US high-tech challenge - knocking down the nuclear ICBM threat

There was a historic encounter over the Pacific Ocean this summer. A rocket fired from Meck Island in the Kwajalein Atoll propelled a homing device more than 100 miles upward toward a dummy ballistic-missile warhead streaking above the atmosphere. The warhead had been fired aloft by a Minuteman missile from Vandenberg Air Force Base 4,200 miles away in California about 20 minutes earlier.

Using an infrared sensor, the ''homing and kill'' device zeroed in on the mock attacker, spread a metal umbrella 15 feet in diameter, and - at a speed of more than 14,000 miles an hour - shattered the warhead (and itself).

It was, as military officers later said enthusiastically, ''like a bullet hitting a bullet.''

Proponents of President Reagan's strategic defense initiative say such tests only hint at ways the United States could protect itself from nuclear attack. They note that lasers carried aboard military test aircraft have burned holes in target drones and hit short-range missiles.

But the Soviet arsenal consists of far more than target drones and missiles or a single dummy warhead. The Soviet Union now has 1,398 land-based intercontinental ballistic missiles (ICBMs) with a total of 5,654 warheads, plus another 980 submarine-based missiles with 2,688 warheads. Under full nuclear attack, the US would face not only thousands of warheads, but tens of thousands of decoys, space chaff, and other ''penetration aids.'' Once the dust from the ballistic missiles had settled, there might be bombers and low-flying cruise missiles to attend to.

For this reason, US strategic nuclear doctrine for two decades has been based on mutual assured destruction (MAD) - the notion that there is nothing either side can do to prevent the other from retaliating with unacceptably devastating force - and a concentrated effort on preserving a sufficient retaliatory force to deter attack.

Still, many scientists and nuclear strategists say there is good reason to at least explore ways of defending against such attack, not only to make the Soviet Union think twice about launching its ICBMs but to save as many people as possible if the warheads start flying.

''The technical means for managing the strategic defensive battle has advanced very substantially over the past dozen years, in all areas,'' Lowell Wood of the Lawrence Livermore National Laboratory told a recent meeting of congressional Republicans. ''Advances in computer speed and compactness, in communications capabilities, and in sensor technologies have been striking in magnitude.''

Defending against intercontinental nuclear attack would be extremely difficult in any case. In five minutes or less, the rocket engines during the fiery, highly visible boost phase burn out. Shortly thereafter the missiles release their warheads and deceptive devices, which hurtle silently through the midcourse ballistic arc for about 20 minutes.

Then the warheads reenter the atmosphere (which strips away the harmless pretenders), for a terminal phase lasting only about one minute.

Nuclear missiles fired from submarines would reach their targets even sooner.

The best place to squelch a nuclear attack is in the boost phase, before the individual warheads pop off.

At this stage, the number of targets a defensive system would have to deal with is at its lowest. But defending at this stage is also the most difficult because of its short duration and because it takes place largely within Earth's atmosphere. Thus, any defense would have to be layered - dealing with each phase of a warhead's flight.

Strategic defense proponents note that if each phase of defense were 90 percent effective, total protection would be more than 99 percent. Even something less than this, they say, could enhance deterrence.

Earlier this year, a study team at the Pentagon, headed by former NASA administrator James C. Fletcher of the University of Pittsburgh, concluded that the technological challenges to missile defense ''are great but not insurmountable.'' This group included a ''red team'' that focused on ways an attacker might counter space-based defenses.

Dr. Fletcher's panel reported that directed-energy devices - lasers, particle beams, and kinetic-energy mechanisms - now offer ''multiple approaches'' to boost-phase intercept.

Such things as laser imaging and millimeter-wave radar now make it possible to discriminate between decoys and warheads in midcourse, the group reported, and advances in computers and signal processing offer the potential for ''enhanced battle management.''

Yet this group of experts cautioned that while near-term demonstrations of such potential capabilities could be made, a decision on whether to actually develop a system of strategic defense could not be made before the 1990s. And it said that ''several critical technological issues'' would require research programs of as much as 20 years in duration.

At the Pentagon and White House, officials are more optimistic about achieving short-term results.

''I, for one, believe that the option to change from today's massive nuclear standoff is not only possible, not only inevitable, but that it's already being offered to us,'' presidential science adviser George A. Keyworth III told a conference on nuclear war at the University of Maryland earlier this month.

Lt. Gen. James A. Abrahamson, head of the Strategic Defense Initiative Organization, says his goal is to demonstrate the feasibility of directed-energy generators by the mid-1980s and their improvement to performance levels necessary for space defense just a few years after that.

Even among proponents of missile defense, there is disagreement over how exotic system would have to be and whether it would in fact be ''star wars,'' that is, a network of armed satellites battling each other as well as nuclear warheads.

At a minimum, according to the Fletcher study group, strategic defense would require full-time global surveillance by satellite. Dr. Wood of the Lawrence Livermore lab says he believes that most missile defense equipment could be ''popped up'' by rocket in wartime. This would include very large mirrors (two or three times the size of today's largest astronomical mirrors) to bounce directed energy beams from target to target every few seconds during an attack.

Retired US Army Lt. Gen. Daniel O. Graham, former director of the Defense Intelligence Agency, says he believes the US could deploy a ballistic missile defense system using ''off-the-shelf'' technology within six years at a cost of spend on strategic defense research over the same period. Now the head of a private organization called High Frontier, General Graham says his system would not have to rely on directed-energy weapons.

Instead, it would fire clouds of pellets at enemy warheads. Such a system, he says, could be 90 percent effective and thereby deter the Soviets from launching a nuclear attack.

Members of the Joint Chiefs of Staff were convinced by Graham that there is potential for missile defense. But one officer working on such systems at the Pentagon says ''we have some real problems'' with the cost, timing, and feasibility of the High Frontier plan.

Stages of an intercontinental ballistic missile's flight BOOST-PHASE Missiles engines are burn ing. Missiles rise to an alt tude of 200 to 300 kilometers. Typically last five minutes for a liquid-fuel rocket and three minutes for a solid-fuel rocket. DEFENSE SCENARIO 1 ASSUMING A DEFENSE EFFECTIVENESS OF: 90% THIS PROPORATION OF ORIGINAL REENTRY VEHICLES WOULD SURVIVE 10% DEFENSE SCENARIO 2 ASSUMING A DEFENSE EFFECTIVENESS OF: 75% THIS PROPORTION OF ORIGINAL REENTRY VEHICLES WOULD SURVIVE: 25% DEFENSE SCENARIO 3 ASSUMING A DEFENSE EFFECTIVENESS OF: 75% THIS PROPORTION OF ORIGINAL REENTRY VEHICLES WOULD SURVIVE: 25%

Stages of an intercontinental ballistic missile's flight A MIDCOURSE B REENTRY A. Post-boost vehicle (PBV), carrying guidance system, warheads, decoys, and other penetration aids separates from booster Warheads reenter the atmo- B. Later, warheads and de- sphere, which strips away coys separate from PBV. decoys. This phase accounts Midcourse trajectory lasts for the final one or two min- from 20 to 25 minutes. utes to flight time. 90% 90% 90% 1% 0.1% 0.01% 75% 75% 75% 6.25% 1.6% 0.4% 90% 90% 90% 2.5% 0.25% 0.025%

Next: But would it work?