Proposed screen for accidental nuclear launches raises problems. `STAR WARS'

The plot is a techno-thriller staple: By accident or madman, unauthorized launch of a single nuclear missile is about to take place. If the strike isn't stopped, an American (or Russian) city will be devastated. World War III is the likely result. So far this scenario exists only in fiction. But whether the United States should try to erect a shield against a real renegade attack is now the focus of a crucial debate over the Strategic Defense Initiative (SDI).

At issue is whether an initial ``star wars'' defense need be only a thin screen capable of stopping a few nuclear missiles. In a larger sense, the argument is also about ``star wars'' progress and promise - or lack thereof.

Protection against accidental launches ``is a very interesting concept and one we're evaluating very seriously,'' said O'Dean P. Judd, the SDI chief scientist, in an interview.

The idea of an accidental launch prevention system (ALPS) was first raised by Sen. Sam Nunn (D) of Georgia, chairman of the Senate Armed Services Committee. It would be, in essence, preschool for ``star wars,'' as ALPS would be set up before the Phase 1 SDI shield now envisioned by the Pentagon, according to Dr. Judd.

A special team of SDI officials is studying the ALPS concept, with a report due in a few months. Judd said that from the beginning, the Pentagon has looked at ``star wars'' as something that would begin with relatively rudimentary defenses and progress in phases to a more complex system with lasers and other exotic weapons.

``When you have the Wright Brothers' Kitty Hawk just getting off the ground, you don't stop doing everything until you get the technology to make a Mach 4 fighter,'' said SDI's chief scientist.

Inasmuch as a screen against accidental launches was first suggested by Senator Nunn, one of the most influential lawmakers on Capitol Hill, there are obvious political reasons for the Pentagon to study the idea.

But the subject also creates a potential problem for Defense Secretary Frank Carlucci. If he recommends that accidental launch prevention be the first ``star wars'' goal, lawmakers and the public may react negatively, as such a system would be a much less capable one than administration SDI rhetoric has promised.

Two defense contractors - McDonnell Douglas and Lockheed - have been promoting ALPS designs on Capitol Hill. Both systems are based on 100 interceptor rockets based at just one site, and thus would not violate the 1972 Antiballistic Missile (ABM) Treaty.

Both designs promise protection against only one attacking missile. Furthermore, McDonnell Douglas officials say that ALPS would not be able to protect US coasts - leaving some 60 percent of the US population still vulnerable.

``Complete defense of the United States cannot be achieved'' while staying within the bounds of the ABM Treaty, McDonnell Douglas program manager D.Dean Hofferth told Congress last week.

More rockets scattered at more sites would provide increased coverage - but also break the ABM pact. Any ALPS design probably would not include space-based interceptors, an important feature of previous ``star wars'' first-phase plans.

Critics say the fact that the Pentagon is considering an accidental launch system shows that it is scaling back its initial SDI goals.

Pentagon officials reply that they are simply thinking about inserting a new, pre-SDI phase in front of old goals, leading to a somewhat metaphysical debate about the definition of ``scaling back.''

Lt. Gen. James Abrahamson, SDI chief, has said in the past that a decision on whether or not to deploy strategic defenses could be made in the early 1990s. In an appearance before Congress this year he said this date had now been put back ``at least a year, maybe two years.'' He blamed budget cuts.

But a new 900-page congressional report will reportedly say that even first-phase SDI systems won't work. The study, produced by the congressional Office of Technology Assessment, says that any ``star wars'' computer system would be so complex as to be unworkable, according to a report in the Washington Post. Dependability of computer software, in particular, would always be a question mark.

The study concluded that an SDI system probably would ``suffer a catastrophic failure'' the first time it was used, according to the Washington Post.

In any case, budget pressures may be pushing the Defense Department to restructure SDI research and include accidental launch protection as a first goal.

All the armed services are being forced by Secretary Carlucci to lower their sights, to bring overall Pentagon budget projections in line with reality. The SDI organization will have to prune its programs, just as the Army, Navy, and Air Force will, according to defense officials.

The armed services tend to view SDI as a financial rival. They worry that ``star wars'' research will draw money away from more conventional military projects in the years ahead.