NASA officials, who have begun to assess their future needs in the wake of the Challenger tragedy, are "not certain yet" that they will need another shuttle, according to sources aware of these preliminary discussions.

" 'What if' scenarios," one source said, show that the National Aeronautics and Space Administration is "not in as bad a shape on future vehicle availablility rates as first appeared."

That is primarily because the shuttle Discovery originally was scheduled to fly only three times in the 20 months between July 1986 and April 1988. In a revised schedule, one NASA source said, Discovery could be made available for launches "every three months" once flights resume. That could mean four additional Discovery flights during that period.

During those same 20 months, Challenger, which exploded Jan. 28, had been scheduled to be launched 10 times.

For the longer term, however, another shuttle and additional unmanned launch rockets may be needed to meet both civilian and national security space needs, according to sources in the administration and on Capitol Hill.

A White House senior interagency group on space, chaired by national security affairs adviser John M. Poindexter, and including NASA, the Defense Department, the Joint Chiefs of Staff, Central Intelligence Agency, Office of Management and Budget and the departments of State, Transportation and Commerce, has begun to work on options for President Reagan. The panel decided last week in an organizational meeting to concentrate on "scheduling remaining orbiters and determining the implications for future launch capabilities," and on the question of whether to build an additional orbiter, according to sources involved in the meeting.

The White House group will not wait for the report of the presidential commission investigating the shuttle accident headed by former secretary of state William P. Rogers, but will "work in parallel with it," one official said.

NASA officials cautioned yesterday that their early estimates of launch availabilities would be revised to take into consideration the Rogers investigation findings, any modifications suggested by those findings and the date when shuttles are cleared to fly. At the NASA working level, however, assumptions are being made that shuttle flight slippage could run from six months to a year or more. Temporary fixes and then more permanent modifications are being built into the forecasts.

No matter what the Rogers commission finds, NASA will have to develop a fix for the seals on the solid booster rocket, sources said. The burn-throughs on these seals, which have been studied for two years, have been a focus of the NASA team and the Rogers investigation. NASA was studying possible changes to correct the problem before the Challenger accident.

As an example of the costs involved and the time needed for possible seal solutions, Roger Cook, the NASA budget analyst whose memo highlighted the problem last July, told his superiors following the Challenger explosion that as a first step, the eight sets of boosters now being manufactured with the old seal system will have to be modified.

That means removing propellants that are already in some casings, a job that Cook wrote could cost "$2 million per flight set." One remedy for the seal problem includes changing steel segments on the booster, Cook reported. That would require a manufacturing lead time of "13 months after successful acceptance testing," he wrote.

Easing some of the early launch pressure on NASA, one source said, was an initial indication that "a lot of commercial customers are not unhappy with delays" because of softness in the space communications business.

The Air Force, for example, has informally told NASA that the first launch of a shuttle from Vandenberg Air Force Base, Calif., now scheduled for late July, can be delayed because it is a test rather than operational flight "and of no particular rush," a NASA official said. The shuttle Discovery, which is to make that first Vandenberg flight, was scheduled to be at Vandenberg by March 20 for a launch test firing, a NASA official said. With Vandenberg running two weeks behind schedule, that trip can be delayed until April, he added.

He said the first Air Force Vandenberg launch of an operational satellite, once scheduled for September, had been unofficially postponed until February before the Challenger accident. The delay was caused by a slip in preparing the cargo for that launch, the new, $1 billion KH12 photo satellite that will give the United States new visual intelligence capabilities over the Soviet Union.

Current NASA thinking calls for two test launches from Kennedy Space Center in Florida before the first military launch goes off from Vandenberg, sources said. That is because NASA wants to resume flights from its normal site, rather than from a new launch pad with a new type of solid rocket booster as is now planned for the Vandenberg shuttle.