Early this week the New York Times carried the obituaries of Ivan Luini and Sergio Savarese, two Italian designers and entrepreneurs in their 40s who had been both successful and highly esteemed. I did not know either of them -- although close friends of mine were very close to Luini, and heartbroken by this news -- but I paid particular attention. The two men died in a crash of exactly the same kind of airplane I had owned and flown for the past six years.

The airplane was the Cirrus SR20, a propeller plane, with one engine, that with its sister SR22 has become the most popular small airplane model in the country. In 1999 I wrote an article about this airplane in the New York Times Magazine, and it was the centerpiece of my book Free Flight, in 2001. In both cases a major point about the plane is that it should be safer than other aircraft, because (among other reasons) of an innovative parachute system, capable of bringing the plane and its occupants to a safe and survivable landing if something would go wrong. The Times obituary noted the existence of the system in the recent crash:

According to Jennifer Kaiser, an air safety investigator for the National Transportation Safety Board, the plane has a parachute system for emergencies, and it deployed at a low altitude before impact. She said the pilot reported icing conditions and turbulence, and an inability to maintain altitude, but it is not known which man was at the controls.

After a crash, the reaction of the piloting world is predictable -- and strong. First there is sympathy for all families affected by the disaster. This is genuine, because everyone who flies airplanes has had "there but for the grace of God" thoughts about being in the same situation. Then there is rationalization -- essentially, the opposite of "there for the grace of God." This involves explaining to oneself why the pilot who suffered the tragic fate made decisions that led to the problem (it is almost always the pilot, not the airplane, that is ultimately responsible for a crash) -- and the instinctive next step of saying, " But of course this would not happen to me." I will give my own reaction as a sample:

At least ninety five pecent of the risk of flying in small airplanes is related to weather. Once you learn the basic skills of flying and landing an airplane, in good weather you can do them almost automatically. In good weather that leaves only the risk of major mechanical failure -- engine stoppage, etc -- which is not unheard of, but is rare.

The accident that killed these two men happened over the Rocky Mountains during bad weather, in the afternoon. No one knows for sure what occurred. But here are the points that went through my mind:

* The SR20 plane these men flew, and that I owned before I sold it when moving to China, is a wonderful craft but relatively underpowered for mountain flying. When I flew mine from the East Coast to Aspen, Colorado last year, I spent 30 minutes circling in the skies over Boulder, gaining altitude to cross the Rockies. In high mountains it is like a moped on a freeway. It has no extra oomph to get out of trouble. (Its SR22 counterpart is much more powerful, and much more expensive.)

* I have flown over the Rockies eight times. For each trip, I have had several rules, based on the plane's limited performance. Only in daylight. Only early in the morning, as soon as possible after dawn, when the air is calmest and the climbing power is greatest, because of the colder, "thicker" air. Never in conditions that would require an instrument flight plan -- which in practice means only in clear weather where you can see exactly where you are relative to the mountains. And never, ever in conditions that indicate either a thunderstorm or icing -- the two threats that small planes cannot endure. (The problem with thunderstorms is obvious. In icing -- basically, being inside a cloud with the temperature below freezing -- the problem is that the plane loses its aerodynamic shape, because the wings are covered with ice, and it stops flying and falls from the sky.)

* At the very time this plane crossed the Rockies, I was stuck for hours in the Salt Lake City airport. Delta, American, United and others had cancelled their flights, because the weather was too dangerous. This was the weather into which the SR20 flew. The pilots apparently reported to air traffic control that they had encountered serious icing. This is the worst possible circumstance in a small airplane. It simply cannot stay in the air when covered with ice.

Most discussion on the pilot internet forums has dwelt on the terrible weather. The stated point -- how risky this was -- has conveyed two usually-unstated ones: Every pilot's half-rationalized assumption that he would not take off into the same conditions, and the related question of how two such brilliant people had decided to do so. The latter has been humbling to many other pilots. If people like these could make the decision....

What about the parachute that was supposed to remove risks? In fact, it has generally been very, very effective. The only other case, beyond this one, in which it appears to have "failed" is when a pilot apparently pulled it at the last minute, during a high-speed nose-drive in even worse weather conditions. A Cirrus pilot (and friend) named Rick Beach, in San Diego, has very carefully analyzed the Cirrus accident history, here.

Sincere sympathies to the pilots' families.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.