OPTIMIZING A DRIFTER CAST STRATEGY WITH A GENETIC ALGORITHM

An experiment design problem-that of drifter cast strategy-is discusse d. Different optimization techniques used as part of preparations for the Semaphore-93 air-sea experiment, during which drifters were deploy ed, are examined. The oceanographic experiment objective was to sample a 500-km-square zone centered at 33 degrees N, 22 degrees W in the Az ores current area, using an average of 25 surface drifters for at leas t one month. We investigate different ''orders of merit'' for determin ing the performance of a particular cast strategy, as well as the meth od of genetic algorithms for optimizing the strategy. Our technique us es dynamic reference knowledge of the area where the simulation takes place. Two reference sets were used: a steady-state field calculated w ith data collected from the Kiel University April 1982 hydrographic ex periment, and data output from a regional quasigeostrophic model assim ilating two years of Geosat altimetric data. The strategies obtained v ia the genetic algorithm method were compared with regular array drift er deployments, which is the intuitive and commonly used approach. It was also found that a cheap objective function gives strategies compar able to one which was more accurate but computationally expensive. Opt imization by the genetic algorithm method is shown to be efficient and robust.