[sacw] Controversy over the yields of India's Nuclear Tests May 1998

[Harsh Kapoor]

South Asians Against Nukes Dispatch
11 January 2000
______________________

(Frontline, January 8, 2000)
No Clear Yield
by M. V. Ramana
The controversy over the yields for India's May 1998 nuclear weapons tests
shows no sign of abating. Faced with continuing scepticism from the
international scientific community, India's nuclear weapons scientists have
tried again to show that the yields they announced after the tests should
be taken seriously. The evidence presented this time is the radioactivity
of samples extracted from the sites after the tests [Frontline, 10
December 1999]. It shall convince no one.
There are questions about the yields of both the explosions of May 11 and
the explosions of May 13. The yield of the May 11 explosions bears on
whether the thermonuclear design succeeded. These questions will be dealt
with in a separate article. Here, only the yields of the two sub-kilotonne
tests of May 13 will be discussed.
=46irst, let us look at the basic features of a nuclear weapon. A modern
fission bomb consists of a core of fissile material highly enriched
uranium or plutonium surrounded by chemical explosive. Fissile materials
are those that absorb (low energy) neutrons and undergo fission, i.e., the
disintegration of heavy nuclei into lighter nuclei and neutrons. When the
chemical explosive is set off, the core of the bomb is compressed. Once the
density of the material in the core exceeds a certain critical value, the
fission process becomes self-sustaining and a chain reaction starts.
As the reactions proceed, the core starts disintegrating from the energy
released by these fission reactions. In order that the explosion releases
as much energy as possible, the weapon is designed to maximize the amount
of fissile material that undergoes fission before the core breaks up. To
aid this, a large number of neutrons are injected into the core a process
known as initiation. The amount of fissile material that is utilized is
crucially dependent upon the time at which the neutrons are injected. An
error of a fraction of a microsecond (one-millionth of a second) in the
timing could result in the yield of the weapon being much less than the
design value. In the case of weapons using plutonium, especially reactor
grade plutonium, such a reduced yield could also result from spontaneous
neutrons emitted by the plutonium itself.
A weapon that explodes with a yield much less than the design yield is
called a fizzle. This term has been widely misused and misunderstood in
the debate surrounding the explosions of May 13. For example, the
=46rontline article misconstrues the claim that the May 13 tests fizzled to
mean that the explosion did not occur at all. With that interpretation,
any evidence for the occurrence of fission would suffice to prove that the
explosion did happen. However, proving that the explosion did not fizzle,
i.e., it resulted in a yield comparable to the design value, requires much
more evidence. BARC is yet to demonstrate that.
Broadly speaking, three different types of methods are used to monitor the
yields of nuclear explosions.
Radiochemical methods make use of the nuclear reactions that occur during
the explosion. By measuring the amounts of some of the radioactive
substances produced as a result of fission, fusion or absorption of
neutrons, the yield of the explosion is estimated.
Hydrodynamic methods make use of the strength of the shock wave produced
by the explosion. By comparing measurements of quantities such as the
velocity or pressure with a theoretical model based on properties of the
surrounding media, an estimate of the yield can be made.
Seismic methods make use of the ground motions caused by the explosion.
These are somewhat similar to the movements produced by an earthquake. By
measuring such ground motions at seismic stations, the yield is estimated
based on the strength of the coupling between the explosion and the ground
motions.
Of these three kinds, the data released in the BARC article (Frontline,
December 10, 1999) pertains only to the radiochemical measurements; in
particular, only gamma radiation measurements are provided. The data
provided, however, is seriously incomplete and cannot be used to infer
anything about the yield.
The first and most obvious problem is that no units are specified for much
of the data. Even elementary physics textbooks emphasize the importance of
specifying the units in which any measurement is expressed. Just saying
that the height of a person is six or that the mass of a watch is fifty is
meaningless. One has to specify that the height of the person is six feet
or that the mass of the watch is fifty grams. Likewise, it is not possible
to conclude anything from Fig. 1 because it does not specify whether the
gamma dose rate at the test site of the 0.5 kilotonne explosion is measured
in Grays, milliGrays (one thousandth of a Gray) or microGrays (one
millionth of a Gray). Neither does it specify the units in which the
height is measured. Without these, it is impossible to conclude anything
about the yield of that test.
=46ig. 3, on the other hand, does mention the units used but does not
provide any markings on the axes. Hence, one cannot determine the
coordinates of any of the data points. Once again, the graph cannot be used
to make any estimate of the yield. There is also the strange feature that
at the zero of the x-axis there are two values for the activity. This is
neither explained nor commented upon.
Perhaps the figure that provides the most useful information for the
purposes of understanding anything about the explosion is Fig. 2. This
shows the gamma activity of three radioactive substances Cesium-137,
Zirconium-95 and Niobium-95. However, as in the other cases, some crucial
pieces of information are not provided. These include the efficiency of
the detector and the size of the sample. The activity of any radioactive
substance per unit mass is inversely proportional to both these quantities.
Since the efficiency is a function of energy, one cannot even find the
ratio of amounts of any two of the radioactive substances. As with the
other figures, this cannot be used to estimate the yield either. (Note:
Even the original version of the article on the BARC web-site
(http://www.barc.ernet.in) has the same problems.)
In the absence of useable radiochemical information, one is forced to rely
upon publicly available seismic measurements to estimate the yield. This
was the case for the 1974 test as well, for which radioactivity
measurements were never made public. The announced yield in that case was
12 kilotons. Privately, however, officials within the Indian nuclear
establishment have admitted that the yields were much lower. (See, for
example, George Perkovich, Indias Nuclear Bomb, University of California
Press, Berkeley, 1999) This lower yield is in better agreement with
seismic estimates made by independent scientists.
The combined yield of the May 13 explosions was so small that it went
undetected even by nearby seismic stations, including the ones at Nilore
and the Kyrgyz network. On the other hand, both these sites recorded the
May 11 explosions very well; for example, the signal-to-noise ratio at
Nilore was over a 1000. Seismologists have also used filtering and
cross-correlation techniques to search the regional data from Nilore for
several hours before and after the announced time of the explosions.
Though several signals from small earthquakes in the Hindu Kush region
were recorded, no signal consistent with a test was found. Accordingly,
they concluded that the seismic magnitude of the May 13 test was at least
500 times smaller than that of the May 11 explosions.
Even allowing for the possibility that the explosions were conducted in a
sand dune, as Indian officials have claimed, these seismologists concluded
that the tests are likely to have been less than 0.3 kilotonnes in size.
(See Brian Barker et al, "Monitoring Nuclear Tests," Science Vol. 281, 25
September 1998) This is only about a third of the combined total of 0.8
kilotonnes claimed by BARC scientists. If these explosions were conducted
in a medium similar to that of the May 11 tests, their yield would have
been less than 0.1 kilotonnes.
If the May 13 explosions were conducted in a sand dune, rather than deep
under the ground, there is a much greater likelihood that radioactivity
would be released to the environment. Such releases have occurred in
several nuclear explosions around the world, damaging human health and the
environment (See, for example, The Hindu Survey of the Environment 99,
June 1999). Even if the tests did not actually release radioactivity, the
risk of that occurring is very real. That nuclear explosions may have been
conducted in such a medium only demonstrates the priorities of the
scientists who planned them.
In conclusion, the publicly available evidence so far suggests that the
yields of the May 13 tests were much smaller than the announced yields.
This would fit the definition of a fizzle, a test whose yield is much
smaller than the design value. In the future, BARC should either provide
raw data that can be independently analyzed or stop producing papers that
pretend to demonstrate their claims while not measuring up to scientific
standards.