Comment by DanielStraight

And clouds, and wave caps, and noise in the input data, and natural variations in ocean color... I think your estimate of the number of false positives is off by several orders of magnitude.