Since the locations of the five gyres are roughly known,
the search for optimum Array positioning was reduced to
these known areas in order to save computational time.
(see Table 2.15). The given boundaries were chosen manually
on the basis of van Sebille’s simulation. Within these
borders, a grid of 1° x 1°, was used; Array lengths were
varied in steps of 50 km, and various Array orientations
were investigated in steps of 15°. A typical optimization
run would test all configurations within the defined gyre
area bounds and pre-defined steps and increments.
For the sake of simplicity, each gyre was handled separately
with only one Array at a time.