13. In lines 251-252, caution should be suggested in adding unknown forcings because these can easily become nothing more than adjustable parameters. Of course, care should also be taken to include all forcings that are quantitatively known.
14. In lines 254-255, it should be noted that while the air-sea interaction can play a role in internal variability, such variability can also occur in the atmosphere alone.
15. In lines 258-261, while water vapor and clouds are indeed critical to the high climate sensitivity of many models, the references cited (Stocker et al., 2001; NRC, 2003) carefully note that water vapor and especially clouds are areas of major uncertainty in models, and even in nature.
16. The discussion of volcanic influence on lines 309-312 should be reworked to include additional work that has been done on this subject. For example, there is more on the effects of volcanoes on European temperatures in Jones et al. (2003) and in Robock and Oppenheimer (2003). The most affected region is Northern Europe—not North America and certainly not Siberia. The two studies cited in lines 309-312 are also basically model studies, and evidence from observations is less convincing.
17. The claim on lines 331-336 should note the substantial uncertainty of such factors as solar variability (Frohlich and Lean, 2004), historical volcanic forcing (Bradley, 1988), and aerosols (Charlson et al., 1992; Anderson et al., 2003).
18. The report appropriately notes that the radiosondes show an abrupt increase in temperature in the troposphere around 1976 and the fact that this is missed in the satellite data which starts in 1979. It has been argued that the surface warming is simply the response to this jump with a delay due the heat capacity of the ocean (Lindzen and Giannitsis, 2002). This is distinctly relevant to the present report.