Does the crash of Air Florida Flight 90 say something about National as a safe, modern airport? You bet it does. Although some simplistic answers may eventually be given by government agencies and airline managements, the real cause is the cumulative sum of National Airport's deficiencies. The real test of a human, a machine or an airport comes at the moment of maximum adverse circumstances. National faced its test Wednesday and failed.

Imagine that two pilots in a plane are sitting near the takeoff end of runway 36 awaiting clearance. In heavy snow or other near- freezing precipitation, they will be concerned about, besides the usual things, ice sticking to the aircraft and ice, snow, slush or standing water on the runway. Jets are prohibited from taking off when slush is a half-inch deep. The pilot, with no sure way of checking this as he moves onto the runway, relies on weather reports. But if the runway has been cleaned 10 to 15 minutes earlier, another half- inch may have accumulated.

As a practical matter, there is at National only one usable runway (18-36)--6,869 feet long, and very substandard for air carrier jets. When snow and slush removal becomes mandatory, National requires that all takeoffs and landings cease. In effect, the airport is closed in the face of perhaps 20 or 30 trusting arrival aircraft. Other modern jet air carrier airports handle this problem by multiple runway operation. This is not possible at National. The result is that operations are often continued during a much higher percentage of marginal conditions.

Our hypothetical pilots are given a hurry- up immediate-takeoff clearance. This pressure is typical and is absolutely necessary since National's 700 to 900 daily trips, between 7 a.m. and 10 p.m., allow only about a 60-second interval between alternating takeoffs and landings--both in the same direction on the single jet runway.

Our pilots wheel onto the runway finishing their "final items" checklist. As they begin to roll, their minds flash to these questions: "Is there any ice or wet, sticking snow on the aircraft that may affect liftoff?" "Am I going to find slush on the last half of the runway?" They know that if either of these becomes apparent on the slippery pavement after they reach approximately 70 mph, to abort means almost certainly sliding into the river.

Though on paper about 100 mph ("V-1") is the speed below which the book guarantees a successful abort, the pilot knows better. So he elects to keep going, hoping things will get better by the end of the runway. At the runway end, he finds he barely has V-1 but does not have "rotate" speed, a few mph higher. He has no choice so he "rotates" anyway, pulls the nose very high, climbs to 100 feet, finds speed decaying, with the aircraft all the time in prestall buffeting.

He knows he must allow a small descent, certainly no more climb, if he is to avoid full stall and certain death. He finds that the aircraft starts slowly to drop in descent profile. He cannot stop until he hits an obstruction. He has worked so intensely hard on these difficult problems that, lo and behold, his landing gear is still down!

There are some other nasty aspects to taking off at National to the north on runway 36. Takeoff minimums are 700 feet Runway Visual Range (RVR). This means takeoff is legally allowed here when the pilot can only see one-tenth of the runway. This is in effect a "0-0" takeoff--zero visibility, zero ceiling.

Our pilot knows that his heading, while rolling down the runway, is straight for the Washington Monument, 550 feet high. He knows that leaving the airport boundary, he must start an immediate left turn northwest, theoretically toward the midpoint of a bridge that he cannot see because he is in the clouds. If there were no clouds, he couldn't see anyway because he has pulled his nose so high.

Most modern airports are located and designed so that Standing Operating Procedure is: climb straight out on runway heading to 500 feet with radio navigation such as "back-course Instrument Landing System" available. Not so at National, where a back-course ILS would take you over the Washington Monument and the White House, which is prohibited.

Can you see the cumulative burdens that the pilot at National in this situation is asked to carry?

In sum, National has:

1.A single substandard runway--and only one is available for air carrier jets. If the combined "footprint' of runway 36 plus the linear ground path to the 14th Street Bridge were superimposed on any of the three Dulles runways, the straight length of Dulles pavement would be longer. Any pilot would feel that a bridge to hurdle at the end of a Dulles runway would be an unusual hazard. Dulles has six different runway ends, three runways 10,000 to 11,000 feet long. Baltimore-Washington International offers four runway ends, all over 9,000 feet.

2.Only one Instrument Landing System. This means that in instrument conditions where landing north is therefore mandatory, many times the wind is from the south or west, requiring dangerous downwind or crosswind landings. No one ever has to land downwind at Dulles where five runway ends offer various forms of ILS. BWI has four forms of ILS. At National, when instrument landings are being made and downwind landings are mandatory, downwind takeoffs are also mandatory because both have to be in the same direction.

3.Poor overrun areas. National's north end 36 has 200 feet with a 10-foot drop into the Potomac. South end has a 375-foot overrun up to approach lights mounted on telephone poles that resemble a tank obstacle course. Dulles and BWI have the FAA's suggested 1,000 foot overruns on all endings.

In my opinion, Flight 90 could have made a more hassle-free departure at Dulles or BWI. It would have had many options throughout the takeoff roll. In any case it would not have had to start an abrupt left turn below 100 feet with no radio aids. There would have been no psychological hazard of a bridge or monument.

Rather, the plane would have proceeded straight out, with back-course ILS for precise guidance available. If the pilot had not exercised his option to accelerate to V-1, then if he found himself in trouble when airborne he could proceed straight out over high-quality overrun, on out over flat fields with no man- made obstructions.

I am convinced that had flight 90 operated from Dulles or BWI, there is an excellent chance that 70-odd people would be living today and that the Washington area and the whole nation would have been spared the horror in everyone's mind today.