Silencing the Bomb Page 19
The past twenty years have seen substantial improvements in monitoring underground testing. Commercial satellite imagery is openly available now with a resolution of better than three feet (one meter). A relatively new technology called INSAR (Interferometric Synthetic Aperture Radar) measures displacements of the Earth’s surface such as those caused by earthquakes and underground explosions. Under the treaty, countries with one or more of these and other technologies, so-called national technical means (NTM), are permitted to use them to verify compliance. They can present those data in requesting an on-site inspection of a suspicious event or a possible nuclear test.
The CTBT Organization interacts with what the treaty calls “member states” through designated national organizations and data centers. Experts from the United States participate in international working groups that set policy, priorities, and technical standards for the International Monitoring System and its data center. The Air Force Technical Applications Center (AFTAC) is the lead data center for the CTBT in the United States. The U.S. National Authority—the lead federal agency for the CTBT—is yet to be designated as its resolution has been stymied by bureaucratic infighting.
Some of the following material is taken from The Comprehensive Nuclear Test Ban Treaty: Technical Issues for the United States (2012), which I helped to write.
MISCONCEPTIONS ABOUT THE ROLE OF THE INTERNATIONAL ORGANIZATION
I would like to clarify two misconceptions I have encountered regarding the CTBT Organization—first, that it is responsible for identifying seismic events as nuclear explosions, earthquakes, or chemical explosions, and second, that it should deal with evasive testing such as decoupled explosions in large underground cavities.
Neither of these is correct. The treaty specifies that the last stages in monitoring—event identification and the possible attribution of a nuclear explosion to a particular country—are the responsibilities of national authorities, not the international organization. The United States insisted on this stipulation in the treaty negotiations. The reason for it is that attribution is a political act—one country asserting or implying that a treaty violation has occurred in another. The United States and other countries do not rely on the CTBT Organization to decide if a particular event was a nuclear explosion. AFTAC’s classified capabilities are better than those of the CTBT Organization. U.S. nuclear monitoring—missions, capabilities, response times for analysis, and countries of concern—are different from those of the international organization.
The treaty and the operational manuals of the IMS specify the numbers and locations of stations, response times, and data quality for its international stations. These stations are distributed globally, without focusing on particular countries, although most of them had to be sited on land. Because existing test sites and countries of concern are located in the northern hemisphere where there is more landmass, the International Monitoring System has better coverage for those areas than, say, for southernmost South America. United States policy makers need to realize that AFTAC and other U.S. capabilities can focus both on better monitoring of regions of special concern to the United States and on global coverage.
OPERATION OF THE INTERNATIONAL MONITORING SYSTEM
Most stations of the International Monitoring System are operating now and are certified for their data quality and integrity (including tampering and data authenticity). The number of IMS stations of all five types grew from a few in 2000 to 83 percent of the full network of 321 stations in 2012 (figure 13.1). The full 100 percent of proposed installations has not been achieved because a few individual countries, such as India, have forbidden deployment of them on their territories.
The seismic network when completed will consist of fifty primary stations and 120 auxiliary stations (figure 13.1). Many of the primary stations are seismic arrays, which, unlike single stations, have the capability to determine the direction (azimuth) from which a seismic wave arrives as well as the distance to its source. Many arrays are very good at detecting small events and the seismic waves that directly follow the P wave such as pP, allowing an event’s depth to be determined.
FIGURE 13.1
Five types of observation stations of the International Monitoring System.
Source: W.-Y. Kim, personal communication.
A number of the stations of the IMS are located in places that previously were not accessible to the United States. Several of them help to monitor a broad swath of countries of concern to the United States, stretching across Russia, China, the Middle East, and southern Asia. Several of the IMS seismic stations and arrays, such as that in Niger in west-central Africa, are among the most sensitive in the world. The very quiet Niger array is located on old crust and the uppermost mantle of the Earth through which seismic waves propagate very efficiently.
One of the main products of the International Data Center is a Reviewed Event Bulletin (REB) for seismic events that is distributed to member countries—those that have signed the treaty—about every ten days. Significant events, like those at the North Korean test site, are distributed within hours.
Figure 13.2 shows the monitoring capabilities of only the primary IMS seismic stations that were operating as of late 2007. In it the globe is contoured for very high likelihood—that is, very high probability—of seismic events of various magnitudes being detected. Countries like the United States that want high confidence in identifying clandestine nuclear testing may pick a very high capability of detection, say 90 percent, as in figure 13.2. A country seeking to test but not wanting to be detected would not chose a 90 percent capability because of the high likelihood of being caught, but would select a smaller number, say only 10 percent or 20 percent probability of being observed.
FIGURE 13.2
Detection capability in late 2007 of thirty-eight operating primary seismic stations of the International Monitoring System. Contours indicate the magnitude of the smallest event that would be detected with high likelihood (90 percent probability).
Prepared by Kværna and Ringdal of NORSAR with yields added for National Academies Report of 2012.
The contours of magnitude in figure 13.2 indicate a very good global and regional detection capability, better than that projected in the 1990s and early 2000s using several computer simulations. Enough stations are now operating that the figure is now based on actual P-wave readings. Detection capabilities for seismic magnitudes and yields are better for countries of obvious concern to the United States than for many areas of the southern hemisphere. This is because more stations are located in the northern hemisphere than in the extensive southern oceans.
Capabilities shown in figure 13.2 are best at magnitude 3.2 for Scandinavia and 3.4 or better for most of Asia, Europe, North America, and North Africa. Those areas include states of obvious concern to the United States and all underground test sites except that of France in the South Pacific, which is now closed. Capabilities are somewhat poorer at magnitudes 3.6 to 3.7 for most of the southern hemisphere. Hydroacoustic waves, which propagate very efficiently to large distances, are much more useful for detection in most areas of the equatorial and southern oceans.
Figure 13.2 also indicates that capabilities for detecting yields of underground nuclear explosions are about 0.1 kilotons (100 tons) for regions with good seismic wave propagation like North Korea and most of Russia and China. Those yield capabilities are about three times worse for regions of poor seismic wave propagation like Nevada, eastern Turkey, and parts of Iran. While the capabilities for determining magnitude and yield for various areas of the world may be hard to take in all at once, they are all exceedingly good. The known yields of nuclear weapons in the U.S. and Russian strategic arsenals that are carried by intercontinental delivery systems are about a thousand times larger. Those capabilities are about a hundred times better than the yield needed to test the trigger for a fusion weapon (hydrogen bomb).
These capabilities do not include the use of additional data from either auxiliary IMS seismic s
tations or the huge number of high-quality stations of the International Federation of Digital Seismograph Networks. Digital recordings from many of those stations are now available in near real time over the Internet. Data are also available from large seismic networks in Canada, Japan, Taiwan, Turkey, the United States, and several European countries. About the same number of P-wave readings for the 2009 and 2013 North Korean nuclear tests were utilized for detection in near real time from those stations as from those of the International Monitoring System. Hence, they are a valuable supplement to international monitoring.
These seismic stations are used as well for other purposes, such as earthquake and tsunami warnings and studies of the Earth’s crust and deep interior. Data from the seismic stations of the IMS also were used to study several great earthquakes such as those off Sumatra in 2004 and Japan in 2011. The radionuclide stations of the IMS have been (and continue to be) valuable in tracking radioactive leakage from the Fukushima, Japan, disaster of 2011.
To summarize, substantial improvements made in the past twenty years in U.S. and international capabilities to monitor underground nuclear testing include the following:
1. High-frequency seismic data from new stations at regional distances (up to 1000 miles or 1600 km) have been tested and implemented for the detection, location, and identification of seismic events, particularly those smaller than magnitude 4.
2. More broadband, high-quality seismic stations and arrays are now transmitting much greater volumes of digital data to data centers in near real time.
3. Major increases in computer power and data storage have led to the use of the entire waveforms of many past seismic events for determining better location and identification of events.
4. India, Pakistan, and North Korea, none of which has signed the treaty, are the only countries that have tested nuclear devices or weapons since the Comprehensive Nuclear Test Ban Treaty was signed in 1996. Tests as small as a fraction of a kiloton can be detected and identified at the Indian and North Korean test sites, which are situated in regions of very good seismic wave propagation. The 2006 North Korean test, which had a yield somewhat less than one kiloton, may well have been a fizzle with a smaller yield than expected.
5. Instrumentation to measure exceedingly tiny amounts of bomb-produced xenon has improved tremendously. The IMS network for monitoring xenon and other noble gases has gone from nearly nonexistent in 2000 to one that provides global coverage today. Several countries, including the United States, make additional measurements of noble gases.
Figure 13.3 shows the great improvements since 1990 in seismic detection of small explosions and earthquakes. The global threshold for detection has improved from about magnitude 4.5 in 1990 to 3.7 in 2007—as small as 0.25 kiloton for regions of good seismic wave propagation and about 0.75 kiloton for areas of poorer propagation. When hydroacoustic capabilities are included for the southern and equatorial oceans, the combined detection limits are better.
FIGURE 13.3
Improvement in seismic monitoring from 1990 to 2016. Since the scales are logarithmic, one unit of magnitude represents about a factor-of-ten improvement in seismic amplitude and yield. Yields assume full coupling in regions with good seismic wave propagation.
Source: National Academies Report, 2012.
Detection thresholds for several regions of particular monitoring interest to the United States are now as good as magnitude 2.8. Capabilities to monitor the test sites of Russia at Novaya Zemlya and China at Lop Nor are better.
U.S. ATTEMPT TO RESTRICT DATA FROM INTERNATIONAL CENTER
The Air Force Technical Applications Center (AFTAC) is the U.S. National Data Center (NDC) for receiving seismic and other information from the International Data Center (IDC) for the CTBT. AFTAC uses the data internally and furnishes it to Department of Defense (DoD) contractors, other governmental agencies, and the weapons labs upon request.
Ralph Alewine of DARPA, who believed that data from the IDC should be restricted and not available in near real time to others, volunteered in May 1997 to draft an interim policy on the release of data in the United States. In June 1998, a White House interagency meeting chaired by Robert Bell of the U.S. National Security Council included a discussion on the release of data from the IDC. Despite skepticism from the National Security Council, Department of Defense officials maintained their position against the release of all data from the IDC, citing concerns about university scientists’ using the data to make independent assessments of compliance issues and, in the event of their reaching a conflicting opinion, undermining the U.S. government’s ability to charge violations or call for on-site inspections.
Interestingly, Alewine used the August 16, 1997, earthquake in the Kara Sea near Novaya Zemlya as an example of his point. It, of course, was the seismic event that was incorrectly identified by the U.S. government and remained so for more than two months. As discussed earlier, several university scientists, including me, identified it as a small earthquake, as did British seismologists, who stated shortly after the seismic event that it was a small earthquake. A number of scientists who studied the event at the U.S. weapons labs also agreed that it was not a nuclear explosion, but they were not allowed to comment publicly about the event for months.
This is a perfect example of why the Department of Defense and the U.S. intelligence agencies cannot be allowed to “bottle up” scientific data such as those from the IMS. Incorrect conclusions made by them have led to false claims about suspicious events of importance to U.S. national security and international relations. Such claims can have very dire consequences.
Fortunately, U.S. scientists can obtain an abundance of openly available seismic data for an event of questionable origin. Because the data are not classified, scientists from countries all over the world obtain data from the International Data Center through their own national authorities.
Although the interagency process normally is one of consensus, Robert Bell at the National Security Council evidently was frustrated that no progress had been made on the issue of data availability. In an extremely unusual move, he called for a vote of persons representing government agencies. Alewine’s proposition was overwhelmingly defeated. Last-minute attempts by DoD to include a delay in the release of data and other restrictions were defeated as well. Hence, it became the position of the United States government in 1998 that all data from the IMS were to be openly available without restriction.
Although a few countries objected to the prompt release of radiological and some other data collected by the IMS, it is now possible for scientists like me in the United States to send a request directly to the IDC in Vienna, open an account, and obtain seismic data. I did just that in 2009 for seismic events near various test sites, as described in the next chapter.
14
MONITORING NUCLEAR TESTS SITES AND COUNTRIES OF SPECIAL CONCERN TO THE UNITED STATES
The Comprehensive Test Ban Treaty Organization (CTBTO) held a conference in Vienna in 2009 that focused on progress made in monitoring the treaty. Six months earlier it had solicited contributions from scientists utilizing data that the organization had collected and analyzed. I made two presentations at the conference using seismic data I received in response to their solicitation.
Meredith Nettles of Lamont and I obtained data from 2000 through 2008 that the CTBTO’s monitoring arm had collected based on their seismic locations within 62 miles (100 km) of six sites used previously for nuclear testing. Identifying events at those sites is of great importance to policy makers. We found that all of those events could be identified as either earthquakes or explosions down to very low magnitudes in China, India, Pakistan, North Korea, and various other countries that are or may be capable of nuclear testing in the future.
Nettles and I examined thirty-eight seismic events of magnitude 3.3 and larger; the International Data Center of the CTBTO usually does not report events that are smaller than this. Most occurred near the former test sites of China, the United State
s, and Pakistan. No events were reported by the center near the Russian site at Novaya Zemlya or India’s test site, both very quiet locations for earthquakes. The identification of small seismic events on or near Novaya Zemlya is described in chapter 12 and is not repeated here. All of the events we studied at the Nevada Test Site (NTS) were earthquakes. We identified all of those in North Korea, including two nuclear explosions and one earthquake.
Identification at magnitude mb > 3.3 for five of the six test sites corresponds to a yield threshold of a small fraction of a kiloton if no serious attempts have been made to evade detection. The identification limit for Pakistan corresponds to about one kiloton, but it likely can be improved by examining high-frequency seismic waves, which our group has not done thus far.
CHINA’S TEST SITE AT LOP NOR
China conducted all of its nuclear explosions at its Lop Nor test site in the northwestern part of the country. Nettles and I found that the Lop Nor site had the greatest number of seismic events within 62 miles (100 km) of it, more than the other five test sites combined. Most of the seismic events that we studied near that site occurred at depths greater than 10 miles (16 km), clearly indicating they were earthquakes. High-frequency seismic waves also identify them and the remainder of the events as earthquakes (figure 14.1).
FIGURE 14.1
Measurements of high-frequency seismic waves for earthquakes and explosions near the Chinese test site at Lop Nor from 2000 to 2008. The log of the amplitude ratio of P to Lg seismic waves is shown on the vertical axis. Triangles denote earlier nuclear explosions at regional stations. Circles indicate earthquakes. Explosions have higher values on the vertical axis than earthquakes for frequencies of 4 and 8 Hz (cycles per second).
Source: Kim, Richards, and Sykes, 2009.
Although authors of several papers noted that the seismic event of March 13, 2003, of magnitude 4.3 to 4.7 was difficult to identify—that it was an anomalous or “problem” event—we found that it could be identified positively as an earthquake by five different methods. The smallest seismic event of magnitude 3.4 that we identified near Lop Nor would correspond to a well-coupled explosion with a yield of about 0.09 kilotons (90 tons). Work by others had indicated an identification capability that was even better, about magnitude 2.5, with a yield of 10 tons (0.01 kilotons).