Research Required to Support Comprehensive Nuclear Test Ban Treaty Monitoring (1997)

Pursuing the research described in the previous chapter will require a sustained federal effort, using the intellectual and technical resources of universities, private consulting companies, National Laboratories, and operational agencies. If the United States is to achieve the stated goal to confidently monitor evasively tested nuclear explosions with a yield of a few kilotons, additional fundamental and directed research will be needed. Simply deploying sensors and bringing the IMS and NTM into operation cannot guarantee technically effective monitoring of the treaty. Further research will enhance the reliability and performance of the CTBT monitoring effort and should lead to cost reductions as the dependence on human analysts decreases and erroneous identifications and On-Site Inspection requests are avoided. The technical knowledge resulting from this research program, from its integration with the efforts of the U.S. NDC and its connections with other agencies and organizations carrying out related research, will also be valuable for addressing claims of CTBT violations made by other nations based on IMS data. In addition, it will benefit other programs of U.S. national interest.

To illustrate the programmatic challenge of the research effort, this chapter describes past and present research programs in support of test ban monitoring. First the chapter establishes the current baseline for support of basic research programs related to CTBT monitoring. The panel then concludes that the current levels are small compared to the needs for supporting the priority research topics in Chapter 3. Consequently, increased funding for basic research will be required to support U.S. efforts to meet national monitoring goals. This chapter also discusses the panel's view of the characteristics of an effective long-term research program and the mechanisms to transition research results to an operational environment. Additional details about the past seismic research programs are provided in Appendix B.

Sections 2.6 and 2.7 have provided an overview of the research structure that is presently in place to support CTBT monitoring. Of particular importance for assessment of the future program structure is the reorganization in DoD programs that took place in 1996 (see Section 2.7). The consolidation of separate lines of funding for CTBT-related research that had existed in the Air Force

Office of Scientific Research (AFOSR), Air Force Phillips Laboratory (AFPL), and ARPA research budgets eliminated the DoD distinctions between basic, applied, and development research programs (i.e., the former 6.1/6.2/6.3 designations). Thus, the programmatic structure and balance to the research effort has been removed and a new approach must be established. The new system should service the immediate operational needs of the NDC and provide the mid- and long-term improvements necessary to maintain confidence in the verifiability of the CTBT. It must do so in the face of changing world situations and a background of numerous nonnuclear events with characteristics similar to nuclear explosions. Meeting these challenges requires a properly organized and managed research program that includes fundamental research conducted by universities, applied and developmental research conducted by private companies, and advanced systems development efforts provided by private companies. These efforts must be coordinated with the efforts and needs of other agencies involved in CTBT monitoring and still other agencies and organizations that use similar technologies.

The FY97 Department of Defense (DoD) external research program is presently funded at a level of $8.8M. In FY 1997 DOE will provide about $0.4M for external research.²² The DoD FY97 budget for development of the IDC system is about $20M, which will need to be sustained at least until the IMS system is deployed. The funding level for basic and applied research for CTBT monitoring from all sources was around $12 million in FY 1995 and FY 1996.²³ If DOE support of external CTBT-related fundamental research decreases as indicated in its current budget projections, the remaining $8.8 million for DoD programs represents about a 27 per cent reduction of the research effort, despite the expanded role of multiple monitoring technologies and the time frame established by the signing of the CTBT in 1996. (At this time, it is difficult to project DOE's effort into the future, largely due to turnovers in management personnel; however, no explicit commitment to sustaining an external research program anywhere near the scale of the FY 1995 and FY 1996 programs has been made.)

Past programs have supported a broad range of research that should be sustained to enhance U.S. CTBT monitoring capabilities. Examples of basic and applied research projects that were supported by AFOSR, AFPL, AFTAC, ARPA, and DOE in FY 1995 and 1996 are summarized in Lewkowicz et al. (1996). These papers describe studies of detection, location, and identification of seismic sources; hydroacoustic and infrasonic wave propagation; radionuclide monitoring; and data processing and analysis. Given the progress to date, it is clear that the challenge of effective CTBT verification requires expansion of this type of research to support effective monitoring of all testing environments (other than space) in the near, mid-, and long-terms.

If the United States is to meet its stated monitoring goals in a timely manner, substantially higher funding than the combined DoD-DOE budget of $12 million/yr for the previous two years will be needed to support the high-priority research issues listed in Chapter 3. The panel's conclusion on the need for increased funding levels is based on the following factors: 1) the geographic breadth of the areas of monitoring interest, 2) the expanded range of monitoring technologies that will be incorporated into the IMS, 3) the lack of previous monitoring experience and global data for some of the new monitoring technologies, 4) the need to carry out new fundamental research to explore the synergies between monitoring technologies, and 5) the importance of basic research programs for training technically competent scientists who will participate in U.S. monitoring operations.

The basic research program being considered is separate from the systems development effort directed at setting up the IDC and also from internal DOE and DoD budgets that support both applied and advanced developmental research. Some proposed undertakings, such as a long-term committed seismological effort to developing a detailed model of the crust and upper-mantle velocity structure under the major land masses of

interest early in the twenty-first century, would require higher funding levels but could result in optimal operational capabilities at that time. Similarly, development of portable rapid radionuclide processing systems could be undertaken if it is decided that this capability is needed, which would also require additional funding levels. Some large-scale field experiments involving controlled explosions and field recordings would likewise require more substantial budgets.

The eventual funding level for R&D programs in CTBT monitoring presumably must be determined after assessments by the Administration and by Congress of the importance of reaching U.S. monitoring goals. The linkage between R&D and operational improvements in monitoring is explicitly recognized in President Clinton's statement of August 11, 1995. Given the CTBT safeguards, it is clear that, at a minimum, current levels of support should be sustained. However, the discussion in Chapter 3 suggests that incremental improvements will not achieve U.S. monitoring objectives, and research is the most likely activity to achieve breakthroughs in monitoring capabilities. If the capabilities of the IMS do not improve with time, there may be false alarms associated with erroneous interpretations of the monitoring data. Such events would be costly if they resulted in an OSI.

Deployment of IMS seismic, hydroacoustic, infrasonic, and radionuclide stations will expand opportunities for data analysis and research progress in a variety of applications other than CTBT monitoring that are now constrained by data availability. If IMS data become available, it will be a valuable contribution to CTBT monitoring if agencies such as NSF, USGS, and NOAA fund research related to multiuse applications of the data stream.

Several important issues relate to program balance for U.S. CTBT research. These include the relative levels of disciplinary efforts; the relative emphasis on fundamental, applied, and advanced developmental research; and mechanisms to coordinate and buffer the competition for resources that can develop between different types of research. The 1995–1997 CTBT research program is the most useful basis for projecting future needs, given that prior to 1995, relatively little emphasis in external fundamental research programs was placed on monitoring technologies other than seismology. For example, as noted above, the total DOE CTBT research program (most of which is for research at four National Laboratories) involved much more support in seismological research ($11.6 million), compared to infrasound research ($1.1 million), hydroacoustic research ($1.2 million), and radionuclide instrumentation development ($3.4 million). The $12 million per year external research programs administered by AFOSR and AFPL for FY 1995 and FY 1996, with support derived from DOE, AFOSR, ARPA, AFTAC, and ACDA, had a similar emphasis on seismology, with modest proportionally smaller efforts in fundamental infrasound, hydroacoustics, and radionuclide research.

Four factors influence the higher priority that has been assigned to seismological research in past programs: (1) seismology has been well suited for monitoring the underground nuclear testing programs of the former Soviet Union, China, and France; (2) seismic monitoring on a global basis presents a well-defined challenge that must be met if U.S. CTBT monitoring objectives are to be achieved; (3) there have been no data sets or CTBT-focused research efforts in the other disciplines, and IMS data are just beginning to accumulate and to clarify operational challenges; and (4) some of the additional technologies are complementary to other NTM. For example, for the United States, infrasound is deemed largely to be a backup system for satellite monitoring as far as atmospheric explosions are concerned. As a result, throughout the U.S. technical discussions during CTBT negotiations, hydroacoustic, radionuclide, and infrasonic monitoring challenges have been considered less profound than those for seismic monitoring. It is important to recognize that the dependence on complementary monitoring capabilities and synergies has resulted in relatively minimal capabilities for some systems (e.g., the IMS hydroacoustic network), making it essential to exploit those synergies optimally within the context of the overall capabilities.

The recommendation above for an increased funding level pertains to the external basic and applied research effort, which is only one element of the overall CTBT research effort. The FY 1997 DoD budget for development of the IDC is about

$20 million, and is clearly a high priority to sustain this effort until the IMS is operational. Some projects funded from this budget line are considered research, but it is a different "flavor" of research from the basic and applied efforts emphasized in Chapter 3. The development and implementation of communications and computer hardware and the development of automated processing software are essential for the IMS and U.S. NDC but will not solve the basic questions that arise in analyzing data from various monitoring systems. Future CTBT research must sustain this type of advanced developmental effort, but it is important to ensure that time urgency and high costs do not squeeze out the fundamental science research program on which future improvements depend so strongly. There was a tendency for this to occur in the former ARPA program, and the new DoD program will have to establish effective firewalls to prevent it. When the fundamental research program budget experiences large fluctuations there is great difficulty in sustaining steady progress on the many research fronts that exist, and the instability can cause university researchers and private companies not to commit to relevant research efforts. This instability means that the best researchers will turn away from CTBT monitoring and take up other fields.

Other research efforts now conducted within government agencies are also separate from the external fundamental research program and should be sustained. For example, the large applied and advanced developmental program of DOE, which addresses all areas of relevance to CTBT monitoring including satellite NTM and On-Site Inspection procedures, should be sustained with budget levels similar to FY 1997 since essential IMS-NTM regionalization and calibration activities are being pursued. Research on atmospheric transport models has been supported for years at NOAA and by DOE at the National Laboratories. The importance of these models for radionuclide backtracking and infrasonic monitoring motivates sustained research in this area. If possible, university research in atmospheric transport modeling should also be encouraged. Instrumentation deployment and operations efforts performed and funded by AFTAC, NSF, and the USGS are also providing essential parts of the IMS and NTM and should be funded with appropriate budget levels. It is important to recognize that government laboratory research programs tend to emphasize applied research, which—if successful—meets some identified operational needs. University and private company projects tend to be less tightly constrained by operational needs and consequently are able to explore areas of research that may have no immediate application but could lead to totally new approaches. The distinctions are not hard and fast, and both groups do research of all types.

With diverse research efforts spread across universities, private companies, and government agencies, it is important to have good coordination of CTBT research. Consolidation of DoD research programs provides an immediate opportunity to enhance such coordination within DoD (at one time it was effectively provided by ARPA). In addition, it offers an opportunity to coordinate with other government and nongovernment agencies doing research and development relevant to CTBT monitoring. DoD and DOE have memoranda of understanding, which effectively bridge the DOE research effort to AFTAC operations, but there appears to be room for improvement in this coordination as well. A better flow of information and assessment of capabilities, upward from the operational environment to the policy and intelligence arenas (ACDA, State Department, and intelligence organizations) and outward to investigators familiar with the various technologies are also important. At present there is little coordination between the explosion monitoring programs of DoD and DOE and the earthquake monitoring programs of the USGS. This is unfortunate because the signals requiring the greatest analysis efforts at the IDC will be earthquake seismograms, often derived from source areas where relevant information can be obtained via the USGS. The USGS also has a key role in the documentation of U.S. seismicity—providing additional data to interpret signals from U.S. blasting and earthquakes, that may be recorded by the IMS and be of some foreign concern. This is an example of an important benefit that would be provided by multiuse of IMS data for research purposes.

Most of the technical disciplines that contribute to CTBT monitoring also contribute to independent

activities of national concern and, therefore, have support from other government agencies. For example, seismology is the cornerstone of earthquake hazard assessment, as well as having a primary role in basic science investigations of the Earth system. Hydroacoustics has long been a mainstay of submarine monitoring and recently has emerged as a potential technology for monitoring global warning. Technical advances in these areas may often benefit CTBT monitoring (e.g., new three-dimensional velocity models or improved wave propagation modeling capabilities in the oceans and solid Earth), and the CTBT research program needs to exploit and possibly coordinate with such efforts. If the IMS data is provided to the scientific community, dual-use applications using these data streams (e.g., for studies of natural hazards) will help to sustain the long-term viability of the CTBT monitoring system. Some of the multiuse applications of seismic data were discussed in NRC (1995). There are also many opportunities for multiuse of the hydroacoustic, infrasonic, and radionuclide data.

Specific research efforts that the CTBT program could pursue in coordination with other agencies include development of improved three-dimensional Earth models for areas of CTBT monitoring interest or a focused effort to determine the gross structure of the crust and lithosphere in Eurasia and other continental areas of interest. These are topics of relevance to global earthquake monitoring and basic science investigations of the Earth supported by NSF and the USGS. There are also fundamental topics such as improved earthquake location methods that span many applications, and the coordination of national research efforts could accelerate progress in these areas and eliminate unnecessary redundancy.

Voluntary international data exchange on calibration events is one of several confidence-building activities discussed in the CTBT. Precisely located earthquakes and details about large mining explosions can greatly accelerate seismic calibration of a region, and it is desirable to pursue such exchanges of data and the compilation of associated ground truth data sets for calibration of the monitoring technologies.

The most important requirement for stability of the CTBT research program is stabilization of the research budget, with a multiyear commitment that firmly establishes its viability for intellectual resources in universities and private companies. Such stability is essential for training technically competent scientists and researchers who will participate in U.S. monitoring operations. Without it, bright young researchers will not enter the fields supportive of CTBT monitoring. To the extent possible, the research program should be buffered from fluctuations imposed by systems development and operational emphases. It should, however, include effective communication of operational needs to the research community.

It is well established that the quality of research programs is enhanced by using peer-review systems, and this is desirable for the CTBT fundamental research program as well. Although some applied and most advanced developmental activities can best be pursued with focused Requests for Proposals (with responses being assessed by government program managers), more flexible announcements for basic research funding need to be made in the external fundamental research program to ensure the influx of innovative ideas and creative approaches to established research areas. Peer review ensures a healthy program that captures cutting-edge approaches and avoids entrenchment. The lack of DSWA experience in supporting basic research programs in universities is a concern, and the lack of a clear distinction between basic and applied research complicates the maintenance of a long-term program.

Annual multidisciplinary workshops involving all of the IMS technologies would provide an important mechanism for communicating research advances and operational needs. Such meetings would also contribute to the development of synergistic monitoring strategies by promoting communication between different research communities. Publication of a comprehensive workshop report would also be a valuable part of this effort. The panel notes that similar workshops were supported by the previous AFOSR program. The last two meetings involved all of the IMS monitoring technologies.

In recent years, AFTAC has enhanced its internal technical expertise by hiring several well-trained seismologists, even when confronted with mandated reductions in manpower. This has proved effective in accelerating the incorporation of new research advances and complex analysis procedures into the operational regime of the U.S. NDC. Increasing the number of Ph.D.-level scientists at the NDC is a proven strategy for abetting technology transfer and is supported by the panel.

An additional means by which research efforts can be transitioned effectively to the operational environment would be the establishment of a CTBT research test bed. Assuming that there is open access to the IMS data, this would require a facility that replicates significant aspects of the IMS and U.S. NDC monitoring system. The facility would have real-time data processing capabilities and historical data archive access. The prototype IDC has operated a limited system of this type, with visitors to the Center for Monitoring Research accessing the IMS data and processing system, but at present there is no clear plan for a broadly accessible test bed system for the long-term. Progress on many of the research issues raised in Chapter 3 will require researchers to analyze actual signals from various monitoring disciplines and give them an opportunity to test proposed analysis methods. This analysis and testing would cultivate the development of new methods in a software environment that is much closer to the operational situation than currently exists at any university or private company, as well as providing realistic constraints on processing. A test bed facility, preferably operated by the U.S. NDC, could also serve as a site for focused investigations of problem events in which experts gather to address technical issues (either in person or by computer link up). This could be designed to be responsive to both short-term and long-term problems that arise in the operational arena. Finally, such a test bed could form the basis for regularly scheduled exchanges between the policy and technical communities, which would make clear the processes, constraints, and uncertainties under which both communities operate. One possibility would be to have the prototype IDC transition into this type of test bed facility once the permanent IDC is established in Vienna. This transition would require substantial funding beyond that described in the basic research budget above.