astrophysics. NASA and NSF should continue to improve cooperation to ensure that critical areas are not neglected and to maximize the nation's scientific return from areas of overlapping agency interest.

The NRC report Supporting Research and Data Analysis in NASA's Science Programs: Engines for Innovation and Synthesis (National Academy Press, Washington, D.C., 1998) highlighted another problem at NASA. NASA currently groups data analysis and missions operations, two very different categories, into a single budget item. Mission operations (MO) involve a team of engineers maintaining and operating a satellite. Data analysis (DA) funds support scientists using the mission data to answer scientific questions. Because the MO and DA numbers are commingled in the NASA budget, the committee found it impossible to track the level of science support at NASA. The committee strongly encourages NASA to separate these two items. Treating the two separately would enable both NASA and outside groups to better evaluate and optimize programs.

One possible path for record keeping might be to use the system described in this report to categorize grants by field and discipline and by the overall categories of instrumentation, technology development, operations, individual investigator research, and construction. Agency databases with annualized grant figures properly categorized not only would help those trying to track changes in the field but also could supply input for broader science policy decision making. In this context, the committee also urges scientific societies to track membership demographics by collecting information directly from their members on discipline, field, and employment status and location.

In compiling the data for this report, the committee came across several areas, such as the study of the Shoemaker-Levy Jupiter collision, in which NASA and NSF cooperated on interesting and significant programs. NASA, DOE, and NSF should continue to improve such cooperation to ensure that critical areas are not neglected and to maximize the nation's scientific return from areas of overlapping agency interest. This includes intra-agency cooperation as well.