Methods of sampling and analysis and our concepts of ocean dynamics

In the last two decades, the increase in multi-nation, multi-ship projects has underlined the importance of intercalibration and intercomparison studies. At the same time, it is becoming obvious that with our present methods of sampling and analysis we can never hope to distinguish the variability in our sampling and analysis techniques from the variability in the universe sampled. The presence of discontinuities in many of the quantities measured suggests that our sampling grid, as maintained by oceanographic vessels, is too limited both in space and in time. Some variety of self-reporting sampling buoy is the obvious answer for physical and chemical parameters, although as yet we do not have detectors for many of the quantities of interest. Even with such buoys, it would seem unlikely that we could construct a grid of such buoys tight enough to measure the local variability of the oceans without the buoys themselves contributing to this variability. The limit finally selected for such a sampling grid will probably be some compromise between scientific capability and economic reality.