SEAFLUX, and Global Energy and Water Cycle Experiment (GEWEX) Radiation Panel efforts to develop technologies for reference radiosondes, and discuss international efforts, not just U.S. efforts.

3. The organization of the chapter is centered around data types, such as “surface”, “tropospheric”, and “reanalyses”. A variety of new issues that do not directly map to the seven recommendations are brought up in these sections (much of which is not relevant or belongs in previous chapters). An alternate organization would be to have seven sections with section headings that are the first sentence of each recommendation. Then the text of each section would tie directly back to a need documented in the earlier chapters, would include discussion of the adequacy of current national and international plans to address this need, and make further specific recommendations for implementation of this recommendation.

4. A substantial amount of new information is introduced for the first time in Chapter 6, including material that should have been introduced in earlier chapters if it is deemed relevant and material that does not directly map to the seven recommendations. The following is specific information that is redundant or should be moved to previous chapters:

• The material in lines 54-71 should be mentioned in the context of Chapters 1 and 5.

• Text on snow and sea ice and sampling inadequacies in lines 179-187 should be moved to Chapter 2.

• For lines 138-177, lines 240-254, and lines 300-317, text on combining surface temperature and dew point temperature is far too wordy, and the main point is lost. This concept should be included in a general recommendation on the need to evaluate and interpret the temperature data in the context of other data sets (e.g., humidity, winds, ocean heat content, etc.) and to understand issues such as the impact of changing land use on temperature trends, as stated in lines 347-355.

• The text in lines 447-494 about recommending specific improved climate model parameterizations is not directly relevant to the present study, although it is appropriate to state in earlier chapters that inadequate parameterizations in numerical weather prediction models contribute to potential problems in using the reanalyses to determine temperature trends.

• The text in lines 98-105 should be moved to Chapters 2 and 4 as these points were not adequately made in those chapters.

5. As far as the current recommendations in Chapter 6 still appear after the chapter is revised, here are comments on each of the current recommendations. The seven recommendations in Chapter 6 have been said numerous times before in other reports. Also, given the relative lack of traceability of these recommendations to the previous five chapters, it may be that a significant recommendation was omitted.