Causation or coincidence? The correlations in time and space of the 2008 eruptions of Cleveland, Kasatochi, and Okmok Volcanoes, Alaska
Abstract
In mid-summer 2008, three significant volcanic eruptions occurred in the Andreanof Islands of the Aleutian Arc, Alaska. Okmok volcano began erupting on July 12, followed by Cleveland on July 21, and then by Kasatochi on August 7. In addition to this temporal correlation, there is also a geographic correlation: the eruptions occurred in a 525 km region representing only about 20% of the arc's length. Given these close proximities in space and time, it is natural to speculate about whether an underlying process is at work. Ultimately, the arc exists because of subduction, but the question remains if a more immediate trigger may be responsible for the concurrence. We began our inquiry into whether a link exists among the three eruptions by posing the following question: What is the probability that, by chance alone, Okmok, Kasatochi and Cleveland could simultaneously erupt? Answering this question requires both a statistical model for eruption frequency and empirical data of where and when eruptions have occurred in the past. We assume that eruptions follow a Poisson distribution, and estimate the expected number of eruptions per time interval for each volcano in the arc from the geologic record and observations contained in the Alaska Volcano Observatory's GeoDIVA database. We then perform a Monte Carlo experiment, simulating 10,000 years of eruptive activity at 30 day intervals. The results of the simulation indicate that the phenomenon of three eruptions beginning in a single month happens about once every 90 years. A spatial constraint requiring that the maximum separation among the volcanoes be less than 525 km increases this interval to about once every 900 years. Though these intervals are not so long as to rule out coincidence, they are long enough to warrant further investigation into the possibility of a common origin. Several candidates for a prospective cause are: (1) the Great Aleutian Earthquake of 1957, which includes the region of the three recent eruptions, may have triggered a period of increased volcanic activity that still persists; (2) a slow slip event, with associated non- volcanic tremor, have may have resulted in static stress changes favorable to volcanic eruptions; or (3) nearby volcanoes may interact with one another in such a way as to increase the chance of clustered eruptions. We consider each of these scenarios (as well as other more remote possibilities) and weigh their relative likelihoods against the probability of random correlation. In the end, no definitive answer emerges, though pure coincidence remains a simple and plausible explanation for this remarkable event.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2008
- Bibcode:
- 2008AGUFM.A53B0278C
- Keywords:
-
- 4255 Numerical modeling (0545;
- 0560);
- 4260 Ocean data assimilation and reanalysis (3225);
- 8178 Tectonics and magmatism;
- 8185 Volcanic arcs;
- 8488 Volcanic hazards and risks