Innovative instrumental methods have been developed that offer us hope of detecting abrupt climate change when it appears. These include both new in situ probes for atmosphere and ocean and remote sensing from satellites. New sensors (for chemical and biological fields as well as physical quantities) and new platforms (moored, drifting, gliding or profiling) are becoming available in large numbers. They provide efficient coverage globally and locally, at lower cost than classical observation methods. In reviewing the abrupt changes seen during the instrumental period, it is obvious how much more would now be understood if just a few well-placed modern instruments had been in place for an extended time.
For abrupt climate change, special emphasis on the dynamics of freshwater in both atmosphere and ocean is needed. Evaporation, precipitation, river flow, surface moisture, cryosphere dynamics, and related upper ocean dynamics are all relatively poorly observed, yet they are as important to climate as are the better observed thermodynamic fields. Cloud and water vapor dynamics in the atmosphere are crucial to climate, yet their representation in coupled numerical models is crude.
Models that are heavily involved in predicting or diagnosing abrupt climate change are known to be inaccurate in representing various high-latitude oceanic processes. Among them are deep convection, sinking and overflow dynamics, flow through narrow passages, over sills and in the descending branch of the oceanic overturning circulation, sea-ice dynamics, and flow and thermodynamics of shallow shelf regions at high latitude. Discrimination between sinking in the Labrador Sea and sinking farther north, with overflow at the Greenland-Scotland ridge system, is not well handled by such models.
For these reasons, sustained observations of the high-latitude ocean and its communication with the Arctic are needed. Hydrographic and chemical-tracer observations must be repeated, quasi-regularly, across important passages and major “stable” water masses. Direct observations of ocean circulation intensity using both in situ and satellite altimetric observations are particularly valuable for inferring meridional overturning variability.
Satellite scatterometer wind fields give us a component of global air/sea interaction with remarkable resolution. Passive radiometric observations give us a wealth of temperature, sea-surface structure and moisture data, and possibly new fields such as surface salinity. Synthetic aperture radar observations from satellites provide high resolution detail of ice fields, upper ocean flows, winds and terrestrial flooding and drainage-basin flows.
Tropical ocean observations near the sea-surface currently provide cov-