"You could be an astronaut floating around and not realize there's a 500-mile-an-hour wind blowing outside your capsule," says Ray Roble.
Now Ray and colleagues have forged the first direct link between upper- and lower-atmosphere GCMs. Late in 1995, they coupled Ray's thermosphere-ionosophere-mesosphere-electrodynamics general circulation model (TIME-GCM) with the NCAR community climate model, version 2 (CCM2). The yet-to-be-published results, says Ray, are "quite spectacular."
Ray's office in the Foothills Lab is lined with auroral pictures of many sizes, shapes, and hues. "It all started in grad school [in aeronomy at the University of Michigan] when I looked at how the atmosphere responded to the aurora. At that time there weren't any GCMs for the upper atmosphere. I came to NCAR and tried to adapt meteorological tools to examine the aurora's influence on the atmosphere.
"When the aurora turns on, how does the atmosphere react to it, chemically and dynamically? How deep into the atmosphere can you trace its effects? Walt Roberts thought maybe it was all the way down to the surface. He did a lot of correlations between solar activity and weather on earth, but he was never able to find physical mechanisms behind the correlations. That's my goal."
Ray joined NCAR in 1970 through the Advanced Study Program. Modeler Bob Dickinson (now at the University of Arizona) was his first mentor. Over the succeeding years, as Ray explored different aspects of the upper-atmosphere modeling problem, he migrated from ASP to three other divisions, the last being HAO. ("I probably hold the record among NCAR scientists for number of divisions," he says.)
Keeping an eclectic approach to his work has helped Ray to better understand the upper atmosphere, because so many different forces compete to affect it. Gravity waves and other perturbations propagate upward from the denser layers below. Infrared radiation flows upward from the earth and lower atmosphere, while ultraviolet (UV) radiation flowing from the sun is intercepted. High-altitude winds blow ions (charged atoms) across the earth's magnetic field, creating a dynamo that moves ions and electrons. Collisions between neutral and charged particles exchange charge and transfer energy and momentum. For a crowning touch, there are the solar storms whose bursts of particles and radiation can almost completely rearrange global thermospheric circulation in a matter of hours.
Each of these factors can vary far more than is typical at lower altitudes. For instance, the upper atmosphere filters out the shortest wavelengths of ultraviolet energy from the sun before it can reach the earth. When aeronomers call these wavelengths the extreme ultraviolet (EUV) range, the word 'extreme' could apply to the variations in UV as well as it does to the wavelengths.
"The total solar radiative output can vary by about 0.1% over an 11-year solar cycle," says Ray. "The ultraviolet varies by a few percent. The extreme ultraviolet and X-rays can vary by factors of 10 to 100, and it can be on a much shorter time scale than the solar cycle."
Recent data on EUV variations has come from satellites and rockets, many with instrumentation developed by HAO scientists and collaborators. One of these is the Upper Atmosphere Research Satellite (UARS), deployed in the early 1990s with contributions by Gary Rottman and many others in HAO. UARS focuses on the total UV range, while rockets deployed by HAO's Tom Woods and colleagues have focused on EUV.
"We've known about atmospheric tides for a long time," says Maura. One of the oldest clues is a long-documented trend in surface air pressures near the equator that shows a small but distinct rise and fall every 12 hours. But the ignorosphere's lack of data prevented tracing details of the tides' influence there--until recently.
To take a closer look at tides, Maura collaborated on an adaptation of a steady-state, two-dimensional model--essentially a vertical, pole-to-pole slice through the atmosphere. Because atmospheric tides move so regularly from east to west and usually persist for days or weeks, the 2-D model allowed Maura to create a climatology of the tides.
"To study the tides with the model, I specify the tidal sources inthe linearized fluid equations and calculate the wind and temperature responses," says Maura. "I think out most significant achievement has been our predictions of the seasonal variability of the diurnal tide in the MLT.
"More and more people are including tides in their models, but among linear models, ours is unique in being able to model the diurnal tide and to include the troposphere, which is the most important source region. These kinds of models complement the first-principles modeling that Ray does." For example, Ray has used the TIME-GCM to study the nonlinear effects of tides on atmospheric chemistry and dynamics.
The TIME-GCM's lower boundary is flux-coupled to the CCM2's upper boundary at about 40 km. Ray typically runs the TIME-GCM on a much shorter time scale than the CCM2 in order to capture the rapid changes induced by solar storms. "An auroral storm might last three to six hours, so you can do a relatively short run. As a result, I can put in a lot of physics and chemistry, the neutral upper atmosphere, the plasma, all the electric currents." HAO's Ben Foster and Cicely Ridley worked with Ray on the model coupling.
What did Ray find so spectacular in the results? "We've found a lot of weather-induced variability that propagates through the model interface high into the thermosphere." This in turn causes considerable variability in the winds, temperature, and composition of the ionosphere. Ray thinks this weather-induced variability interacts with solar and auroral-induced variability, making it an important key to understanding solar-terrestrial coupling. "I'm getting more confident now, especially since some of the structures that appear in the coupled model are being observed by instruments aboard the UARS satellite."
A paper on the results should appear at the spring meeting of the American Geophysical Union. In the meantime, says Ray, "There's a lot of work to do to make the models blend together." Ray also is curious about how the coupled model will handle a doubled-carbon dioxide scenario. Last year he published a paper, "Major greenhouse cooling (yes, cooling): The upper atmosphere response to increased CO2" in the Reviews of Geophysics. It outlines a Venus-like response in which carbon-dioxide increases cause greater radiation to space from the upper atmosphere and a resultant cooling of as much as 50íC at heights of 400 km.
Like any good modeler, though, Ray keeps an eye on the real world--and on his first aeronomic love. He eagerly anticipates the launch of a new satellite by NASA, tentatively scheduled for 2 February, as part of its International Solar-Terrestrial Program. It will gather high-resolution photos of auroras in the visible and ultraviolet spectra. "It'll provide us a great deal of input. Together with data of atmospheric response measured by the UARS satellite, it will be a wonderful opportunity to test our model's performance in predicting the atmospheric response to solar and auroral variability.
"Also, we'll be seeing a lot of beautiful auroral images." --BH