In 1987 Daniel Davis of Stony Brook University and I published “The Yields of Soviet Strategic Weapons” in Scientific American. We tried to ascertain if the Soviet Union possessed more powerful nuclear weapons and more capable intercontinental-range (strategic) delivery systems than the United States. To do so, we determined yields from seismic waves generated by large Soviet explosions and combined them with published information on so-called throw weights—how many tons (or kilograms) of nuclear warheads their missiles could carry. We found the two countries were at near parity in intercontinental delivery systems and the nuclear weapons they carried.
We matched the occurrences of various large nuclear explosions with published information on the dates the delivery systems were first tested and then deployed. We used calculations of yields of Soviet explosions at Novaya Zemlya, their arctic test site, that I had obtained with Graham Wiggins, a student who worked for me the summer before, using a combination of seismic body and surface waves.
The Scientific American article was motivated by reports in the media of hugely inflated estimates of the capabilities of Soviet missiles and the yields of each of the multiple nuclear weapons they carried. Senator Robert Dole, a high-ranking Republican, announced that the yield of each of the eight to ten nuclear warheads carried by the large Soviet missile, the SS-18, was fourteen times larger than what we subsequently concluded in 1987. Dole’s huge overestimates had a great impact on U.S. defense policy and fueled fears about the consequences of a Soviet nuclear attack.
For each of the Soviet intercontinental land-based ballistic missiles (ICBM) known by NATO as the SS-17, SS-18, and SS-19, two major versions (mods) existed—one with a single very large warhead and a second with multiple, independently targetable reentry vehicles (MIRVs). The capabilities of the very large SS-18 missile understandably were of great concern to the United States. At the negotiations for the Threshold Test Ban Treaty (TTBT) in June 1974, General Edward Rowny, who represented the U.S. Joint Chiefs of Staff, had focused on the large payload of nuclear weapons of the SS-18 and not much on the testing of yields of warheads the United States should seek to prohibit under the TTBT.
If the United States had insisted in 1974 on a TTBT with a lower yield limit of, say, 50 kilotons, tests as large at 150 kilotons would not have been permitted after March 31, 1976, and the USSR might well not have been able to fully test multiple warheads for its SS-17, SS-18, and SS-19 strategic missiles. At that time, the United States was ahead of the USSR in testing and deploying land- and sea-based missiles with multiple warheads. The TTBT allowed very large tests to be conducted for nineteen months after it was signed, which gave the Soviet Union time to catch up with the United States in testing multiple warheads.
Paul Nitze, who served as a U.S. high-level official and as a special adviser to the president on arms control, testified before Congress in July 1979 about the SS-18 and other Soviet long-range delivery systems. His estimate of 750 kilotons for each of the multiple warheads on the SS-18, as published in the Congressional Record—Senate, was much smaller than Senator Dole’s. Nitze expressed great fear that the Soviet Union would obtain a substantial lead over the United States in the nuclear arms race with its larger missiles and their many warheads of high yield.
Interestingly, on May 31, 1979, journalist Walter Pincus wrote a front-page story in the Washington Post stating that the U.S. intelligence community had substantially downgraded its estimate of the explosive yields of each of the eight to ten weapons carried by the SS-18, from the previous year’s 1200 kilotons to 600 kilotons in 1979. He reported that analysts had successively lowered their previous estimates of the yield of each of its warheads from 3000 to 2000 to 1200 and thence to 600 kilotons. If correct, this represented a reduction in yield by a factor of five and was close to the yield Davis and I published in 1987. If, as originally estimated, the SS-18 missile carried ten warheads of 3000 kilotons each, its total payload would have been 30,000 kilotons. If it carried eight warheads of 600 kilotons each, the total yield for each SS-18 would have been 4800 kilotons, a huge difference. In a sense, this was a continuation of the “yield wars” described in chapter 10.
Pincus stated, “This sharp reassessment of the SS-18’s yield could play a major role in the coming Senate debate on the U.S.-Soviet strategic arms limitation treaty (SALT II), according to sources on Capitol Hill.” President Jimmy Carter and Soviet Premier Leonid Brezhnev signed that bilateral treaty on June 18, 1979, but the U.S. Senate failed to ratify it. Debate over the capabilities of Soviet missiles and their warheads, along with the Soviet invasion of Afghanistan, contributed to the Senate’s inaction. Nevertheless, the two countries honored the terms of SALT II until the Reagan administration withdrew from the treaty in 1986 after accusing the USSR of violating it.
Pincus said that the six warheads on the Soviet’s second largest missile, the SS-19, in 1979 were each rated as 500 kilotons instead of the previously estimate of 800 kilotons. Davis and I also obtained 500 kilotons for the yield of each of its multiple warheads.
It is unfortunate the United States focused so much effort on the yields of tests and alleged Soviet cheating at Eastern Kazakhstan rather than on the much larger nuclear explosions that were set off at Novaya Zemlya. The yields of weapons deployed by the Soviet Union in the 1970s on its strategic delivery vehicles were large enough that full-yield tests had to be conducted at Novaya Zemlya between 1966 and 1975. It is for that reason that Davis and I decided to work on those tests and their implications for the capabilities of Russia’s strategic weapons.
The United States was not permitted to monitor any of the nuclear explosions on Novaya Zemlya, nor was it allowed to conduct a Joint Verification Experiment there. Unlike Eastern Kazakhstan and sites of several peaceful nuclear explosions, the Russians did not publish exact yields for any of their explosions at Novaya Zemlya until twenty-two years after they signed the Threshold Test Ban Treaty.
In 1996 and 2000, the ministries of Atomic Energy and Defense of the Russian Federation finally published some information on the numbers and yields of nuclear tests. Many of the tests at Eastern Kazakhstan, as well as peaceful explosions, consisted of single explosions. Exact yields published by the Russian Federation for several of those explosions agree with estimates made by Western scientists, including me, and in 1992 by Ringdal, Marshall, and Alewine.
Nevertheless, the Russian information is more difficult to interpret for Novaya Zemlya explosions. It indicated that many large tests between 1966 and 1976 at the northern site consisted of multiple explosions that were detonated very close in time, in what are called salvos. In contrast, they listed three of the four explosions at the southern Novaya Zemlya test site from 1973 to 1975 as single events.
Why did the Soviet Union test two or more nuclear devices in a salvo? The northern site at Novaya Zemlya is located farther north than the northernmost part of Alaska. Working and testing in the far arctic at Novaya Zemlya were limited to times of less severe weather, ice conditions, and quite possibly by the amount of sunlight. Most of those underground tests occurred during September and October, with only a handful in August, November, and December. Hence, the Soviets had little time each year to conduct large explosions at Novaya Zemlya, especially in the period between the signing of the TTBT in July 1974 and its start date in early 1976.
Russian publications in 1996 and 2000 listed the sum of the yields tested per year at Novaya Zemlya but gave only a broad range of yields for underground explosions on a particular date. For example, on August 23, 1975, they listed a salvo consisting of eight explosions—four with yields of 150 to 1500 kilotons, two of 20 to 150 kilotons, and two of 0.001 to 20 kilotons. Thus, the individual yields of the four largest explosions in that salvo are very uncertain.
Many people in the United States, including Davis and me, were not aware in 1986 that large numbers of explosions at the northern test site at Novaya Zemlya were fired at nearly the same time. Subsequently, U.S. and British scientists identified several other events that consisted of more than single explosions.
Since the Russian government published the cumulative yield per year for Novaya Zemlya, one of the better comparisons with their data is the summed yields we published. The sum of the yields in the Russian lists for the two test sites on Novaya Zemlya between 1966 and 1975 is about 1.7 times greater than what we published. Since many of the salvos consisted of two or more individual large explosions, our published yields of weapons on Soviet delivery systems probably were not as uncertain as a factor of 1.7.
Why did we do poorly for yields at Novaya Zemlya but so well for Eastern Kazakhstan and Soviet peaceful explosions? We made two incorrect assumptions when we calculated Soviet yields at Novaya Zemlya in 1987. One was that seismic waves passing through the upper mantle beneath those large tests did so in the same way as beneath Eastern Kazakhstan. The other was that events on a given date were single explosions.
I now believe that our calculated yields for salvos at the northern site on Novaya Zemlya were too small. Most explosions at the two arctic sites occurred at northern Novaya Zemlya. Our magnitude corrections (mb bias) with respect to Nevada were like those we calculated for Eastern Kazakhstan—that is, about 0.35 magnitude units. We applied that correction in calculating yields for the two Novaya Zemlya sites using explosions of known yield and mb in Nevada and the Aleutians.
The Russian lists of 1996 and 2000 stated that their largest underground nuclear explosion occurred on September 12, 1973, at the northern Novaya Zemlya site. Davis and I had concluded, however, that it was the second largest Soviet explosion and that one at southern Novaya Zemlya on October 27, 1973, was larger. Two Russian references give the yield of the single great event of 1973 at southern Novaya Zemlya as 3500 kilotons. Subtracting that yield and that of a much smaller event from the Russian sum of 7820 kilotons for 1973 gives a yield of about 4200 kilotons for the largest Soviet underground test on September 12, 1973, at the northern site. Some Russian experts told my colleagues that event was smaller, around 3500 kilotons. In any case, we clearly underestimated its yield at 1830 kilotons.
In 1987 Davis and I calculated an average yield of 3450 kilotons for the single explosion at southern Novaya Zemlya in October 1973, which was nearly identical to the published Russian yield. Thus, I conclude that the magnitude correction we used was about right, 0.35 mb units, for southern Novaya Zemlya but was too large for the northern site. The explosion on September 12, 1973, was not the largest Soviet underground test as they claim unless the magnitude correction or bias for the northern site on Novaya Zemlya is smaller than that for the southern site.
In 1979 Marshall and colleagues used the speed of P waves at the top of the Earth’s mantle as an indirect indication of the magnitude corrections to be made for various test sites. They concluded that northern Novaya Zemlya was more attenuating than Eastern Kazakhstan. I now concur with their judgment.
Davis and I estimated the yields of the single very large warheads carried by the SS-17 and SS-19 missiles as 2000 and 3500 kilotons based on our calculated yields for the two largest underground Soviet tests of September 12 and October 27, 1973. My revised estimates for those two are about 3500 and 4200 kilotons. The larger yield is comparable to that of the largest U.S. underground test, Cannikin, which was intended for the U.S. Spartan anti-ballistic missile. The Cannikin device, however, was not deployed as a weapon, whereas the Soviet nuclear devices almost certainly were.
On April 5, 1985, Jack Anderson and Dale van Atta reported in the Washington Post that a warhead gap existed between the number of Soviet warheads estimated by the CIA, 6500, and by the U.S. Defense Intelligence Agency (DIA), 8500. They stated, “But at the last minute, the DIA chickened out and allowed publication of the CIA’s less scary estimate.” They went on to say, “A Soviet SS-18 could hold two dozen warheads. On that both agencies agree.” They then concluded, “In this case, we’d be inclined to lean toward the DIA estimate.” This is the only instance I know of in which the SS-18 has been credited with carrying more than eight to ten warheads.
Jack Anderson had published incorrect estimates of Soviet yields in the Washington Post in August 1982, stating, “the Soviets appear to have exceeded the 150-kiloton limit [of the Threshold Treaty] at least 11 times since 1978. One test in September 1980 was clocked at a likely size of 350 kilotons, according to my sources.” We now know that his estimates of those eleven yields at Eastern Kazakhstan were too large. I think Anderson and van Atta most likely gave inflated estimates for the SS-18.
In summary, I think the better estimates are that the yield of the individual (MIRVed) warheads on the Russian SS-18 intercontinental missile is about 500 to 600 kilotons and the yield of those on the SS-19 is about 500 kilotons. Both of those missiles were still deployed as of mid-2017. While Russia has more long-range warheads deployed on land-based missiles, the United States has more of its assets at sea. The published yield of the W-88 warhead on U.S. Trident II submarines is 475 kilotons.
STRATEGIC DELIVERY SYSTEMS AND NUCLEAR WEAPONS OF THE UNITED STATES AND RUSSIA
The number of long-range nuclear weapons maintained on high alert by the Russian Republic and the United States has been reduced through various bilateral treaties since our paper in 1987 (figure 17.1). Nevertheless, Russia and the United States each still have many more nuclear weapons than other countries.
FIGURE 17.1
Numbers of long-range (strategic) nuclear weapons and their delivery vehicles permitted under bilateral treaties between the United States and Russia.
Source: U.S. State Department.
Figure 17.2 indicates that the number of U.S. nuclear warheads reached a peak of about 31,255 in 1967. Table 17.1 lists the numbers of offensive (long-range) nuclear arms of Russia and the United States at the end of 2013. The number of warheads that could be delivered rapidly by the United States and Russia is still huge—1688 and 1400—as are the numbers of facilities, cities, and people that could be targeted. A major exchange of weapons by the two nuclear superpowers could result in the immediate deaths of hundreds of millions of people in the two countries as well as comparable numbers in other countries in the Northern Hemisphere. The longer-term consequences, which are huge, are examined in the last chapter.
FIGURE 17.2
U.S. nuclear stockpile, 1945–2014.
Source: Redrawn from fact sheet, U.S. Department of State, 2014.
TABLE 17.1 Bureau of Arms Control, Verification, and Compliance Fact Sheet, January 1, 2014: Numbers of offensive (strategic) nuclear arms of Russia and the United States
| CATEGORY OF DATA |
UNITED STATES OF AMERICA |
RUSSIAN FEDERATION |
| Deployed ICBMs, deployed SLBMs, and deployed heavy bombers |
809 |
473 |
| Warheads on deployed ICBMs, on deployed SLBMs, and nuclear warheads counted for deployed heavy bombers |
1688 |
1400 |
| Deployed and non-deployed launchers of ICBMs, deployed and non-deployed launchers of SLBMs, and deployed and non-deployed heavy bombers |
1015 |
894 |
ICBMs denote intercontinental land-based ballistic missiles; SLBMs, submarine-launched long-range ballistic missiles. Data are from the biannual exchange of data required by the treaty and are declared current as of September 1, 2013.
Source: U.S. Department of State.
Russia and the United States each have a different sense of what constitutes “strategic delivery systems” for nuclear weapons. The United States, separated from Russia, China, and other nuclear states by large oceans, understandably makes a distinction between intercontinental delivery systems and others of intermediate and shorter range. Russia is faced with nuclear states at varying distances. Thus, it does not make as strong a distinction between intercontinental delivery systems and those for possible use within Asia and Europe. Various bilateral treaties between Russia and the United States do specify the distance ranges for various nuclear delivery systems.
India and Pakistan each have about a hundred nuclear weapons, and their numbers are growing. A major exchange between them could well involve several hundred million causalities. While India is Pakistan’s sole major adversary, India faces threats from both Pakistan and China, which has been an ally of Pakistan for decades. A major exchange involving the three countries could cause more than half a billion fatalities, and the conflict might spread to other nuclear powers. India and Pakistan may well not have weapons that are as safe as those of the United States, Russia, and China. In addition, their ability to monitor actions by other countries is likely not as great.
The world has been fortunate that nuclear weapons have so far not been used by one state against another since 1945, but this could end at a moment’s notice. I do not derive much comfort from comments by some analysts that nuclear weapons have prevented a world war since 1945.
U.S. Secretary of Defense Robert McNamara noted during the Cuban Missile Crisis of 1962 that it did not matter that the United States had many times more warheads than the Soviet Union. If the Russians had delivered just ten warheads on the territory of the United States, it would have been a disaster beyond what we had ever experienced. For that reason, he said, the United States, even in 1962, would not have carried out a preemptive strike against the Russians for fear of what they could do to us in return. The United States and Russia continued to keep their strategic arsenals on a par with one another for decades after 1970.
