Adaptation to unknown territory: potential and limitations of observation-based projections

The scientific consensus about the reality and seriousness of anthropogenic global warming does not include the feasibility of long-term projection of regional and local climate by the conventional physics-based climate models. Such projections are critic al elements in any successful strategy for adaptation to climate change. A main objective of this project is to explore new, alternative prediction schemes and to establish more precise limits to predictability. We explore the potential and limitations of unconventional types of climate projection which is mainly based on historic and reconstructed data on climatic variables and forcing, and different scenarios for future forcing. It is expected that this can contribute to the convergence to a scientific consensus on this issue. The R&D challenges encompass implementation of a host of quantitative methods; statistical analysis of climate records, dynamic-stochastic modeling of climatic and socio-economic variables and their interactions, forecasting of trends, uncertainty, and extreme-value statistics. New methods are developed, and old methods are applied in new areas. Emphasis is on observation-based dynamic-stochastic models rather than physics-based general circulation models. Another class of re search challenges is to explore how knowledge about the potential and limitations of climate projections can be conveyed to end users in such a way as to be of value to industry, policymakers and wider society. The critical organizational challenge is to establish a well-functioning interdisciplinary national and international network of researchers and professionals with high credentials in the fields of mathematical finance, market analysis, quantitative sociology, time-series analysis and forecasting, dynamic-statistical modeling, and climate dynamics.