Following the U.S. declaration of war on December 8, 1941, Arrow, who was certain to be drafted, enlisted in the hope of securing an officer’s commission in the U.S. Army Corps where he believed he would have a chance to use his mathematical and statistical training. He was quickly approved to attend an aviation training program at New York University in October 1942, taking “active duty” breaks from classes for rifle drill, which he and his colleagues thought rather silly. Nonetheless, he came out of that program in September 1943 commissioned as a weather officer with the rank of second lieutenant and was assigned to a weather research facility in Asheville, North Carolina; in July 1945 he was transferred to the weather division headquarters of the Army Air Force. It was during that time in Asheville that he wrote a memorandum that later, in 1949, became his first professional paper (“On the Use of Winds in Flight Planning” in the Journal of Meteorology). That paper presented an algorithm for taking advantage of winds aloft to save fuel on North Atlantic air crossings, an idea that was not acted upon by the military at that time but became the canonical practice for North Atlantic flight paths in the postwar period.
That is from the new, excellent, and consistently interesting Finding Equilibrium: Arrow, Debreu, McKenzie and the Problem of Scientific Credit, by Till Düppe and E. Roy Weintraub. Unlike many history of economic thought books, this one tells you “what actually happened,” such as how an Econometrica editor (Robert Strotz) decided to publish the McKenzie paper before the Arrow-Debreu paper, when he had both in hand.