Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
5 ISMS AND U.S. NATIONAL VERIFICATION RESEARCH AND DEVELOPMENT INFRASTRUCTURE In this chapter, the third charge is considered, which concerns the research and development infrastructure required to support the proposed ISMS. This section presumes that the data quality and data access issues discussed in the preceding chapters will be adequately addressed; it focuses on the research and development infrastructure and knowledge transfer that wait provide technical support for the ISMS and U.S. CTBT monitoring operations. Monitoring a CTBT poses many unprecedented technical and scientific challenges, and there will be a continuing need for basic and applied research and advanced technology and automated systems development in all of the disciplines that contribute to the monitoring system (OTA, ~ 988~. Given the imminent implementation of a prototype CTBT monitoring system, it is critical to have an integrated and reviewed program that carries out basic research, tests the results in operational settings, and implements useful, cost-effective advances in the operational system. This holds for both the ISMS and the U.S. monitoring systems. While CTBT monitoring is intrinsically an arms-control issue, primary responsibility for seismological nuclear test monitoring and research has historically resided within the Department of Defense (DOD) and the Department of Energy (DOE). Given the current climate of shifting organizational roles, it is not clear which government organization is ultimately most suitable for overseeing CTBT monitoring efforts. it is assumed that the mission for monitoring research will continue to reside within DOD and DOE, with supporting activities by the USGS and seismological research community. it is important that if these agency roles change, the basic seismological research effort be maintained by those responsible for the functions of monitoring, verification, and hazard reporting. This chapter describes basic mechanisms 53
54 Comprehensive Test Ban Monitoring System by which participating agencies can implement an effective research and development program in seismology servicing the operational needs ofthe CTBT monitoring effort. The actions recommended below will ensure that an integrated research and development system consisting of basic, applied and advanced development elements supports the CTBT monitoring efforts of both the TSMS and the U.S. nuclear monitoring systems. More effective coordination of the overall research program and more efficient transfer of technological advances into the operational regime will result from implementation of these recommendations. They will also support the development of personnel with appropriate expertise and capabilities. Introduction and Background The panel was asked to consider the following charge: Research Feedback. An important aspect ofthe GSE concept is that the system can evolve. This includes regular improvement of the processing capabilities (e.g., travel-time and amplitude path corrections, enhancement of phase identification and event location, and new processing techniques). What is the best way to implement promising basic and applied seismic research within the GSE system? To what stage must research be taken (e.g., publication, algorithms, or finished software) to most expeditiously and reliably implement it within the system? What are the long-term national research and development programs required to support the envisaged monitoring system? The pane! issued a broad request for feedback on research and development infrastructure issues. The responses highlight several weaknesses ofthe past and present infrastructure related to research and development for seismological monitoring of nuclear testing treaties. A wide range of issues surfaced, involving program guidance, funding issues, need for a research teethed, and basic structural problems within participating agencies. It also became clear that the issue of research in support of the GSE system or ISMS system is only part of the broader issue of how seismological research should be organized to service the U.S. national verification effort, particularly since it is not clear that the ISMS will have any event identification responsibilities. We now provide some background on the existing research and development infrastructure. The research and development efforts in seismology that support nuclear test monitoring date back to the 1959 Berkner Pane] report "The Need for Fundamental Research in Seismology". That report provided a rationale for fundamental research in seismology as an integral part of any successful treaty monitoring system. Many of the original arguments remain valid today. While the intervening 35 years have brought
VERIFICATION R&D INFRASTRUCTURE 55 great advances in our knowledge of earth structure, global seismicity, seismic wave propagation, and characteristics of nuclear explosion and earthquake signals, the technical requirements for monitoring a sequence of nuclear testing treaties (the ~ 963 Limited Test-Ban Treaty, the 1968 Treaty on the Non-Proliferation of Nuclear Weapons, the 1974 Threshold Test-Ban Treaty, and the 1976 Peaceful Nuclear Explosion Treaty) have kept pace with, and even temporarily exceeded the seismological capabilities. We are on the threshold of another quantum jump in the required monitoring capabilities, as negotiations progress toward eventual signing and entry into force of a Comprehensive Test-Ban Treaty. Many technical challenges confront the new CTBT monitoring effort, and continued seismological research is essential to ensure adequate U.S. national verification capabilities. Effective verification of a CTBT will require detection, location, and identification efforts using regional and teleseismic data from sources and stations distributed worldwide. The problems to be faced involve critical issues such as how to distinguish the small seismic vibrations from large quarry blasts from the vibrations produced by a small nuclear explosion. Given the many regions ofthe world for which there is little or no familiarity with the crustal effects on seismic waves, the accuracy of event locations and the confidence in event identifications will be unacceptable until research efforts calibrate each region. These efforts will require seismologically trained personnel, monitoring systems capable of processing large numbers of events, and new algorithms for detection, location, and identification tuned to specific regions of Earth. Availability of the personnel, algorithms, databases and systems that are needed requires a program that trains personnel and provides an orderly transition for concepts and algorithms from their conception to implementation in the operational systems. An integrated program that supports a continuum of efforts from basic research to the operating system will provide the United States with effective CTBT verification capabilities. The continuum includes the following categories. · Basic research efforts that tend to be focused in the universities. These efforts train personnel, develop new theories and relevant concepts, carry out investigations in areas of interest, and provide useful data. Long-term research efforts on basic seismic wave propagation, source theory, and techniques to extract information from waveforms are concentrated in the basic research area. . Applied research efforts that tend to be concentrated in the efforts of the private contractors, with some contributions from universities. These efforts look to both basic research and operational needs to determine their priorities. They develop the concepts identified by the basic research into functional algorithms and expand and organize the data bases begun there. They interact with the operating systems to develop
56 Comprehensive Test Ban Monitoring System algorithms and carry out studies to acidress specific problems encountered by the operating system in the course of its day-to-day functions. Finally, they develop concepts for the operating system and develop elements that can be incorporated into it. The contractors are a major part of the employment environment that provides job opportunities for the students and post-doctoral fellows who have gained experience in the basic research efforts. Sustaining this job market should help to attract university researchers to this area. . Advanced development research that tends to be concentrated in the contractors and the operating agencies. These efforts develop new operating systems that integrate hardware and software for data acquisition, communications, data archives, processing, analysis, and display. Although past efforts spanned these functions, the current demand for functional CTBT monitoring systems tends to stress advanced development and applied research, and most research funding should be oriented in these areas. However, the long-term challenges of CTBT monitoring wall require the development of new detection, location, and identification concepts and the availability of highly trained personnel. These assets will only be available if the research program includes a strong basic research element as part of the overall integrated program. Given the fact that many of the scientists trained in the seismic research programs of the 1960s are nearing retirement age, the pane! is concerned that serious degradation of the basic research program will result in a lack of qualified, knowledgeable verification seismologists at the time that the CTBT enters into force. The pane! acknowledges the fiscal considerations but is vehement in its support for the continuing need for a strong research effort. Over the past several decades' the Department of Defense (DOD) has held the primary role in supporting basic seismic research and development for nuclear monitoring. Actual monitoring operations have been conducted by the Air Force Technical Applications Center (AFTAC). It is essential to sustain research, development, and transfer to the operations within DOD. In principle, the DOD nuclear monitoring research program supporting this operation has a standard DOD subdivision into basic research (the so-called 6. ~ program), exploratory development research (6.2 program), and advanced development research (6.3 program) components. The DOD research effort is currently organized under the Air Force Office of Scientific Research (AFOSR) (6. i), the Air Force Phillips Laboratory (AFPL) (6.2), the Air Force Technical Applications Center (AFTAC) (6.3), and the Advanced Research Projects Agency (ARPA). in practice, AFOSR and ARPA have, at various times, had dominant cross-cutting roles that have combined the 6. ~ and 6.2 functions.
VERIFICATION R&D INFRASTRUCTURE 57 The DOD 6.1 basic research program, currently administered by AFOSR, has an annual funding level of about $4 million for external research. This program supports fundamental relevant research investigations, with a current emphasis on small-event discrimination. The 6.2 exploratory development program has the goals of identifying and extending promising research results from the 6.l (AFOSR) program and transitioning these to AFTAC, the operational client. The 6.2 exploratory development program was sponsored by ARPA in the past (with administrative support primarily through AFPL), but as of FY95, the responsibility for this effort was transferred to DOE. Currently, a scaled-back 6.2 program resides at AFPL, with no direct funding for external support. In ~ 995, AFPL began to administer some of the exploratory develop- ment projects funded by AFTAC and DOE. A 6.3 program has been administered by AFTAC and is intended to develop advanced systems and mature technologies for the operational environment. For the next two years, AFTAC will be supporting about $3.2 million/yr of external research Finding for efforts in the 6.2 area, with a corresponding reduction ofthe normal 6.3 effort that AFTAC would support. The total AFTAC CTBT verification budget for operations, research and development in seismology is $24 million for FY95 (DOD, ~ 994~. In the past, ARPA was a primary supporter of all levels of seismological research, with an increasing emphasis over the last 10 years or so on applied and advanced developmental research. ARPA has been involved in developing seismic arrays for use in nuclear monitoring, but starting in FY95, responsibility for deploying and operating new arrays was transferred to AFTAC. Currently, ARPA's seismological effort is focused on development of a prototype ISMS International Data Center, with a total ARPA CTBT verification research and operations budget of $13.8 million for FY94 (DOD, ~ 9944. Over the next two years, ARPA may provide some funding to the 6.2 research program administered by AFPL, using funds provided to bridge a phase-out of ARPA research and development in nuclear monitoring seismology. Beginning in FY95, the Department of Energy (DOE) was assigned the mission "to carry out research and development necessary to provide U.S. government agencies responsible for monitoring andior verifying compliance with a CTBT with technologies, algorithms, hardware, and software for integrated systems to detect, locate, identify, and characterize nuclear explosions at the thresholds and confidence levels that meet U.S. requirements in a cost-effective manner" (DOE, 1994~. Much of this mission had previously resided with ARPA, but the establishment of the DOE program was not intended to preclude continued DOD research efforts, and indeed most of DOD's 6. I, 6.2, 6.3 and DOE's programs are now closely coordinated. The external funding provided by the DOE program is about $4 million/yr, with the program (DOE, ~ 994) being strongly oriented along the lines of exploratory and advanced development research. At DOE's request, AFPL administers the DOE external grants in this program.
58 Comprehensive Test Ban Monitoring System The annual DOE budget for CTBT monitoring research in areas other than space is about $24 million (DOD, 1994~. This program builds on long-standing internal research programs on test-ban verification at the DOE national laboratories. It is obvious that DOD and DOE agencies engaged in CTBT monitoring activities will do their work with an emphasis on the U.S. needs to monitor the rest of the world. A somewhat separate issue is the importance of U.S. help in building up international programs for CTBT monitoring as well. This activity is of specific interest to policy agencies such as the U.S. Department of State and the U.S. Arms Control and Disarmament Agency, which is in charge of coordination with other U.S. agencies and with the conduct of the negotiations of the Conference on Disarmament in Geneva. The National Science Foundation (NSF) and the U. S. Geological Survey (USGS) fund basic research and instrument deployments in seismology that have substantial potential relevance to the nuclear monitoring arena, but these agencies do not have explicit missions regarding nuclear monitoring operations Development of earth models, basic research on earthquake source physics, and studies of regional tectonics are a few of their areas that intersect the nuclear monitoring arena. There are plans to incorporate many ofthe NSF/USGS seismic stations in the ISMS, as well. The USGS is helping the U.S. ISMS-NDC to provide data from stations located in the United States to the TSMS-~. The United States must have the capability of prompt and detailed reporting for seismic sources on U.S. territory (earthquakes, routine and unusual mine blasting to the extent that the consequent signals are detected by the ISMS-~, accidents With explosions, and large-scale explosions in military programs). The USGS is the existing U.S. agency best able to document seismic activity in U.S. territory, as its seismological monitoring effort involves a great number of stations in North America. The U.S. Bureau of Mines may also help in this effort. The research efforts mentioned above have had varying degrees of coordination over time. For 17 years AFPL, AFOSR, and ARPA programs have held an annual Seismic Research Symposium that has brought together the basic and applied research communities. These are distributed across universities, private companies, and federal agencies. In ~ 994, the AFPL and AFOSR Seismic Research Symposium and a meeting of ARPA contractors were held separately. in ~ 995 ARPA again organized a separate meeting in May for its contractors and European leaders of the GSE, while AFTAC, AFPL, AFOSR and DOE met jointly in September. These separate meetings serve to exacerbate the growing rift between the basic and applied programs. There is a general perception that communications between the operations and research communities have not been effective. A Seismic Review Pane] provides advice to AFTAC, but is not charged with coordinating the research needs of the operational environment and the basic research programs. The challenge of monitoring a CTBT will heighten the need
VERIFICATION R&D INFRASTRUCTURE 59 for efficient development and transfer oftechnologies from basic research to operational capability. The technical difficulties and research program strategies for monitoring a CTBT have been thoroughly addressed in several recent documents (AFTAC, 1994; Blandford et al., 1992, DOD, 1994; DOE, 1994; van der Vink et al., 19941. Appendix B tabulates some of the seismological research efforts that are being developed and conducted in the DOE research program. In brief, the major change from past monitoring efforts is that CTBT monitoring will require global identification of events down to low seismic magnitudes, whereas previous Threshold Test-Ban Treaty monitoring emphasized yield estimation for large explosions at Soviet and Chinese test sites. The need to detect, locate, and identify relatively low-magnitude events mandates the use of short-period seismic wave energy recorded at arrays and the use of high- frequency data recorded at regional (< ~ 200 km) distances. Although research on regional seismic wave propagation is a rapidly developing area, fundamental issues remain regarding event discrimination with regional signals that are the topic of many current investigations. A direct consequence of the need to analyze small-event signals is that very large numbers of events, perhaps 1 00 to 300, must be analyzed on a daily basis. This requires enhanced automation and computer assisted decision-making in the operational environment. Nuclear tests must be discriminated from large quarry blasts, earthquakes, and rock bursts. The number and characteristics of these events pose major challenges to seismic analysis procedures. The very mature methodologies of yield estimation for large explosions developed for monitoring the Threshold Test-Ban Treaty are not directly relevant to monitoring a CTBT. The technical challenges are such that continued basic and applied research is vital to the long-term success of a CTBT monitoring effort. U.S. Research and Development Infrastructure The effective development and transfer of new seismological advances into the CTBT monitoring operation require a well-coordinated, effective research and development infrastructure. We will first focus on the large-scare characteristics of the necessary infrastructure, and then provide specific suggestions as to how to achieve effective research and development support ofthe operational effort. The exact structure of the U.S. TSMS National Data Center is not yet worked out. For the purposes of this report we assume that the primary responsibility for U.S. nuclear test monitoring operations will continue to reside with the Air Force, while DOE will continue to have extensive seismic and nonseismic research and development activities relevant to CTBT
60 Comprehensive Test Ban Monitoring System monitoring and the USGS will assist in providing data on seismic sources on U.S. territory. Although, the pane] recommendations are given in the context of continued DOD and DOE involvement, the basic rationale underlying the recommendations is not agency specific. in addition, the recommendations of this section are not restricted to the ISMS, but are perhaps even more relevant to the U.S. monitoring system. In order to meet the challenges of CTBT verification, it is critical to establish stable and sufficient levels of funding for the overall CTBT monitoring research effort. Upheavals and annual crises in the DOD funding sources (ARPA, AFOSR, AFPL, and AFTAC) over the past 5 years have reduced the number of Ph.D. students in seismology who are working on relevant problems and have caused many leading university seismologists and consulting groups to back away from the program. While the present situation is somewhat more stable, continued involvement of the university anc! contractor resources requires that the agencies (DOD and DOE) funding research in this area commit to funding relevant efforts for clearly defined times. This commitment is vital for the training of a new generation of seismologists committed to nuclear monitoring research and operations, as well as providing employment opportunities in the contracting companies. It will also provide the entry point for new computer technologies, largely driven by university research, to be brought to bear on treaty monitoring applications. These will be important for CTBT monitoring in a resource-limited environment. The various agencies that are contributing to CTBT monitoring efforts all recognize the technical demands that these efforts will place on seismological capabilities. While seismological methods are quite advanced in general, development of the necessary global understanding of regional seismic wave propagation, the fundamental nature of regional wave discriminants, and regional earthquake and quarry blast source characteristics requires resolution of fundamental research issues. While there must be a substantial emphasis in the overall research program on applied research and advanced systems development, there will be a continuing, probably long-term need for a fundamental research program relevant to the nuclear monitoring effort. This will ensure that technological and theoretical advances in the rapidly progressing field of seismology are brought to bear on nuclear monitoring issues. Operations under a CTBT will require highly trained analysts and decision makers. The universities can make important contributions both in training and in appropriate research. The contracting companies that participate in the applied research and advanced systems development programs will provide one ofthe major job markets for university-trained seismologists. AssurIiing that DOD will continue to have a major operational CTBT monitoring function, the pane! feels that it is critical that DOD sustain a research and development program in seismology with an integrated basic, applied In . in. . , and advanced development extort. -i As program should be based on the standard DOD hierarchical structure of 6. ~ L 3
VERIFICATION R&D INFRASTRUCTURE 61 6.2, and 6.3 research efforts, each of which has a natural home within the existing Air Force structure. Any departure from the recognized DOD hierarchy weakens the overall effort and can lead to instability when uncoordinated actions are taken in different units. The DOE research and development program and the DOD effort should be closely coordinated to avoid redundancy. The coexistence of DOD and DOE research and development programs will best exploit the distributed expertise within the agencies, and as long as there is effective coordination, the combined effort should service the monitoring needs. The existing DOD basic research (6.~) program administered by AFOSR contributes fundamental research in regional wave propagation, event location, and source discrimination. In the short-term it would be valuable to augment this program to enable additional exploration of the basic physics underlying the regional seismic wave discriminants that are being proposed for the operational environment. Long-term stability of this program should be established in recognition of the heightened challenges of CTBT monitoring and the difficulties of anticipating what research directions will advance operational capabilities. (Applied research programs tend to lack the innovation and breakthrough discoveries that are common in basic research efforts.) However, there is concern within the operational effort at AFTAC regarding the effectiveness of the 6.1 program. it appears that there is a need to enhance communications across the DOD research hierarchy to ensure that the basic research program does emphasize the problems relevant to the operations. The panel recommends that the basic research 66.~) program in seismology, currently administered by theAFOSR, should be sustained and expanded to maintain an influx of researchers andfundamental research on long-term problems associated with seismological monitoring of a CTBT. Within the standard DOD framework for research, the promising developments of a basic research (6. ~ ) program are farther developed in an exploratory development, or applied research (6.2), program. Typically, this involves a greater percentage of private company contractors than university contractors. The present situation for DOD seismic research is not optimal, in that AFPL, the Air Force branch explicitly identified as having responsibility for applied seismology research, has no funding of its own for external research in this area. This is a consequence of the past few years of unsettled and politicized budgets. External funding in the 6.2 program of $2 million/yr has evaporated. In the standard DOD model, AFPL would conduct research that directly follows up on the AFOSR basic research program. At present' AFPL does administer contracts for directed research efforts sponsored by DOE' AFTAC' and (ephemerally) ARPA, but these are not designed to ensure effective technology transfer Mom the
62 Comprehensive Test Ban Monitoring System AFOSR program, as the funding sources are quite focused on enhancing short-term operational capabilities. While the existing directed research effort appears to be functioning very well, the AFOSR program is undesirably isolated from the operational environment. It is possible to improve the current situation by providing stable funds for external research to AFPL, with the explicit charge of effectuating the desired technology transfer, as long as strong communications are put in place to bridge the diverse components ofthe DOD research effort. The 6.2 funding should be coordinated with and should complement the existing directed research programs sponsored by DOE and AFTAC. The establishment of stable 6.2 funding would remove the need for AFTAC to divert some of its advanced development resources, as is occurring at present, and a stronger 6.3 program could be established. The pane! recommends that theAir Force exploratory development research 66.29 program in seismology, currently administered by the AFPL, should be provided with a stable base for externalfunding to enable effective development and transfer of promising research and lechnologiesirom theAir Force basic research program to fAeAir Force operational environment. AFPL should continue to administer file external DOE program, as participation in this join' effort pro vices a natural mecha- nism for the various elements to coordinate their research effort The CTBT monitoring emphasis on regional waves in diverse areas of the world requires extensive application of basic seismological techniques to characterize wave propagation and source characteristics in many regions. Sufficient funding to conduct field studies and calibration efforts to perform systematic regional characterization, to pursue new basic research developments, and to integrate the results into a standard format and/or database should be provided to the applied research effort. AFTAC has supported advanced development research (6.3) for many years, and this continues to make excellent sense as long as AFTAC is the operational arm of the nuclear monitoring effort. The pane' recommends that the advanced development research 663) programs in seismology, currently administered by the Air Force Technical Applications Center, should be sustained. The large seismic databases to be collected by the TSMS and the requirement of analyzing vast numbers of small events creates large computational demands for the CTBT monitoring system. For several years, ARPA has overseen the development of advanced computational platforms that can efficiently perform the data analysis. This effort should be sustained. Exploration of new computer technologies and intelligent computer systems is within the purview of ARPA and is an element of the DOE
VERIFICATION R&D INFRASTRUCTURE 63 program. it is clear that communications technologies are going to continue to advance at a remarkable rate and that entirely new monitoring capabilities will emerge in the future. A natural, and important, role for ARPA is to be looking forward to the next generation monitoring system, once it is no longer operating the prototype TDC. New strategies for event location and identification will be enabled as the number of seismic data streams increases, and it is realistic to imagine real-time access to hundreds, or perhaps thousands, of seismic stations from global regional networks as well as arrays and broadband stations. The communications, data processing, and data archival challenges will be significant. Meeting them will require advanced computer technologies. The pane! recommends that the clevelopment of the prototype ·SMSInterna- tional Data Center, currently being performed by ARPA, should be sustained, the exploration of new computer technologies and intelligent computer systems by ARPA an`DOE be coordinated, and the results incorporated in the US monitoring system, to the extent appropriate One of the major challenges in the DOD research structure is to ensure that it actually functions, with relevant projects funded under the 6. ~ program being picked up by the 6.2 program, and, if promising, passed to the 6.3 program. At present no formal overview of all levels exists. Conventional mechanisms for research documentation, such as reports and meeting presentations need to be reconsidered to enhance the technology transfer from one program to the next. Below, we discuss the concept of a research test bed, which might provide a more effective means for moving research developments into the operational system. The pane! recommends that the Air Force basic research 66 ]), exploratory development 66 29, ar'd advanced development 66 3) programs should he coordinated by a group or organization that is aware of the monitoring requirements, the operational needs, and current and past research efforts In response to its recently acquired responsibility for providing research and development efforts to support U.S. monitoring agencies, the DOE has developed a detailed, focused research plan (DOE, 1994) for its internal and external CTBT monitoring research program. This program draws upon the DOE national laboratories and supports applied research and some advanced development efforts at universities and in the private sector. It provides limited support to basic research. DOE and AFTAC are engaged in a continuing exchange in order to minimize impediments to technology transfer across agency boundaries. This interaction needs to be maintained and expanded. The program appears to be functioning well and has an appropriate level of support.
64 Comprehensive Test Ban Monitoring System The pane! recommends that the directed research and, advanced development research programs in seismology currently administered by the Department of Energy should be sustained. A knowledgeahie, responsible advisory mechanism should oversee the combined DOD/DOE research effort to ensure relevance and continued coordination of the programs. Given a stable funding structure and a closely coordinated basic, applied, and advanced development research program, it is still of major importance to that researchers be well informed of the concerns of the monitoring environment. This has not been achieved very successfully in the past. There have been, and will continue to be, issues associated with restricted access to classified data sources and procedures. For example, while the TSMS will be an open system, with unclassified data, the U.S. monitoring effort watt involve additional classified data sources. Private companies have commonly performed classified research for ARPA and AFTAC, and it is appropriate for such activities to continue as needed. Universities, and many private companies, perform unclassified research. It is important that classification not be an unnecessary impediment to performance of relevant research. To the extent consistent with national security issues, the research community funded by DOD and DOE programs should be provided with information about the operational methodologies and the operating system, along with access to relevant data (such as the ISMS data). This is critical for focusing the research efforts on relevant issues. Where security issues intervene on critical topics, appropriate clearances should be provided to promising researchers. Feedback from the operational system to the research environment could improve relevant problem solving. For example, providing lists of problem events and an indication ofthe nature ofthe ambiguity could focus research on the most important operational problems. Other forms of outreach, such as newsletters from the monitoring community and site visits to both funded and unfunded seismic research universities and companies, could focus research attention on relevant problems. Clearly, national . security concerns impose some limits on communication, but there is no question that communications can be improved. The annual Seismic Research Symposium provides a good opportunity for researchers, program managers, and operations personnel to define and discuss the specific monitoring problem that each applied research effort addresses, its context, and long-term concerns that might be addressed by basic research. The fragmentation of this symposium in ~ 994 and ~ 995 should be reversed, as this undermines communication across the participating research levels. Continuing the current Broad Agency Announcement (BAA) process is important to attracting a wide range of researchers and new technological approaches to the nuclear monitoring programs.
VERIFICATION R&D INFRASTRUCTURE 65 A valuable feedback to the monitoring effort would be a systematic comparison of results from the U.S. monitoring system with the results of earthquake monitoring operations. This would involve comparison of event catalogs and the associated location and size estimates. Earthquake monitoring information such as fault mechanisms and source characteristics may also be useful for calibration efforts in the monitoring operation. This feedback is an integral part of assessing the system performance and is a means for validating the operational procedures. Numerous organizations have contributed, and will continue to contribute, to our understanding of regional seismicity independent of the nuclear monitoring community. The results of these efforts should be exploited in nuclear monitoring activities. The panel recommends that improved communication between the DOD operational environment and researchers in the basic and applied programs be fostered. Release of i~zformalior, about operational methodologies and procedures, lists of problem events, and comparisons of seismic bulletins from differer~t communities are among the activities that could enhance responsiveness of the research community to the operational requirements. Communicatio'2 across the various elements of the monitoring and research communities should be fostered. Symposia, site visits, and advisory panels should be part of this communications effort. Focused experiments, involving broad communities, should be conducted to concentrate effort on important issues. One of the most difficult aspects of all research program infrastructures is the interface with the operational environment. This holds equally true for the ]:SMS and the U.S. monitoring operations. As research ideas develop and mature in the basic and applied research programs, it is critical that they be tailored to the operational needs. This has been performed inefficiently in the past, primarily by AFTAC assigning some of its limited number of internal personnel to implement technologies emerging from the ARPA and AFOSR programs. Given the advanced computer environment required for the CTBT monitoring effort, it is important for researchers to be able to develop and test their research products in an unclassified test system. To the extent possible, consistent with national security considerations' a readily accessible test system should have the same (unclassified) data bases, data streams, and operating system as the operational environment. Thorough documentation of the software interfaces for the system should be provided so that new procedures can be interfaced and examined under realistic conditions. The establishment of a test bed system should be accompanied by a ground-truth data base for assessing the performance of new methods and for deciding which approaches should actually be incorporated into the operating system. Research intended for the operational system should be documented' preferably with open
66 Comprehensive Test Ban Monitoring System publications, and the statistical perfonnance on the grour~-truth data base should be established. This evaluation should include information about false alarms and the probabilities of missed violations. Software that has been exercised in the test bed should be readily exported to the operational environment once it has been thoroughly validated. A test bed ofthis type will be an essential element ofthe effort to incorporate the most useful developments from the basic and applied research efforts. It will free the operations personnel from maintaining the system, while providing a platform for them to interact with the external research program. The pane! recommends that, to the extent possible, consistent with national security considerations, an unclassif ed experimental test bedfacility that replicates the basic U.S. and ISMS analysis procedures be established to enable new deve/top- ments to be tested in a realistic environment, enhancing transfer of applied" research results into the operational systems. Ground-truth data bases should be included to determine performance of new methods. Regionalized data bases relevant to detection and discrimination efforts should be establish ec' and made available to the research programs through the test bedfacili~. A special event data base should be maintained and made available to all researchers to eliminate laborious "rediscovery" of valuable data. This data base could be organized on a regional basis along the lines of the research program regionalization. This would include data for historical explosions, ground-truth events, and compilations of useful signals assembled by contractors in the course of their funded research. This data base could be made available through the test bed described above, or through an agency capable of maintaining and providing access to an expanding data base of this type. AFTAC may find it useful to maintain a data base with cIassified/restricted data for access by researchers with clearances. The pane! recommenars that a research data base of important seismic recordings be assembled and maintained. There are many research problems in seismology that are relevant to earthquake monitoring as well as nuclear monitoring operations. These include event catalog determination, with attendant technical issues such as improved association algorithms, multiple-phase location procedures, and event location in heterogeneous models. interagency working groups bridging the earthquake and nuclear test monitoring agencies should help to coordinate and foster research with dual applications. This could help to activate relevant research activities in-refined catalog preparation. The pane! recommends that major research efforts that have potential benefit Is for teeth nuclear test and earthquake monitoring, such as enhanced" association algorithms, new regional event locatiorl procedures, and event location procedures
VERIFICATION R&D INFRASTRUCTURE 67 in three-dimensional models, should lee coordinated through interagency working groups. The panel recommends that a program that supports posidocloralfellows and visiting researchers should be established at both the International Data Center and the U.S. National Data Center. This will enhance communication and research coordination between the operational and research environments. The primary focus of the above recommendations has been the U.S. research program, but there is clearly valuable research conducted internationally as well. Historically, a significant portion of this has actually been funded through the ARPA program, and maintaining that program will sustain some of the international effort. Annual seismological symposia have provided a useful forum for interaction between U.S. researchers and their international colleagues, and that would be sustained by the foregoing recommendations. Presumably, some international research activity will be associated with different national data centers, and technological advances will feed back to the TSMS via those centers. International collaborations on field deployments that characterize regional crustal structures and seismic wave-propagation characteristics in different regions provide a natural means of mutually enhancing treaty verification capabilities, and such efforts should be supported by the DOD 6.2 and DOE programs. This chapter has outlined a Tong-term research and development infrastructure to support the operational system that will monitor a CTBT. A balanced, well-organized program involving relevant basic research, applied research, and advanced development research is the logical approach. A guiding principle for a stable research effort is that the program be optimized within existing agency hierarchies, avoiding any gaps in the structure that may impede technology transfer. To ensure that relevant research is conducted and brought to operational capabilities, enhanced lines of feedback from the operational system and knowledgeable advisory panels need to be established, along with a realistic research test bed with ground-truth and relevant data bases. Implementing the recommendations of this chapter will ensure that CTBT monitoring efforts continue to have the critical influx of research innovations and technical developments vital to an effective monitoring operation. It should also be noted that many international research activities on nuclear monitoring problems are directly supported by the DOD. While this chapter has emphasized the needs of the U.S. program, the international efforts will be similarly strengthened if the recommendations are implemented.
