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

Radionuclide source terms for a variety of test scenarios determine the percentage of radioactive particulates and noble gases available for global atmospheric transport. DOE experts at Lawrence Livermore and other National Laboratories have considered the various possible uncertainties associated with each source term. There appears to be considerable uncertainty in these data, and research is needed to upgrade them for use in CTBT modeling.

The following is a list of scenarios considered in this study. The range of source terms for gaseous and particulate releases is given in Table G.1.

These are tests conducted in the stratosphere or above the transition zone in the troposphere. The troposphere runs from the Earth's surface up to a range from 10 to 13 km (6 to 8 miles) high where the stratosphere begins. The troposphere is marked by decreasing temperature with height. The stratosphere is a nearly isothermal layer that has its upper boundary at about 50 km (30 miles) above the Earth's surface.

Free-air bursts occur at sufficiently high altitude that no surface debris, soil, or water is drawn up into the fireball. Essentially all of the condensable nuclear debris is in the form of small particulates having radii between 0.01 and 1.0 micrometers (µm). Therefore, little local fallout occurs near the geographic site of the test. The debris deposited in the lower regions forms tropospheric fallout that will reach the surface over a month's time in the general latitude of the test site. The finer particulates deposited in the upper atmosphere form stratospheric fallout that may continue for years with a nearly worldwide distribution.

Radioactive noble gases do not form particulates and will mix and move along with the atmospheric air. Because of the small size of airborne particulates, gases formed by the decay of precursors in the particulates will be able to diffuse out and contribute to the atmospheric noble gas concentration. As shown in Table G.1,

the source term assumes that almost all of the radioactive particulates and all of the radioactive noble gas are released into the atmosphere.

For high-altitude (stratospheric) tests, it may take several months before any radioactive material can reach the ground. By that time, most of the radionuclides of interest will have decayed. For confirmation of this type of test, stratospheric sampling would be required.

For low-altitude (tropospheric) tests, the debris will quickly reach the surface. Because of radioactive decay, the chemical properties of different fission products, and the fallout rate of different-sized particulates, the radioisotopic composition of the debris will change rapidly. The refractory fission products, including radioisotopes of the lanthanides, rapidly form stable oxides, condensing onto the dust particulates where they fallout more quickly than the more volatile fission fragments. These radioisotopes are useful for providing attribution information. Radionuclides collected far from the explosion are usually highly enriched with the more volatile elements and are only marginally useful for attribution.

The one significant way to conduct an evasive atmospheric test in a manner that will reduce the

likelihood of detection of the radioactive products is to detonate the device during a large and intense rainstorm. Depending on conditions, the amount of airborne particulates will be reduced by a factor of 10 to 100. Although the noble gases will not rain-out, the washing out of noble gas precursors such as the isotopes of iodine can reduce the amount of noble gas subsequently released to the atmosphere (Glasstone, 1957). Depending on test conditions, as much as 80 per cent of the nonvolatile fission products may be in the local fallout, including the noble gas precursors, primarily the fission products of tin, antimony, and tellurium.

For a 1-kiloton (kt) yield burst exploded in the aboveground transition zone (less than 100 m above the surface of the soil), the strong updraft produced by the explosion will cause large amounts of dirt and debris to be sucked into the atomic fireball and injected into the atmosphere. Depending on test conditions, a large fraction of the fission products will be released into the atmosphere, but much of the particulate matter and some of the noble gas precursors, such as isotopes of tin, antimony, and tellurium, will be removed quickly as local fallout.

Rain-out events, as previously mentioned for tropospheric tests, can affect any of the soil-burst tests that release radioactive fission products promptly into the atmosphere. Again, depending on conditions, the amount of airborne particulates will be reduced by a factor of 10 to 100.

Many of the phenomena and effects of a nuclear explosion occurring on the Earth's surface are similar to those associated with aboveground transition zone airbursts. In surface bursts, however, an even larger amount of rock, soil, and other material will be vaporized and taken up into the fireball. Thus, there will be even larger amounts of local fallout to tie up fission products and keep them from being released.

When a nuclear explosive is detonated under the ground, a fireball is formed consisting of extremely hot gases, including vaporized rock, soil, and bomb residue, at high pressure. If detonation takes place at too shallow a depth (less than about 70 m for a 1 kt burst), the gases will break through the surface and carry up large quantities of rock and debris into the atmosphere. The results are similar to an aboveground burst, but because of the presence of a larger volume of absorbing material, the amounts of particulates and radioactive noble gas are reduced. It is primarily the more volatile species of the fission products that are released into the atmosphere.

For underground nuclear events that are unable to contain gases especially well, xenon-133 may be detected for 25 to 30 days at a distance greater than 300 m and for 30 to 70 days at distances less than 300 m. Thus, there is the potential for use of a mobile noble gas monitoring system to aid in identifying the detonation site.

The most significant evasive way (from the standpoint of radionuclide monitoring) to conduct an underground test is to detonate it so that there is containment of the hot gases generated by the detonation to keep them from venting at the surface. For a 1 kt explosion a burial depth of greater than about 100 m will contain these gases. Past tests indicate that the burial depth for containment will vary with the cube root of the explosion yield. The most likely release, if any, would be radioactive noble gases. These gases would be released through cracks or fissures penetrating to background sources not of interest to a maximum of the iodine and xenon prompt fission yields and would be released over a period of a few days. Tellurium and antimony precursors formed under these conditions do not readily release their xenon decay products. However, the likely result, using

modern published containment practices, is that no gases will be released for a period of weeks to months. By this time, all of the radioactive noble gases except krypton-85 (10.7-year half-life) may have decayed to undetectable levels. At these times, argon-37 (35-day half-life), made from fast neutron transmutation of calcium-40 in the soil, along with krypton-85 may still be detected.

A nuclear device detonated above water in the transition zone will vaporize and carry water up into the fireball. At high altitudes this water will condense to form water droplets, which in turn will form a radioactive cloud similar to ordinary atmospheric clouds. As cooling continues, much of the water, with its suspended radioactive particulates and dissolved fission product ions, will gradually fall back to the surface as rain, spreading radioactivity over a large area of the ocean. It was assumed in this study that no xenon gas would be released from precursor decay in water. This assumption has not been proven.

Rain-out events, as previously described for tropospheric tests, can affect any type of ocean burst test that releases radioactive fission products promptly into the atmosphere. Again, depending on conditions, the amount of airborne particulates will be reduced by a factor of 10 to 100, with particulates being removed in preference to the more volatile fission products.

When a nuclear device is exploded at or near the surface of the water, the results will be similar to the above-water transition zone tests except that the amount of water drawn into the fireball will increase dramatically.

Underwater tests may offer one of the best ways to avoid radionuclide detection and/or tion. A device could be detonated and monitored by aircraft and/or surface and underwater vessels that can be long gone before the event can be investigated. In underwater nuclear tests the fireball will be smaller than that formed in airbursts. The resulting bubble of hot gases remains essentially intact until it reaches the surface. At this point, the gases, carrying some liquid and most of the radioactivity, are expelled into the atmosphere. As the pressure of the bubble is released, water rushes into the cavity, forming a hollow column of spray. The radioactive contents of the gas bubble are vented through this hollow column and form a cauliflower-shaped cloud at the top. About 20 seconds after the detonation there will be a massive water fallout that returns much of the radioactivity to the ocean surface. The descending water will form a continuous mass of mist—from the top of the nuclear cloud down to the surface—that eventually is dispersed by the wind. The deeper the point of detonation, the lower will be the amount of radiation released. For a 1 kt device detonated less than about 300 m deep, it is likely that at least some of the volatile fission products will be ejected from the water.

Underwater tests will leave a highly radioactive pool of water above the ocean's thermocline layer. This pool may contain from 25 per cent to nearly 100 per cent of the radioactive debris that disperses slowly compared to atmospheric debris clouds. The thermocline is a thermally stable layer of water that exists in the oceans. A widespread permanent thermocline layer exists beneath the surface layer from depths of about 230 to 900 m. A seasonal thermocline at a much shallower depth forms during summer as a result of solar heating. Thus, it would be important to locate and sample a radioactive pool before it disperses.

Currently no data are available for an underwater test so deep that the gas bubble collapses before reaching the surface. In this case the bubble of hot gases would experience repeated oscillations

in both diameter and elevation before collapsing. It appears that a point is reached beyond which the total release of radioactive products does not decrease with depth. Airborne debris from these events consists mostly of radioactive noble gases.

The report also includes the following important conclusions:

1. The particulate source term can be severely reduced or eliminated in many scenarios, especially when rain-out is considered.

2. Gaseous radionuclides are more difficult to conceal than particulates in virtually all scenarios (except possibly, high-altitude events). Thus, a xenon detection system may have the capability to detect nuclear explosions in scenarios where particulate detection fails.

3. In some scenarios, neither gas nor particulate sampling systems will detect an explosion, specifically an underground nuclear explosion that is emplaced using modern containment practices.

4. Remote gas sampling will probably not provide useful attribution information because of the limited information contained in the xenon isotopic signatures.

5. High-quality samples (with minimum fractionation) are the key to attribution. Radionuclides may be especially important for attribution of events detonated over international territory. Aircraft collection of atmospheric debris or collection of water samples from the broad ocean area is necessary to obtain such samples.

6. On-Site Inspection using gas sampling is feasible in the case of moderate leaks but will require precise location information (i.e., distances less than two to three depths of burial) for well-contained events.