What we have learned about field programs

First, it is the people most of all who make a field program successful, and cooperation and collegial consensus building are vital during all phases; good health and a sense of humor both help make this possible. Second, although numerical models are now playing a central role in all phases of a field program, not paying adequate attention to the observations can lead to problems. And finally, it cannot be overemphasized that both funding agencies and participants must recognize that it takes several years to fully exploit the datasets collected, with the corollary that high-quality datasets should be available long term.