NASA's feud with the National Oceanic and Atmospheric Administration over who gets to carry the nation's meteorological satellites into space is nearing an end. NOAA will abandon a scheme to use surplus Air Force missiles to launch the so-called Metsats into orbit and the National Aeronautics and Space Administration will give NOAA a discount for carrying them into orbit aboard the space shuttle.

The feud began last February when NASA Administrator James M. Beggs criticized the Air Force and NOAA for a plan to launch NOAA's satellites on Titan 2 missiles instead of from the space shuttle.

"I don't want to suggest that anything dark is going on here, but some people think this whole affair may be a heavy-handed scheme by the Air Force to give the shuttle a black eye," Beggs said at the time. "I don't like it. It doesn't help NASA any."

Replied Anthony J. Calio, then NOAA's acting administrator and just last week confirmed administrator by the Senate: "The Air Force has offered me a deal that's going to save me $90 million and I'm going to take it."

The deal never went through, in large part because NASA and NOAA began working behind the scenes to bury the hatchet. Just last week, a letter went out from Beggs to Calio outlining a deal governing three Metsat launches starting in 1989.

NASA agreed to share with NOAA the $80 million cost of modifying the Metsat for a shuttle launch, agreed to put a Metsat on a shuttle within three months after NOAA asks for it and agreed to shave $6.5 million off the $105 million in fees NASA would have charged NOAA for the three launches. Beggs said, "We gave them a good price. They're on the shuttle at what essentially is our cost for launching them."

SMARTER SHUTTLES . . . Late next year, the space agency expects to take delivery of the first "flight model" of a new IBM computer for the space shuttle that will work at half the cost and have more than twice the memory and three times the speed of the old IBM model that has been used to navigate the shuttle since it first flew in April 1981.

The new computer fits in one box instead of two and is half the weight. IBM says the new model is more reliable, in part because it uses fewer parts. IBM predicts a performance time of 10,000 hours before the first failure and a mean time of 6,000 hours between failures. NASA has ordered 26 of the new computers at a cost of $45 million.

NASA expects to fly the new model on the shuttle in May 1987, which will be none too soon. There have been six computer hardware failures aboard the shuttle. Three caused launch delays. NASA flies five computers aboard the shuttle, four to navigate the $1.2 billion spaceliner and the fifth as a backup in case the others fail at the same time.

REACHING AFAR IN '86 . . . When he stood up to speak at this week's staff meeting, NASA's Deputy Associate Administrator Samuel W. Keller said: "We expect 1986 will be the most demanding year we've ever had."

In January, Voyager encounters the planet Uranus. In March, the space shuttle flies a mission dedicated to observing Halley's comet. In May NASA will launch Ulysses, which will fly to Jupiter and use its gravity to "slingshot" itself around the sun, and Galileo, which will orbit Jupiter. In August, the space telescope blasts off aboard the space shuttle. The three spacecraft cost $2 billion, not counting launch costs.

Keller's toughest time will occur in May when Ulysses and Galileo will be in the cargo bays of two space shuttles sitting on their launch pads at the same time. NASA will try to launch Ulysses May 15 and Galileo four days later. If anything delays the May 15 launch, NASA has 24 more days to get the shuttles into space, a "launch window" that allows the two spacecraft to fly to Jupiter with the least amount of fuel. Keller said, "I'll be glad when 1986 is over."