Quirin Schiermeier has an excellent article in Nature which looks at four areas of significant dispute within climate science circles: regional climate forecasts, precipitation forecasts, aerosols and palaeoclimate data.
Regional forecasts are a critical tool for local governments and it seems to me that we are awaiting good approaches to closely-grained data to justify the costs of action on adaptation in some cases:
"Downscaled climate models face particular uncertainty problems dealing in regions with complex topography, such as where mountains form a wall between two climatically different plains. Another potential source of error comes from projections concerning future greenhouse-gas emissions, which vary depending on assumptions about economic developments."
With the levels of uncertainty high, we need to mount potent multi-factorial arguments to ensure that we capture all of the benefits - as I suspect that the costs will have a large in-built variability for some time.
