U.S. Dept. of Commerce / NOAA
/ OAR / GFDL
*Disclaimer
To study the dynamical interaction between large-scale wave disturbances and the general circulation of the atmosphere.
To identify and elucidate the physical and dynamical mechanisms which maintain climate and cause its variation, and to examine their generality in the context of paleoclimate and the atmospheres of other planets.
To evaluate the impact of human activities on climate.
2.1 BACKGROUND FOR COUPLED CLIMATE
MODELING AT GFDL CLIMATE CHANGE
This section summarizes the general characteristics of two distinct coupled atmosphere-ocean climate models, both of which are in active use in research on global warming and other aspects of climate variability and sensitivity.
The R30 coupled model, which has been used extensively during the past year, has an atmospheric horizontal resolution of 3.75° longitude and 2.25° latitude, with 14 levels in the vertical. It is coupled to an ocean model with 2° horizontal resolution, a simple free-drift ice model, and a "bucket" land hydrology. The coupled model is run with adjustments to the air-sea fluxes of heat and fresh water to prevent climate drift. Two versions of the R30 coupled model have been developed, differing in their oceanic sub-grid scale diffusivities and initialization procedures. Using nomenclature taken from the IPCC Third Assessment Report, these two versions are identified as GFDL_R30_b and GFDL_R30_c. (A different nomenclature was used in A99/P00, in which these two versions were identified as R30V1 and R30V2, respectively. A more detailed description of the differences between GFDL_R30_b and GFDL_R30_c appears in A99/P00.) Several long control integrations have been performed with these models, and numerous simulations of all or part of the period 1865-2089.
The R15 coupled model has only half of the horizontal resolution of R30 in both the atmosphere and ocean. The low resolution and simplicity of this model, as well as the flux adjustment strategy, are designed to allow integrations that would otherwise be computationally prohibitive. This model has been invaluable throughout the past decade for initial studies of the variability and sensitivity of the coupled atmosphere-ocean system.
Both the R15 and R30 models have been used to simulate the evolution of the earth's climate due to past changes in greenhouse gases and sulfate aerosols, as well as to project climate change into the 21st century. In these "global warming scenario" integrations, an equivalent CO2 concentration is used to represent changes in all of the trace greenhouse gases, and changes in aerosol loading are modeled by changing the surface albedo. The runs begin in either 1765 or 1865, depending on resolution, using historical greenhouse gas and aerosol concentrations. The runs proceed into the late 21st century with equivalent CO2 increasing at a rate of 1% per year after 1990. All of these projections are clearly uncertain. The radiative forcing for the 20th century is also uncertain, due to uncertainties in aerosol loading (both anthropogenic and volcanic), indirect effects of aerosols on clouds, and possible changes in solar irradiance. Any climate projections must be considered with these uncertainties in mind.
2.2.1 Sea Ice and Global Warming
Decreases in late 20th century Arctic sea ice coverage simulated in an R15 version of the GFDL coupled climate model that included transient greenhouse gas and sulfate aerosol forcing has been shown to agree well with the observations (1657). Similar Arctic sea ice trends exist in more recent, higher resolution R30 climate model simulations (GFDL_R30_c). Fig. 2.1 summarizes the evolution of sea ice distributions calculated in three R30 climate change scenario experiments. (Averaging the results of three experiments makes it easier to see the greenhouse gas-induced signal among the year-to-year noise of interannual variability.) By the year 2000, the total volume of Arctic sea ice is between 75 and 80 percent of that which was simulated to exist in the 1950s. In these experiments, the total volume of Arctic sea ice continues to decrease through the 21st century, so that only about half of that which was present in the 1950s remains in the year 2050.
2.2.2 Water Vapor Feedback and Global Warming
The question of the relative importance of different regions for water vapor feedback is a source of some confusion in the literature. This question has been reexamined as part of a comprehensive review of the current state of science and of the controversies surrounding water vapor feedback (mh). Fig. 2.2 shows an estimate of the individual contributions by water vapor (top) and temperature (bottom) to changes in the outgoing longwave radiation resulting from a spatially-uniform 1 K perturbation in temperature under the assumption of constant relative humidity. In performing these calculations, the atmosphere was divided into 10 vertical layers of equal mass, using temperature and humidity data from the European Centre for Medium Range Weather Forecasting, and cloud data from the International Satellite Cloud Climatology Project. The results presented here are zonal averages for July only.
The sensitivity of outgoing longwave radiation to temperature perturbations (Fig. 2.2, bottom) is strongly affected by the cloud distribution. Where upper level clouds are prevalent (e.g., 0°-10°N), the outgoing infrared radiation is most sensitive to temperatures at the level of these emitting surfaces, and is relatively insensitive to temperatures deeper in the atmosphere. Where skies are clearer (e.g., 10°S-20°S), lower tropospheric temperatures control the outgoing flux.
Figure 2.2 (top) shows that the subtropical dry zones are somewhat more important than the moister zone in the deep tropics for the strength of the fixed relative humidity water vapor feedback. This feature is a consequence of the presence of clouds. If clear skies are assumed to exist everywhere, the maximum in this figure shifts to the moister regions in the tropics. The claim that the subtropical dry zones always dominate water vapor feedback is not supported by this analysis.
If temperature changes are uniform and relative humidities remain unchanged as the climate warms, these results show that the humidity response in the free troposphere above 800 mb is responsible for almost all of the infrared water vapor feedback, with only 10% contributed by the boundary layer. Roughly 55% of the total is due to the tropical free troposphere (30°N-30°S), and 35% to the extratropics. Of this tropical contribution, about two-thirds, 35% of the total, is due to the upper half of the troposphere, from 100-500 mb.
2.2.3 Sea Level Rise and Global Warming
For the period 1910 to 1990, the time averaged rate of global mean sea level rise due to thermal expansion simulated by the different models varied between 0.3 and 0.8 mm/yr (Fig. 2.3). The observed rate for the same period is estimated to be somewhat higher (1 to 2 mm/yr), but includes contributions from melting land ice, land storage changes, etc., in additional to the ocean thermal expansion term. From 1990 to 2100, the average rate of projected global mean sea level rise due to thermal expansion accelerates, varying from 2.0-3.7 mm/yr among the different AOGCMs. Thus, the nine models simulate that the warming of ocean waters will be responsible for a 0.2 m to 0.4 m rise in global mean sea level over the period 1900 to 2100. The projected acceleration in the rate of sea level rise is consistent with increases in warming rates seen in the models' global mean surface air temperature responses. The nine models' projected 21st century warming rates vary from 2.6°C to 3.9°C.
Due to the slow penetration of heat into the oceans interior, the response time scale for sea level is very long (order of 1000 years). Therefore, sea level can be expected to continue to rise long after atmospheric GHG levels are stabilized.
2.2.4 Response of Southern Hemisphere Winds to Global Warming
The response bears some resemblance to observed trends of the last 30 years in the Southern Annular Mode. This suggests that these trends are part of the atmosphere's response to greenhouse-gas loading. However, the observed trends are much stronger than the model response for the corresponding time period in the simulation. The reasons for this discrepancy are poorly understood. Possible causes are low-frequency variability in the atmospheric record, relatively coarse resolution in the stratosphere, and an absence of radiative-forcing effects associated with ozone depletion in the greenhouse warming integrations.
2.2.5 Discriminants of 20th Century Changes in Surface Temperature
The first discriminants indicate a relatively uniform and steady warming of the ocean surface, with a weak cooling over the North Atlantic being the most prominent exception. Over the continents, these patterns show regions of cooling, especially in July and typically of 1,000-2,000 km size. Most of these are in industrial regions where sulfate aerosol loadings are high. The cooling in Central Africa may be related to biomass burning. The cooling in July over North America appears to be too widespread as compared to the expected aerosol distribution centered on the Eastern states, and may be related to land surface changes. Most of the cooling centers are on scales smaller than the 5,000 km and larger scales considered in many detection studies. New detection/attribution techniques need to be devised to distinguish climate change on these smaller spatial scales from natural variability, possibly based on the comparison of observational discriminants with ensembles of discriminants from simulations.
2.2.6 Response of Climate to Natural and Anthropogenic Forcings
2.2.7 Climate Scenarios for IPCC Third Assessment Report
The latest integrations using the A2 and B2 scenarios show that, during the next 100 years the global mean surface air temperature (SAT) increase relative to today's conditions is about 3.5°C for the A2 scenario and 2.4°C for the B2 scenario (Fig. 2.4). Despite very different
emissions for the GHG and aerosols, the increase in the global mean SAT is quite similar over the first 50 years of the integrations. The model response to the A2 scenario is also similar to the previous IPCC scenario (Is92A) over the entire 100 year simulation. These similarities in the model response are due to compensation between lower GHG concentrations and reduced aerosol loading in the newer A2 scenario, as compared with the older (Is92A) scenario. The model output from these integrations is available from the IPCC Data Distribution Center and is being used in various studies of the societal impact of climate change by the impacts community.
2.3 CLIMATE VARIABILITY AND DYNAMICS
2.3.1 Response of the North Atlantic Climate System to the Arctic/North Atlantic Oscillation
The Arctic Oscillation denotes a tendency for out of phase sea level pressure variations between the Arctic and midlatitudes, with associated changes in the midlatitude westerly winds. An index of the Arctic Oscillation is used which is related to the difference in sea level pressure between the Arctic and midlatitude North Atlantic. This index is quite similar to the North Atlantic Oscillation (NAO) index, and hence we view the AO and NAO as effectively identical for this work.
The trend in the AO/NAO is introduced in the model by adding anomalous surface fluxes of heat, water, and momentum to the ocean component of the coupled model. These fluxes have the spatial signature of the AO/NAO, as deduced from a long control integration of the model. The amplitude of the anomalous flux pattern increases linearly over a 30 year period, corresponding to a 9 mb increase in the AO index (similar to the observed increase in the AO over the last 30 years). After that point, the amplitude of the forcing is held fixed at the elevated level.
Our primary focus is on the response of the North Atlantic thermohaline circulation (THC) to the AO/NAO trend. It is found that the stronger winds over the North Atlantic associated with the positive AO/NAO extract more heat from the ocean, thereby cooling and increasing the density of the upper ocean in the subpolar North Atlantic and thus strengthening the THC. For this model, changes in heat flux dominate over water and momentum fluxes.
Most studies of GHG-induced climate change suggest a weakening of the THC in the North Atlantic in response to increased freshening and warming in the subpolar region (1634). As described above, a sustained upward trend in the AO/NAO tends to oppose this THC weakening. The competition between these tendencies has been explored using additional ensembles of numerical experiments (Fig. 2.5). One ensemble is forced by increasing GHGs and sulfate aerosols; the second ensemble adds to this forcing the effects of the positive trend
in the AO/NAO. Differences between these two ensembles identify the influence of the AO/NAO trend. It is found that a sustained upward trend in the AO/NAO could effectively delay the GHG-induced weakening of the THC by several decades. This result is of particular importance if the positive trend in the AO/NAO is a response to increasing greenhouse gases, as has been recently suggested.
2.3.2 Observed and Simulated Multidecadal Variability in the North Atlantic
In the first study (jn), a direct comparison was made between multidecadal variability simulated by the R15 coupled model and observed variability as inferred from analyses of both instrumental and proxy records. The model variability is characterized by a large-scale pattern of sea surface temperature (SST) anomalies in the North Atlantic (Fig. 2.6), with a time scale (distinct from red noise) of 40-80 years. The instrumental record shows a spatial pattern of SST anomalies quite similar to that in the model, and with an indication of a multidecadal time scale. The shortness of the instrumental record, however, makes a determination of the timescale of variability difficult. Therefore, the simulated variability was compared to multicentury proxy records of climate variations, as deduced from tree rings, ice cores, and other indicators. Analyses of the proxy records reveal a pattern of variability in the North Atlantic which is quite similar to the simulated variability. This observed pattern has a distinct time scale (approximately 70 years) which is also similar to the model results.
In the second study (1695), experiments were conducted to investigate possible mechanisms for the simulated multidecadal variability. Specifically, the question is whether the variability involves strong, two-way coupling between the atmosphere and ocean (as is the case for ENSO), or an oceanic response to atmospheric forcing. A suite of experiments was conducted in which the ocean component of the R15 coupled model was forced with time series of surface fluxes from both the fully coupled model and from an atmosphere-only model using a prescribed seasonal cycle of SSTs. These experiments reveal that the simulated multidecadal variability is best viewed as an oceanic response to low frequency (multidecadal) atmospheric surface flux forcing. In particular, the spatial pattern of fluxes associated with the North Atlantic Oscillation is crucial in generating the oceanic variability. Interactions between the atmosphere and ocean do play a key role, however, in determining the amplitude of the simulated variability by modifying the magnitude of the air-sea heat flux variations.
dynamics of the variability will be studied in a series of idealized runs in which the model adjusts to a sudden imposition of NAO-like forcing, which appears to play a key role in the multidecadal variability.
2.3.3 Partitioning of Poleward Heat Transport Between Ocean and Atmosphere in the Tropics
2.4.1 Coupled Climate Model Development
The current-generation coupled model incorporates the set of physical parameterizations that have been the basis for numerous climate variability and climate change studies over the past two decades. The development of this model within FMS will insure continuity with previous work using the existing GFDL coupled climate model, facilitate systematic testing of new model physics, and provide a very efficient model code for extended paleoclimate experiments and other applications.
The next-generation coupled model will incorporate a number of important changes in physical parameterizations, including a diurnal cycle, improved boundary layer, improved land surface and sea ice models, a more comprehensive treatment of radiative climate forcings, and enhanced ocean model resolution and physics. This model will be critical to GFDL's future climate change research program, including the laboratory's contributions to future climate change assessments.
During FY00 the development and testing of key components of these models has been a primary focus of the climate group. This effort has extended to include numerous collaborators from several groups within the laboratory. Efforts have continued in the integration and testing of model components of the coupled model framework. Through FMS, the groundwork has been laid for porting the new climate models to massively parallel computing systems. This critical task has been undertaken in preparation for the arrival of the next-generation large-scale computing system at the laboratory.
Related information on various coupled model components, including the FMS framework itself, the radiation code, the land model, the sea ice model, and the ocean model are contained elsewhere in this report.
2.4.2 Development and Testing of the Land Dynamics Model
The Land Dynamics (LaD) model is an extension of an earlier scheme with a record of successful application in climate modeling. The most significant changes from the original model include: 1) introduction of non-water-stressed stomatal control of transpiration, in order to correct a tendency toward excessive evaporation; 2) conversion from globally constant land parameters (with the exception of vegetation-dependent snow-free surface albedo) to more complete vegetation- and soil-dependence of all surface parameters, in order to provide a more realistic representation of geographic variations in water and energy balances and to enable model-based investigations of land-cover change; 3) introduction of soil sensible heat storage and transport, in order to move toward realistic diurnal-cycle modeling; 4) a groundwater (saturated-zone) storage reservoir, in order to provide more realistic temporal variability of runoff; and 5) a rudimentary runoff-routing scheme for delivery of runoff to the ocean, in order to provide realistic fresh-water forcing of the ocean component of a coupled global climate model.
The performance of the new land dynamics model was evaluated in stand-alone mode, using the International Satellite Land Surface Climatology Project Initiative I global dataset and a recently produced observation-based water-balance dataset for major river basins of the world. The model performance was evaluated by comparing computed and observed runoff ratios from many major river basins of the world. Special attention was given to distinguishing between two components of the apparent runoff-ratio error: the part due to intrinsic model error and the part due to errors in the assumed precipitation forcing. The pattern of discrepancies between modeled and observed runoff ratios is consistent with the precipitation-error estimates that were produced as part of a companion study. The new model is tuned by adjustment of a globally constant scale factor for non-water-stressed stomatal resistance (Fig. 2.7). After tuning, significant overestimation of runoff is found in
environments where an overall arid climate includes a brief but intense wet season, and this error may be explained by the neglect of upward soil-water diffusion from below the root zone during the dry season. With the exception of such basins, and in the absence of precipitation errors, the model predicts annual runoff ratios with a root-mean-square deviation from the observations of about 0.05. The new model matches observations better than its predecessor, which has a negative runoff bias and greater scatter.
2.5.1 Effects of Changes in Earth's Orbit on Climate
The climate of
the orbital forcing experiment responds in a manner that is qualitatively
consistent with these paleodata. A period of diminished sea ice cover at
10 ka is followed by increasing sea ice toward the present, with the most
rapid increase occurring from 8 to 4 ka and a more gradual increase thereafter
(Fig. 2.8). The periodic changes in obliquity (i.e., tilt of Earth's
axis) and the precession of the equinoxes combine to yield this trend,
as evident from a statistical decomposition of the model response into
components associated with obliquity and precession. Obliquity peaked most
recently at ~10 ka and has decreased thereafter. Because Arctic temperature
is high when obliquity is high, decreases in obliquity cool the Arctic
from 10 ka through the present. Because of sea ice-albedo-temperature feedbacks,
Arctic temperatures are highest and sea ice least extensive when the smallest
Earth-Sun distance (i.e., perihelion) occurs in late spring, as
it did at ~14 ka. Thus, precession also favors increasing Arctic ice from
14 ka through ~3 ka, as perihelion moved from late spring through summer
into late autumn. Thus, the rapid increase in sea ice concentration from
8 to 4 ka results from the superposition of cooling effects from both obliquity
and precession, and the subsequent slowing of the growth in sea ice extent
occurs as the effects of precession begin to oppose the obliquity-driven
cooling.
2.5.2 Simulation of Last Glacial Maximum with HadCM3 Coupled Climate Model
To initialize the model, a strategy was devised that employed output from a LGM simulation in which a slab ocean was substituted for the oceanic component of HadCM3. Beginning from modern observations of ocean temperature and salinity, the SSTs in HadCM3 were relaxed toward the SSTs simulated by the slab ocean model for a period of 70 years. This initial period was intended to facilitate a more rapid cooling toward a glacial state. At the conclusion of the first 70 years, the relaxation is discontinued, leaving the model free to determine a climatic state that is in equilibrium with the glacial boundary conditions. As of this writing, the model has run more than 600 additional years, during which slow changes in its dynamic and thermodynamic state have occurred.
Although HadCM3
has yet to fully equilibrate with the LGM boundary conditions, there is
very clear evidence of important interactions between the atmosphere and
ocean. Surface temperatures in the coupled run are substantially different
from those simulated by a slab ocean coupled to the same atmosphere, particularly
in the North Atlantic, as evident in Fig.2.9, where a "tripole" anomaly
pattern is found. Shifts in the position of the Gulf Stream and the regions
of deep water formation are primarily responsible for this altered anomaly
pattern. In the eastern tropical Pacific, enhanced equatorial cooling (relative
to the slab model) has emerged during the course of the integration.
2.5.3 Effects of Fresh Water Discharge on Climate
In both experiments, the external fresh water flux makes the surface waters in the northern North Atlantic more buoyant, inhibiting oceanic convection. This weakening of the convection isolates the surface waters from the deeper waters leading to further freshening. The weakening of the convection also leads to a cooling of the ocean surface, since convection typically brings warmer subsurface water to the ocean's surface. The surface temperature anomalies are locally up to 12°C and the cooling covers most of the northern North Atlantic. In addition, as the surface waters become more buoyant the thermohaline circulation (THC) weakens (Fig. 2.10).
In comparing the THC response for the two models, one notes that the initial weakening is very similar in the two experiments both in terms of rate and magnitude of the weakening. However, the THC recovery in the R30 experiment is much faster than in the earlier
R15 experiment. The reasons for this difference are unclear. The rapid recovery in the R30 experiment applies not only to the THC, but also to sea surface temperature.
Figure 2.11 shows the CO2-induced response of annual mean runoff for the 184 separate basins. According to this figure, runoff increases for approximately 70 percent of the river basins in response to the quadrupling of CO2. From an analysis of the geographical distribution of CO2-induced change of annual mean runoff, almost half of the negative values are produced from river basins located in various places in the southern half of the U.S. where the model indicates a general region of drying. Other river basins exhibiting negative runoff change include those located in equatorial regions of South America and Africa, India, southern Indochina and southern Europe. In general, river basins located in middle to higher latitudes in the Northern Hemisphere show an excess of river basin runoff, a result which is consistent with previous greenhouse warming studies of CO2-induced change of land-surface hydrology.
2.6.2 Land-Process Influences on Monthly River Discharge Variability
conversion of soil-water excess to river discharge. The first filter causes a snowmelt-related amplification of high-frequency variability in those basins receiving significant snowfall. The second filter causes a relatively constant reduction in variability across all frequencies and can be predicted well using a semi-empirical water-balance relation. The third filter, associated with groundwater and surface-water storage in the river basin, causes a strong reduction in high-frequency variability of many basins. The strength of this reduction can be quantified by an average residence time of water in storage, which is typically on the order of 20-50 days. However, the residence time is demonstrably influenced by freezing conditions in the basin, fractional cover of the basin by lakes, and runoff ratio (ratio of mean runoff to mean precipitation). Large lake areas enhance storage and can greatly increase total residence times (100 to several hundred days). Freezing conditions appear to cause bypassing of subsurface storage, leading to smaller residence times (0 to 30 days). Small runoff ratios tend to be associated with regions where most of the runoff is produced by processes that bypass the (deep) saturated zone, leading to relatively small residence times for such basins (0 to 40 days, Fig. 2.12).
2.6.3 Modeling Land
Influences on Variability of Macro-Scale Water
and Energy Fluxes
It has been found that leaving out information on global variations in these parameters leads to a significant degradation of the ability of the model to predict the annual runoff ratio. An additional set of optimization experiments in which the parameters are examined individually reveals that spatial variations in stomatal resistance add the most predictive power to the model in stand-alone mode. Further single-parameter experiments with surface roughness length, available water capacity, thermal conductivity, and thermal diffusivity show very little sensitivity to global variations in these parameters. Finally, it is found that the model performance exceeds that of the Budyko and generalized Turc-Pike water-balance equations, implying that the model benefits not only from information on global variations in land characteristics, but also from information on the temporal structure of the forcing.
2.6.4 A Minimalist Probabilistic Description of Root-Zone Soil Water
The main problem in defining Jupiter's meteorology, particularly the character of the jets and vortices, comes from the fact that the nature and extent of the motions are not generally known for the region below the clouds. To develop a theory for the circulation, 3-D primitive equation models are used to investigate the formation and coexistence of the various turbulent and coherent phenomena for hypothetical vertical structures. The main hypotheses involved - that the atmospheric circulation occurs within a relatively thin upper layer and is driven by horizontal temperature gradients - have been examined with a wide range of models. Present studies are concerned with the role of the vertical structure on the formation of vortices and jets.
In particular, studies of the dynamical response of thin atmospheric layers have been extended to examine the genesis and equilibration of multiple anticyclonic vortex sets in a Jovian context (ni). Current modeling efforts focus on the three main groups seen on Jupiter, namely, the Great Red Spot, the four Large Ovals, and the dozen or so Small Ovals that occur at latitudes of -21', -33', -41', respectively. The generation and equilibration of the vortices associated with long solitary baroclinic Rossby waves in a stratified fluid are examined numerically using a primitive equation model with Jovian parameters subject to simple heating functions. Following earlier findings (1400,1454), the motions are confined to thin upper layers by exponential vertical structures that favor absolute vortex stability. The selective character of the Rossby vortices also provides a theoretical probe of the vertical structure of the atmosphere over a wide range of latitudes.
From a wide range of calculations, it has been found that vortex sets resembling the three main Jovian groups in scale, form, and number can be simultaneously generated and maintained in a steady configuration provided that the heating components that drive the alternating jets are carefully chosen so as to make the easterly jets marginally unstable. Otherwise, more complex arrangements evolve. The vortices decrease rapidly in size and become more numerous with latitude due to reductions both in the widths of the containing jets and in the propagation speeds. The amplitudes of the vortices depend on the strength of the baroclinic instability of the easterly jets that generate them and on the weak heating that maintains them. Vortices generally merge in the same way as those in the unheated system (1400) to explain the singularity of the Great Red Spot, while the Ovals remain multiple because of their smaller scale and mutual similarity.
Evolution of features in the model, however, are more complex in the heated system because the generation of new storms offsets the tendency to merge into fewer vortices. Regenesis can occur continuously if the jet instabilities are restored too quickly by the heating. Otherwise, weak regenesis produces weak eddies that are absorbed by existing vortices. A steady configuration requires a balance between a weak heating and a weak dissipation. Intrazonal interactions between vortices from different sets also occur and can in some situations lead to the mutual destruction of anticyclones which model the Great Red Spot and a Large Oval. They can also produce a novel form of coherence in which a large warm anticyclone occurs within a cool cyclonic zone whose flow it blocks; such storms are energized by a continuous inflow of eddy energy from the anticyclonic zone and act as a fluid dynamical "black hole".
Numerical constraints on diffusion and resolution suggest that the small eddies also contribute to the dynamical balance of the vortices and should be represented explicitly. Furthermore, the heating forms required to produce the easterly jets that generate the vortices imply that the existing heat imbalance may be more local than global and, as such, controlled by the albedo. This, in turn, implies that dynamics alone cannot explain all aspects of Jupiter's circulation, that physics must also be involved.
*Portions of this document contain material that has not yet been formally published and may not be quoted or referenced without explicit permission of the author(s).