There are a variety of views about the success of interdisciplinary climatic change research caricatured, at the extremes, as either a totally dysfunctional or fully harmonious system. The former view focusses on results provided by climate modellers which have not been accompanied by adequate descriptions or evaluations of reliability; the delay between generation of climate simulations and their use in impact assessment and policy development; and the erection of barriers to communication and information flow. On the other hand, it is possible to argue that the climate community functions well; that impacts researchers choose to use 'agreed' rather than more recent, and still debated, results; and that policy makers are well-informed about the reliability of climate models and impact evaluations.
Those who believe everything is fine cite the Intergovernmental Panel on Climate Change (IPCC) (e.g. Tegart et al., 1990; Houghton et al., 1990Houghton et al., , 1992) ) and case studies such as those by Rosenberg et al. (1991) and Rosenzweig and Parry (1994) as evidence and so, interestingly, do those who contend that the climate 'industry' is beset by multifaceted divisions including technology, training, information access, understanding, terminology, economic development, and social and political perceptions and flexibility (e.g. Cubasch et al., 1994;Henderson-Sellers and Howe, 1996). Whatever the true state of climate research, the levels of funding received, for example in the USA climate science has been termed the most successful sales job by scientists since the 'War on Cancer' in the early 1970s (Victor and Salt, 1994), demand a description of the societal benefits derived. One of these might be enhanced climatic-change information sharing: the focus of this essay.
The case for a more transparent process of information exchange has an historical perspective. Originally, climate-model predictions were of interest only to modellers, providing one means of model evaluation not widely exercised until recently (e.g. Cess et al., 1991;Gates, 1992). Today, model projections, grouped into scenarios, have importance for both research and policy communities. For the research community, scenarios are necessary organizing devices for integrated assessment of the climate issue (e.g. Rotmans et al., 1994). For the policy community, scenarios describe alternative possible futures: costs and benefits associated with particular scenarios are one contribution to the assessment of policy options.
Despite this increasingly important role for scenarios, the climate modelling community continues to generate results of simulations in formats that are largely not useful for consumers from other disciplines nor for policy purposes (e.g. Skiles, 1995). The proliferation of poorly documented climate scenarios is seen by some