cipitation is stored for highly variable periods in surface water, groundwater, glaciers, and ground ice and permafrost. Changes in the total flux of precipitation and in the balance among the components of the hydrological cycle can affect ocean circulation, vegetation types and productivity, and the availability of freshwater for human and ecosystem needs. Despite the obvious importance of the hydrological cycle, surprisingly little is known about total groundwater storage and water quality, trends in recharge and discharge, permafrost evolution (including feedbacks on greenhouse gases), feedbacks related to the generation and persistence of drought, how drought is eventually broken, and related topics. A concerted effort should be made to monitor essential components of the hydrological cycle, as this system is likely to be greatly affected by abrupt changes in climate.
Sophisticated systems are available to measure atmospheric properties and circulation, to assimilate new and old data, and thus to characterize the behavior of the atmosphere. Some key data sets, such as those from weather ships, remain vulnerable to funding cuts, and additional sampling in some regions would be useful (NRC, 1999d). Reliable assimilation of well-calibrated data sets is required to characterize gradual and abrupt climate change observed in instrumental records. Furthermore, additional analyses of the instrumental records are needed to identify climatic modes and their long-term behavior.
Understanding of how abrupt climate change will affect economic and ecological systems requires improved monitoring, detection, and measurements of critical components of nonmarket and ecological systems. Monitoring might provide early warning of climate-change effects and help minimize their influence. On the whole, the market economy in developed nations is well monitored in most sectors through national economic measures and accounts. Measures for nonmarket and ecological systems are much less systematic. Various research and monitoring activities should be considered:
Efforts to identify plant and animal species should continue because this information is essential for determining extinction rates.
High-frequency monitoring of terrestrial ecosystems is needed be