IPCC procedures state that if a consensus is not reached, "differing views shall be explained and, upon request, recorded." All comments made during the process are listed and their resolution noted. For example, you can go to the Harvard Environmental Science and Public Policy Archives (click Take Aways and References for a link) and see all the comments and responses on the first and second drafts of the Working Group I Fourth Assessment Report. Of course, this doesn’t mean that every scientist is entirely satisfied with everything in the final document. However, it does mean that the IPCC process is open. More about dissenting positions is discussed in the section, "How Sure Are Scientists?"

Of the IPCC reports, only the Summary for Policymakers receives a word-by-word review by international diplomats. The lead scientific authors of the report are present to ensure that any changes are scientifically valid. The diplomats have a say in how the summary is worded, but the scientists have the last word on what is said.

No. Climate models aren’t fit to a particular trend in temperature. However, some climate processes are represented by approximations (called parameterizations), and those formulas may be fit to best match observed data. There are actually only a few of these that are adjusted, and once they are set, the model results are not constrained to fit observations or expected results.

Regional climate models look at just a small area of the globe, but at higher spatial resolutions. They do this by creating many smaller grid boxes within part of the larger global grid. This higher resolution can capture more of the terrain details that greatly influence local climates. However, this also greatly increases the number of calculations. To keep the computations to a manageable level, the models are run for a limited area, such as the Arctic, the U.S., or New England.

Regional models are improving and can currently provide realistic simulations of smaller features like hurricanes that the global climate models can’t. However right now, the models can only make general projections about average temperature and precipitation patterns, especially where the grid sizes are still too large to adequately depict local complex terrain.