The Aqua-Planet Experiment (APE) was first proposed by Neale and Hoskins (2000a) as a benchmark for atmospheric general circulation models (AGCMs) on an idealised water-covered Earth. The experiment and its aims are summarised, and its context within a modelling hierarchy used to evaluate complex models and to provide a link between realistic simulation and conceptual models of atmospheric phenomena is discussed. The simplified aqua-planet configuration bridges a gap in the existing hierarchy. It is designed to expose differences between models and to focus attention on particular phenomena and their response to changes in the underlying distribution of sea surface temperature.
Climate simulations by 16 atmospheric general circulation models (AGCMs) are compared on an aqua-planet, a water-covered Earth with prescribed sea surface temperature varying only in latitude. The idealised configuration is designed to expose differences in the circulation simulated by different models. Basic features of the aqua-planet climate are characterised by comparison with Earth. The models display a wide range of behaviour. The balanced component of the tropospheric mean flow, and mid-latitude eddy covariances subject to budget constraints, vary relatively little among the models. In contrast, differences in damping in the dynamical core strongly influence transient eddy amplitudes. Historical uncertainty in modelled lower stratospheric temperatures persists in APE. Aspects of the circulation generated more directly by interactions between the resolved fluid dynamics and parameterized moist processes vary greatly. The tropical Hadley circulation forms either a single or double inter-tropical convergence zone (ITCZ) at the equator, with large variations in mean precipitation. The equatorial wave spectrum shows a wide range of precipitation intensity and propagation characteristics. Kelvin mode-like eastward propagation with remarkably constant phase speed dominates in most models. Westward propagation, less dispersive than the equatorial Rossby modes, dominates in a few models or occurs within an eastward propagating envelope in others. The mean structure of the ITCZ is related to precipitation variability, consistent with previous studies. The aqua-planet global energy balance is unknown but the models produce a surprisingly large range of top of atmosphere global net flux, dominated by differences in shortwave reflection by clouds. A number of newly developed models, not optimised for Earth climate, contribute to this. Possible reasons for differences in the optimised models are discussed. The aqua-planet configuration is intended as one component of an experimental hierarchy used to evaluate AGCMs. This comparison does suggest that the range of model behaviour could be better understood and reduced in conjunction with Earth climate simulations. Controlled experimentation is required to explore individual model behaviour and investigate convergence of the aqua-planet climate with increasing resolution.
This paper explores the sensitivity of Atmospheric General Circulation Model (AGCM) simulations to changes in the meridional distribution of sea surface temperature (SST). The simulations are for an aqua-planet, a water covered Earth with no land, orography or sea-ice and with specified zonally symmetric SST. Simulations from 14 AGCMs developed for Numerical Weather Prediction and climate applications are compared. Four experiments are performed to study the sensitivity to the meridional SST profile. These profiles range from one in which the SST gradient continues to the equator to one which is flat approaching the equator, all with the same maximum SST at the equator. The zonal mean circulation of all models shows strong sensitivity to latitudinal distribution of SST. The Hadley circulation weakens and shifts poleward as the SST profile flattens in the tropics. One question of interest is the formation of a double versus a single ITCZ. There is a large variation between models of the strength of the ITCZ and where in the SST experiment sequence they transition from a single to double ITCZ. The SST profiles are defined such that as the equatorial SST gradient flattens, the maximum gradient increases and moves poleward. This leads to a weakening of the mid-latitude jet accompanied by a poleward shift of the jet core. Also considered are tropical wave activity and tropical precipitation frequency distributions. The details of each vary greatly between models, both with a given SST and in the response to the change in SST. One additional experiment is included to examine the sensitivity to an off-equatorial SST maximum. The upward branch of the Hadley circulation follows the SST maximum off the equator. The models that form a single precipitation maximum when the maximum SST is on the equator shift the precipitation maximum off equator and keep it centered over the SST maximum. Those that form a double with minimum on the equatorial maximum SST shift the double structure off the equator, keeping the minimum over the maximum SST. In both situations only modest changes appear in the shifted profile of zonal average precipitation. When the upward branch of the Hadley circulation moves into the hemisphere with SST maximum, the zonal average zonal, meridional and vertical winds all indicate that the Hadley cell in the other hemisphere dominates.
We examine the results of the Aqua-Planet Experiment Project (APE), focusing mainly on the structure of equatorial precipitation in the subset of participating models for which the details of model variables are available. Despite the unified setup of the APE, the Hovmöllor plots of precipitation in the models exhibit a considerable degree of diversity, presumably as a result of the diversity among the implementation of the various physical processes. Nevertheless, the wavenumber-frequency spectra of precipitation show certain similarities, and the power spectra can be divided into Kelvin, westward inertio-gravity, and “advective” components. The intensity of each of these three components varies significantly among different models. The composite spatial structures corresponding to these three components are produced by performing regression analysis with space-time filtered data. The composite horizontal structures of the Kelvin and westward inertio-gravity components are similar among the models, and resemble those expected from the corresponding equatorial shallow-water wave modes. These resemblances degrade at the altitude levels where the value of phase velocity approaches the zonal mean zonal wind speed. The horizontal structure of the advective component diverges significantly among the models. The composite vertical structures are strongly model dependent for all three components. The comparison of the vertical and horizontal structures associated with convective and stratiform heating of the composite disturbances indicates that the diversity of the vertical structures originates from differences in the implementation of the physical processes, especially the implementation of cumulus parameterization.
In this paper, we examine the steady state responses of models participating in the Aqua-Planet Experiment Project (APE) to the zonal asymmetry of equatorial sea surface temperature (SST) anomalies (SSTAs). Experiments were performed using three different SSTA distributions, i.e., two localized SSTAs with a common shape but different intensities, and an SSTA that varied with zonal wavenumber one. The structure of the responses obtained differs significantly among the models; however, some common features are also present. The principal features of the responses to the localized SSTAs are a positive precipitation anomaly over the warm SSTA, a widespread negative precipitation anomaly along the intertropical convergence zone, a pair of Rossby wavetrains along the equatorward flanks of mid-latitude westerly jets originating from a pair of upper tropospheric anticyclones that develop to the east of the warm SSTAs, and zonally wavelike precipitation and geopotential anomalies along the baroclinic zones. The structure of the tropical responses differs considerably from the Matsuno-Gill pattern, and the magnitude of the responses is almost proportional to the intensity of the localized SSTA in each of the models. The responses to the zonal wavenumber one SSTA are dominated by zonal wavenumber one structures. Around the longitudes of the warm (cold) SSTA, tropical precipitation increases (decreases). At longitudes east of the positive precipitation anomaly, the region of nearly zero absolute vorticity near the equator in the upper troposphere expands polewards, and the midlatitude westerly jets become narrower and stronger. To the west of the positive precipitation anomaly, the upper tropospheric region of nearly zero absolute vorticity shrinks, and the mid-latitude jets become weaker but broader, so that the regions of westerly winds extends to the equator, which results in the development of a zonal mean westerly wind anomaly around the equator. The longitudinal shift of the upper tropospheric westerly zonal wind anomaly relative to the precipitation anomaly is in marked contrast to that associated with the Walker circulation and the convection center around the Maritime Continent.
The effect of meridional variation of sea surface temperature (SST) on tropical atmospheric circulation is analyzed using Aqua-planet Experiment (APE) simulations.The meridional SST gradient around the narrow SST peak in CONTROL simulation favours a strong and single equatorial Intertropical Convergence Zone (ITCZ, defined by the maximum of zonally averaged total precipitation) in all APE models. In contrast, flat equatorial SST peak (FLAT simulation) favours split/double ITCZs flanking the SST maximum, in the majority of the APE models. Although there is reasonable agreement for SST sensitivity of ITCZ among the APE models in CONTROL, there exists disparity among them in FLAT case. Similarly, while the total and convective precipitation responses are consistent among the models, the large-scale precipitation response shows considerable inter-model variations in FLAT case. The APE intercomparison indicates that the occurrence and positioning of the ITCZ are primarily related to boundary layer moisture convergence as a response to the meridional variation of SST. Furthermore, the meridional gradient of tropospheric temperature is found to be an important factor that can influence the positioning of ITCZ. FLAT SST distribution is found to be similar to the observed distribution over the Indian region during summer season. Models that yield double ITCZs in this case simulate an easterly jet over the equatorial region (˜15° equatorward of the ITCZ). This is analogous to the Tropical Easterly Jet (TEJ), which is a unique feature observed over the Indian region during summer monsoon season, with its core at 12°N, equatorward of the seasonal convergence zone centered along 25°N. In these models, positive meridional temperature gradient and the associated easterly shear in the atmosphere strengthened by moisture convergence penetrate up to the upper troposphere, with which TEJ is in thermal wind balance.
An aqua-planet simulation using the Nonhydrostatic ICosahedral Atmospheric Model (NICAM) shows a diurnal precipitation cycle with a minor maximum in the afternoon, even though sea-surface temperature is constant during the simulation. The present study explores the factors that control the afternoon precipitation peak, making use of the simulation results. The temperature in the lower troposphere shows a minor minimum in the afternoon, coinciding with the precipitation peak. It is suggested that the “squeezing through temperature reduction” (whereby condensation is enhanced and more water vapor is squeezed within a cloud due to reduced temperature) is the most important factor in explaining the afternoon precipitation peak. The temperature minimum is associated with a dynamical process (not a diabatic process), and its relationship with the atmospheric tide is discussed.
The convergence of the zonal averaged equatorial precipitation with increasing vertical resolution in simulations with Community Atmosphere Model (CAM3) Eulerian spectral transform and finite volume dynamical cores is considered. The cores are both coupled to the standard CAM3 parameterization package. With the standard CAM3 26 level grid, the two versions converge to different states when the horizontal resolution alone is refined; the spectral transform to a single precipitation maximum and the finite volume to a double. With increasing vertical resolution both converge to a double structure. However, in the subsidence regions the high vertical resolution simulations have a very different climate balance and parameterized forcing than the lower resolution simulations and thus they do not represent the expected climate associated with the lower resolution dynamical cores. The cause of the different parameterized forcing is studied by considering the evolution of the 60-level model starting from a state created by the 26-level model. The cause is shown to be the discrete approximations in the shallow convection. When the 60-level model is presented with an initial state interpolated from a 26-level model state, the columns are stable by the discrete test in the shallow convection, even though they are unstable when the discrete calculation is based on the coarser 26-level grid. The Planetary Boundary Layer parameterization pumps water vapor into the lower troposphere, low clouds increase to unrealistic levels and force strong longwave radiative cooling. This destabilizes the column until the discrete test is satisfied on the 60-level grid and the shallow convection becomes active again. However the simulated state is by then very different and unlike the earth’s atmosphere. Similar unrealistic behavior has been seen in earth-like simulations.
Potential influence of a midlatitude SST frontal zone on the zonally symmetric variability in the extratropical atmosphere is assessed through idealized aqua-planet experiments with a general circulation model. In one experiment with a midlatitude frontal SST gradient as sharp as that observed in the Southwestern Indian Ocean, the annular mode is well reproduced in the model summer hemisphere. As actually observed in the Southern Hemisphere, the model annular mode represents a north-south seesaw in westerly wind speed around the climatological joint axes of a midlatitude westerly jet and storm track, with intensified (weakened) midlatitude westerlies under the enhanced (reduced) eddy momentum transport. This essential feature of the annular mode is retained also over the winter hemisphere, although its structure is somewhat distorted owing to the seasonal intensification of a subtropical jet (STJ). In the other experiment, elimination of the frontal SST gradient results in an equatorward shift of westerly wind anomalies associated with the annular mode in the summer hemisphere, in step with a shift of the mean joint axes of the jet and storm track. More importantly, both the amplitude and persistence of the mode are substantially reduced. In the winter hemisphere, the elimination also results in a marked weakening of the annular variability of midlatitude westerlies. Due to the weakening of the near-surface baroclinicity from lack of the frontal SST gradient, anomalous eddy momentum transport is also reduced markedly. Unlike the observed annular mode, the dominant mode of variability primarily represents STJ variability. Though idealized, these model experiments suggest the potential importance of a midlatitude oceanic frontal zone for the year-round dominance and robustness of the annular mode signal against the wintertime intensification of a STJ, by enhancing storm-track activity.
The responses of the equatorial zonal wind and the Hadley circulation to the equatorial zonal wavenumber one sea surface temperature (SST) anomaly, Ts*, are examined in an atmospheric general circulation model (AGCM) with an aquaplanet condition. The Hadley cell is weakened as the magnitude of Ts* increases, balancing with a decrease in the zonal-mean diabatic heating over the tropics. The decrease of heating reflects a nonlinear relationship between precipitation and SST; deep convection, such as a super cloud cluster, is significantly suppressed over cold Ts*, whereas is slightly enhanced over warm Ts*. The effective suppression of deep convection is accomplished by the stable boundary layer and the dry subsidence anomaly associated with the Walker cell which is excited by the SST anomaly. And the decreased convection acts to further reinforce the subsidence via thermodynamic balance. Therefore, this positive feedback between large-scale circulation and deep convection determines the nonlinear relationship and controls the strength of the Hadley cell. In terms of the energetics of the tropical circulation, the Hadley cell has to be weakened to compensate for the lack of energy supply caused by an increase of tropical radiative cooling due to the effective suppression of deep convection over cold Ts*. We compared the results of our AGCM with that of other 15 aquaplanet AGCMs integrated with the same SST distribution. While the Hadley cell is weakened in all AGCMs when Ts* is added to the zonal uniform SST, there is a large diversity in the strength. This suggests that the difference in the physical parameterization causes a different sensitivity of the Hadley cell response to zonally asymmetric SST. The magnitude of weakening is approximately proportional to the decreased (increased) amount of the deep convective precipitation (the radiative cooling) over the tropics. This strong relationship suggests that the positive feedback also works in other AGCMs. It is considered that the feedback is also important for understanding the formation of a real tropical climate.