After four months, the question of whether an atomic explosion took place the night of Sept. 22 in the skies near South Africa is the subject of hot debate inside the U.S. government.

The technical experts at places like the Department of Energy and the Los Alamos Scientific Laboratory in New Mexico need no convincing. They fully believe that somebody triggered a small atomic explosion about five miles up in the atmosphere over the most remote regions of the southern Indian Ocean east of South Africa at 3 o'clock in the morning on Sept. 22. i

Officials at the White House, the State Department and the Central Intelligence Agency aren't so sure.Later this week, the White House Office of Science Policy will release a report under the name of science adviser Frank Press that says just that and no more, namely that maybe there was a nuclear explosion and maybe there wasn't.

The "Press Report," as it's already being called, will conclude that what a Vela reconnaissance satellite may have seen the night of Sept. 22 was a small meteorite or mirrored satellite reflecting sunlight back into the photosensors of the Vela and confusing it into thinking it had seen an atomic explosion. This is the latest in a series of explanations that started with a "superbolt" of lightning or a freak combination of a lightning strike and entry of a meteor into the atmosphere at precisely the same time.

"We have no reason to disagree with the Press Report," a CIA spokesman said yesterday. "We have no evidence that there was an atomic test."

Scientists at the Los Alamos Scientific Laboratory, in the Department of Energy and even a few technical people at the State Department question why there should be such equivocating. They point out that every sign they've seen identifies the event of Sept. 22 as a nuclear explosion, from the original sighting of the fireball by the Vela satellite to the faint but unmistakable echoes of the explosion's sonic boom on early warning radar.

"Look, the Vela satellite picked up a signature like this 41 times before," a State Department official said. "in every one of those 41 instances, there was never any question about the fact that a nuclear test had taken place. Each of those 41 was undeniably a nuclear explosion. This one was, too."

The Press Report will question why nobody has seen radioactive fallout that would be conclusive evidence that an explosion took place. Scientists counter that there was a hint of fallout in the rainwater in New Zealand and that an explosion as small as the one detected by the Vela satellite might not leave much fallout behind anyway.

"In a high-yield test, the rising fireball and hot air mass would get into predictable high altitude winds and carry the fallout all over the Earth," one scientist at Los Alamos said. "A low altitude and low-yield explosion like this one seemed to be doesn't have enough energy to carry it up to high altitudes. The fallout would have come down with local winds and rains and disappeared into the ocean in 24 hours."

Los Alamos scientists, who designed and built the Vela satellite so it would never mistake an atomic fireball for one of nature's events, scoff at the suggestion that what it saw was a lightning strike or any other such phenomenon.

They say that the fireball "signature" of an atomic explosion is unmistakably its own. No matter what its size or whether it uses uranium or plutonium, a nuclear fireball generates a pulse of light that briefly disappears when the shock wave blots it out and then reappears as a second pulse of light that is exactly 99 times more intense than the first pulse.

"This is what the Vela saw the night of Sept. 22," one Los Alamos scientist said. "If it was an accidental reflection of sunlight off some satellite mirrors, one of the mirrors would have had to be exactly 99 times bigger than the other mirror. Not much chance of that."

This lightning superbolt explanation makes even less sense to Los Alamos scientists. First, as superbolt produces only a single flash. Second it generates a flash almost 400 times les intense and 100 times shorter in duration than even a one-kiloton explosion. The explosion of Sept. 22 was believed to be four kilotons.

The White House, State Department and CIA are reluctant to call the event of Sept. 22 a nuclear explosion in part because they can't point a finger at which country may have pulled the trigger. Countries that have been discussed include South Africa, Israel, India and the Soviet Union.

State Department officials tend to dismiss India, whose government at the time of the test disavowed nuclear weapons. The Soviets were dismissed too, in part because there was no reason for them to test in that part of the world and no reason to believe they had any aircraft, missiles or submarines in that part of the world at that time.

"It basically comes down to a choice between South Africa and Israel or both," a State Department source said.

"It's conceivable that a defector or somebody who was involved will show up someday," another official said. "That's the only kind of information that might reasonably change the picture."

There's another reason for shying away from calling the Sept. 22 event an atomic explosion. It would point up the fact that the United States was lucky to spot the explosion with a single Vela satellite and is poorly equipped to police a comprehensive ban against testing of nuclear weapons anywhere in the atmosphere on Earth.