1. Introduction
Clouds perturb the net top-of-the-atmosphere (TOA) irradiance from that of clear-sky conditions (e.g., Ramanathan et al. 1989). The spatial distribution of radiation deposited to the earth generates the temperature gradient. General circulations reduce the meridional temperature gradient by transporting the energy poleward. Because the dynamics is driven by the temperature gradient, it is postulated that poleward energy transport by dynamics is likely to be affected by the presence of clouds. Stuhlmann and Smith (1988) showed, using climatological low- and midlevel clouds, that both cloud types contribute to an increase in the rate of generation of zonal available potential energy in the midlatitudes and the polar regions. Because only a part of the available potential energy is converted to kinetic energy (Peixoto and Oort 1992), this does not necessarily means that clouds increase the rate of poleward energy transport. Zhang and Rossow (1997) estimated the effects of clouds on meridional energy transport from the International Satellite Cloud Climatology Project’s (ISCCP; Schiffer and Rossow 1983) global radiative flux dataset. They concluded that clouds enhance the meridional energy transport by the atmosphere and reduce the transport by the ocean. Zhang and Rossow (1997), however, did not consider the surface latent heat and sensible heat flux effects in estimating cloud effects on the meridional energy transport. Sohn and Smith (1992) discussed the effects of cloud type on the meridional circulation but they did not explicitly compute the zonal mean atmospheric cloud effect. Randall et al. (1989) used a general circulation model to estimate the impacts of the atmospheric cloud radiative effect. They concluded that the atmospheric cloud radiative effect over oceans intensifies the Hadley circulation, tropical easterlies, and subtropical westerly jets.
In this study, we use TOA and surface shortwave and longwave irradiances from Clouds and the Earth’s Radiant Energy System (CERES; Wielicki et al. 1996) data and estimate the zonal mean cloud radiative effect at the TOA and surface and to the atmosphere. We also include the cloud effect on the surface latent and sensible heat fluxes (enthalpy flux) using the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis (Kalnay et al. 1996) to analyze the zonal mean atmospheric cloud effects on the meridional energy transport by the atmosphere quantitatively. We then investigate the atmospheric cloud effect by cloud height and optical thickness so that the impacts on the meridional energy transport, especially from the midlatitudes to the polar regions, can be understood.
2. Method
a. CERES data and daily and monthly averaging
CERES data from the Terra satellite taken from March 2000 through February 2003 are used in this study. TOA and surface irradiances are computed with a two-stream model (Fu and Liou 1993; Kato et al. 2005) by forcing the agreement of the TOA irradiance with that derived from CERES radiance measurements by angular distribution models (Loeb et al. 2005; Kato and Loeb 2005). Inputs to the two-stream model are Moderate Resolution Imaging Spectroradiometer (MODIS)-derived cloud and aerosol properties, MODIS-derived surface skin temperature, ozone amounts (Yang et al. 2007), and temperature and relative humidity profiles from the Goddard Earth Observing System (EOS) Data Assimilation System (GEOS-4; Bloom et al. 2005). Six-hour and 1° × 1° GEOS-4 maps are linearly interpolated in space and time for the CERES overpass time and location. When MODIS-derived aerosol properties are not available, modeled aerosol properties obtained from a transport model (the Model of Atmospheric Transport and Chemistry, MATCH; Collins et al. 2001) are used. The ocean spectral surface albedo is from Jin and Stamnes (1994). Broadband land surface albedos are inferred from the clear-sky TOA albedo derived from CERES measurements (Charlock et al. 2006). More detailed descriptions of irradiance computations in the CERES project are found in Charlock et al. (2006). TOA-modeled irradiances are constrained to match CERES-derived irradiances by altering mainly cloud properties by the method described in Rose et al. (1997). Those instantaneous irradiances are included in a CERES product called the Cloud and Radiative Swath (CRS). The edition 2B, which is available from the National Aeronautics and Space Administration (NASA) Langley Atmospheric Science Data Center, is used in this study.
We convert instantaneous irradiances to daily mean irradiances by the method described in appendix A. The underlying assumption in the process is that meteorological conditions do not change over the course of a day. The error in the daily value, therefore, depends on the actual diurnal variation of the albedo and transmittance. For example, a systematic diurnal variation of water vapor or cloud amount that is not sampled at the overpass time (≈1030 LT) is not considered in our estimate. In addition, the solar zenith–dependent transmittance is estimated by averaging the instantaneous transmittance computed for the overpass time as a function of the solar zenith angle and scene type. Because Terra is on a sun-synchronous orbit, the transmittance for large solar zenith angles is from high-latitude regions. The surface irradiance near sunrise and sunset at low-latitude regions is, therefore, computed using transmittance derived from high-latitude regions. As a consequence, the daily downward surface irradiance in low-latitude regions is overestimated. In addition, no effort is made to correct the instantaneous longwave irradiance for the daily value; we simply averaged daytime and nighttime values weighted by the day–night fraction. We estimate the error in shortwave and longwave surface irradiances by comparing with surface observations in sections 3b and 3e.
3. Results
a. Zonal net irradiance
The net shortwave irradiance of the atmosphere Fatmsw, which is the shortwave absorption by the atmosphere, is primarily a function of solar zenith angle (Fig. 1a). The seasonal mean net shortwave irradiance of the atmosphere shown in Fig. 1a increases toward the tropics. However, the absorbed irradiance by the atmosphere in the summer hemisphere is nearly constant with latitude and rarely exceeds 100 W m−2 (Fig. 1a). When Fatmsw is divided by the daily mean insolation at the TOA, the seasonal mean shortwave absorptance of the atmosphere is between 0.2 and 0.25, except in the polar regions (Fig. 1d). When the absorbed irradiance is averaged, the global and annual mean shortwave absorptance of the atmosphere is 0.217. This is significantly larger than the absorptance (0.196) estimated by Kiehl and Trenberth (1997) and the absorptance (0.208) estimated by Raschke et al. (2005) from ISCCP-FD data (from 1991 through 1995). As mentioned above, our estimate is based on a two-stream model using cloud properties derived from MODIS and water vapor profile from GEOS-4, while Kiehl and Trenberth (1997) used a standard atmosphere with 49% low-level, 5% middle-level, and 20% high-level clouds. Aerosols are included in our estimate but excluded in the estimate by Kiehl and Trenberth (1997). The difference is then attributed to the difference in cloud properties (optical thickness, height, droplet size, phase), aerosol properties, and water vapor amount. The net longwave irradiance of the atmosphere Fatmlw is more negative in the tropics than in polar regions (Fig. 1b). Here, Fatmlw shows a larger seasonal variation in the Northern Hemisphere than the Southern Hemisphere (Fig. 1b). Because the larger seasonal variation in the Northern Hemisphere also appears in the clear-sky net longwave TOA irradiance (Fig. 1e), it is likely caused by the seasonal variation in the water vapor amount. The net longwave irradiance, Fatmlw, over the Antarctic is significantly less negative than Fatmlw over the Arctic because of the lower temperature and water vapor amount over Antarctica.
When the net shortwave and longwave atmospheric irradiances are added, Fatmrad is negative (Fig. 1c) because the magnitude of Fatmlw is larger than the magnitude of Fatmsw. The value is about −100 W m−2 for all latitude but Fatmrad is more negative in the Southern Hemisphere than in the Northern Hemisphere. Our estimate of the annual mean atmospheric net irradiance is similar to that given by Zhang and Rossow (1997) from ISCCP-FC data, although their estimate does not show the difference between the Northern and Southern Hemispheres.
b. Comparison with surface observations
As mentioned earlier, our estimate of the surface irradiance is based on a radiative transfer model. Assumptions made in the estimation process introduce errors in the daily and monthly mean shortwave and longwave surface irradiances. In this section, we compare modeled surface irradiances with observations for the error estimate. Note that the modeled shortwave and longwave TOA irradiances are tuned to agree with CERES irradiances. As a result, the daily global annual mean TOA irradiance agrees with those derived from CERES measurements to within 0.7% for shortwave and to within 0.2% for longwave.
We use observations taken at three sites: the province of Manus, Papua New Guinea (2°S, 147°E, tropics); the southern U.S. Great Plains (36°N, 97°W, midlatitudes); and Barrow, Alaska (71°N, 156°W, Arctic). Radiation data at Manus and the southern Great Plains (SGP) were taken as a part of the Atmospheric Radiation Measurement (ARM; Stokes and Schwartz 1994; Ackerman and Stokes 2003) program run by the Department of Energy. The National Oceanic and Atmospheric Administration/Climate Monitoring and Diagnostics Laboratory (NOAA/CMDL) Solar and Thermal Radiation (STAR) group took radiation data at Barrow, Alaska. Observed shortwave and longwave irradiances taken from March 2000 through February 2003 are averaged to compute monthly mean values. Figure 2 shows a comparison of monthly mean shortwave and longwave surface downward irradiances and Table 1 lists the annual mean shortwave and longwave irradiances for the three sites. The modeled annual mean downward shortwave irradiance over three 1° × 1° grid boxes, which contain the three observation sites, are 7.8% greater than the corresponding observations when the differences from the three sites are averaged. Similarly, the mean relative difference of the modeled annual mean downward longwave irradiance over three 1° × 1° grid boxes containing the three observation sites and corresponding observed irradiances is 1.1%, where the modeled irradiances at all three sites are smaller. When modeled shortwave and longwave irradiances are compared with observations at 25 Baseline Surface Radiation Network (BSRN; Ohmura et al. 1998) sites, including above the three sites, the mean relative difference is 6.5% for shortwave and −3.2% for longwave, and the standard deviations are 5.8% and 5.1%, respectively. Because the difference from the three sites is within the standard deviation of all BSRN sites, the above three sites are representative for all BSRN sites. Therefore, we use values from the three sites to estimate the errors in the modeled atmospheric irradiance below and in section 3(e). While the reason for the difference in the modeled and observed downward surface irradiances, especially for shortwave radiation, needs to be investigated in the future, we treat the difference as the error in the model in this study.
Upward shortwave and longwave irradiances are more problematic because measurements over a small area do not represent a larger area that CERES instruments observe. This is especially true for the Manus site since most of the area within the 1° × 1° grid box is ocean. The mean relative difference of the modeled and observed upward shortwave irradiances from the SGP and Barrow sites is 1.7% (modeled irradiance is greater) and the mean relative difference of the upward longwave irradiances of the two sites is 0.9% (modeled irradiance is greater). These differences lead to a 16.1% underestimation of the net shortwave irradiance of the atmosphere and a 4.4% overestimation of the net longwave irradiance of the atmosphere when the differences over the three sites are averaged. The relative difference of the shortwave plus longwave net irradiance of the atmosphere is −10.8%. Table 1 summarizes the differences between the modeled and observed annual mean irradiances at the three sites.
c. Comparison with IceSat-derived cloud cover
Because the surface irradiance is computed with MODIS-derived cloud properties, it is affected by retrieval errors such as those in cloud fraction, optical thickness, cloud height, droplet size, and phase. A part of the error in the surface irradiance shown in the previous section is, therefore, caused by the error in the retrieved cloud properties. Even though comparisons of these properties with those derived from more accurate methods are required to fully understand the cause of the error, we expect that the surface irradiance is most sensitive to the cloud fraction among cloud properties. We, therefore, only compare the cloud fraction with that derived from more accurate measurements in this section.
Figure 3 shows the meridional cloud fraction derived from MODIS (Minnis et al. 2003) and from the Geoscience Laser Altimeter System (GLAS) on board the Ice, Cloud, and Land Elevation Satellite (IceSat; Zwally et al. 2002). The GLAS cloud fraction is derived from a 1064-nm laser (medium resolution, GLA09 release 26) and the cloud optical thickness is derived from a 532-nm laser (medium resolution, GLA11 release 24). The maximum optical thickness that GLAS can derive is 5. The monthly mean cloud fractions over the Northern and Southern Hemisphere midlatitudes as derived from GLAS are 0.73 and 0.78, respectively. The corresponding cloud fractions from MODIS derived by the CERES cloud algorithm are 0.58 and 0.79, respectively. The threshold of cloud detection of the CERES cloud algorithm is 0.3 (Minnis et al. 2003). The fraction of clouds having an optical thickness greater than 0.3 derived from GLAS is 0.69 for the Northern Hemisphere midlatitudes and 0.75 for the Southern Hemisphere midlatitudes. If we neglect the sampling time difference between the two satellites, the error in the MODIS-derived cloud fraction is less than 0.1 in most regions when the cloud fraction from the GLAS-derived optical thickness greater than 0.3 is compared.
d. Zonal mean cloud effect
To understand cloud effects on the energy budget of the atmosphere, we compute the zonal mean cloud radiative effect at the top of the atmosphere and the surface, as well as the zonal mean cloud radiative effect to the atmosphere. The cloud radiative effects are computed as the net irradiance under all-sky conditions minus the net irradiance under clear-sky conditions (Ramanathan 1987; Ramanathan et al. 1989). A positive value indicates a warming effect by clouds. To avoid the influence of the water vapor amount and aerosol property differences between clear-sky and all-sky conditions (Li and Trishchenko 2001), we use computed irradiances with and without clouds for the estimate. In addition, this approach provides the cloud radiative effect for all samples as opposed to the approach using the difference between observed all-sky and clear-sky irradiances. Note that both shortwave and longwave irradiances computed with and without clouds are included in the CRS CERES product. The zonal mean shortwave effects at the TOA and the surface are both negative (Fig. 4). The zonal mean shortwave effect to the atmosphere is generally positive, which indicates that the zonal mean cloud radiative effect at the surface is more negative than that at the TOA. A large radiative effect to the atmosphere occurs in midlatitudes in spring and summer in both hemispheres. The zonal mean atmospheric cloud radiative effect is slightly negative (less than 1 W m−2) in the tropics and polar regions (10 W m−2). Note that the negative zonal mean shortwave cloud effect in the tropics is within the error estimated in the previous section (the model underestimates by ≈10 W m−2), if all bias error occurs under cloudy conditions. The zonal mean atmospheric shortwave cloud effect is, therefore, mostly positive. Low-level clouds increase the shortwave absorption in the atmosphere by increasing the photon pathlength while high-level clouds reduce the transmittance to absorbing water vapor layers at a lower part of the atmosphere. Figure 4 indicates, therefore, that increasing absorption by low-level clouds is larger than decreasing absorption by high-level clouds when the effects are weighted by the corresponding probability of cloud occurrence.
The zonal mean cloud longwave effect is shown in Fig. 5. When viewed from the TOA, clouds reduce the effective temperature by reducing the atmospheric transmittance and increasing the emission from the altitude where clouds are located. The effective temperature difference from the clear-sky value decreases as the surface temperature and cloud height decrease so that the zonal mean TOA cloud longwave effect decreases with latitude (Fig. 5a). A smaller zonal mean TOA longwave cloud effect near 20°N and 20°S is due to a small amount of clouds in these regions. When viewed from the surface, clouds increase the emissivity of the atmosphere and effective temperature. The emissivity difference from a clear-sky value increase as the water vapor amount decreases so that the zonal mean cloud radiative effect at the surface increases with latitude (Fig. 5b). The drop in the zonal mean surface cloud effect over Antarctica is also due to a smaller cloud amount compared to the cloud amount near 60°S. When the zonal mean surface effect is subtracted from the zonal mean TOA effect, the zonal mean atmospheric longwave effect is positive in the tropics and almost linearly decreases with latitude to negative values in polar regions (Fig. 5c). This simple linear relation with latitude is caused by the combination of the latitudinal dependence of the difference between the mean cloud-top height and the water vapor effective emission height determined from the surface and the latitudinal dependence of the cloud fraction (appendix B).
When zonal mean shortwave and longwave cloud radiative effects are combined, clouds reduce the net irradiance at the TOA except in the midlatitudes in the winter hemisphere (Fig. 6a). The zonal mean surface cloud effect is negative in the tropics and midlatitudes, and positive in polar regions except in the Arctic summer (Fig. 6b). The shortwave effect, therefore, dominates in the cloud radiative effect at the TOA and surface in the summer hemisphere. Because the shortwave effects at the TOA and at the surface are both cooling and nearly equal, the longwave effect dominates in the atmosphere (i.e., Fig. 6c closely resembles Fig. 5c). Subsequently, when the zonal mean shortwave and longwave effects are combined, a gradient in the meridional atmospheric cloud radiative effect exists—positive in the tropics and negative in polar regions. The meridional gradient persist for all seasons. Because the sign changes between low and high latitudes, the global mean radiative cloud radiative effect to the atmosphere is only 2.2 W m−2 when the zonal mean atmospheric cloud radiative effect is averaged. Although the sign of the cloud radiative effect is different, this small annual and global mean net atmospheric cloud radiative effect is consistent with the result of Raschke et al. (2005), who showed that the global mean value from ISCCP data is −1.3 W m−2.
The cloud radiative effect depends on the cloud properties. To understand the atmospheric cloud radiative effect dependence on cloud properties, Fig. 7 shows the net cloud radiative effect as a function of cloud optical thickness and cloud-top pressure. Only single-layer clouds are included in Fig. 7 to show the dependence clearly. Cloud optical thickness primarily affects the shortwave irradiance while cloud height primarily affects the longwave irradiance. Note that the atmospheric cloud radiative effect shown in Fig. 7 (second row) is the effect in the entire atmospheric column as a function of cloud-top height and cloud optical thickness; it is not the vertical profile of the effect. Figure 7 shows that the atmospheric cooling effect in the polar regions is caused by low-level clouds, which occur most frequently in polar regions (Fig. 7, bottom row). The atmospheric effect of high-level clouds is warming for all regions (Fig. 7). The warming effect is larger in the tropics than in the polar regions because the frequency of occurrence of thick high-level clouds is larger and the mean cloud-top height of high-level clouds is higher in the tropics.
The parameterization of the surface enthalpy flux suggests that the flux is a function of wind speed and the local gradient between the sea surface and the atmosphere (Washington and Parkinson 1986, p. 122; Fairall et al. 1996). We need to test, therefore, if there is any relation between the net surface irradiance and the surface enthalpy flux by using observations. For this purpose, we used the surface enthalpy flux from the Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite monthly data (HOAPS-3, 0.5° global grid; Grassl et al. 2000). Figure 8 shows the surface enthalpy flux anomalies divided by the monthly mean value as a function of the net surface irradiance anomalies divided by the monthly mean value from March 2000 between 20°N and 20°S. The anomalies are defined as the deviation of the zonal mean values from the monthly mean value over the 4-yr period from March 2000 through December 2004. Although, Fig. 8 indicates that the surface enthalpy flux change is positively correlated with the surface net irradiance change, the correlation coefficient is 0.16. If ocean surface property changes are caused by radiation and the surface flux is balanced by the surface net irradiance, the surface enthalpy flux is likely to correlate with the surface net irradiance. A weak correlation, as shown in Fig. 8 indicates, however, that the surface enthalpy flux anomalies are not predominately driven by the net surface irradiance anomalies over tropical oceans on a monthly time scale. We, therefore, consider that the cloud effect on the surface enthalpy flux estimated with (5) is the maximum cloud effect on the surface enthalpy flux. The minimum effect is simply no cloud effect on the surface enthalpy flux. The uncertainty envelope of the atmospheric cloud effect is then given with and without the assumption (5) on the surface enthalpy flux. The uncertainty in the cloud effect on the surface enthalpy flux is large in the tropics where the surface enthalpy flux is large. When the surface enthalpy flux is small, such as in the midlatitudes and polar regions, the uncertainty envelope is also small.
When the cloud effect on the surface enthalpy flux is included with the assumption that the cloud effect on the enthalpy flux is proportional to the net surface irradiance, the positive effect in the tropical atmosphere is almost eliminated (Fig. 9). Therefore, the uncertainty in the cloud effect on the surface enthalpy flux is too large to determine the atmospheric cloud effect observationally over the tropics. However, the meridional gradient of the atmospheric cloud effects between the midlatitudes and the polar regions remains because of a stronger cooling caused by clouds in polar regions and a smaller net enthalpy flux at the surface.
Our zonal mean atmospheric cloud radiative effect in the tropics is smaller than that estimated by Randall et al. (1989, Fig. 20) who ran a general circulation model with and without clouds over oceans to estimate the zonal mean atmospheric cloud radiative effect. Their results indicate that the cloud radiative effect acts to intensify the Hadley circulation and tropical easterlies. These patterns are caused by a larger tropical boundary layer wind speed, stronger surface evaporation, and a larger water vapor amount in the atmosphere. In their study, therefore, the cloud effect on the surface enthalpy flux is to increase the flux, further warm the atmosphere, and cool the surface instead of offsetting the cloud radiative effect in the tropics. This is opposite in effect of the cloud effect estimated with our assumption. Therefore, their zonal mean atmospheric cloud effect is larger than our estimate in the tropics. This suggests that the cloud effect on the surface enthalpy flux by (5) is too large and the gradient of the zonal mean cloud effect between the tropics and midlatitude might exist. Our zonal mean atmospheric cloud radiative effect over midlatitudes and polar regions is, however, similar to that estimated by Randall et al. Our result for the meridional gradient of cloud effect in the midlatitudes and the polar regions is, therefore, consistent with the results of Randall et al. (1989) that clouds intensify subtropical westerly jets because the meridional gradient of the atmospheric cloud effect between the midlatitudes and the polar regions can enhance the vertical shear of the geostrophic winds through the thermal wind relation (e.g., Holton 1992, p. 75).
e. Meridional gradient error estimate
If we take the mean net irradiance of the atmosphere from three sites used for the comparison, the relative difference of −10.8% estimated in section 3b corresponds to a 9.5 W m−2 bias error. As shown in Table 1, the net irradiance of the atmosphere over the SGP site is biased high by 3.1 W m−2 and that over the Barrow site is biased low by 9.7 W m−2. If we assume that all the error occurs in cloudy conditions and the error in clear-sky irradiances are negligible, the bias error in the zonal mean net atmospheric radiative cloud effect is −9.5 W m−2 and the uncertainty in the irradiance difference between midlatitude and polar regions is 12.8 W m−2. This error estimate in the zonal mean cloud effect is the upper limit because the error in the shortwave daily mean irradiance is partly due to the underestimated water vapor amount in the atmosphere at a large solar zenith angle in the low latitudes, which affects both the all-sky and clear-sky downward shortwave irradiances. A comparison with IceSat-derived cloud fraction shows that the MODIS-derived cloud fraction is biased low at Northern Hemisphere midlatitudes and biased high at 70°N. Because the atmospheric cloud effect in the polar regions is more cooling than that in the midlatitudes, the cloud fraction error at two sites is consistent with the error estimate of the meridional gradient of the cloud effect. Note that the error estimate derived from a comparison with surface observations includes the effects of errors in the cloud retrieval since the modeled surface irradiances were computed with MODIS-derived cloud properties.
Zhang and Rossow (1997) estimated that the uncertainty in the zonal mean surface enthalpy flux is at least 20 W m−2. This corresponds to about 13% of the net surface irradiance over tropical oceans. This leads to a 13% uncertainty in the surface cloud effect by (6), which gives about a 5 W m−2 uncertainty in both the zonal mean surface and the atmospheric cloud effect. Because of a smaller surface enthalpy flux in the midlatitudes and the polar regions than in the tropics, we expect that the uncertainty in the zonal mean atmospheric cloud effect due to the uncertainty in the surface enthalpy flux is smaller than 5 W m−2 in the midlatitudes and polar regions. Therefore, the gradient of the cloud effect between the midlatitudes and the polar regions shown in Fig. 9 is significant even when the modeling error and uncertainty in the surface enthalpy flux are considered. More importantly, the meridional gradient of the cloud effect is physically plausible as described in section 3b. We therefore conclude that the meridional gradient of the cloud effect in the midlatitudes and the polar regions is significant and we subsequently estimate the cloud effect on the meridional energy transport in section 4.
4. Discussion
Clouds present in polar regions cool the atmosphere more than those in the midlatitudes and tropics cool their residing atmosphere. The meridional gradient of the atmospheric cloud effect enhances the temperature gradient in the atmosphere from that found under clear-sky conditions. Because the mean zonal available potential energy is generated by heating warm air at low latitudes and cooling cold air at high latitudes (Lorenz 1955; Peixoto and Oort 1992, p. 377), this indicates that clouds increase the rate of generation of the mean zonal available potential energy in the atmosphere (Stuhlmann and Smith 1988).
Energy transport to the polar regions by dynamics can compensate for cooling by clouds if the zonal mean available potential energy is converted to kinetic energy. If the energy transport by the dynamics in the atmosphere does not compensate the cooling by clouds, the atmosphere is simply colder than the atmosphere with no clouds and the meridional temperature gradient in the atmosphere can intensify with time without changing the TOA net irradiance significantly. Because clouds alter the net TOA irradiance as shown in Fig. 6, it is postulated that clouds affect the rate of meridional energy transport (e.g., Zhang and Rossow 1997; Moore and Vonder Haar 2001; Weaver 2003). In this paper, we therefore, estimate the rate of energy transport by the atmosphere equivalent to the zonal mean cloud effect using a CERES dataset. We then examine, based on the zonal mean atmospheric effect by cloud type, whether or not the dynamics can indeed utilize the meridional gradient of the cloud effect in transporting the energy poleward.
Figure 10 shows the rate of the meridional energy transport in the atmosphere 2πRhFc cosθ separated by seasons. Positive values indicate northward transport and negative values indicate southward transport. Two lines that provide the uncertainty envelope are shown in each plot. The solid line is the total cloud effect integrated along the latitude and the dash–dot line considers only the radiative effect. As the effect of radiation on the surface enthalpy flux increases, 2πRhFc cosθ approaches the line that includes the surface enthalpy change in the cloud effect. The energy transport has a maximum (minimum) in the Northern Hemisphere (Southern Hemisphere) midlatitudes and it is mostly positive (negative) in the midlatitudes to the polar regions, especially in the fall and winter hemisphere. This estimate neglects dynamical feedbacks such as the enthalpy flux change due to the wind speed change suggested by Randall et al. (1989). When the cloud indirect effect on the surface enthalpy flux is included, however, the study by Randall et al. suggests that the cloud effect on the meridional enthalpy transport might be larger in the tropics. These two estimates (dash–dot and solid lines in Fig. 10), therefore, can be considered to be an envelope of the rate of the meridional atmospheric energy transport equivalent to the zonal mean direct cloud effect. This result indicates that the meridional gradient of the cloud effect increases the rate of poleward energy transport from the midlatitudes to the polar regions by the dynamics in the atmosphere.
5. Conclusions
We have estimated the zonal mean radiative cloud effect to the atmosphere using 3 yr of CERES data. Although it is mostly positive for all four seasons, the zonal mean shortwave atmospheric cloud effect is small. A large effect occurs near the midlatitudes in spring and summer, presumably due to a large cloud fraction. Clouds, therefore, increase the shortwave absorption in the atmosphere. The net zonal mean atmospheric cloud radiative effect is, however, dominated by the longwave effect. The zonal mean longwave effect results in a warming in the tropics and an approximately linear decrease with latitude to a cooling effect in the polar regions. The meridional-dependent atmospheric cloud radiative effect is a result of decreasing the TOA cloud effect and increasing the surface cloud effect with latitude. This meridional dependence of the cloud effects is caused by the latitudinal difference of the mean cloud-top height and water vapor effective emission height determined from the surface and the latitudinal variation in the cloud fraction.
We consider the upper limit of the cloud effect of the surface enthalpy flux with the assumption that the cloud effect of the surface enthalpy flux is proportional to the net irradiance at the surface. The uncertainty envelope of the atmospheric cloud effect is between the upper limit and the lower limit, which is simply given by no cloud effect on the surface enthalpy flux. While the uncertainty in the cloud effect on the surface enthalpy flux is large in the tropics, the meridional gradient between the midlatitudes and the polar regions exists even when uncertainties in the cloud effect on the surface enthalpy flux and in the modeled irradiances are taken into account.
The atmospheric cooling effect of clouds in the midlatitudes and the polar regions is caused by low-level clouds. Frequently occurring low-level clouds and their stronger cooling effect in the polar regions lead to the meridional gradient of the cloud effect between the midlatitudes and the polar regions. The meridional gradient of the cloud effect increases the rate of generation of the mean zonal available potential energy. Because the cooling effect over the polar regions is caused by relatively stationary low-level clouds, the meridional gradient of the cloud effect increases the rate of the meridional energy transport. Middle- and high-level clouds, which tend to be advected by winds have a warming effect on the atmosphere. This leads to an increase in the rate of conversion of the mean to eddy available potential energy. Clouds in polar regions then warm the surface, except in the Arctic in summer.
Because clouds form as a consequence of the dynamics, these results suggest a possible feedback process (e.g., Weaver 2003) among cloud type, meridional energy transport, and surface temperature in polar regions. If we consider that about 50% of the energy emitted to space by polar regions is provided from the midlatitudes (Kato et al. 2006), understanding the link between energy transport and the zonal mean cloud effect is just as important as understanding the local cloud feedback processes in polar regions.
Acknowledgments
We thank Drs. Bruce Wielicki, Dennis Hartmann, Tony Slingo, and Norman Loeb for useful discussions; Dr. Christopher Weaver for useful comments; and the NOAA/CMDL Solar and Thermal Radiation (STAR) group for providing surface radiation measurements. NCEP–NCAR reanalysis data were provided by the NOAA/OAR/Advection PSD, Boulder, Colorado, from their Web site (http://www.cdc.noaa.gov/). HOAPS data were obtained from their Web site (http://www.hoaps.zmaw.de/).
ARM data were made available through the U.S. Department of Energy as part of the Atmospheric Radiation Measurement program. The CERES data were obtained from the NASA Langley Atmospheric Science Data Center. The work was supported by a grant from the CERES and NASA Energy and Water Study (NEWS) projects (NNL04AA26G and NNL07AA00C) and a NASA grant (NNG04GL94G).
REFERENCES
Ackerman, T. P., and G. M. Stokes, 2003: The Atmospheric Radiation Measurement program. Phys. Today, 56 , 38–45.
Bloom, S. A., and Coauthors, 2005: Documentation and validation of the Goddard Earth Observing System (GEOS) data assimilation system version-4. NASA Tech. Rep. Series on Global Modeling and Data Assimilation NASA/TM-2005-104606, Vol. 26, 181 pp.
Charlock, T. P., F. G. Rose, D. A. Rutan, Z. Jin, and S. Kato, 2006: The global surface and atmosphere radiation budget: An assessment of accuracy with 5 years of calculations and observations. Preprints, 12th Conf. on Atmospheric Radiation, Madison, WI, Amer. Meteor. Soc., 10.5. [Available online at http://ams.confex.com/ams/pdfpapers/112984.pdf.].
Collins, W. D., P. J. Rasch, B. E. Eaton, B. V. Khattatov, J-F. Lamarque, and C. S. Zender, 2001: Simulating aerosols using a chemical transport model with assimilation of satellite aerosol retrievals: Methodology for INDOEX. J. Geophys. Res., 106 , 7313–7336.
Fairall, C. W., E. F. Bradley, D. P. Rogers, J. B. Edson, and G. S. Young, 1996: Bulk parameterization of air-sea fluxes for Tropical Ocean Global Atmosphere Coupled Ocean–Atmosphere Response Experiment. J. Geophys. Res., 101 , 3747–3764.
Fu, Q., and K-N. Liou, 1993: Parameterization of the radiative properties of cirrus clouds. J. Atmos. Sci., 50 , 2008–2025.
Grassl, H., V. Jost, R. Kumar, J. Schulz, P. Bauer, and P. Schluessel, 2000: The Hamburg Ocean-Atmosphere Parameteres and Fluxes from Satellite Data (HOAPS): A climatological atlas of satellite-derived air-sea-interaction parameters over the oceans. Max Planck Institute for Meteorology Rep. 312, Hamburg, Germany, 15 pp.
Holton, J. R., 1992: An Introduction to Dynamical Meteorology. 3rd ed. Academic, 511 pp.
Jin, Z., and K. Stamnes, 1994: Radiative transfer in nonuniformly reflecting layered media: Atmosphere–ocean system. Appl. Opt., 33 , 431–442.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437–471.
Kato, S., and N. G. Loeb, 2003: Twilight irradiance reflected by the earth estimated from Clouds and the Earth’s Radiant Energy System (CERES) measurements. J. Climate, 16 , 2646–2650.
Kato, S., and N. G. Loeb, 2005: Top-of-atmosphere shortwave broadband observed radiance and estimated irradiance over polar regions from Clouds and the Earth’s Radiant Energy System (CERES) instruments on Terra. J. Geophys. Res., 110 .D07202, doi:10.1029/2004JD005308.
Kato, S., F. G. Rose, and T. P. Charlock, 2005: Computation of domain-averaged irradiance using satellite-derived cloud properties. J. Atmos. Oceanic Technol., 22 , 146–164.
Kato, S., N. G. Loeb, P. Minnis, J. A. Francis, T. P. Charlock, D. A. Rutan, E. E. Clothiaux, and S. Sun-Mack, 2006: Seasonal and interannual variations of top-of-atmosphere irradiance and cloud cover over polar regions derived from the CERES data set. Geophys. Res. Lett., 33 .L19804, doi:10.1029/2006GL026685.
Kiehl, J. T., and K. E. Trenberth, 1997: Earth’s annual global mean energy budget. Bull. Amer. Meteor. Soc., 78 , 197–208.
Li, Z., and A. Trishchenko, 2001: Quantifying uncertainties in determining SW cloud radiative forcing and cloud absorption due to variability in atmospheric conditions. J. Atmos. Sci., 58 , 376–389.
Loeb, N. G., N-M. Smith, S. Kato, W. F. Miller, S. K. Gupta, P. Minnis, and B. A. Wielicki, 2003: Angular distribution models for top-of atmosphere radiative flux estimation from the Clouds and the Earth’s Radiant Energy System instrument on the Tropical Rainfall Measuring Mission Satellite. J. Appl. Meteor., 42 , 240–265.
Loeb, N. G., S. Kato, K. Loukachine, and N. M. Smith, 2005: Angular distribution models for top-of-atmosphere radiative flux estimation from the Clouds and the Earth’s Radiant Energy System instrument on the Terra satellite. Part I: Methodology. J. Atmos. Oceanic Technol., 22 , 338–351.
Lorenz, E. N., 1955: Available potential energy and the maintenance of the general circulation. Tellus, 7 , 157–167.
Minnis, P., D. F. Young, S. Sun-Mack, P. W. Heck, D. R. Doelling, and Q. Trepte, 2003: CERES cloud property retrievals from imagers on TRMM, Terra, and Aqua. Remote Sensing of Clouds and the Atmosphere VIII, K. P. Schaefer et al., Eds., International Society for Optical Engineering (SPIE Proceedings, Vol. 5235), 37–48.
Moore, R. W., and T. H. Vonder Haar, 2001: Interannual variability of cloud forcing and meridional energy transport for the Northern Hemisphere winter from 1984 to 1990. J. Climate, 14 , 3643–3654.
Ohmura, A., and Coauthors, 1998: Baseline Surface Radiation Network (BSRN/WCRP): New precision radiometry for climate change research. Bull. Amer. Meteor. Soc., 79 , 2115–2136.
Peixoto, J. P., and A. H. Oort, 1992: Physics of Climate. American Institute of Physics, 520 pp.
Ramanathan, V., 1987: The role of Earth radiation budget studies in climate and general circulation research. J. Geophys. Res., 92 , 4075–4095.
Ramanathan, V., R. D. Cess, E. F. Harrison, P. Minnis, R. B. Barkstrom, E. Ahmad, and D. Hartmann, 1989: Cloud-radiative forcing and climate: Results from the Earth Radiation Budget Experiment. Science, 243 , 57–63.
Randall, D. A., D. A. Dazlich, and T. G. Gorsetti, and Harshvardhan, 1989: Interactions among radiation, convection and large-scale dynamics in a general circulation model. J. Atmos. Sci., 46 , 1943–1970.
Raschke, E., A. Omura, W. B. Rossow, E. B. Carlson, Y. C. Zang, C. Stubenrauch, M. Kottek, and M. Wild, 2005: Cloud effects on the radiation budget based on ISCCP data (1991 to 1995). Int. J. Climatol., 25 , 1103–1125.
Rose, F., T. P. Charlock, D. A. Rutan, and G. L. Smith, 1997: Tests of a constraint algorithm for the surface and atmospheric radiation budget. Preprints, Ninth Conf. on Atmospheric Radiation, Long Beach, CA, Amer. Meteor. Soc., 466–499.
Schiffer, R. A., and W. B. Rossow, 1983: The International Satellite Cloud Climatology Project (ISCCP): The first project of the World Climate Research Programme. Bull. Amer. Meteor. Soc., 64 , 779–784.
Sohn, B-J., and E. A. Smith, 1992: The significance of cloud-radiative forcing to the general circulation on climate time scales—A satellite interpretation. J. Atmos. Sci., 49 , 845–860.
Stokes, G. M., and S. E. Schwartz, 1994: The Atmospheric Radiation Measurement Program: Programatic background and design of the Cloud and Radiation Testbed. Bull. Amer. Meteor. Soc., 75 , 1201–1221.
Stuhlmann, R., and G. L. Smith, 1988: A study of cloud-generated radiative heating and its generation of available potential energy. Part II: Results for a climatological zonal mean January. J. Atmos. Sci., 45 , 3928–3943.
Trenberth, K. E., and J. M. Caron, 2001: Estimates of meridional atmosphere and ocean heat transports. J. Climate, 14 , 3433–3443.
Trenberth, K. E., and D. P. Stepaniak, 2003: Covariability of components of poleward atmospheric energy transports on seasonal and interannual timescale. J. Climate, 15 , 3691–3705.
Washington, W. M., and C. L. Parkinson, 1986: An Introduction to Three-Dimensional Climate Modeling. University Science Books, 422 pp.
Weaver, C. P., 2003: Efficiency of storm tracks and important climate parameter? The role of cloud radiative forcing in poleward heat transport. J. Geophys. Res., 108 .4108, doi:10.1029/2002JD002756.
Wielicki, B. A., B. R. Barkstrom, E. F. Harrison, B. B. Lee III, G. Louis Smith, and J. E. Cooper, 1996: Clouds and the Earth’s Radiant Energy System (CERES): An Earth Observing System experiment. Bull. Amer. Meteor. Soc., 77 , 853–868.
Yang, S-K., S. Zhou, and A. J. Miller, cited. 2007: SMOBA: A 3-dimensional daily ozone analysis using SBUV/2 and TOVS measurements. [Available online at www.cpc.ncep.noaa.gov/products/stratosphere/SMOBA/smoba_doc.html.].
Zhang, Y-C., and W. B. Rossow, 1997: Estimating meridional energy transport by the atmospheric and oceanic general circulations using boundary fluxes. J. Climate, 10 , 2358–2373.
Zwally, H. J., and Coauthors, 2002: ICESat’s laser measurements of polar ice, atmosphere, ocean, and land. J. Geodyn., 34 , 405–445.
APPENDIX A
Daily Mean Irradiance Estimate
To convert instantaneous irradiances to a daily value, we need to obtain the TOA albedo, transmission, and surface albedo as a function of the solar zenith angle. We sort and average the instantaneous TOA albedos estimated from the CERES radiance as a function of the solar zenith angle and scene type using data taken from the Tropical Rainfall Measuring Mission (TRMM) satellite from January 1998 through August 1998 and March 2000. Note that the effect of a spherical earth that leads to a nonnegligible TOA shortwave irradiance over regions where the solar zenith angle is greater than 90° is included based on the work of Kato and Loeb (2003). Similarly, we sort and average transmissions and surface albedos as a function of solar zenith angle and scene type using 1 yr of Terra CRS from March 2000 through February 2001. The increment of the solar zenith angle bin is 10° for the TOA albedo and 5° for the surface transmission and albedo. The number of scene types that contain no snow and sea ice is 590 (Loeb et al. 2003). The number of snow and sea ice scenes is 60 (Kato and Loeb 2005).
APPENDIX B
Zonal Mean Atmospheric Cloud Radiative Effect
(a) Seasonal and zonal mean net shortwave irradiances, (b) net longwave irradiance, and (c) net shortwave and longwave irradiance of the atmosphere under all-sky conditions. (d) The atmospheric shortwave absorptance is computed from (a) and divided by the zonal mean solar constant. The (e) net longwave irradiance and (f) net shortwave plus longwave irradiance of the atmosphere under clear-sky conditions are also shown.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
(a) Seasonal and zonal mean net shortwave irradiances, (b) net longwave irradiance, and (c) net shortwave and longwave irradiance of the atmosphere under all-sky conditions. (d) The atmospheric shortwave absorptance is computed from (a) and divided by the zonal mean solar constant. The (e) net longwave irradiance and (f) net shortwave plus longwave irradiance of the atmosphere under clear-sky conditions are also shown.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
(a) Seasonal and zonal mean net shortwave irradiances, (b) net longwave irradiance, and (c) net shortwave and longwave irradiance of the atmosphere under all-sky conditions. (d) The atmospheric shortwave absorptance is computed from (a) and divided by the zonal mean solar constant. The (e) net longwave irradiance and (f) net shortwave plus longwave irradiance of the atmosphere under clear-sky conditions are also shown.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Comparison of the (top) monthly mean downward shortwave, (middle) longwave, and (bottom) net atmospheric irradiances for the Manus (TWP), Southern Great Plains (SGP), and Barrow, AK (NSA), sites. Open circles and closed squares indicate modeled irradiances and observations derived from March 2000 through February 2003, respectively. The error bars indicate the maximum and minimum observed values during the 5-yr period (March 2000–February 2005) for the surface downward shortwave and longwave and 4-yr period (March 2000–February 2004) for the atmospheric net.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Comparison of the (top) monthly mean downward shortwave, (middle) longwave, and (bottom) net atmospheric irradiances for the Manus (TWP), Southern Great Plains (SGP), and Barrow, AK (NSA), sites. Open circles and closed squares indicate modeled irradiances and observations derived from March 2000 through February 2003, respectively. The error bars indicate the maximum and minimum observed values during the 5-yr period (March 2000–February 2005) for the surface downward shortwave and longwave and 4-yr period (March 2000–February 2004) for the atmospheric net.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Comparison of the (top) monthly mean downward shortwave, (middle) longwave, and (bottom) net atmospheric irradiances for the Manus (TWP), Southern Great Plains (SGP), and Barrow, AK (NSA), sites. Open circles and closed squares indicate modeled irradiances and observations derived from March 2000 through February 2003, respectively. The error bars indicate the maximum and minimum observed values during the 5-yr period (March 2000–February 2005) for the surface downward shortwave and longwave and 4-yr period (March 2000–February 2004) for the atmospheric net.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
(a) Seasonal zonal mean cloud fraction derived from MODIS measurements by the CERES cloud algorithm (Minnis et al. 2003). Data from March 2000 through February 2004 are averaged. (b) Monthly zonal mean cloud fraction derived from October 2003 GLAS 1064-nm laser data for all clouds (dotted line), those for which the optical thickness was greater than 0.3 (solid line), and for when the optical thickness was greater than 1 (dash–dot line). The cloud optical thickness is derived from GLAS 532-nm laser data. The cloud fraction derived from MODIS radiances for the same month is shown by the dashed line. (c) Difference between the MODIS- and GLAS-derived cloud fractions. The fraction of clouds of which the optical thickness is less than 0.3 is excluded from the GLAS-derived cloud fraction.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
(a) Seasonal zonal mean cloud fraction derived from MODIS measurements by the CERES cloud algorithm (Minnis et al. 2003). Data from March 2000 through February 2004 are averaged. (b) Monthly zonal mean cloud fraction derived from October 2003 GLAS 1064-nm laser data for all clouds (dotted line), those for which the optical thickness was greater than 0.3 (solid line), and for when the optical thickness was greater than 1 (dash–dot line). The cloud optical thickness is derived from GLAS 532-nm laser data. The cloud fraction derived from MODIS radiances for the same month is shown by the dashed line. (c) Difference between the MODIS- and GLAS-derived cloud fractions. The fraction of clouds of which the optical thickness is less than 0.3 is excluded from the GLAS-derived cloud fraction.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
(a) Seasonal zonal mean cloud fraction derived from MODIS measurements by the CERES cloud algorithm (Minnis et al. 2003). Data from March 2000 through February 2004 are averaged. (b) Monthly zonal mean cloud fraction derived from October 2003 GLAS 1064-nm laser data for all clouds (dotted line), those for which the optical thickness was greater than 0.3 (solid line), and for when the optical thickness was greater than 1 (dash–dot line). The cloud optical thickness is derived from GLAS 532-nm laser data. The cloud fraction derived from MODIS radiances for the same month is shown by the dashed line. (c) Difference between the MODIS- and GLAS-derived cloud fractions. The fraction of clouds of which the optical thickness is less than 0.3 is excluded from the GLAS-derived cloud fraction.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Seasonal and zonal mean cloud shortwave radiative effects (a) at the TOA, (b) at the surface, and (c) to the atmosphere.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Seasonal and zonal mean cloud shortwave radiative effects (a) at the TOA, (b) at the surface, and (c) to the atmosphere.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Seasonal and zonal mean cloud shortwave radiative effects (a) at the TOA, (b) at the surface, and (c) to the atmosphere.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Seasonal and zonal mean cloud longwave radiative effects (a) at the TOA, (b) at the surface, and (c) to the atmosphere.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Seasonal and zonal mean cloud longwave radiative effects (a) at the TOA, (b) at the surface, and (c) to the atmosphere.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Seasonal and zonal mean cloud longwave radiative effects (a) at the TOA, (b) at the surface, and (c) to the atmosphere.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Seasonal and zonal mean cloud radiative effects (shortwave plus longwave) (a) at the TOA, (b) at the surface, and (c) to the atmosphere.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Seasonal and zonal mean cloud radiative effects (shortwave plus longwave) (a) at the TOA, (b) at the surface, and (c) to the atmosphere.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Seasonal and zonal mean cloud radiative effects (shortwave plus longwave) (a) at the TOA, (b) at the surface, and (c) to the atmosphere.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Contours of the daily mean cloud shortwave plus longwave effects (top row) at the TOA, (second row) to the atmosphere, and (third row) to the surface for the tropics (30°N–30°S), Northern Hemisphere midlatitudes (30°–60°N), Southern Hemisphere midlatitudes (30°S–60°N), the Arctic (60°–90°N), and the Antarctic (60°–90°S) as a function of the cloud optical thickness (τ) and cloud-top height in pressure coordinates estimated from July 2002 data. Only single-layer clouds are used. Contours in the fourth row indicate the logarithm (base 10) of the 2D normalized histogram of cloud occurrence. Daily mean irradiances are computed by the method discussed in appendix A except that daytime and nighttime longwave irradiances are weighted by the number of samples.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Contours of the daily mean cloud shortwave plus longwave effects (top row) at the TOA, (second row) to the atmosphere, and (third row) to the surface for the tropics (30°N–30°S), Northern Hemisphere midlatitudes (30°–60°N), Southern Hemisphere midlatitudes (30°S–60°N), the Arctic (60°–90°N), and the Antarctic (60°–90°S) as a function of the cloud optical thickness (τ) and cloud-top height in pressure coordinates estimated from July 2002 data. Only single-layer clouds are used. Contours in the fourth row indicate the logarithm (base 10) of the 2D normalized histogram of cloud occurrence. Daily mean irradiances are computed by the method discussed in appendix A except that daytime and nighttime longwave irradiances are weighted by the number of samples.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Contours of the daily mean cloud shortwave plus longwave effects (top row) at the TOA, (second row) to the atmosphere, and (third row) to the surface for the tropics (30°N–30°S), Northern Hemisphere midlatitudes (30°–60°N), Southern Hemisphere midlatitudes (30°S–60°N), the Arctic (60°–90°N), and the Antarctic (60°–90°S) as a function of the cloud optical thickness (τ) and cloud-top height in pressure coordinates estimated from July 2002 data. Only single-layer clouds are used. Contours in the fourth row indicate the logarithm (base 10) of the 2D normalized histogram of cloud occurrence. Daily mean irradiances are computed by the method discussed in appendix A except that daytime and nighttime longwave irradiances are weighted by the number of samples.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Normalized 2D histogram sorted by the net surface irradiance (shortwave + longwave) anomaly divided by the monthly mean value and the surface enthalpy (latent heat and sensible heat) flux anomaly divided by the monthly mean value. The latent heat and sensible heat fluxes are taken from HOAPS data (Grassl et al. 2000). The anomalies are defined as the deviation of the zonal mean values from the averaged value over the 4-yr period from March 2000 through December 2004. One month of data (March 2000) between 20°N and 20°S were used for the plot. The solid line indicates the linear regression fit. The slope of the regression line is 0.36 and the correlation coefficient is 0.16.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Normalized 2D histogram sorted by the net surface irradiance (shortwave + longwave) anomaly divided by the monthly mean value and the surface enthalpy (latent heat and sensible heat) flux anomaly divided by the monthly mean value. The latent heat and sensible heat fluxes are taken from HOAPS data (Grassl et al. 2000). The anomalies are defined as the deviation of the zonal mean values from the averaged value over the 4-yr period from March 2000 through December 2004. One month of data (March 2000) between 20°N and 20°S were used for the plot. The solid line indicates the linear regression fit. The slope of the regression line is 0.36 and the correlation coefficient is 0.16.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Normalized 2D histogram sorted by the net surface irradiance (shortwave + longwave) anomaly divided by the monthly mean value and the surface enthalpy (latent heat and sensible heat) flux anomaly divided by the monthly mean value. The latent heat and sensible heat fluxes are taken from HOAPS data (Grassl et al. 2000). The anomalies are defined as the deviation of the zonal mean values from the averaged value over the 4-yr period from March 2000 through December 2004. One month of data (March 2000) between 20°N and 20°S were used for the plot. The solid line indicates the linear regression fit. The slope of the regression line is 0.36 and the correlation coefficient is 0.16.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Seasonal and zonal mean cloud effects to the atmosphere. The effects include that on the shortwave and longwave irradiances, as well as on the surface latent and sensible heat fluxes (enthalpy flux).
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Seasonal and zonal mean cloud effects to the atmosphere. The effects include that on the shortwave and longwave irradiances, as well as on the surface latent and sensible heat fluxes (enthalpy flux).
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Seasonal and zonal mean cloud effects to the atmosphere. The effects include that on the shortwave and longwave irradiances, as well as on the surface latent and sensible heat fluxes (enthalpy flux).
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Seasonal and zonal mean atmospheric cloud effects converted to the rate of meridional energy transport by the atmosphere. The solid line indicate the equivalent meridional energy transport including the cloud effect on the net irradiance, the surface latent heat, and the sensible heat fluxes. The dash–dot line indicates the equivalent meridional enthalpy transport including only the cloud effect on the net irradiance. The positive and negative value indicates, respectively, northward and southward transports.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Seasonal and zonal mean atmospheric cloud effects converted to the rate of meridional energy transport by the atmosphere. The solid line indicate the equivalent meridional energy transport including the cloud effect on the net irradiance, the surface latent heat, and the sensible heat fluxes. The dash–dot line indicates the equivalent meridional enthalpy transport including only the cloud effect on the net irradiance. The positive and negative value indicates, respectively, northward and southward transports.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Seasonal and zonal mean atmospheric cloud effects converted to the rate of meridional energy transport by the atmosphere. The solid line indicate the equivalent meridional energy transport including the cloud effect on the net irradiance, the surface latent heat, and the sensible heat fluxes. The dash–dot line indicates the equivalent meridional enthalpy transport including only the cloud effect on the net irradiance. The positive and negative value indicates, respectively, northward and southward transports.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Fig. B1. Monthly and zonal mean cloud-top pressure derived from daytime July 2002 MODIS data by the CERES cloud algorithm (Minnis et al. 2003). Only single-layer clouds are considered.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Fig. B1. Monthly and zonal mean cloud-top pressure derived from daytime July 2002 MODIS data by the CERES cloud algorithm (Minnis et al. 2003). Only single-layer clouds are considered.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Fig. B1. Monthly and zonal mean cloud-top pressure derived from daytime July 2002 MODIS data by the CERES cloud algorithm (Minnis et al. 2003). Only single-layer clouds are considered.
Citation: Journal of Climate 21, 17; 10.1175/2008JCLI1982.1
Observed and modeled annual mean irradiances.
