EVALUATING RECURRENCE INTERVALS FOR MONOGENETIC ERUPTIONS IN THE SOUTHWESTERN USA (Invited Presentation)

The southwestern United States (south of Snake River Plain, east of Sierra Nevada/Cascades)has >40 volcanic fields with ~1400 Quaternary basaltic monogenetic volcanoes. Basaltic eruptions can produce lava flows and high ash plumes that pose hazards for aviation and infrastructure. Knowing the likelihood of an eruption can help put the hazards into perspective, but the quality of dates on these monogenetic volcanoes is variable. A simple Poisson temporal hazard model using all available dates for the monogenetic volcanism in the region yields annual probabilities of 1.4 x 10^-3– 9.9 x 10^-4, with a 50% chance of an eruption (and new volcano) within 500-700 years. The overall average repose interval (710-1010 years) is comparable to those at Cascade stratovolcanoes, but with much less certainty on the location. These interpretations are based upon dates of uneven quality and non-systematic location distribution. Older K/Ar dates commonly are too old due to inherited Ar, and dating techniques (e.g. Ar/Ar, cosmogenic, OSL) have improved in precision in the past 20 years. To show how this affects the overall recurrence frequency and interval, we focus on recent dates in the San Francisco volcanic field (SFVF). Five volcanoes are thought to have erupted in the past 100000 years in the SFVF. The age of Doney Crater was thought to be 251 ± 8 ka (Ar/Ar) but is now 68-70 ka based on cosmogenic He, Strawberry Crater was 51 ± 46 ka (K/Ar) but is now 50 ± 1 ka (cosmogenic He), SP Crater was ~70 ka (K/Ar) but is now 50-59 ka (cosmogenic He, OSL). The date on the Merriam Crater cluster was revised from 150 ka to 20 ± 2 ka based on a variety of techniques. Sunset Crater has been dated using paleomagnetic secular variation, dendro-chronology and -chemistry, and archaeological context, at 1085 CE. Using the estimates of 600 volcanoes erupting in the SFVF in 6 million years, an average recurrence interval is 10000 years, but the past 100000 years show a recurrence interval of 20000 years. The inherent difficulty in the method is illustrated here – three eruptions may have occurred within 15-20 kyr, and two more occurred at ca. 20 ka and 1 ka. Such complications probably occur at the 40 fields in the study area and elsewhere. Real recurrence frequency and recurrence interval are not consistent, and the dates used to calculate estimates of them may not be reliable either.