climate system to human activities, and to develop improved capabilities for applying and evaluating these predictions.
The climate research of the past few decades drives the requirements for future research by focusing our attention on the remaining uncertainties and on the importance of climatic research for society:
• Climate variability, such as El Niño, can be characterized by significant economic and human dislocations. Modeling studies over the past two decades suggest that aspects of this climate variability may be predictable. In cases where El Niño/Southern Oscillation (ENSO) events were predicted in advance, immediate practical benefits were realized through human response and adaptation.
• Analyses of historical records have revealed a number of interesting cases of longer-period fluctuations for North America and other parts of the world, while model studies have demonstrated that ocean-atmosphere and land-bio-sphere-atmosphere interactions are plausible mechanisms to explain decade-to-century variability. Historical and paleoclimatic data, as well as coupled models, indicate the potential for significant climate variability on long time scales. Such changes can be expected to occur in the future, irrespective of human impacts on climate. Current observational capabilities and practice are inadequate to characterize many of the changes in global and regional climate. An enhancement of current observational capability and improved knowledge of the coupled Earth system will therefore likely increase our understanding of climate variability on all time scales and lead to a greater realization of practical benefits.
• The effort to predict the climate response to increases in greenhouse gases has both demonstrated the importance of this problem to society and focused attention on many of the most important limitations of current climate models. Increased concentrations of greenhouse gases and changes in land use and land cover are directly and indirectly tied to human activities. Current model projections based on increases in greenhouse gases and aerosols and on land cover change indicate the potential for large, and rapid, climate change relative to the historical and paleoclimatic records, with concomitantly large influences on human activities and ecosystems. Although remarkable progress in developing these climate models has occurred over the past two decades, current climate models are characterized by a great number of uncertainties. Improved predictive capability is likely to have a positive impact on economic vitality and national security because of its potential to minimize risk and maximize benefit associated with the impacts of any climate change.
A comprehensive analysis of the remaining scientific questions and uncertainties and of the societal drivers for climate research leads us to four major imperatives for the twenty-first century. Each imperative is associated with a series of basic requirements: