hirty years of exploration of the inner and outer planets of the solar system has taught us that no planet is like any other. The known geophysics of Earth barely begins to help us understand the formation, composition, and conditions on any other planet. Planetary missions again and again have underscored the truth that simple sets of physical laws can interact to produce an infinite variety of material incarnations.
Some planets have little or no atmosphere while others have substantial gaseous envelopes. Venus and Jupiter are two examples of planets with substantial atmospheres. They're not exactly vacation spots, and chances are that the human race will not visit either planet for a long, long time--certainly not without a great deal of high-tech protective gear. But Venus, like the Earth, has an atmosphere, an envelope of gases surrounding the rocky planetary body. And while Jupiter's "surface" may have to be defined at some arbitrary interior level, it, too, has a gaseous envelope.
That's where resemblances end and mind-bending puzzles begin. Gerald Schubert and David Baker, computational geophysicists, have been modeling the atmospheres of Venus and Jupiter, using data from the various space missions that have investigated the "air" on both planets. Their work is helping to explain mysterious data from the missions--21 to Venus, six to Jupiter--already undertaken, and it will aid in the design of future Venusian and Jovian robotic explorers.
Schubert is a distinguished professor of geophysics and planetary physics in the department of Earth and Space Science at UCLA's Institute of Geophysics and Planetary Physics. He has been a major investigator for a number of Venusian probes and for the recent Galileo mission to Jupiter. Baker obtained his Ph.D. under Schubert's direction in October 1997 and is now working for the Universities Space Research Association in the Mesoscale Atmospheric Processes branch at the NASA Goddard Space Flight Center in Maryland.
Figure 1: The Venusian Atmosphere
Results from a Venus atmospheric convection simulation at the subsolar point from 12 to 60 kilometers altitude. Perturbation potential temperature is shown with cooler colors indicating relatively cold potential temperatures and warmer colors indicating relatively warm potential temperatures. The lower convection layer spans from 18 to 30 kilometers altitude and the cloud-level convection layer extends from 47 to 56 kilometers altitude. Cloud-level convection--characterized by cold, narrow downdrafts--penetrates the underlying stable layer near 45 kilometers altitude and generates gravity waves at that altitude. Convectively generated gravity waves freely propagate in the stable layer from 30 to 47 kilometers altitude.
Work on the Venusian atmosphere was the basis of Baker's dissertation. Results were presented at the Fall American Geophysical Union (AGU) meeting in San Francisco last December, and Baker, Schubert, and a colleague, Philip W. Jones of the Los Alamos National Laboratory, are preparing a paper for submission to the Journal of Atmospheric Sciences.
"High-speed winds observed near and within the clouds are perhaps the most puzzling characteristic of Venus' atmosphere," Baker said. "We've used a two-dimensional mesoscale model of the Venusian atmosphere to learn whether waves generated by convection may speed up or slow down the high winds."
The Venusian year (orbital period) is 225 Earth days long, while the Venusian "day" (rotation period; the planet rotates from east to west) is 243 Earth days--longer than the planet's year. But while the planet moves slowly, the Venusian atmosphere is rotating around it from east to west every four (Earth) days, driven by winds as high as 100 meters per second (220 mph). "Planetary scientists have long been seeking the causes of this atmospheric superrotation," Baker said.
The "air" on Venus is 96 percent carbon dioxide. The planet's surface temperature is about 740 K, lapsing smoothly to 200 K above the upper cloud layer at a height of 65 km. The clouds, comprised of fine sulfuric acid droplets, extend from roughly 45 to 65 kilometers above the surface. Embedded within the cloud layer is a highly turbulent convection layer. "The clouds absorb incoming solar energy which drives convective motions," Baker said. In addition, convection likely occurs in two layers closer to the surface, from 0 to 5 kilometers altitude and from 18 to 30 kilometers altitude. Convection from these layers may penetrate into adjacent regions and generate gravity waves. These waves, in turn, may influence the atmospheric superrotation.
To model convection in the Venusian atmosphere, Schubert and Baker used a 2-D representation of the atmosphere as a perfect compressible gas. The model represents a vertical slice through the atmosphere at the equator, centered on the subsolar point and extending 60 kilometers to either side (Figure 1). "It's an extremely high-resolution mesoscale model," Baker said, "with about 850 grid points horizontally, for a resolution of 140 meters, and more than 400 gridpoints vertically, resolving 120 meters." The model has been run on the CRAY T3E supercomputer at SDSC.
"We were looking for convectively generated gravity waves in the lower atmosphere, and we found them," Baker said. "We thought that these waves might be absorbed in the superrotating zonal winds, thus speeding up the winds by dumping energy into them. But, to our surprise, gravity waves below 48 kilometers altitude exerted a dramatic slowing effect on the zonal winds, enough to slow the winds to a halt after about a week if the superrotation were not being driven by some much stronger process."
Along the way, the model provided much information about convective penetration into adjacent atmospheric layers, "but we still have a mystery here," Schubert said. "Our work suggests that it's a bigger mystery than we originally thought."
A simulated convective downdraft in Jupiter's atmosphere. Perturbation density is shown here with cooler colors indicating more dense air and warmer colors indicating less dense air. Arrows indicate the direction and magnitude of the winds. Less dense air is forced into the deep atmosphere by the strong (more dense) convective downdraft.
"When the Galileo probe entered the Jovian atmosphere in December 1995, it dropped into a 'hot spot,' a volume containing almost no water vapor," Baker said. "The first question was, how could such a thing be formed or sustained in the atmosphere of Jupiter?"
On Jupiter, the "air" is mainly hydrogen gas. On Earth, a parcel of moist air is lighter than the same volume of dry air, but moist "air" (H2 + H2O) on Jupiter would sink--so how was the less dense, dry air found to go so deeply into the Jovian atmosphere? And what prevented the "hole" from closing up and sending what should have been a "bubble" of H2 to the top of the atmosphere?
Schubert and Baker used a high-resolution mesoscale model of the Jovian atmosphere very similar to the Venusian model--except for the boundary conditions. The model box encompassed about 180 kilometers vertically over a transect of 520 kilometers at the latitude of the dry hole, with a horizontal resolution of about 600 meters and vertical resolution of 500 meters. The vertical pressure ran from about 0.1 bars at the top to 22 bars at the bottom (Figure 2). "Our model indicates that cool downwellings in the Jovian atmosphere could create deep, dry holes of the sort the probe entered," Baker said.
Schubert and Baker caution that their models, which are computationally intensive, are still 2-D and do not necessarily represent the physics of each situation as accurately as may be necessary. "We started the Jupiter computations in May and finished them in January," Baker said, "while the Venusian work extended over a couple of years. Faster computers and networks would be required for us to do equally well in three dimensions." --MM