study coupled dynamics on a timescale of several weeks. However, significant progress can be made using a novel computational approach that couples cloud-scale and large-scale dynamics within a single dynamical framework. This approach uses a two-dimensional (slab symmetric and horizontally periodic) cloud-resolving model (with horizontal grid spacing of about 1 km) within each column of a large-scale model (with horizontal grid spacing of several hundreds of kilometers) to represent cloud-scale and mesoscale dynamics, and their coupling with radiative and surface processes. This approach, referred to as cloud-resolving convection parameterization (CRCP; see Grabowski and Smolarkiewicz 1999; Grabowski 2001, hereafter G01; Grabowski 2002, hereafter G02; and Grabowski 2003, hereafter G03), was applied to idealized large-scale convection organization on a rotating constant–sea surface temperature (SST) aquaplanet. These simulations created large-scale equatorially trapped disturbances resembling the Madden–Julian Oscillations (MJO). Recently, the same approach was applied as a “superparameterization” in the National Center for Atmospheric Research (NCAR) Community Climate System Model by Khairoutdinov and Randall (2001).

CRCP is well suited to study the impact of cloud-scale processes on climate because, unlike traditional approaches, it allows dynamical feedback to changes in cloud parameters to occur at physically reasonable spatial scales associated with the cloud and mesoscale dynamics. For instance, the indirect impact of atmospheric aerosols on climate, that is, the impact on processes concerning formation and growth of cloud droplets and ice crystals inside clouds, can be addressed with confidence using CRCP.

The paper is organized as follows. The next section presents numerical simulations discussed in this paper. Section 3 discusses the results, starting with the prescribed radiation simulations and finishing with the interactive radiation. Conclusions are drawn in section 4.

2. The model: Cloud microphysics and simulations performed

Model simulations presented in this paper consider convective–radiative quasi equilibrium on an idealized constant-SST (30°C) aquaplanet, with the same size and rotation as the earth. The global model is the anelastic nonhydrostatic two-time-level nonoscillatory forward-in-time Eulerian/semi-Lagrangian model in spherical geometry (Smolarkiewicz et al. 2001; Grabowski and Smolarkiewicz 2002). The Eulerian version of the model is used. The global model applies a low horizontal resolution (32 × 16), with 51 levels in the vertical and a uniform grid length of 0.5 km. The global model time step is 12 min.

The two-dimensional cloud-resolving model embedded in each column of the global model has horizontal periodic domain of 200 km with a 2-km grid length and it is aligned along the zonal (E–W) direction (see section 4 of G01). The vertical grid is the same as in the global model. The fact that the global model and cloud-scale models solve exactly the same equations and have the same vertical grid simplifies the coupling between the two models. The cloud model time step is 20 s. In addition, a gravity wave absorber is used in the uppermost 9 km of each cloud model with an inverse characteristic timescale increasing linearly from 0 at the bottom of the absorber to 1/600 s−1 at the top of each model domain.

The initial thermodynamic profiles as well as the reference profiles are the 0000 UTC 1 September 1974 Global Atmospheric Research Programme (GARP) Atlantic Tropical Experiment (GATE) sounding (i.e., as in Grabowski et al. 1996b, 1998). The atmosphere is assumed to be initially at rest. Free slip lower boundary conditions are applied (i.e., no surface friction). The thermodynamic fields within CRCP domains are initiated by applying instantaneous cloud-scale fields from a 2D convective–radiative quasi-equilibrium single CRCP simulation model to all columns (see section 4 in G01). The global model is initiated by averaging the CRCP domain fields and applying small-amplitude perturbations (0.3 K) to the temperature field.

The simulations apply either prescribed radiative cooling (as in G01 and G02), or an interactive radiation transfer model. The prescribed radiation simulations are instructive by separating the impacts on the quasi-equilibrium cloud dynamics from the impact on radiative processes. The prescribed radiative cooling is as in G01 and G02 (i.e., −1.5 K day−1 below 12 km, linearly decreasing to 0 between 12 and 15 km, and 0 above 15 km). The prescribed radiation simulations are run for 80 days.

The interactive radiation simulations apply the NCAR Community Climate Model radiation code (Kiehl et al. 1994; see also section 4 in G00). A diurnal cycle of solar radiation is not considered, the solar constant is reduced to 436 W m−2 (i.e., the nominal solar constant at an equinox divided by π), and a zero zenith angle is assumed over the entire aquaplanet. Although highly idealized, this is consistent with the “Tropics everywhere” design of aquaplanet simulations. The radiative calculations are performed within CRCP domains once every global model time step (i.e., every 12 min). The condensed water fields for radiative transfer calculations include cloud water, cloud ice, and snow, but not rain. The effective radius for cloud water droplets is assumed to be 10 μm, whereas for ice particles the effective radius depends on the ice water content based on measurements in tropical anvils reported by McFarquhar and Heymsfield (1997) as in (2) of G00. We stress that the radiative transfer applies cloud-scale fields supplied by CRCP as in traditional cloud-resolving simulations (e.g., Wu et al. 1999) and it does not involve any subgridscale assumptions concerning cloud structure and overlap. This is different from Khairoutdinov and Randall (2001) who used the cloud-resolving model only to replace the deep convection parameterization in the climate model. The interactive radiation simulations are run for 60 days.

The emphasis herein is on cloud microphysics, so a summary is helpful (see Grabowski 1998 for details). The microphysical scheme considers just two classes of condensed water: cloud condensate and precipitation. They assume the liquid state (cloud water, rain) for temperatures warmer than a threshold temperature Tw = 268 K, and solid (cloud ice, snow) for temperatures colder than a threshold of Ti = 253 K. Clouds are always at vapor saturation with respect to water for temperatures warmer than Tw, and with respect to ice for temperatures colder than Ti. Linear interpolations are used for intermediate temperatures. The requirement of saturation inside clouds prescribes the condensation/evaporation rate of the cloud condensate (cloud water, cloud ice). Precipitation (rain, snow) forms from the cloud condensate, grows by accretion of cloud condensate and by diffusion of water vapor (in the case of snow) inside clouds, and evaporates outside clouds. Conversion from cloud water to rain depends on the assumed cloud condensation nuclei (CCN) spectrum. Conversion from cloud ice to snow is a strong function of the temperature and it is based on the rate of growth of ice crystals. It is the same in all experiments.

Results from two categories of sensitivity simulations are presented following G00. Simulations that assume small cloud droplets (concentration of 2000 cm−3) and small precipitation particles as prescribed by the intercept parameter of the exponential particle size distribution, No = 109 m−4 for both rain and snow, is referred to as SP0 (with prescribed radiation) and SP (with interactive radiation), where SP stands for “small particles.” Respectively, simulation assuming large cloud droplets (concentration of 50 cm−3) and large precipitation particles (No = 105 m−4 for both rain and snow) are referred to as LP0 and LP, where LP stands for “large particles.”

The concentration of cloud droplets impacts the autoconversion from cloud water to rain in Berry's scheme as described in Grabowski (1998). The intercept parameter No affects the simulations through the sedimentation rate, as well as on the rate of growth/evaporation of precipitation particles for the same environmental conditions. As illustrated in Fig. 1 in G00, the sedimentation velocity for snow changes by a factor of 2 between simulations SP0/SP and LP0/LP; for rain, it changes by a factor of about 3. The growth rates for snow change by a factor of 2–3 (with the accretion rate decreasing and the deposition rate increasing when No changes from 105 to 109 m−4). For rain, the growth rate by accretion increases by a factor of 3–4, and the rate of evaporation increases by more than an order of magnitude.

As discussed in Grabowski et al. (1999; hereafter GWM99), the change of the evaporation rate of raindrops has a profound effect on the strength of convective and mesoscale downdrafts and, consequently, on the coupling between convection and surface processes. The change of concentration of cloud droplets from 50 to 2000 cm−3 decreases the autoconversion rate by more than two orders of magnitude for a cloud water mixing ratio of 0.5 g kg−1, and more than an order of magnitude for a cloud water mixing ratio of 5 g kg−1 (cf. Table 2 of GWM99). The changes in the representation of warm rain physics applied in the simulations SP0/SP and LP0/LP are extreme, whereas modification of the ice physics are more moderate, arguably within the limits of our knowledge of ice processes in tropical deep convection. In interactive radiation simulations, the effective radius of ice particles for the simulation SP is half that predicted by (2) in G00. For the LP, it is twice the value therein.

3. Results

a. General features

As far as large-scale organization of convection is concerned, the results from the four simulations are similar to those discussed in G01, G02, and G03. Convection lacks large-scale organization outside the equatorial waveguide and scattered convective clouds provide the heating required to balance radiative cooling. Within the waveguide, on the other hand, equatorially trapped disturbances develop and they resemble the MJO observed in the Tropics. This is illustrated in Figs. 1 and 2, which show Hovmöller diagrams of the surface precipitation, at the equator and in midlatitudes, for the entire length of simulations SP0 and LP0, as well as spatial distribution of surface precipitation and surface fluxes at day 80 and at latitudes corresponding to the Hovmöller diagrams. In both SP0 and LP0, MJO-like coherent structures within the equatorial waveguide are associated with highly localized surface precipitation and surface fluxes. A notable difference between SP0 and LP0 is a faster development of the wavenumber-1 MJO-like coherent structure in SP0. This is likely related to the impact of cloud microphysics on convective and mesoscale downdrafts and consequently on the convective enhancement of surface fluxes (cf. GWM99). As shown in G03 (section 5b), convective enhancement of surface fluxes plays an important role in the development of MJO-like coherent structures. As far as features other than these shown in Figs. 1 and 2 are concerned (e.g., zonal and meridional distribution of the temperature, moisture, wind field, etc.), the results are similar in SP0 and LP0 (not shown). Moreover, results for the simulations using the radiation transfer model, SP and LP are qualitatively similar to those in Figs. 1 and 2 and are not shown (cf. Fig. 5 in G03).

Figures 3 and 4 show evolution of the tropospheric density-weighted temperature and precipitable water for all four simulations. The figures show that the convective–radiative planetary quasi-equilibrium solutions are approximately attained close to the end of simulations. Weak temperature and moisture fluctuations are apparent in the quasi equilibrium, and a small temperature trend is still present at the end of simulations SP and LP. In general, the impact of cloud microphysics resembles results of cloud-resolving modeling of GATE cloud systems discussed in GWM99. For instance, the prescribed radiation simulations have lower density-weighted temperature and lower precipitable water than simulations using interactive radiation. Moreover, simulations featuring large cloud and precipitation particles (LP0 and LP) are colder (and thus more unstable) than corresponding simulations with small particles.

Table 1 summarizes model results by showing convective available potential energy (CAPE), lower-tropospheric cloud fraction, upward vertical velocity and convective velocity, and updraft and downdraft cloud mass fluxes. These are averages over the entire aquaplanet and over the last 10 days of the simulations. CAPE is calculated using a reversible parcel including ice physics, but the condensate loading is limited to 3 g kg−1. Cloud fraction is calculated within CRCP domains and is defined as a ratio between the number of cloudy grid boxes (total condensate, cloud plus precipitation, larger than 0.01 g kg−1) and the total number of grid boxes at a given level. The number shown in Table 1 comes from averaging cloud fraction across the lower troposphere (2–6 km) to avoid complications due to stratiform anvil clouds. The upward vertical velocity and convective velocity are defined at any level as a mean vertical velocity in cloudy grid boxes and in grid boxes that are cloudy and feature vertical velocities larger than 1 m s−1, respectively (cf. Robe and Emanuel 1996). Finally, the updraft and downdraft cloud mass fluxes are defined as mass fluxes associated with cloudy grid boxes (cf. section 4b in GWM99). The velocities and mass fluxes represent averages over model levels between 3 and 13 km.

Results shown in Table 1 are consistent with Figs. 3 and 4 and they echo GWM99. For instance, the warmer temperature profiles in small-particle simulations result in the lower quasi-equilibrium CAPE values. The interactive radiation simulations are warmer (cf. Figs. 3 and 4) and they feature lower CAPE. This is consistent with weaker radiative cooling when compared to the one prescribed in SP0 and LP0 (cf. Fig. 12 to be discussed in section 3c). However, the range of CAPE values is large and this might be an artifact of low vertical resolution of CRCP, which is not adequate to resolve boundary layer processes. As expected, the higher CAPE simulations feature larger velocity of both the mean updraft and the mean updraft core. Small-particle simulations feature larger cloud fractions and larger downdraft cloud mass fluxes. Note that the inversely proportional relationship between CAPE and cloud mass fluxes (see discussion in section 4b of GWM99 and Fig. 7 therein) holds for prescribed radiation simulations, but not for the interactive radiation. This is likely related to the small-scale cloud–radiation interactions, an aspect further discussed in section 3c.

b. Prescribed radiation simulations

This section discusses the impact of cloud microphysics on convective–radiative quasi equilibrium when radiative cooling is prescribed, similar to Emanuel (1991) and Robe and Emanuel (1996). This helps separate the impact of cloud microphysics on cloud dynamics from that associated with radiative processes. However, one should distinguish between the impact of cloud microphysics on the dynamics of a single cloud system (not what we address here), from the impact on gross characteristics in convective–radiative quasi equilibrium (i.e., on ensembles of clouds and their interaction with the cloud-free environment).

Figure 5 shows profiles of the temperature, water vapor mixing ratio, relative humidity (defined with respect to water saturation), and cloud fraction (defined as a fraction of points within CRCP domains with total condensate larger than 0.01 g kg−1 at a given level), all globally averaged during the last 10 days of simulations. The figure shows that the simulation with small cloud and precipitation particles (SP0) is warmer and more moist than LP0. Cloud fractions in SP0 are about twice as large as in LP0. The potential temperature difference between SP0 and LP0 is largest in the upper troposphere (the maximum is about 7 K at 13 km). Somewhat surprising, but similar to GWM99, is the similarity of the relative humidity profiles, with SP0 being only about 10% higher than LP0.

The warmer troposphere in SP0 than in LP0 is consistent with GWM99 (cf. their Fig. 4). The warmer climate in simulations featuring small cloud and precipitation particles was interpreted in GWM99 as a result of enhanced surface fluxes. This was explained as the impact of cloud microphysics on the strength of convective and mesoscale downdrafts and, in turn, the impact of downdrafts on atmosphere–ocean exchange. Since SP0 and LP0 apply the same radiative cooling, the sum of surface sensible and latent fluxes must balance the prescribed radiative cooling in the convective–radiative quasi equilibrium. Before the quasi equilibrium is reached, however, the impact on surface fluxes explains the observed temperature and moisture trends.

The impact of surface fluxes on the temperature and moisture is best illustrated through the evolution of profiles of the equivalent potential temperature (θe). Figure 6 shows the difference between θe profiles for simulations SP0 and LP0, averaged over two 5-day periods: days 1–5 (i.e., during the approach to quasi equilibrium) and days 76–80 (i.e., in the quasi equilibrium). During the approach to quasi equilibrium, θe profiles in the free troposphere differ less then in the equilibrium, and the difference changes sign near the surface, for example, the warmer simulation SP0 has a colder boundary layer. This is reminiscent of results discussed in GWM99 (cf. their Fig. 4) and leads to enhanced (reduced) surface θe flux for SP0 (LP0). The difference in the surface θe fluxes results in corresponding modification of the tropospheric θe, which is the only way (if the exchange with stratosphere can be neglected) the difference between θe profiles can be changed in simulations with prescribed radiation. In the quasi equilibrium, the difference between θe profiles for SP0 and LP0 is almost constant between 2 and 14 km, with SP0 being about 7–8 K warmer than LP0. Near the surface, however, SP0 is about half a degree warmer than LP0. Because SP0 features stronger gustiness of surface winds due to the impact of cloud microphysics on convective and mesoscale downdrafts, the higher θe near the surface in SP0 is consistent with the same quasi-equilibrium surface θe flux as in LP0.

Because quasi-equilibrium temperature profiles in SP0 and LP0 feature dramatically different CAPE, it is worthwhile to highlight differences in processes responsible in maintaining the temperature profiles. The evolution of the domain-averaged potential temperature profile θ is deduced by horizontally averaging the temperature equation over the horizontally periodic computational domain (cf. Sui et al. 1994, their section 2b; Grabowski et al. 1996a, their section 4; Grabowski and Moncrieff 2002, their section 4a):
i1520-0442-16-21-3463-e1
where the overbar represents horizontal average over the entire planet at a given level; ρo(z) is the base-state anelastic density profile; primed variables represent deviations from the horizontal average; Lυ is the latent heat of vaporization; cp is the specific heat at constant pressure; c and e stand for rates of condensation/deposition and evaporation/sublimation,1 respectively; and Fθ represents the vertical flux of potential temperature due to processes parameterized by the model (such as the surface fluxes, subgrid-scale turbulent fluxes, gravity wave absorber in the upper part of the domain, etc.). The four terms on the right-hand side of (1) are the convective (or eddy), radiative cooling, latent heating, and subgrid-scale transport term, respectively.

Equation (1) shows that the evolution of the mean temperature at any level depends on the balance between radiative cooling, latent heating, convective transport, and parameterized subgrid-scale fluxes. The quasi-equilibrium temperature profile gets established once the sum of all terms on right-hand side of (1) becomes zero. The turbulent fluxes are typically small compared to other terms, except near the surface and near the tropopause, and radiative cooling is the same in LP0 and SP0. Consequently, the evolution of the average temperature depends to the first order on the interaction between latent heating and convective transport. In the upper troposphere, the latent heating is typically small and convective transport dominates (e.g., Sui et al. 1994, their Fig. 15; Grabowski et al. 1996a, their Fig. 12; Grabowski et al. 2000, their Fig. 11a).

Figure 7 shows global mean of the quasi-equilibrium (10-day average) latent heating for LP0 and SP0. The latent heating is larger than the prescribed radiative cooling between 2 and 12 km in both simulations. The convective term (together with subgrid-scale processes) is responsible for the transport of the excess heating toward the surface (where the surface sensible heat flux contributes to the budget as well) and toward the upper troposphere. In the upper troposphere, where a significant imbalance between radiative cooling and latent heating exists (between 11 and 15 km), the convective temperature flux divergence [the first term on the rhs of (1)] fills the gap.

The difference of the latent heating between LP0 and SP0 is relatively small (a few tenths of a degree per day), but has a distinct pattern: latent heating in SP0 is smaller in the midtroposphere, where latent heating outpowers radiative cooling, and it is larger in the upper troposphere, where the convective flux divergence term dominates. It follows that, in the layer between 11 and 15 km, the convective temperature flux divergence has to be smaller in SP0 than in LP0 by a few tenths of a degree per day. The opposite must be true in the midtroposphere. This is indeed observed in the model data (not shown).

In summary, cloud microphysics is capable of exerting significant impact on the mean temperature and moisture profiles even in the case where radiative impacts are neglected. The key is the impact of cloud microphysics on the coupling between convection and the ocean surface during the approach to quasi equilibrium. The next section examines the combined dynamical and radiative impacts on the global convective–radiative quasi equilibrium.

c. Interactive radiation

As far as the large-scale organization of convection is concerned, simulations with interactive and prescribed radiation are similar (cf. Figs. 1 and 5 in G03). However, the interactive radiation simulations (SP and LP herein) are considerably warmer and moister as documented in Fig. 4. This is further illustrated in Fig. 8, which shows the globally averaged profiles in the quasi equilibrium (last 10 days) for SP and LP, in the same format as Fig. 5. Figure 8 shows that the simulation with small cloud and precipitation particles (SP) is again warmer and moister than LP. Because the arguments in the previous section apply to the interactive simulations as well, the warmer climate of these simulations is a combination of the impact of cloud microphysics on the quasi-equilibrium cloud dynamics, and on radiative processes. The latter aspect will be discussed in more detail below. The potential temperature difference between SP and LP is largest in the upper troposphere (the maximum is about 12 K at 13 km). The differences in the relative humidity profiles are even smaller than in the prescribed radiation simulations, especially in the upper two-thirds of the troposphere. All these results are in general agreement with GWM99.

An intriguing impact of interactive radiation is revealed by comparing the cloud fraction profiles shown in Figs. 5 and 8. In the lower half of the troposphere, the differences between prescribed and interactive radiation profiles are similar and feature cloud fractions 2–3 times higher in SP0 and SP than in LP0 and LP. In the upper troposphere, on the other hand, the cloud fractions in the interactive radiation simulations are similar in SP and LP (in stark contrast to the differences between SP0 and LP0), except that clouds in SP are deeper. The latter is likely a result of slower sedimentation of small particles in SP (cf. Wu et al. 1999, their Fig. 9), although the interaction between dynamics and radiation, leading to the lofting of stratiform anvils as simulated in Donner et al. (1999), may contribute to the overall result as well. Arguably, the cloud–radiation interactions lead to faster dissipation of upper-tropospheric anvils in SP because the upper-tropospheric quasi-equilibrium cloud fractions in SP and LP are similar.

The warming and moistening accompanying the approach to the quasi equilibrium can again be illustrated by the evolution of θe profiles, as in Fig. 6. The profiles, averaged over days 1–5 and 55–60, are shown in Fig. 9. The warming of the mean θe profile due to the surface fluxes is similar to the prescribed radiation simulations (cf. Fig. 6). An important difference is evident in the upper troposphere where the interactive radiation in SP initially results in cooling, and later changes sign to make the upper troposphere warmer, thus adding to the relative warming due to the impact of cloud microphysics on surface fluxes.

The evolution of the upper-tropospheric temperature due to radiative processes is directly related to the changes of the upper-tropospheric cloudiness during the approach to quasi equilibrium. This is illustrated in Figs. 10 and 11, which show profiles of cloud fraction and radiative cooling averaged over two periods, days 1–5 and 11–15. During the approach to quasi equilibrium, the cloud fractions tend to decrease in the upper troposphere and increase in the lower troposphere. These changes seem small in LP, but they impact the mean radiative cooling considerably (cf. Fig. 11). In SP, the changes are significant, especially in the upper troposphere where the quasi-equilibrium cloud fractions are about half of the early values. Such a strong reduction of the upper-tropospheric cloudiness in SP is likely a combination of cloud–radiation interactions and weak radiative destabilization of the atmospheric column. Cloud–radiation interactions are dominated by the destabilization of upper parts of anvil clouds by strong radiative cooling (cf. Fig. 11) and are evident in cloud mass fluxes which show upper-tropospheric enhancements in both SP and LP (not shown). At the same time, however, small radiative cooling across the entire troposphere results in weak convection and consequently reduced cloudiness. These feedbacks illustrate an intricate balance among convection, clouds, surface processes, and radiative transfer, where cloud condensate and precipitation are essential components of small- and mesoscale circulations, and radiative transfer affects the strength of these circulations through the mean destabilization of the troposphere. It is important to stress that, both in nature and in the global model with CRCP, these interactions take place at the cloud scale.

Interactive radiation simulations are considerably warmer than the prescribed radiation ones, especially in the upper half of the troposphere. This is consistent with the estimates of CAPE shown in Table 1 and also with the fact that radiative cooling is weaker in SP and LP than prescribed in SP0 and LP0. Figure 12 shows global-mean quasi-equilibrium (10-day average) profiles of radiative cooling rates for LP and SP, partitioned into shortwave and longwave components. The cooling rates are similar in the lower troposphere, but in the upper half of the troposphere radiative cooling is significantly weaker in SP. The longwave and shortwave contributions show that the net impact is associated with stronger longwave cooling in LP that reaches the maximum at about 12 km, above the maximum shortwave heating. In SP, the longwave cooling and shortwave heating extend farther up, in agreement with the quasi-equilibrium cloud fraction profiles shown in Figs. 8 and 10.

Table 2 summarizes the differences between SP and LP in terms of the vertical energy fluxes, averaged over the entire planet for the last 10 days. The difference between net radiative fluxes at z = 18 km and at the surface (referred to as the net radiative flux divergence in the discussion below) illustrates the magnitude of radiative cooling averaged over the entire troposphere. The net radiative flux divergence is −80 and −92 W m−2 in SP and LP and it can be compared to about −140 W m−2 in SP0 and LP0. In quasi equilibrium, this energy loss has to be balanced by the sum of surface sensible and latent heat fluxes. This is not exactly true in Table 1—the vertical transport across the tropopause is the likely culprit. The imbalance (1–2 W m−2) is consistent with the eddy temperature flux at z = 18 km (not shown). The impact of cloud microphysics on surface fluxes is opposite to that observed in GWM99, that is, the average surface fluxes are smaller in the simulation with small cloud and precipitable particles (SP). This is because the total surface heat flux has to balance the radiative cooling in the quasi equilibrium, which is smaller in SP. It is likely, however, that the impact of microphysics on local surface fluxes (e.g., in the vicinity of convective systems) agrees with results of GWM99.

The changes in the flux divergence and the total surface heat flux between SP and LP are almost the same, that is, 12 W m−2 for the change of the flux divergence (from −80 to −92) and 13 W m−2 for the change of the total surface flux (from 78 to 91). The most significant difference is in the net radiative flux into the ocean, which changes from 196 W m−2 for SP to 245 W m−2 for LP. This change is 4 times larger than the impact of cloud microphysics on the atmospheric column. The strongly positive net surface energy flux (i.e., the net radiative flux minus the total surface heat flux; 154 W m−2 for LP and 118 W m−2 for SP) implies that the ocean should warm, as illustrated in idealized swamp–ocean simulations discussed in section 4 of G00.

The fact that the surface energy budget is strongly positive requires comment. First, the observed budget in the Tropics is rather small, about 20 W m−2 [e.g., 17.5 W m−2 for the tropical western Pacific warm pool during Tropical Ocean Global Atmosphere Coupled Ocean–Atmosphere Response Experiment (TOGA COARE) Weller and Anderson 1996, their Table 4]. This fits the estimate of the oceanic heat transport out of the Tropics as deduced by climate model simulations (cf. section 4 in G00). Nevertheless, the “Tropics everywhere” global model setup using the interactive radiation and a zero zenith angle across the entire aquaplanet is unrealistic. Moreover, such a setup overestimates the solar energy input because the top-of-the-atmosphere (TOA) downward solar flux is 436 W m−2 across the entire aquaplanet in current simulations, whereas the globally averaged value (i.e., including the equator-to-pole variations) is about 340 W m−2 on earth. It is also highly likely that the 2-km horizontal grid length applied in CRCP models excludes shallow convection and/or boundary layer cloudiness. These clouds have a strong cooling effect on the surface energy budget because they reflect solar radiation back to space but have a small impact on the longwave radiation (the cloud-top temperatures are close the surface temperature). In addition, the simulations do not include large-scale forcing due to Hadley and Walker circulations, which are associated with meridional and zonal SST gradients. Such large-scale circulations would tend to increase the cloud amount over warm SSTs and thus limit the downward solar flux at the surface. An attempt to take into account the impact of these large-scale circulations was included in idealized simulations of convective–radiative quasi equilibrium over the swamp–ocean discussed in section 4 of G00. Including “large-scale forcings” across the entire aquaplanet in the present simulations would be physically inconsistent, but may significantly reduce the surface energy budget. Since the SST is constant herein, the surface energy budget has no effect—it is discussed here only to highlight possible impacts of cloud microphysics on the ocean surface.

In conclusion, the results from interactive radiation simulations suggest that the effect of cloud microphysics can be understood as a combination of the impact on cloud dynamics (discussed in the previous section) and the impact on radiative processes. Apparently, these two work together, that is, they both lead to warmer mid- and upper troposphere in simulations with smaller cloud and precipitation particles. However, the main influence of cloud microphysics seems to be on the net energy budget at the ocean surface, with the net flux smaller in simulations with smaller cloud and precipitation particles, in agreement with G00.

4. Discussion and conclusions

This paper is a next step in the author's long-term goal to quantify the role of cloud microphysical processes in atmospheric general circulation and climate. Herein, a relatively simple system is considered (when compared to the earth's climate): a constant-SST “Tropics everywhere” aquaplanet with the same size and rotation as the earth. The word “climate” is used in this paper in a rather narrow sense of mean thermodynamical profiles of temperature, water vapor, relative humidity, and cloudiness, although the impact (or lack thereof) on climate fluctuations on intraseasonal timescales is also illustrated. It is argued that the answer to such a question necessarily requires cloud-resolving modeling. The complex interactions between cloud dynamics and microphysics on the one hand, and large-scale flow, radiation, and surface processes on the other, cannot be reliably studied using parameterized convection. The novel approach, which involves application of the cloud-resolving convection parameterization (CRCP, i.e., the “superparameterization”), addresses this issue with more confidence.

The cornerstone of CRCP is the application of a two-dimensional cloud-resolving model to represent cloud- and mesoscale processes in every column of a large-scale model. Past studies suggest that two-dimensional framework does offer a meaningful representation of cloud- and mesoscale processes compared to fully three-dimensional simulations (e.g., Lipps and Hemler 1986; Tao and Soong 1986; Tao et al. 1987; Grabowski et al. 1998; Tompkins 2000; see also a discussion in G01). However, Donner et al. (1999) noted potentially important differences between two- and three-dimensional cloud-resolving simulations when interactive radiation was applied. It would be useful from the point of view of the current study to compare convective–radiative quasi equilibria obtained using two- and three-dimensional cloud-resolving models with the emphasis on cloud microphysics. Regardless of the outcome of such a comparison, however, model simulations using CRCP have to be viewed only as an intermediate step, before cloud-resolving modeling of atmospheric general circulation becomes practical.

This paper discusses results from two sets of simulations that reach global convective–radiative quasi equilibria, one with prescribed radiation (SP0 and LP0) and one using an interactive radiation transfer model (SP and LP). Such an approach allows separating the impact of cloud microphysics on convective dynamics from the combined impact on convective dynamics and radiative transfer. In general, results discussed herein echo those of GWM99 and G00. Simulations with small cloud and precipitation particles result in global quasi-equilibrium temperature and moisture profiles that are warmer and more humid (by several degrees and a few grams per kilograms of water vapor) than simulations with large cloud and precipitation particles. This is explained as the impact of cloud and mesoscale downdrafts on surface energy fluxes during the approach to quasi equilibrium. In the quasi equilibrium, the mean potential temperature profile (and thus the overall stability of the atmospheric column) is maintained through the interaction between radiative cooling and convective heating. Convective heating is dominated by two effects: 1) the release of latent heat due to phase changes of water, and 2) convective heat transport (the divergence of the eddy temperature flux). The latter effect dominates heat budget in the upper troposphere. Changes to the cloud microphysics imposed in simulations discussed herein change the partitioning of the total convective heating into latent heating and the eddy flux divergence, with the magnitude of a few tenths of degrees Kelvin per day (cf. Fig. 7).

With the interactive radiation, the approach to quasi equilibrium involves the impact of cloud microphysics on surface processes, but also gradual evolution of mean cloud fraction profiles and thus the mean radiative cooling. This effect is particularly strong in the simulation featuring small cloud and precipitation particles (SP). In the quasi equilibrium, the upper-tropospheric temperature difference between SP and LP is larger than the difference between SP0 and LP0 because the simulation featuring small cloud and precipitation particles (SP) has weaker radiative cooling there. Interactive radiation directly affects the dynamics of the upper-tropospheric anvil clouds: the upper-tropospheric cloud fractions, similar in SP and LP, are in stark contrast to different values in SP0 and LP0. This illustrates a potentially crucial advantage of CRCP over traditional approaches where dynamical interactions between convection and radiative transfer are only possible on scales resolved by the large-scale model.

The marked result, especially in view of the discussion in Emanuel 1991 (cf. Fig. 6 therein), is that the changes of the temperature profiles invoke changes in the water vapor profiles so that the relative humidity profile remains approximately the same. This was also seen in GWM99, although 1-week-long simulations therein did not permit deducing whether such a result would apply to convective–radiative quasi equilibrium.

Results herein reinforce the conjecture of Grabowski (2000) that the most important impact of cloud microphysics on the tropical climate seems to occur through the ocean temperature. However, the constant-SST aquaplanet results in the surface energy budget being far higher than in reality. Consequently, the current results should be treated with caution. Work is under way to extend the constant-SST modeling setup into a rotating aquaplanet with realistic meridional SST distribution as, for example, in Hayashi and Sumi (1986). This framework should shed light on the impact of cloud microphysics not only on the tropical climate, but also on coupling between the Tropics and extratropics (poleward water and energy transports in particular).

Finally, simulations presented herein illustrate the potential of the CRCP framework in studies of the role of cloud microphysics in climate and climate change. Cloud microphysics parameterizations in cloud-resolving models need to be further improved before these models can provide reliable measures of the anthropogenic impact of atmospheric aerosols on cloud processes. This is especially true for ice processes. However, any new approaches developed and tested in cloud-resolving models can be immediately transfered into the CRCP framework. This is one of the reasons why CRCP is viewed as a promising technique as far as climate and climate change research is concerned.

Acknowledgments

Numerical simulations were performed on NCAR's prospect and blackforest parallel computers. Comments on the manuscript by M. Moncrieff are acknowledged, as is the editing of the manuscript by K. Sandoval. Constructive comments by the reviewers lead to the final version of this paper. This work is supported by NCAR's Clouds in Climate Program (CCP).

REFERENCES

  • Donner, L. J., 1993: A cumulus parameterization including mass fluxes, vertical momentum dynamics, and mesoscale effects. J. Atmos. Sci., 50 , 889906.

    • Search Google Scholar
    • Export Citation
  • Donner, L. J., , C. J. Seman, , and R. S. Hemler, 1999: Three-dimensional cloud-system modeling of GATE convection. J. Atmos. Sci., 56 , 18851912.

    • Search Google Scholar
    • Export Citation
  • Emanuel, K. A., 1991: A scheme for representing cumulus convection in large-scale models. J. Atmos. Sci., 48 , 23132335.

  • Emanuel, K. A., , and M. Živković-Rothman, 1999: Development and evaluation of a convection scheme for use in climate models. J. Atmos. Sci., 56 , 17661782.

    • Search Google Scholar
    • Export Citation
  • Fowler, L. D., , and D. A. Randall, 1996: Liquid and ice microphysics in the CSU General Circulation Model. Part III: Sensitivity to modeling assumptions. J. Climate, 9 , 561586.

    • Search Google Scholar
    • Export Citation
  • Fowler, L. D., , and D. A. Randall, 1999: Simulation of upper tropospheric clouds with the Colorado State University general circulation model. J. Geophys. Res., 104 , 61016121.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., 1998: Toward cloud resolving modeling of large-scale tropical circulations: A simple cloud microphysics parameterization. J. Atmos. Sci., 55 , 32833298.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., 2000: Cloud microphysics and the tropical climate: Cloud-resolving model perspective. J. Climate, 13 , 23062322.

  • Grabowski, W. W., 2001: Coupling cloud processes with the large-scale dynamics using the Cloud-Resolving Convection Parameterization (CRCP). J. Atmos. Sci., 58 , 978997.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., 2002: Large-scale organization of moist convection in idealized aquaplanet simulations. Int. J. Numer. Methods Fluids, 39 , 843853.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., 2003: MJO-like coherent structures: Sensitivity simulations using the Cloud-Resolving Convection Parameterization (CRCP). J. Atmos. Sci., 60 , 847864.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , and P. K. Smolarkiewicz, 1999: CRCP: A cloud resolving convection parameterization for modeling the tropical convecting atmosphere. Physica D, 133 , 171178.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , and M. W. Moncrieff, 2002: Large-scale organization of tropical convection in two-dimensional explicit numerical simulations: Effects of interactive radiation. Quart. J. Roy. Meteor. Soc., 128 , 23492375.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , and P. K. Smolarkiewicz, 2002: A multiscale anelastic model for meteorological research. Mon. Wea. Rev., 130 , 939956.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , M. W. Moncrieff, , and J. T. Kiehl, 1996a: Long-term behavior of precipitating tropical cloud systems: A numerical study. Quart. J. Roy. Meteor. Soc., 122 , 10191042.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , X. Wu, , and M. W. Moncrieff, 1996b: Cloud resolving modeling of tropical cloud systems during Phase III of GATE. Part I: Two-dimensional experiments. J. Atmos. Sci., 53 , 36843709.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , X. Wu, , M. W. Moncrieff, , and W. D. Hall, 1998: Cloud resolving modeling of tropical cloud systems during Phase III of GATE. Part II: Effects of resolution and the third spatial dimension. J. Atmos. Sci., 55 , 32643282.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , X. Wu, , and M. W. Moncrieff, 1999: Cloud resolving modeling of tropical cloud systems during Phase III of GATE. Part III: Effects of microphysical parameterizations. J. Atmos. Sci., 56 , 23842402.

    • Search Google Scholar
    • Export Citation
  • Grabowski, W. W., , J-I. Yano, , and M. W. Moncrieff, 2000: Cloud-resolving modeling of tropical circulations driven by large-scale SST gradients. J. Atmos. Sci., 57 , 20222039.

    • Search Google Scholar
    • Export Citation
  • Hayashi, Y-Y., , and A. Sumi, 1986: The 30–40 day oscillations simulated in an “aqua planet” model. J. Meteor. Soc. Japan, 64 , 451467.

    • Search Google Scholar
    • Export Citation
  • Kelly, M. A., , D. A. Randall, , and G. L. Stephens, 1999: A simple radiative-convective model with a hydrological cycle and interactive clouds. Quart. J. Roy. Meteor. Soc., 125 , 837869.

    • Search Google Scholar
    • Export Citation
  • Khairoutdinov, M. F., , and D. A. Randall, 2001: A cloud resolving model as a cloud parameterization in the NCAR Community Climate System Model: Preliminary results. Geophys. Res. Lett., 28 , 36173620.

    • Search Google Scholar
    • Export Citation
  • Kiehl, J. T., , J. J. Hack, , and B. P. Briegleb, 1994: The simulated Earth radiation budget of the National Center for Atmospheric Research community climate model CCM2 and comparisons with the Earth Radiation Budget Experiment (ERBE). J. Geophys. Res., 99 , 2081520827.

    • Search Google Scholar
    • Export Citation
  • Lipps, F. B., , and R. S. Hemler, 1986: Numerical simulation of deep tropical convection associated with large-scale convergence. J. Atmos. Sci., 43 , 17961816.

    • Search Google Scholar
    • Export Citation
  • McFarquhar, G. M., , and A. J. Heymsfield, 1997: Parameterization of tropical cirrus ice crystal size distributions and implications for radiative transfer: Results from CEPEX. J. Atmos. Sci., 54 , 21872200.

    • Search Google Scholar
    • Export Citation
  • Rennó, N. O., , K. A. Emanuel, , and P. H. Stone, 1994: Radiative-convective model with an explicit hydrologic cycle. 1: Formulation and sensitivity to model parameters. J. Geophys. Res., 99 , 1442914441.

    • Search Google Scholar
    • Export Citation
  • Robe, F. R., , and K. A. Emanuel, 1996: Moist convective scaling: Some inferences from three-dimensional cloud ensemble simulations. J. Atmos. Sci., 53 , 32653275.

    • Search Google Scholar
    • Export Citation
  • Smolarkiewicz, P. K., , L. G. Margolin, , and A. Wyszogrodzki, 2001: A class of nonhydrostatic global models. J. Atmos. Sci., 58 , 349364.

    • Search Google Scholar
    • Export Citation
  • Sui, C-H., , K. M. Lau, , W-K. Tao, , and J. Simpson, 1994: The tropical water and energy cycles in a cumulus ensemble model. Part I: Equilibrium climate. J. Atmos. Sci., 51 , 711728.

    • Search Google Scholar
    • Export Citation
  • Tao, W-K., , and S-T. Soong, 1986: A study of the response of deep tropical clouds to mesoscale processes: Three-dimensional numerical experiments. J. Atmos. Sci., 43 , 26532676.

    • Search Google Scholar
    • Export Citation
  • Tao, W-K., , J. Simpson, , and S-T. Soong, 1987: Statistical properties of a cloud ensemble: A numerical study. J. Atmos. Sci., 44 , 31753187.

    • Search Google Scholar
    • Export Citation
  • Tompkins, A. M., 2000: The impact of dimensionality on long-term cloud-resolving model simulations. Mon. Wea. Rev., 128 , 15211535.

  • Weller, R. A., , and S. P. Anderson, 1996: Surface meteorology and air–sea fluxes in the western equatorial Pacific warm pool during the TOGA Coupled Ocean–Atmosphere Response Experiment. J. Climate, 9 , 19591990.

    • Search Google Scholar
    • Export Citation
  • Wu, X., , W. D. Hall, , W. W. Grabowski, , M. W. Moncrieff, , W. D. Collins, , and J. T. Kiehl, 1999: Long-term behavior of cloud systems in TOGA COARE and their interactions with radiative and surface processes. Part II: Effects of cloud microphysics on cloud–radiation interaction. J. Atmos. Sci., 56 , 31773195.

    • Search Google Scholar
    • Export Citation
Fig. 1.
Fig. 1.

Results for simulation SP0 at (left) the equator and (right) at 51°S. (top) Hovmöller diagrams of the surface precipitation for the entire simulation. Precipitation intensity larger than 0.2 and 5 mm h−1 are shown using gray and black shading, respectively. (middle and bottom) Zonal distributions of the total surface heat flux (sensible plus latent) and the surface precipitation, respectively, at latitudes of the Hovmöller diagrams and on day 80

Citation: Journal of Climate 16, 21; 10.1175/1520-0442(2003)016<3463:IOCMOC>2.0.CO;2

Fig. 2.
Fig. 2.

As in Fig. 1, but for simulation LP0

Citation: Journal of Climate 16, 21; 10.1175/1520-0442(2003)016<3463:IOCMOC>2.0.CO;2

Fig. 3.
Fig. 3.

Evolution of the globally averaged (top) density-weighted temperature and (bottom) precipitable water for LP0 (dashed lines) and SP0 (solid lines)

Citation: Journal of Climate 16, 21; 10.1175/1520-0442(2003)016<3463:IOCMOC>2.0.CO;2

Fig. 4.
Fig. 4.

As in Fig. 3, but for LP and SP

Citation: Journal of Climate 16, 21; 10.1175/1520-0442(2003)016<3463:IOCMOC>2.0.CO;2

Fig. 5.
Fig. 5.

Quasi-equilibrium globally averaged profiles of the temperature, water vapor mixing ratio, relative humidity, and cloud fraction for SP0 (dashed lines) and LP0 (solid lines)

Citation: Journal of Climate 16, 21; 10.1175/1520-0442(2003)016<3463:IOCMOC>2.0.CO;2

Fig. 6.
Fig. 6.

The difference between globally averaged profiles of the equivalent potential temperature θe for SP0 and LP0. The profiles shown represent deviations from the average of the two θe profiles. Dashed (solid) lines show profiles for SP0 (LP0). The profiles marked “1” (thin lines) are for days 1–5, whereas those marked as “2” (thick lines) show the quasi-equilibrium profiles

Citation: Journal of Climate 16, 21; 10.1175/1520-0442(2003)016<3463:IOCMOC>2.0.CO;2

Fig. 7.
Fig. 7.

Quasi-equilibrium globally averaged profiles of the latent heating in SP0 and LP0 for days 71–80. The prescribed radiative cooling QR is shown for a reference

Citation: Journal of Climate 16, 21; 10.1175/1520-0442(2003)016<3463:IOCMOC>2.0.CO;2

Fig. 8.
Fig. 8.

As in Fig. 5, but for SP and LP

Citation: Journal of Climate 16, 21; 10.1175/1520-0442(2003)016<3463:IOCMOC>2.0.CO;2

Fig. 9.
Fig. 9.

As in Fig. 6, but for SP and LP

Citation: Journal of Climate 16, 21; 10.1175/1520-0442(2003)016<3463:IOCMOC>2.0.CO;2

Fig. 10.
Fig. 10.

Evolution of globally averaged cloud fraction profiles in LP and SP. Thin solid lines show profiles for days 1–5, thin dashed profiles are for days 11–15, and thick lines are the quasi-equilibrium profiles (days 56–60)

Citation: Journal of Climate 16, 21; 10.1175/1520-0442(2003)016<3463:IOCMOC>2.0.CO;2

Fig. 11.
Fig. 11.

Globally averaged radiative cooling profiles in LP and SP for days 1–5 (solid lines) and days 11–15 (dashed lines)

Citation: Journal of Climate 16, 21; 10.1175/1520-0442(2003)016<3463:IOCMOC>2.0.CO;2

Fig. 12.
Fig. 12.

Quasi-equilibrium globally averaged profiles of (left) the radiative cooling for LP and (right) SP. Thin solid, thick solid, and dashed lines represent total, longwave, and shortwave radiative cooling, respectively

Citation: Journal of Climate 16, 21; 10.1175/1520-0442(2003)016<3463:IOCMOC>2.0.CO;2

Table 1.

Selected properties of the quasi-equilibrium states in four simulations discussed in this paper. The quantities are averages over the entire aquaplanet and over the last 10 days of each simulation. The first column identifies the simulation; the second column shows CAPE calculated as explained in the text; the third column shows the cloud fractions; the fourth and fifth columns show mean upward vertical velocity and mean convective velocity; the sixth and seventh columns show updraft and downdraft cloud mass fluxes

Table 1.
Table 2.

Vertical energy fluxes averaged over the entire aquaplanet and over the last 10 days of the simulations SP and LP. The first column identifies the simulation; the second column shows the net radiative flux divergence across the troposphere; the third and fourth columns show surface sensible and latent heat fluxes; the fifth column shows the net radiative flux into the ocean; the last column shows the net energy flux into the ocean

Table 2.
*

The National Center for Atmospheric Research is sponsored by the National Science Foundation.

1

Note that the parameterization of microphysics applied in the simulations does not distinguish between latent heating of condensation and deposition (see Grabowski 1998 for a discussion).

Save