satellite mission. Future developments in satellite oceanography promise more of the same at ever-increasing accuracy, coupled with the deployment of new satellite-borne instruments. Yet sea-truth is essential and we envisage in situ observations that will be made by an unprecedented class of autonomous instruments and probes. The ability to manipulate these tools in mid-mission is developing. While we are making enormous strides in sampling the global ocean better, we still have far to go for truly adequate spatial and temporal sampling, though the era of grossly undersampling the global ocean is dead.

A national effort to support sustained high-quality global observations over decades is needed. Measurements of air-sea fluxes of heat, fresh water, and gases, of surface and sub-surface temperature, salinity, and velocity, are all necessary to meet new scientific challenges and practical needs. Looking beyond the equatorial TOGA-TAO array, long-term subsurface measurements spanning the global ocean are required.

Given the rapid increase in Lagrangian measurements by drifting and profiling floats, and the parallel increase in geochemical tracer data, an intense approach to Lagrangian analysis of advection and diffusion is warranted; our existing base of theoretical tools and concepts is not worthy of the observations that we are about to receive.

Many emerging physical oceanographic issues concern connections between large-scale and small-scale motions; for example, the relation between small-scale turbulent mixing and the large-scale meridional overturning circulation. Analogous connections and interactions between scales are arising in issues of societal concern, often centered around the increasing recognition that many issues previously regarded as regional now require a global perspective. Anthropogenic pollutants have reached the open ocean and are known to be transported far from their sources. A better understanding is needed of small-scale processes and small-scale aqueous systems (estuaries, wetlands, coral reefs) and their impacts on global issues. For example, the growth of plankton populations, which affect carbon dioxide levels and thus may be important in global warming scenarios, is dependent on details of circulation at fronts, sea-ice, and mixed-layer boundaries.

In most coastal regions, the strongest persistent gradients in properties (for example, salinity, temperature, nutrients or suspended materials) are found in the cross-shelf direction. This is because cross-shelf flow is often inhibited by topography and because the coastal ocean is the contact zone between terrestrial influences, such as river runoffs, and oceanic influences characterized by nonlinear physical dynamics and oligotrophic biological conditions. Progress has certainly been made on some aspects of the flows that determine cross-shelf transports, especially those related to surface and bottom boundary layer processes. A good deal more has yet to be learned about exchanges that occur in the interior of the water column. The problem is difficult because it often appears that the processes that are relevant for the dominant alongshore flows do not apply to cross-shelf flows. For example, it is likely that instabilities and topographic influences may dominate the exchange process. The exchange itself needs to be understood if we are to address issues such as the control of biological productivity in the coastal ocean, or the removal of contaminants from the near-shore zone.

In addition to cross-shelf exchange processes themselves, there is the question of how the coastal ocean couples to its surroundings on both the landward and seaward sides. Estuarine processes are important for determining the quantity and quality of terrestrial materials that reach the open shelves. The oceanic setting, including eddies, filaments, and boundary currents, in turn determines how effectively coastal influences can spread offshore, or how the oceanic reservoir will affect shelf conditions. Consequently, the study of the continental shelf demands consideration of both offshore and near-shore (estuarine and surf zone) dynamics.

Our understanding of inland waters, such as estuaries, wetlands, tide flats, and lakes, will be aided by the same observational and computational technologies that promise progress on the general circulation problem. This work will afford exciting opportunities for interdisciplinary research blending physical oceanography with biology, geochemistry, and ecology. Examples are tidal flushing through the root system of a wetland, and the physical oceanography of coral reefs.

Lakes can be useful analogs of the ocean, with wind and thermally driven circulations, developing coastal fronts, and topographically steered currents. Lakes are important as model ecosystems that are simpler and more accessible than ocean ecosystems. Significant progress can be foreseen in the coming decades in limnology, helped by the tools and ideas developed for the ocean.

The expertise of the physical oceanography community should make possible substantial advances in the understanding of all these shallow systems. Because of the major roles played by turbulence and complex topography, these systems pose impressive and fascinating challenges to physical oceanography.

Past achievements in quantifying small-scale turbulent mixing in the main thermocline, coupled with exciting re