Nuclear Tests Are Blameless--for Quakes : Seismic shocks: Suspect what you will, the data just don’t correlate.

In the wake of the most powerful quake to strike the state in 40 years, people are wondering if it really was mere coincidence that the seismic spasms of June 28 were preceded just five days earlier by an underground nuclear blast in southern Nevada. Or was there perhaps a more sinister connection, as suggested by Paul Conrad’s “Chain Reaction?� cartoon of July 1?

I am an ardent environmental activist who has attended more than a dozen demonstrations at the Nevada Test Site in the last six years, and have been arrested several times. So one might think that I would be more than willing to join with those seeking to blame our tectonic tremors on the tests, particularly given my perspective as a senior seismic analyst at Caltech.

While unease about earthquakes is certainly understandable, my own research leads me to conclude that the relationship between California quakes and nukes in Nevada is essentially nonexistent. Furthermore, those in the anti-nuclear movement who take advantage of the situation by pandering to public fears and misperceptions risk compromising their own credibility, and they may end up diverting attention away from far more compelling and legitimate reasons to oppose continued testing--of which I would cite the threat of further proliferation first and foremost.

The problem is that few people have any real understanding of what constitutes a significant correlation between two sets of events that may be linked through cause and effect, let alone any insight into statistics or probability theory. Just because a quake happens to follow close after a nuclear test does not necessarily a correlation make.

Quakes have been with us since the dawn of time and will no doubt continue to inconvenience us after the last nuke is ancient history. To establish a correlation between the two, one would have to demonstrate that the incidence of quakes following nuclear tests has been substantially higher than what the long-term average seismicity would indicate. Other plausible explanations would also have to be ruled out, including random chance, given enough combinations to examine.

The fact is that the seismic record, when studied objectively and with the proper statistical methodology, simply does not support a contention that nuclear testing in Nevada, or anywhere else, has any effect of triggering earthquakes beyond the immediate area of the tests themselves--out to distances of at most 20 to 30 miles for the larger blasts. This is true even for tests at or above the 150-kiloton limit we have observed since 1976 (equivalent to a seismic magnitude of 5.8 to 5.9), which presumably would have the most pronounced effect--if there actually were any effect.

The rather modest 3.9 magnitude of the June 23 test (suggesting a yield of perhaps less than a half-kiloton) and the even smaller 3.0 test June 9 make it doubly dubious that either could have played any part in helping to initiate our current seismic sequence, which in hindsight almost certainly began with the 6.1 Joshua Tree quake April 22.

The lack of any demonstrated statistical correlation between nukes and quakes could allow one to maintain with equally vacuous validity that earthquakes “cause� nuclear tests, rather than the other way round. Indeed, a rather tenuous case could be made that the time of most intense testing in Nevada, from the mid-1960s to 1976 (when weapons up to the megaton range were routinely detonated), actually corresponded with one of the quieter periods of Southern California seismicity--notwithstanding the 1971 Sylmar quake. I trust that even the most vociferous advocate of continued nuclear testing would hesitate to suggest that we set off even more nukes in an ultimately futile effort to try to keep the Big One at bay.