• Bergman, J. W., and H. H. Hendon, 1998: Calculating monthly radiative fluxes and heating rates from monthly observations of cloud cover. J. Atmos. Sci.,55, 3471–3491.

  • ——, and ——, 2000: The impact of clouds on the seasonal cycle of radiative heating over the Pacific. J. Atmos. Sci.,57, 545–566.

  • Bony, S., Y. Sud, K. M. Lau, J. Susskind, and S. Saha, 1997: Comparison and satellite assessment of NASA/DAO and NCEP–NCAR reanalyses over tropical ocean: Atmospheric hydrology and radiation. J. Climate,10, 1441–1462.

  • Cess, R. D., and Coauthors, 1995: Absorption of solar radiation by clouds: Observations versus models. Science,267, 496–499.

  • ECMWF Research Department, 1992: Research manual 1: ECMWF data assimilation scientific documentation. ECMWF, 88 pp. [Available from ECMWF, Shinfield Park, Reading, Berkshire RG2 9AX, United Kingdom.].

  • Gill, A., 1980: Some simple solutions for heat induced tropical circulation. Quart. J. Roy. Meteor. Soc.,106, 447–462.

  • ——, 1982: Atmosphere–Ocean Dynamics. Academic Press, 662 pp.

  • Hoskins, B. J., H. H. Hsu, I. N. James, M. Matsutani, P. D. Sardeshmukh, and G. H. White, 1989: Diagnostics of the global atmospheric circulation based on ECMWF analyses 1979–1989. WCRP-27, WMO/TD-326, 217 pp. [Available from WMO, Case Postale No. 2300, CH-1211 Geneva 2, Switzerland.].

  • Kalnay, E. M., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc.,77, 432–471.

  • Kessler, W. S., L. M. Rothstein, and D. Chen, 1998: The annual cycle of SST in the eastern tropical Pacific diagnosed in an ocean GCM. J. Climate,11, 777–799.

  • Kiehl, J. T., J. J. Hack, M. H. Zhang, and R. D. Cess, 1995: Sensitivity of a GCM climate to enhanced shortwave cloud absorption. J. Climate,8, 2200–2212.

  • ——, ——, G. B. Bonan, B. A. Boville, B. P. Briegleb, D. L. Williamson, and P. J. Rasch, 1996: Description of the NCAR community climate model (CCM3). NCAR Tech. Note NCAR/TN-420+STR, National Center for Atmospheric Research, Boulder, CO, 152 pp.

  • Lau, N.-C., 1979: The observed structure of tropospheric stationary waves and the balance of vorticity and heat. J. Atmos. Sci.,36, 996–1016.

  • Li, Z., L. Moreau, and A. Arking, 1997: On the solar energy disposition: A perspective from observation and modeling. Bull. Amer. Meteor. Soc.,78, 53–70.

  • Liu, A. Z., M. Ting, and H. Wang, 1998: Maintenance of circulation anomalies during the 1988 drought and 1993 floods over the United States. J. Atmos. Sci.,55, 2810–2832.

  • Ma, C.-C., C. R. Mechoso, A. W. Robertson, and A. Arakawa, 1996:Peruvian stratus clouds and the tropical Pacific circulation: A coupled ocean–atmosphere GCM study. J. Climate,9, 1635–1645.

  • Mechoso, C. R., and Coauthors, 1995: The seasonal cycle over the tropical Pacific in coupled ocean–atmosphere general circulation models. Mon. Wea. Rev.,123, 2825–2838.

  • Mitchell, T. P., and J. M. Wallace, 1992: The annual cycle in equatorial convection and sea surface temperature. J. Climate,5, 1140–1156.

  • Newell, R. E., J. W. Kidson, D. G. Vincent, and G. J. Boer, 1974: The General Circulation of the Tropical Atmosphere. Vol. 2. The MIT Press, 371 pp.

  • Nigam, S., 1994: On the dynamical basis for the Asian summer monsoon rainfall–El Niño relationship. J. Climate,7, 1750–1771.

  • ——, 1997: The annual warm to cold phase transition in the eastern equatorial Pacific: Diagnosis of the role of stratus cloud-top-cooling. J. Climate,10, 2447–2467.

  • Ramanathan, V., 1987: The role of Earth radiation budget studies in climate and general circulation research. J. Geophys. Res.,92, 4075–4095.

  • ——, and W. C. Collins, 1991: Thermodynamic regulation of ocean warming by cirrus clouds deduced from observations of the 1987 El Nino. Nature,351, 27–32.

  • ——, B. Subasilar, G. J. Zhang, W. Conant, R. D. Cess, J. T. Kiehl, H. Grassl, and L. Shi, 1995: Warm pool heat budget and shortwave cloud forcing: A missing physics? Science,267, 499–503.

  • Randall, D. A., Harshvardhan, D. A. Dazlich, and T. G. Corsetti, 1989: Interactions among radiation, convection, and large-scale dynamics in a general circulation model. J. Atmos. Sci.,46, 1943–1970.

  • Ridout, J. A., and T. E. Rosmond, 1996: Global modeling of cloud radiative effects using ISCCP cloud data. J. Climate,9, 1479–1496.

  • Rossow, W. B., and R. A. Schiffer, 1991: ISCCP cloud data products. Bull. Amer. Meteor. Soc.,72, 2–20.

  • ——, and L. C. Garder, 1993: Validation of ISCCP cloud detections. J. Climate,6, 2370–2393.

  • ——, A. Walker, D. Beuschel, and M. Roiter, 1996: International Satellite Cloud Climatology Project (ISCCP) documentation of new cloud dataset. WMO/TD-No. 737, World Meteorological Organization, 115 pp. [Available online at http://isccp.giss.nasa.gov/documents.html].

  • Sherwood, S. C., V. Ramanathan, T. P. Barnett, M. K. Tyree, and E. Roeckner, 1994: Response of an atmospheric general circulation model to radiative forcing of tropical clouds. J. Geophys. Res.,99, 20 829–20 845.

  • Slingo, J. M., and A. Slingo, 1991: The response of a general circulation model to cloud longwave radiative forcing: II Further studies. Quart. J. Roy. Meteor. Soc.,117, 333–364.

  • Stephens, G. L., 1996: How much solar radiation do clouds absorb? Science,271, 1131–1133.

  • Weare, B. C., 1997: Comparison of NCEP–NCAR cloud radiative forcing reanalyses with observations. J. Climate,10, 2200–2209.

  • Webster, P. J., 1994: The role of hydrological processes in ocean–atmosphere interactions. Rev. Geophys.,32, 427–476.

  • Yanai, M., S. Esbensen, and J. H. Chu, 1973: Determination of bulk properties of tropical cloud clusters from large-scale heat and moisture budgets. J. Atmos. Sci.,30, 611–627.

  • View in gallery

    Comparison of the surface circulation from the NCEP–NCAR reanalysis data with the calculated response of the linear model at 975 mb to NCEP total diabatic heating rates. Shown are 1984–90 averages for (a) JF reanalysis data, (b) AS reanalysis data, (c) JF calculated

  • View in gallery

    (Continued) fields, and (d) AS calculated fields. The left-hand side in each panel show the zonal mean surface winds. The right-hand side show the eddy winds (arrows) and eddy geopotential height (m, contoured and shaded).

  • View in gallery

    A comparison of the JF 1984–90 deep equatorial circulation from the NCEP–NCAR reanalysis data averaged over 5°S–5°N with calculated fields. (a) Reanalysis zonal mass flux (109 kg s−1), (b) reanalysis zonal and vertical wind

  • View in gallery

    (Continued) vectors, (c) calculated zonal mass flux, and (d) calculated zonal and vertical wind vectors. Vertical winds in (b) and (d) have been enhanced by a factor of 1000.

  • View in gallery

    Response of the linear model at 975 mb to (a) latent heating rates and (b) radiative heating rates from NCEP for JF 1984–90. The left-hand side shows the zonal mean surface winds. The right-hand side shows the eddy winds (arrows) and eddy geopotential height (m, contoured and shaded).

  • View in gallery

    Zonal mass flux (109 kg s−1), averaged over 5°S–5°N from the linear model forced by (a) latent heating and (b) radiative heating from NCEP for JF 1984–90.

  • View in gallery

    Response of the linear model at 975 mb to (a) ISCCP radiative heating rates and (b) cloud radiative forcing for JF 1984–90. The left-hand side shows the zonal mean surface winds. The right-hand side shows the eddy winds (arrows) and eddy geopotential height (m, contoured and shaded).

  • View in gallery

    Response of the linear model at 975 mb to (a) LW cloud radiative forcing and (b) SW cloud radiative forcing for JF 1984–90. The left-hand side shows the zonal mean surface winds. The right-hand side shows the eddy winds (arrows) and eddy geopotential height (m, contoured and shaded).

  • View in gallery

    Convective cloud mask applied to CRF for JF 1984–90.

  • View in gallery

    Response of the linear model at 975 mb to (a) convective cloud radiative forcing and (b) subtropical cloud radiative forcing for JF 1984–90. The left-hand side shows the zonal mean surface winds. The right-hand side shows the eddy winds (arrows) and eddy geopotential height (m, contoured and shaded).

  • View in gallery

    Response of the linear model to residual diabatic heating from NCEP for JF 1984–90. (a) Surface circulation: the left-hand side shows the zonal mean surface winds. The right-hand side shows the eddy winds (arrows) and eddy geopotential height (m, contoured and shaded). (b) Zonal mass flux (109 kg s−1) averaged over 5°S–5°N.

  • View in gallery

    Response of the linear model at 975 mb to ISCCP cloud radiative forcing for JF 1989. The left-hand side shows the zonal mean surface winds. The right-hand side shows the eddy winds (arrows) and eddy geopotential height (m, contoured and shaded). (a) Control calculation. (b) Enhanced SW cloud absorption. (c) Using ISCCP-D data.

  • View in gallery

    The 1984–90 seasonal cycle averaged over 85°–105°W in the eastern Pacific. (a) Meridional winds (m s−1, contoured and shaded) and SST (K, thick contours) from the reanalysis data. (b) Meridional winds (contoured and shaded) calculated from NCEP total diabatic heating rates (thick contours).

  • View in gallery

    The 1984–90 seasonal cycle of meridional winds (m s−1) averaged over 85°–105°W in the eastern Pacific calculated from (a) NCEP latent heating rates and (b) NCEP radiative heating rates.

  • View in gallery

    The 1984–90 seasonal cycle of meridional winds (m s−1) averaged over 85°–105°W in the eastern Pacific calculated from (a) ISCCP radiative heating rates and (b) cloud radiative forcing.

  • View in gallery

    The 1984–90 seasonal cycle of meridional winds (m s−1) averaged over 85°–105°W in the eastern Pacific calculated from (a) convective cloud radiative forcing and (b) subtropical cloud radiative forcing.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 125 125 6
PDF Downloads 6 6 2

Cloud Radiative Forcing of the Low-Latitude Tropospheric Circulation: Linear Calculations

View More View Less
  • 1 NOAA–CIRES Climate Diagnostics Center, Boulder, Colorado
© Get Permissions
Full access

Abstract

The role of clouds for low-latitude atmospheric circulations is examined in a linearized calculation forced by diabatic heating rates. A comparison of the circulation calculated from total diabatic heating, obtained from reanalysis data, with observed fields determines which aspects of the calculation are realistic and which are not. The role of clouds is quantified by the circulation calculated from atmospheric cloud radiative forcing, which, in turn, has been calculated with the National Center for Atmospheric Research radiative transfer model using cloud properties observed in the International Satellite Cloud Climatology Project.

In general, cloud radiative forcing contributes about 20% to the magnitude of low-latitude circulations. It typically reinforces the circulation that is driven by convective latent heating. Cloud radiative forcing tends to have a stronger influence in the lower troposphere than at upper levels. It influences local circulations more than remote ones. In particular, cloud radiative forcing from local low cloud cover is the dominant source of diabatic heating influencing subtropical circulations over the eastern oceans. Cloud radiative forcing from low clouds is also found to be important for seasonal variations of meridional winds over the cold tongue in the eastern Pacific. This indicates that atmospheric cloud radiative forcing, and not just surface forcing, is important for ocean–atmospheric coupling there.

Additional calculations are performed that test the sensitivity of the atmospheric circulation to different sources of diabatic heating rates. These sources include radiative heating rates that have been calculated from different cloud data, different cloud overlap assumptions, and enhanced cloud short-wave absorptivity. The principal conclusions of this investigation are unchanged by these calculations. However, enhanced short-wave absorption by clouds systematically reduces the impact of clouds on atmospheric circulations.

Corresponding author address: John W. Bergman, CIRES–NOAA Climate Diagnostics Center, R/E/CD1, 325 Broadway, Boulder, CO 80309-0449.

Email: jwb@cdc.noaa.gov

Abstract

The role of clouds for low-latitude atmospheric circulations is examined in a linearized calculation forced by diabatic heating rates. A comparison of the circulation calculated from total diabatic heating, obtained from reanalysis data, with observed fields determines which aspects of the calculation are realistic and which are not. The role of clouds is quantified by the circulation calculated from atmospheric cloud radiative forcing, which, in turn, has been calculated with the National Center for Atmospheric Research radiative transfer model using cloud properties observed in the International Satellite Cloud Climatology Project.

In general, cloud radiative forcing contributes about 20% to the magnitude of low-latitude circulations. It typically reinforces the circulation that is driven by convective latent heating. Cloud radiative forcing tends to have a stronger influence in the lower troposphere than at upper levels. It influences local circulations more than remote ones. In particular, cloud radiative forcing from local low cloud cover is the dominant source of diabatic heating influencing subtropical circulations over the eastern oceans. Cloud radiative forcing from low clouds is also found to be important for seasonal variations of meridional winds over the cold tongue in the eastern Pacific. This indicates that atmospheric cloud radiative forcing, and not just surface forcing, is important for ocean–atmospheric coupling there.

Additional calculations are performed that test the sensitivity of the atmospheric circulation to different sources of diabatic heating rates. These sources include radiative heating rates that have been calculated from different cloud data, different cloud overlap assumptions, and enhanced cloud short-wave absorptivity. The principal conclusions of this investigation are unchanged by these calculations. However, enhanced short-wave absorption by clouds systematically reduces the impact of clouds on atmospheric circulations.

Corresponding author address: John W. Bergman, CIRES–NOAA Climate Diagnostics Center, R/E/CD1, 325 Broadway, Boulder, CO 80309-0449.

Email: jwb@cdc.noaa.gov

1. Introduction

This investigation examines the role of clouds in a linearized calculation of the low-latitude tropospheric circulation forced by specified diabatic heating rates. These heating rates are obtained from reanalysis data and from radiative transfer calculations using observed cloud properties. For reasons discussed below, we focus primarily on the role of clouds for the seasonal cycle of surface winds in the eastern Pacific. To put those results in a more general context, we examine the role of clouds for the surface circulation at low latitudes throughout the globe. In addition, we examine the zonal circulation along the equator (i.e., the “Walker circulation”), through which deep convective heating in the western Pacific influences surface winds across the Pacific basin.

Our approach, because it is linear, allows us to quantify the relative impact of different components of diabatic heating on the circulation. We calculate the steady response of the circulation to total diabatic heating and to the individual components: latent heating and radiative heating. The role of clouds is quantified by the circulation driven by “cloud radiative forcing” (CRF), which is calculated by subtracting clear-sky radiative heating rates from the net radiative heating rates. We further diagnose the relative contributions by deep convective clouds, for example, in the intertropical convergence zone (ITCZ), and by shallow cloud fields at subtropical locations.

This study is motivated by the need to understand, more precisely, the role of clouds in the coupled ocean–atmospheric climate system. Clouds are important because they both affect and are affected by variations of sea surface temperature (SST). For example, in the Tropics, deep convective clouds occur preferentially where SST is high because of the influence of moist, warm low-level air on the tropospheric stability. Thus, cloud variability is a function of SST. Clouds affect SST by redistributing radiative energy throughout the climate system. This directly affects SST beneath the clouds via radiative fluxes at the surface, or “surface CRF.” In fact, Ramanathan and Collins (1991) suggest that surface cloud radiative forcing in the western Pacific is the primary thermostatic control that constrains SST to maximize at about 302 K. Clouds also affect SSTs indirectly by altering atmospheric heating rates, via “atmospheric CRF.” That source of atmospheric heating influences surface winds, which influence wind-driven surface fluxes, which influence SST. Such an indirect mechanism has been proposed, in previous studies, to be important in the equatorial Pacific (e.g., Slingo and Slingo 1991; Webster 1994). In this case, the contrast of atmospheric CRF in the convective western Pacific, where clouds are distributed throughout the troposphere, compared to CRF in the nonconvective eastern Pacific, where clouds are primarily confined to the lower troposphere, could be strong enough to have a significant impact on the Walker circulation. The subsequent impact on surface winds affects latent and sensible heat fluxes at the surface as well as ocean upwelling in the cold tongue, thus impacting SSTs across the Pacific basin.

Studies with ocean GCMs, atmospheric GCMs, and coupled ocean–atmospheric GCMs all show that simulated tropical circulations are sensitive to cloud radiative forcing (e.g., Randall et al. 1989; Slingo and Slingo 1991; Sherwood et al. 1994; Ma et al. 1996; Kessler et al. 1998). These results are important for two reasons. First, they show that cloud variability is an important component of climate, with potentially many types of feedbacks operating. The second is a practical consideration. GCMs, particularly coupled GCMs, are becoming increasingly important for climate change research and for forecasting. To be effective in those endeavors, it is necessary that GCMs accurately simulate observed climate variability from physically motivated parameterizations.

Unfortunately, coupled GCMs do not adequately simulate climatological conditions (e.g., Mechoso et al. 1995). It is common for SSTs simulated by these models to be too warm in the subtropical eastern Pacific and too cold along the equatorial Pacific cold tongue. To explore the role of “low clouds” (principally stratocumulus and stratus) for this systematic bias, Ma et al. (1996) prescribed enhanced low cloud fractions in the subtropical eastern Pacific in a coupled GCM experiment. That change corrected the warm bias in the southeastern Pacific but exacerbated the cold bias along the cold tongue. Those results suggest that clouds are indeed important for ocean–atmospheric coupling in the eastern Pacific and that better cloud parameterizations might improve simulations by coupled GCMs.

Linear models are well suited for analyzing the impact of clouds on circulations at tropical and subtropical locations. A successful conceptual model of the atmospheric circulation at low latitudes is that of a linear response to diabatic heating (e.g., Gill 1980). For example, in a study that is closely related to the current study, Nigam (1997) calculated the linear response of the tropical atmosphere to diabatic heating rates derived from the thermodynamic equation using data from the European Centre for Medium-Range Forecasts (ECMWF) analyses. By examining, separately, the atmospheric responses to different components of the forcing, Nigam found that radiative cooling associated with low clouds south of the equator is important for seasonal variations of cross-equatorial surface winds in the far eastern Pacific. Those winds, according to Mitchell and Wallace (1992), are important contributors to the seasonal cycle of SST there.

Nigam (1997) inferred the importance of cloud radiative forcing from the spatial collocation of a maximum of diabatic cooling with large fractional coverage of low clouds. In support of this inference, Bergman and Hendon (2000) found that cloud radiative forcing is, indeed, the primary contributor to seasonal variations of radiative cooling in the atmosphere at low latitudes. In that study, 7 yr (1984–90) of monthly radiative fluxes and heating rates initially calculated by Bergman and Hendon (1998) were utilized to examine the impact of clouds for seasonal variations of radiative heating over the Pacific. Cloud radiative forcing was found to account for 20%–50% of the seasonal variations of surface fluxes and diabatic heating in the atmosphere. An important aspect of atmospheric CRF is that it reinforces latent heating from tropical deep convection. That is, large-scale spatial patterns of anomalies for latent heating rates and CRF are similar. Furthermore, seasonal variations of those two sources of heating are, for the most part, in phase. Those relationships make it straightforward to interpret the role of clouds for atmospheric heating. Nevertheless, because of the complex dependence of the circulation on the spatial distributions of the heating rates, those relationships only indicate what the impact of cloud radiative forcing on low-latitude circulations might be. We follow up on that research by explicitly calculating the impact of atmospheric cloud radiative forcing for seasonal variations of the atmospheric circulation at low latitudes. Section 2 describes the calculations and the diabatic heating rates used to force the linear model. Section 3 examines the steady response at low latitudes to specified heating rates. It quantifies the relative importance of cloud radiative forcing to other sources of diabatic heating and provides a critical analysis of the shortcomings of our approach. Section 4 explores, in finer detail, the impact of clouds on the seasonal cycle of surface winds in the equatorial eastern Pacific. Section 5 provides concluding remarks and interpretation.

2. Methodology

a. Linear model

The atmospheric model is a baroclinic pressure coordinate model with 20 vertical levels and T20 horizontal resolution. It solves the linearized primitive equations
i1520-0469-57-14-2225-e1
where
i1520-0469-57-14-2225-eq1a
The overbar represents basic-state quantities, which are specified, and primed quantities are the perturbation quantities being calculated. The subscript h refers to the horizontal component of vector quantities, and F represents frictional forces. Other symbols represent traditional thermodynamical and dynamical quantities. Frictional forces F′ are simulated by fourth-order horizontal diffusion and linear dissipation with rate ν. The diabatic heating rate Q′ contains a specified component q′ as well as linear damping with rate α. The coefficient of horizontal diffusion is specified at K = 2 × 1016 m4 s−1 throughout. Dissipation rates follow a piecewise linear pressure dependence with a timescale of 1 day at the surface, increasing to 10 days at 850 mb, constant from there to 100 mb, and then decreasing linearly to 3 days at p = 0.

We wish to understand the role of clouds for as much of the tropical circulation as possible and to minimize ambiguities in our results that occur from the use of a complex basic state. We, therefore, consider perturbations to a basic state that is motionless with thermal structure specified from the climatological vertical distribution of potential temperature, averaged over the latitude band 30°S–30°N, from the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis data (Kalnay et al. 1996). The steady-state solution is estimated by the average over the last 5 days of a 20-day integration initiated from a zero perturbation state and forced by monthly mean diabatic heating rates. The choices of averaging period and integration length are essentially arbitrary. However, the steady-state solution varies little for integration times greater than about 5 days and averaging periods greater than 2 days. To obtain seasonal variations, separate calculations are performed for each month, using the corresponding monthly mean thermal field and diabatic heating rates.

b. Specified heating fields

Heating rates q′ used to force the atmospheric model are obtained from monthly data in a 7-yr (1984–90) composite seasonal cycle. Two primary sources of data are utilized. 1) Total heating rates, latent heating rates, and radiative heating rates from the NCEP–NCAR reanalysis data allow quantification of the relative importance of those sources of atmospheric heating to the low-latitude circulation. 2) Radiative heating rates calculated from observed cloud fields with a radiative transfer model (Bergman and Hendon 1998) provide an independent source of radiative heating and allow detailed analysis of the role of cloud radiative forcing.

The NCEP–NCAR reanalysis data provides a convenient source of atmospheric heating fields. There are six components to that data representing vertical diffusion of potential temperature, latent heating from deep convection, latent heating from shallow convection, latent heating from large-scale moisture convergence, longwave (LW) radiative heating, and shortwave (SW) radiative heating. While convenient to use, heating rates in the reanalysis are not directly linked to observations, but rather are calculated by physical parameterizations in a GCM that has been initialized by observed dynamical fields. Those physical parameterizations have weaknesses that are evident in reanalysis fields that are related to heating rates, for example, precipitation and cloud properties (e.g., Kalnay et al. 1996; Bony et al. 1997). So, despite the assimilation of observations into the reanalysis process, the accuracy of NCEP–NCAR heating rates is uncertain. The reliability of results based on reanalysis heating rates is tested in section 3c.

It is difficult to quantify the effects of clouds on the atmospheric circulation from NCEP–NCAR radiative heating rates because the vertical distribution of cloud radiative forcing is not available in that dataset. Furthermore, cloud properties from the reanalysis, like heating rates, are derived by GCM physical parameterizations and have inaccuracies (e.g., Bony et al. 1997; Weare 1997; Bergman and Hendon 1998). To examine the role of cloud radiative forcing in detail, monthly radiative fluxes have been calculated from monthly averaged atmospheric data at 2.5° horizontal resolution and at 45-mb vertical resolution in the troposphere. The calculation outlined here is described in detail by Bergman and Hendon (1998). Cloud properties and surface temperature for the calculations are obtained from the International Satellite Cloud Climatology Project (ISCCP) C2 dataset (Rossow and Schiffer 1991). Atmospheric temperature and humidity are obtained from the ECMWF analyses (ECMWF Research Department 1992). The radiative transfer model was developed at NCAR for the general circulation model CCM3 (Kiehl et al. 1996) and was modified by Bergman and Hendon (1998) to incorporate observed cloud properties. Those heating rates were calculated only in the latitude band ϕ = 40°S–40°N because cloud detection in ISCCP deteriorates at high latitudes, particularly over land (Rossow and Garder 1993). For consistency, the meridional structures of all specified heating rates are modified by the following cosine-tapered window:
i1520-0469-57-14-2225-e2
This choice of window does not adversely affect our results, which only concern low latitudes. In fact, circulations in the latitude band 30°S–30°N that are calculated from global (not shown) and windowed heating rates are virtually identical.

3. Low-latitude circulations

a. Assessment of model response

To interpret the results of the linear calculations effectively, it is important to know which aspects of the circulation are realistically reproduced by the model and which are not. Here, we compare the calculated circulation to NCEP–NCAR reanalysis data during January–February (JF), when convection in the northeastern Pacific ITCZ is suppressed, and during August–September (AS), when convection there is active. These two periods allow us to contrast summertime with wintertime circulations over subtropical and extratropical locations.

Figures 1a and 1b show the low-latitude surface circulation from the reanalysis data for JF and AS, respectively. The right-hand side displays the “eddy” (i.e., deviations from the zonal mean) circulation in terms of winds vectors and geopotential height (contoured and shaded). Zonal mean wind vectors are shown in the left hand side. Important features to note are seasonal contrasts of: 1) the subtropical anticyclones in the eastern oceans, which tend to be strongest in the summer hemisphere, and 2) the cross-equatorial flow in the eastern Pacific, which is strongest during AS when convection in the ITCZ is active.

To diagnose the deep equatorial circulation in the longitude–height plane, we use the “zonal mass flux” as defined by Newell et al. (1974):
i1520-0469-57-14-2225-e3
where re is the radius of the earth, Δϕ = 10° is a latitude increment, and 〈u′〉 is the eddy zonal wind averaged over the latitude band 5°S–5°N. At locations where the meridional divergence is weak, M measures the strength of the Walker circulation. The mass flux is a scalar field and is easily contoured. It, therefore, allows straightforward comparisons between data sources while containing essential information about the structure of the Walker circulation. The zonal mass flux for JF from the reanalysis data is displayed in Fig. 2a. The deep equatorial circulation, of which the Walker circulation is the divergent component, is displayed in Fig. 2b in terms of u′ and w′. Here, the magnitude of w′ has been enhanced by a factor of 1000. The eddy flow follows approximately the contours in Fig. 2a, with positive values of M indicating clockwise circulations. Extrema of M indicate where u′ changes sign. There are three major circulation features: counterclockwise flow over the western Pacific and Indian Ocean, clockwise flow over the eastern Pacific, and clockwise flow over South America.

The “surface”1 response of the model to total diabatic heating from the reanalysis (Figs. 1c,d) produces a reasonable meridional structure of mean winds. It reproduces the observed eddy fields best over open ocean, at low latitudes, and in the summer hemisphere. Poor reproductions of tropical circulations in the northwestern Pacific during JF (Fig. 1c) and in the western Indian Ocean during AS (Fig. 1d) are, perhaps, the most obvious exceptions. The seasonal changes of circulation features noted previously for the observed circulation are also evident in the calculated fields. In particular, the positions of the summertime subtropical anticyclones (west of Australia, South America, and Africa in Fig. 1c; west of North America and Europe in Fig. 1d) are well reproduced by the calculation. The strengths of those features are within 30% of observed values in the reanalysis data. The calculated fields degrade over locations where orography, nonlinearities, and the nonzero basic state become important (cf. Lau 1979). Thus, they are most unrealistic at extratropical locations during winter.

The zonal mass flux and deep equatorial circulation calculated from total diabatic heating rates for JF are shown in Figs. 2c and 2d, respectively. In addition, the eddy component of the diabatic heating rates is shaded in Fig. 2c. Heating rates have deep vertical profiles, with strong deviations from the zonal mean occurring throughout the depth of the troposphere. Positive deviations occur where convection is active (e.g., 60°–180°E; 30°–80°W) and are collocated with upward eddy motion. Downward motion occurs where heating deviations are negative (e.g., 90°–150°W; near 0°). The three large-scale features of the zonal mass flux seen in Fig. 2a are reproduced in the linear model (Fig. 2c). However, the circulations calculated by the model are too strong, with the circulation centers (i.e., the extrema of the mass flux) located too high in the troposphere. This interpretation based on the mass flux is verified in a comparison of the eddy circulation produced by the model (Fig. 2d) with the reanalysis data (Fig. 2b).

The analysis above reveals significant weaknesses in the linear calculation, which must be considered in subsequent experiments. However, there are certain aspects of the observed circulation that the linear model consistently reproduces. In particular, and of import to this investigation, circulation patterns are faithfully reproduced over the eastern oceans (e.g., subtropical anticyclones). To best exploit the strengths of the calculation, we will focus on the impact of cloud radiative forcing relative to other sources of diabatic heating in the atmosphere. This will minimize the impact of inaccuracies in the calculated amplitudes of circulation features on our results. We will also search for aspects of the circulation driven by cloud radiative forcing that are typical of a broad category of circulation features (e.g., aspects that are common to all subtropical anticyclones) instead of examining only isolated features.

b. Role of clouds

Figure 3 shows the surface response during JF to (a) latent heating (the combination of deep convective, shallow convective, and large-scale condensation) and (b) net (LW + SW) radiative heating from NCEP. Latent heating dominates tropical heating rates (e.g., Yanai et al. 1973), and that dominance is evident in the model response. Note that contour intervals are halved and arrow lengths doubled in Fig. 3b with respect to those in Fig. 3a. Radiative heating is most important in the Southern (i.e., summer) Hemisphere, particularly near the subtropical anticyclones. At those locations, the circulation driven by radiative heating is often stronger than that driven by latent heating. The deep equatorial circulation, as quantified by the zonal mass flux, is dominated by latent heating (Fig. 4a) at all longitudes. The strength of the circulation driven by radiative heating (Fig. 4b) is typically less than 20% of that driven by latent heating. However, the relative importance of radiative heating increases in the lower troposphere, where zonal fluctuations of that heating source are strongest. Overall, the impact of radiative heating in the lower troposphere reinforces the circulation driven by latent heating. That is, winds are typically in the same direction and geopotential heights of the same sign in Figs. 3a and 3b, while Figs. 4a and 4b exhibit mass fluxes of the same sign below 600 mb. Thus, radiative heating tends to strengthen low-latitude circulations, a finding consistent with GCM studies of cloud radiative forcing (e.g., Sherwood et al. 1994).

Radiative heating rates, calculated by Bergman and Hendon (1998, hereafter “ISCCP radiative heating rates”), are derived from satellite observations of cloud properties. Cloud observations are not utilized to produce the reanalysis data and, thus, ISCCP and NCEP radiative heating rates represent independent data sources. The low-latitude circulation driven by ISCCP radiative heating (Fig. 5a) is similar to that driven by NCEP radiative heating (Fig. 3b) in both magnitude and spatial pattern. However, circulations driven by the ISCCP radiative heating rates tend to be somewhat stronger than those driven by NCEP radiative heating rates and there are substantial differences in the northern subtropics. The deep equatorial circulation driven by these two sources of radiative heating (not shown) have larger discrepancies than the surface circulations, but they agree on the most important point: that radiative heating contributes only a small fraction (i.e., about 20%) to that circulation.

The similarity between Figs. 3b and 5a seems, at the outset, remarkable in light of previous studies that find substantial errors in radiative fluxes from the NCEP–NCAR reanalysis data (Bony et al. 1997; Weare 1997;Bergman and Hendon 1998). However, those errors result primarily from faulty SW reflectivity of the reanalysis clouds, with cloud distributions playing a secondary role. Cloud reflectivity in the reanalysis does not completely undermine the atmospheric radiative heating rates because SW cloud radiative forcing of atmospheric heating rates is much smaller than LW cloud radiative forcing (e.g., Ramanathan 1987; Ridout and Rosmond 1996; Bergman and Hendon 1998, 2000).

The clear-sky component of radiative heating is large. However, spatial variations of radiative heating, which are important to the eddy circulation, are dominated by cloud radiative forcing, particularly over the oceans (e.g., Bergman and Hendon 2000). As a result, most of the circulation driven by radiative heating can be attributed to cloud radiative forcing (Fig. 5b). Cloud radiative forcing is less important during winter (i.e., in the north) and over land, where LW emissions from the surface have a strong impact on spatial variations of radiative heating. Cloud radiative forcing is dominated by the LW component, which can either be a source of atmospheric heating or cooling depending on the depth of the cloud distribution. The response to LW CRF (Fig. 6a) is similar to the response to net CRF (Fig. 5b) and is about a factor of 5 stronger than the response to SW CRF (Fig. 6b). An important aspect of the individual LW and SW components of CRF is that they are systematically of opposite sign (e.g., Bergman and Hendon 1998, 2000). That relationship is also found in the calculated circulations. For example, in the subtropics, where LW CRF forces anticyclonic motion, SW CRF typically forces cyclonic motion. So estimates of the impact of cloud radiative forcing based only on the LW component systematically overestimate the role of clouds for atmospheric circulations by about 20%.

In regions of deep tropical convection, CRF is a source of deep atmospheric heating. In contrast, over subtropical locations dominated by low clouds, CRF is a source of shallow, atmospheric cooling (e.g., Bergman and Hendon 1998). To separate the individual contributions by these two cloud categories, cloud radiative forcing is subjected to the following geographical mask
i1520-0469-57-14-2225-e4
where Chi is the monthly mean high cloud fractional coverage from the ISCCP dataset. The geographical distribution of C for JF (Fig. 7) highlights areas of deep tropical convection. Multiplying CRF by C yields the cloud radiative forcing from those locations (i.e., “convective” CRF); multiplying CRF by 1 − C yields “subtropical” cloud radiative forcing, which emanates from areas dominated by low clouds.

Cloud radiative forcing primarily affects the local circulation: convective CRF dominates the eddy circulation near convective locations (Fig. 8a) and subtropical CRF dominates the circulation near the subtropical anticyclones (Fig. 8b). There are, however, locations where remote effects are obvious. For example, over tropical South America, winds driven by subtropical CRF are nearly 50% as large as those driven by convective CRF despite C = 1 over that entire region. These results indicate an important role for low clouds in low-latitude circulations because subtropical CRF dominates the circulation driven by radiative heating at locations where radiative heating dominates the energy budget.

c. Critical analysis

The atmospheric model is idealized and neglects important interactions in the atmosphere. Furthermore, heating rates are not directly observed in the atmosphere and estimates of heating rates are subject to substantial error. As a result, we have limited our conclusions to results that appear robust and apply to the whole category of circulation features. In this section, we examine the response of the atmospheric model to different sources of diabatic heating rates in order to further test the reliability of our conclusions.

We first examine the model’s response to a different source of total diabatic heating data. Heating rates in the NCEP–NCAR reanalysis data are calculated by GCM parameterizations and are subject to considerable uncertainty. An alternative source of total diabatic heating is obtained from the residual of the thermodynamical equation, in which all components of that equation, except diabatic heating rates, are calculated explicitly from reanalysis data. This method suffers from the coarse time resolution of the data and from uncertainties in the vertical velocity. Nevertheless, residual heating rates are calculated more directly from observed quantities than are the reanalysis heating rates and have been the preferred source of diabatic heating in previous studies (e.g., Hoskins et al. 1989; Nigam 1994, 1997; Liu et al. 1998). Whether or not this residual provides a more accurate estimate of the diabatic heating rates is not the issue here. Instead, we use residual diabatic heating rates as an independent estimate of diabatic heating rates, which helps us to test the reliability of our results.

A comparison of the JF circulation calculated from the residual diabatic heating rates (Fig. 9) to observed circulations (Figs. 1a and 2a) does not alter any of the conclusions drawn in section 1(c), which compared the circulation calculated from NCEP–NCAR heating rates (Figs. 1c and 2c) to observed fields. In fact, with the exception of small-scale features over locations where the linear calculation is suspect (e.g., the strong geopotential anomaly over the Himalaya in Fig. 9a that is not found in Fig. 1c), the circulations from the two calculations resemble each other more than either resembles the observed circulation. It appears that errors in the calculated circulations result more from unrealistic aspects of the linear model than from errors in the specified diabatic heating rates.

We next examine the sensitivity of the calculated circulation to changes in the specified cloud distributions used for the ISCCP radiative heating rates. Bergman and Hendon (1998) examined the sensitivity of the ISCCP radiative heating rates to the choice of cloud data, to the assumptions about cloud vertical distributions, and to the SW cloud optical properties. They found that uncertainties within the ISCCP cloud detection algorithms and uncertainties of the SW absorptivity of clouds were the dominant sources of error. Errors generated by uncertainties in assumptions used to create the vertical distributions of clouds (e.g., cloud overlap and position of the lowest cloud base) were relatively minor. Here, we assess the response of the linear model to similar uncertainties.

Figure 10 shows the surface response to JF 1989 radiative heating rates from three different radiative transfer calculations. The radiative heating rates used for Fig. 10a are the same as those described in section 2b. The differences between the circulation shown in Fig. 10a and that in Fig. 5a, then, results from the different time-averaging periods.

The circulation in Fig. 10b was forced by heating rates calculated with enhanced SW cloud absorption. For this calculation, the specified cloud SW single-scattering albedo was reduced as described by Kiehl et al. (1995) to simulate enhanced cloudy-sky SW absorption. The comparison of Figs. 10a and 10b examines the sensitivity of calculated circulation to this controversial aspect of cloud radiative properties (e.g., Cess et al. 1995;Ramanathan et al. 1995; Stephens 1996; Li et al. 1997). The strength of the eddy circulation in Fig. 10b is reduced compared to Fig. 10a because variations of atmospheric cloud radiative forcing are systematically reduced by this form of enhanced cloud absorption (Bergman and Hendon 1998).

The radiative heating rates for Fig. 10c are calculated in the same manner as described in section 2b, except that cloud properties are obtained from the ISCCP-D data instead of ISCCP-C data. The most significant difference between these two datasets is that ISCCP-D data has larger fractions of high clouds in the Tropics than ISCCP-C data (Rossow et al. 1996). As a result, CRF calculated from ISCCP-D data at tropical convective locations is enhanced in the upper troposphere and reduced in the lower troposphere compared to CRF from ISCCP-C data (Bergman and Hendon 1998). The resulting surface circulation (Fig. 10c) has substantial differences compared to the control heating rates (Fig. 10a). Nevertheless, the general features that have been discussed in previous sections, and the conclusions drawn from them, are common to both calculations.

Similar sensitivity calculations were also performed using different assumptions regarding cloud overlap and the height of the lowest cloud base (cf. Bergman and Hendon 1998). As with the calculated heating rates, circulations (not shown) are not as sensitive to these alterations as they are to the choice of observational data or SW absorptivity. So, while the choice of heating rate data significantly alters the magnitude of the specific features of calculated circulations, the general conclusions of sections 3a and 3b seem robust.

4. Seasonal cycle in the east Pacific

Figure 11a displays the observed seasonal cycles of meridional winds (shaded) and SST (thick contours) in terms of a 7-yr composite monthly time series. Values have been averaged over 85°–105°W, in the far eastern Pacific. Focusing on 0°–10°S, where SST variations are strongest, time mean meridional winds (left side) are northward at about 2.5 m s−1. That magnitude is only slightly larger than the maximum seasonal deviation and so the seasonal minimum of meridional winds during February and March indicates the near collapse of the cross-equatorial winds. The weakest northward winds between the equator and 10°S occur approximately one month before the seasonal maximum of SST and the strongest occur during August–October, just prior to the seasonal minimum of SST. Those relationships led Mitchell and Wallace (1992) to suggest that cross-equatorial winds, associated with the seasonal cycle of convection in the ITCZ, force the seasonal cycle of SST in the eastern Pacific cold tongue via Ekman upwelling.

The relationship between seasonal variations of convection and meridional winds, alluded to by Mitchell and Wallace (1992), is illustrated in Fig. 11b. This figure displays the seasonal cycle of meridional winds (shaded) calculated by the linear model from the seasonal cycle of total diabatic heating rates (thick contours) in the reanalysis data. The calculated winds that reproduce much of the observed behavior exhibit the classic response to zonally uniform tropical heating (e.g., Gill 1982). Seasonal variations of heating and meridional winds are either in phase or 180° (6 months) out of phase. Meridional structures are in quadrature: convergent winds at the surface are associated with heating maxima and divergent winds are associated with heating minima.

Figure 12 compares the meridional winds driven by the latent (Fig. 12a) and radiative (Fig. 12b) components of diabatic heating from the NCEP–NCAR reanalysis. North of the equator, latent heating is the dominant forcing for the seasonal cycle of meridional winds, but south of the equator, where the seasonal cycle of SST is the strongest, radiative heating rates are as, if not more, important.

The role of clouds is illustrated in the analysis of meridional winds calculated from ISCCP radiative heating rates. That seasonal cycle (Fig. 13a) exhibits substantial differences compared to the response calculated from NCEP–NCAR radiative heating rates (Fig. 12b). Seasonal variations in Fig. 13a are stronger than those in Fig. 12b, they have a 1–2-month phase lead in the latitude band 10°S–10°N, and there is nearly a 6-month phase difference at 20°S. These differences are expected for such detailed analysis using quantities as uncertain as heating rates. Nevertheless, Fig. 13a reinforces our conclusions from Fig. 12 that radiative heating is important for the seasonal cycle of meridional winds over the eastern Pacific cold tongue.

The seasonal cycle of meridional winds calculated from cloud radiative forcing (Fig. 13b) dominates the response to radiative heating (Fig. 13a). Of that response, convective CRF (i.e., from the ITCZ) contributes most of the response north of the equator (Fig. 14a), while subtropical CRF (i.e., from the subtropical eastern Pacific and cold tongue) contributes most of the response south of the equator (Fig. 14b). These results support earlier results by Nigam (1997), which indicated that cloud-top radiative cooling by low clouds in the southeastern Pacific is important to the enhancement of cross-equatorial winds that occurs from March to May.

5. Conclusions

This study investigated the role of clouds for low-latitude atmospheric circulations in a linearized calculation forced by specified diabatic heating rates. To the extent that observed circulations were reproduced by the calculation, this approach allowed us to quantify the relative contributions to the circulation by different components of the specified diabatic heating. Here, the role of clouds was quantified by the circulation forced by cloud radiative forcing, calculated from cloud properties in ISCCP.

Important aspects of the tropical and subtropical circulation were reproduced by the calculation when forced by total diabatic heating rates, particularly over the oceans. At these locations, observed spatial patterns of surface winds, geopotential height, and the deep equatorial circulation were realistically represented in the calculated circulation. The magnitudes of those features were typically within 30% of the observed magnitudes. Calculated circulations were not realistic at some locations at higher latitudes, especially during winter and over land, where orography, transient eddies, and variations of the basic state are known to be important. In light of these results, the role of cloud radiative forcing relative to other sources of atmospheric heating was analyzed for maritime tropical and subtropical circulations.

In general, cloud radiative forcing was found to contribute about 20% to the low-latitude maritime circulations, both at the surface and in the deep troposphere. Cloud radiative forcing tends to have a stronger influence in the lower troposphere than at upper levels. An important aspect of the cloud influence is that it reinforces circulations driven by latent heating rates from tropical convection. As a result, cloud radiative forcing systematically strengthens low-latitude circulations over the oceans. These results are consistent with, but indicate a smaller role of clouds than, what has been found in GCM simulations (e.g., Sherwood et al. 1994). However, in GCM studies, the dynamical impact of cloud radiative forcing is typically investigated by completely removing it from the GCM and comparing the results to a control simulation. That method does not separate the first-order impact of large-scale atmospheric CRF on the large-scale atmospheric circulation (which we calculate here) from the second-order impact. The second-order impact results from the influence of atmospheric CRF on the local stability, which then influences convection and, thus, latent heating of the atmosphere in the GCM (e.g., Randall et al. 1989). Because large-scale latent heating rates from tropical convection dominate diabatic heating rates in the Tropics, the second-order impact of atmospheric CRF on the large-scale tropical circulation might be more important than the first-order impact in the vicinity of strong convective activity. The first-order impact is probably more important in the subtropics.

For the most part, atmospheric cloud radiative forcing influences local circulations more than remote ones. In particular, the strong influence of clouds near the subtropical anticyclones, where CRF accounts for better than half of the observed strength of circulations, results from subtropical cloud radiative forcing. The influence of subtropical CRF is also evident for seasonal variations of meridional winds over the cold tongue in the eastern Pacific. If those winds are indeed important for the seasonal cycle of SST in the eastern Pacific cold tongue, as suggested by Mitchell and Wallace (1992), then our results support the conclusions of Nigam (1997). That is, cloud radiative forcing from low clouds in the eastern Pacific affects SSTs there, not only by reducing SW absorption at the surface, but also by enhancing atmospheric cooling rates in the atmosphere.

Acknowledgments

ISCCP cloud data were obtained from the NASA Langley Research Center EOSDIS Distributed Active Archive Center. The linear model was provided, along with helpful advice, by J. Whitaker of CDC. Heating rates calculated from the residual of the thermodynamical equation were provided by M. Newman of CDC. We wish also to acknowledge helpful comments by our colleagues at CDC and by anonymous reviewers. This research was supported in part by a Pan American Climate Studies Grant from NOAA’s Office of Global Programs.

REFERENCES

  • Bergman, J. W., and H. H. Hendon, 1998: Calculating monthly radiative fluxes and heating rates from monthly observations of cloud cover. J. Atmos. Sci.,55, 3471–3491.

  • ——, and ——, 2000: The impact of clouds on the seasonal cycle of radiative heating over the Pacific. J. Atmos. Sci.,57, 545–566.

  • Bony, S., Y. Sud, K. M. Lau, J. Susskind, and S. Saha, 1997: Comparison and satellite assessment of NASA/DAO and NCEP–NCAR reanalyses over tropical ocean: Atmospheric hydrology and radiation. J. Climate,10, 1441–1462.

  • Cess, R. D., and Coauthors, 1995: Absorption of solar radiation by clouds: Observations versus models. Science,267, 496–499.

  • ECMWF Research Department, 1992: Research manual 1: ECMWF data assimilation scientific documentation. ECMWF, 88 pp. [Available from ECMWF, Shinfield Park, Reading, Berkshire RG2 9AX, United Kingdom.].

  • Gill, A., 1980: Some simple solutions for heat induced tropical circulation. Quart. J. Roy. Meteor. Soc.,106, 447–462.

  • ——, 1982: Atmosphere–Ocean Dynamics. Academic Press, 662 pp.

  • Hoskins, B. J., H. H. Hsu, I. N. James, M. Matsutani, P. D. Sardeshmukh, and G. H. White, 1989: Diagnostics of the global atmospheric circulation based on ECMWF analyses 1979–1989. WCRP-27, WMO/TD-326, 217 pp. [Available from WMO, Case Postale No. 2300, CH-1211 Geneva 2, Switzerland.].

  • Kalnay, E. M., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc.,77, 432–471.

  • Kessler, W. S., L. M. Rothstein, and D. Chen, 1998: The annual cycle of SST in the eastern tropical Pacific diagnosed in an ocean GCM. J. Climate,11, 777–799.

  • Kiehl, J. T., J. J. Hack, M. H. Zhang, and R. D. Cess, 1995: Sensitivity of a GCM climate to enhanced shortwave cloud absorption. J. Climate,8, 2200–2212.

  • ——, ——, G. B. Bonan, B. A. Boville, B. P. Briegleb, D. L. Williamson, and P. J. Rasch, 1996: Description of the NCAR community climate model (CCM3). NCAR Tech. Note NCAR/TN-420+STR, National Center for Atmospheric Research, Boulder, CO, 152 pp.

  • Lau, N.-C., 1979: The observed structure of tropospheric stationary waves and the balance of vorticity and heat. J. Atmos. Sci.,36, 996–1016.

  • Li, Z., L. Moreau, and A. Arking, 1997: On the solar energy disposition: A perspective from observation and modeling. Bull. Amer. Meteor. Soc.,78, 53–70.

  • Liu, A. Z., M. Ting, and H. Wang, 1998: Maintenance of circulation anomalies during the 1988 drought and 1993 floods over the United States. J. Atmos. Sci.,55, 2810–2832.

  • Ma, C.-C., C. R. Mechoso, A. W. Robertson, and A. Arakawa, 1996:Peruvian stratus clouds and the tropical Pacific circulation: A coupled ocean–atmosphere GCM study. J. Climate,9, 1635–1645.

  • Mechoso, C. R., and Coauthors, 1995: The seasonal cycle over the tropical Pacific in coupled ocean–atmosphere general circulation models. Mon. Wea. Rev.,123, 2825–2838.

  • Mitchell, T. P., and J. M. Wallace, 1992: The annual cycle in equatorial convection and sea surface temperature. J. Climate,5, 1140–1156.

  • Newell, R. E., J. W. Kidson, D. G. Vincent, and G. J. Boer, 1974: The General Circulation of the Tropical Atmosphere. Vol. 2. The MIT Press, 371 pp.

  • Nigam, S., 1994: On the dynamical basis for the Asian summer monsoon rainfall–El Niño relationship. J. Climate,7, 1750–1771.

  • ——, 1997: The annual warm to cold phase transition in the eastern equatorial Pacific: Diagnosis of the role of stratus cloud-top-cooling. J. Climate,10, 2447–2467.

  • Ramanathan, V., 1987: The role of Earth radiation budget studies in climate and general circulation research. J. Geophys. Res.,92, 4075–4095.

  • ——, and W. C. Collins, 1991: Thermodynamic regulation of ocean warming by cirrus clouds deduced from observations of the 1987 El Nino. Nature,351, 27–32.

  • ——, B. Subasilar, G. J. Zhang, W. Conant, R. D. Cess, J. T. Kiehl, H. Grassl, and L. Shi, 1995: Warm pool heat budget and shortwave cloud forcing: A missing physics? Science,267, 499–503.

  • Randall, D. A., Harshvardhan, D. A. Dazlich, and T. G. Corsetti, 1989: Interactions among radiation, convection, and large-scale dynamics in a general circulation model. J. Atmos. Sci.,46, 1943–1970.

  • Ridout, J. A., and T. E. Rosmond, 1996: Global modeling of cloud radiative effects using ISCCP cloud data. J. Climate,9, 1479–1496.

  • Rossow, W. B., and R. A. Schiffer, 1991: ISCCP cloud data products. Bull. Amer. Meteor. Soc.,72, 2–20.

  • ——, and L. C. Garder, 1993: Validation of ISCCP cloud detections. J. Climate,6, 2370–2393.

  • ——, A. Walker, D. Beuschel, and M. Roiter, 1996: International Satellite Cloud Climatology Project (ISCCP) documentation of new cloud dataset. WMO/TD-No. 737, World Meteorological Organization, 115 pp. [Available online at http://isccp.giss.nasa.gov/documents.html].

  • Sherwood, S. C., V. Ramanathan, T. P. Barnett, M. K. Tyree, and E. Roeckner, 1994: Response of an atmospheric general circulation model to radiative forcing of tropical clouds. J. Geophys. Res.,99, 20 829–20 845.

  • Slingo, J. M., and A. Slingo, 1991: The response of a general circulation model to cloud longwave radiative forcing: II Further studies. Quart. J. Roy. Meteor. Soc.,117, 333–364.

  • Stephens, G. L., 1996: How much solar radiation do clouds absorb? Science,271, 1131–1133.

  • Weare, B. C., 1997: Comparison of NCEP–NCAR cloud radiative forcing reanalyses with observations. J. Climate,10, 2200–2209.

  • Webster, P. J., 1994: The role of hydrological processes in ocean–atmosphere interactions. Rev. Geophys.,32, 427–476.

  • Yanai, M., S. Esbensen, and J. H. Chu, 1973: Determination of bulk properties of tropical cloud clusters from large-scale heat and moisture budgets. J. Atmos. Sci.,30, 611–627.

Fig. 1.
Fig. 1.

Comparison of the surface circulation from the NCEP–NCAR reanalysis data with the calculated response of the linear model at 975 mb to NCEP total diabatic heating rates. Shown are 1984–90 averages for (a) JF reanalysis data, (b) AS reanalysis data, (c) JF calculated

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

Fig. 1.
Fig. 1.

(Continued) fields, and (d) AS calculated fields. The left-hand side in each panel show the zonal mean surface winds. The right-hand side show the eddy winds (arrows) and eddy geopotential height (m, contoured and shaded).

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

Fig. 2.
Fig. 2.

A comparison of the JF 1984–90 deep equatorial circulation from the NCEP–NCAR reanalysis data averaged over 5°S–5°N with calculated fields. (a) Reanalysis zonal mass flux (109 kg s−1), (b) reanalysis zonal and vertical wind

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

Fig. 2.
Fig. 2.

(Continued) vectors, (c) calculated zonal mass flux, and (d) calculated zonal and vertical wind vectors. Vertical winds in (b) and (d) have been enhanced by a factor of 1000.

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

Fig. 3.
Fig. 3.

Response of the linear model at 975 mb to (a) latent heating rates and (b) radiative heating rates from NCEP for JF 1984–90. The left-hand side shows the zonal mean surface winds. The right-hand side shows the eddy winds (arrows) and eddy geopotential height (m, contoured and shaded).

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

Fig. 4.
Fig. 4.

Zonal mass flux (109 kg s−1), averaged over 5°S–5°N from the linear model forced by (a) latent heating and (b) radiative heating from NCEP for JF 1984–90.

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

Fig. 5.
Fig. 5.

Response of the linear model at 975 mb to (a) ISCCP radiative heating rates and (b) cloud radiative forcing for JF 1984–90. The left-hand side shows the zonal mean surface winds. The right-hand side shows the eddy winds (arrows) and eddy geopotential height (m, contoured and shaded).

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

Fig. 6.
Fig. 6.

Response of the linear model at 975 mb to (a) LW cloud radiative forcing and (b) SW cloud radiative forcing for JF 1984–90. The left-hand side shows the zonal mean surface winds. The right-hand side shows the eddy winds (arrows) and eddy geopotential height (m, contoured and shaded).

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

Fig. 7.
Fig. 7.

Convective cloud mask applied to CRF for JF 1984–90.

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

Fig. 8.
Fig. 8.

Response of the linear model at 975 mb to (a) convective cloud radiative forcing and (b) subtropical cloud radiative forcing for JF 1984–90. The left-hand side shows the zonal mean surface winds. The right-hand side shows the eddy winds (arrows) and eddy geopotential height (m, contoured and shaded).

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

Fig. 9.
Fig. 9.

Response of the linear model to residual diabatic heating from NCEP for JF 1984–90. (a) Surface circulation: the left-hand side shows the zonal mean surface winds. The right-hand side shows the eddy winds (arrows) and eddy geopotential height (m, contoured and shaded). (b) Zonal mass flux (109 kg s−1) averaged over 5°S–5°N.

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

Fig. 10.
Fig. 10.

Response of the linear model at 975 mb to ISCCP cloud radiative forcing for JF 1989. The left-hand side shows the zonal mean surface winds. The right-hand side shows the eddy winds (arrows) and eddy geopotential height (m, contoured and shaded). (a) Control calculation. (b) Enhanced SW cloud absorption. (c) Using ISCCP-D data.

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

Fig. 11.
Fig. 11.

The 1984–90 seasonal cycle averaged over 85°–105°W in the eastern Pacific. (a) Meridional winds (m s−1, contoured and shaded) and SST (K, thick contours) from the reanalysis data. (b) Meridional winds (contoured and shaded) calculated from NCEP total diabatic heating rates (thick contours).

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

Fig. 12.
Fig. 12.

The 1984–90 seasonal cycle of meridional winds (m s−1) averaged over 85°–105°W in the eastern Pacific calculated from (a) NCEP latent heating rates and (b) NCEP radiative heating rates.

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

Fig. 13.
Fig. 13.

The 1984–90 seasonal cycle of meridional winds (m s−1) averaged over 85°–105°W in the eastern Pacific calculated from (a) ISCCP radiative heating rates and (b) cloud radiative forcing.

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

Fig. 14.
Fig. 14.

The 1984–90 seasonal cycle of meridional winds (m s−1) averaged over 85°–105°W in the eastern Pacific calculated from (a) convective cloud radiative forcing and (b) subtropical cloud radiative forcing.

Citation: Journal of the Atmospheric Sciences 57, 14; 10.1175/1520-0469(2000)057<2225:CRFOTL>2.0.CO;2

1

Here, we use the circulation from the lowest model level (975 mb) to represent the surface response of the model.

Save