View the original post on the IUSSCAA message board here

What US Seismologists Concluded About the Loss of the Russian SSGN KURSK in 2000

By Bruce Rule - Aug 26, 2014


US seismologists were correct when they assessed the 3-Hz KURSK-associated seismic-wave signal to have brne “bubble-pulse” from a major explosive event. That signal was seismically detected at a range of 3100 statute miles (5000km) by a sensor in Alaska; however, their conclusion that missile explosions were responsible for the disaster was in error. The first event was the explosion of HPT torpedo fuel that, in 135 seconds, “cooked off” the entire KURSK torpedo warhead load, and nearly blew off the bow of the submarine. Their estimated yield for the main event (more than 8000 lbs (3600kg) of TNT) probably was close to the mark. The Russians finally acknowledged the source of the explosions but only after imagery of the wreck falsified their initial claim that a collision with a US submarine was responsible.

In special cases, such as the KURSK event, seismic data is the “other hand” of acoustic data. An understanding of what can and cannot be derived from seismic data can be useful to System analysts.


The Kursk Explosion by Brian Savage and Don V. Helmberger

Abstract: On 12 August 2000 two explosions damaged the Russian submarine, the Kursk. The largest event was well recorded at seismic networks in northern Europe, which we then modeled. We developed a hybrid method based on generalized ray theory that treats an explosive source embedded in a fluid and recorded along continental paths. Matching record sections of observations with synthetics, we obtain an estimate of explosive size of slightly over 8000 lbs (3600 kg). Several earth models determined previously, K8 and a Baltic model, were used to assess accuracy. These results are in general agreement with other investigators using more empirical methods. Knowing the conventional missile yield and the explosion size allows for an estimate of approximately five missiles exploded in the second larger explosion on-board the Kursk.


On 12 August 2000, two events were located in a seismically inactive region, less than 10 degrees from Novaya Zemlya, where a large number of nuclear tests were conducted over the decades. These events were located close together in the Barents Sea region at shallow depths. A few days after the events occurred, news agencies reported that a Russian submarine, the Kursk, had been damaged during exercises in the Barents Sea. The seismic events were located in the same region and at the same time as when the Kursk reportedly sunk.

The Kursk is a Russian nuclear submarine, Oscar II class, with the ability to carry 24 anti-ship cruise missiles. While the missiles can be nuclear in nature, 500 kt, it is more likely that those carried on the Kursk were a conventional explosive size of 750 kg (Bellona Foundation, 2000), as the nuclear warheads have been placed into storage. A single missile of this size, exploded underwater, is easily capable of generating seismic waves that can travel local distances. Given low attenuation, the same missiles can be detected regionally, while multiple missile explosions can generate energy that has the ability to travel much further. Being that the time and spatial extent of the events match those of the damaged Kursk, and that there was an explosive capability most likely present at the time of the events, leads us to the conclusion that the two events recorded seismically and the damage to the Kursk are related.

Our main purpose in this article is to get a better understanding of what happened onboard the Kursk in terms of the size of the explosions. A few techniques have been utilized to determine the size of underwater explosions, which include using various magnitude scales or character- istics in the amplitude spectra (Gitterman and Shapira, 1994; Baumgardt and Der, 1998; Gitterman et al., 1998; Koper. et al., 200. (Comment: it was Koper who, in 2001, told the writer that (quote) The sensitivity of underwater acoustics is legendary in the seismic community. (end quote)

However, the approach taken in this article will be to use synthetic waveforms to understand more about the seismic signatures in the time domain. The observations (short-period, 0.2–20 Hz, vertical field) are displayed in Fig- ure 2, plotted as a reduced section which aligns the Pn. We have included some reference lines to aid in the phase iden- tification. Since we see frequencies of 3 Hz (Comment: this was the bubble-pulse) at distances over 900 km, it is most likely the case that attenuation does not play a very important factor. Additionally, there is little to no apparent frequency shift as we proceed out in distance, reinforcing this point. Moreover, no large amplitude varia- tion with azimuth is seen, as is characteristic of an explosion.

In this study, we address the extended P-wave motions because most of our knowledge about crustal structure is for P waves. A discussion of the relationship between P and S waves in our data will be reserved until later, as they are intimately related to the water in which they are produced. In order to compute synthetics for an underwater explosion, we will first characterize an underwater explosive source, and then describe seismic-wave propagation within a water layer. Finally, we will compare the synthetics to the wave- form data in an attempt to estimate explosion size.

Underwater Explosive Sources

Following the descriptions by Weston (1960), Arons (1954), and Cole (1948), the time dependence of pressure of an underwater explosion can be formulated. Note that a single explosion in water is comprised of a series of overpres- sures and underpressures caused by the pulsation of the gas volume in the water. Examples of the gas volume pulsation can be seen in the article by Helmberger (1968). These over- pressures, or bubble pulses, essentially act as new sources,

The over-pressures decay as an ex-potential where the relaxation time of the initial pulse is defined.

The under-pressures occur over a much greater time period and return the pressure of the region back to a hydrostatic level. As will be shown later, as the yield increases, the effect of the later pulses on the final waveforms decreases dra- matically. To simplify the calculation, only the first pulse will be used in the characterization of the source. Later, we will add the bubble pulses as secondary effects. The under-water source description, classically defined in units of pressure, needs to be defined in units of displacement potential for synthetics to be computed.