The impact of new data on an expert elicitation of the probability of volcanic intersection of the nuclear waste repository at Yucca Mountain
Abstract
An expert elicitation completed in 1996 estimated the frequency of intersection by a dike of the proposed high-level radioactive waste repository at Yucca Mountain. That assessment resulted in a mean annual frequency of intersection of 1.7e -8, or about 1 chance in 60 million per year. The expert panel primarily considered the past 5 million years of volcanic history when estimating the probability of intersection, and included in their assessment the possibility that at least a few volcanic centers were buried in alluvial basins near Yucca Mountain. The results of a regional aeromagnetic survey completed in 1999 suggested the potential for a higher number of buried volcanic centers than previously considered in the 1996 elicitation. Therefore, a new expert elicitation was convened in 2004 with the primary purpose of assessing the impact of buried volcanic centers on probability estimates. A major data-gathering effort completed in 2006 provided high-resolution aeromagnetic, drilling, and geochronology results that provided information on the number, location and age of buried basaltic centers near Yucca Mountain. Most buried basalt was determined to be of Miocene age. The youngest group of buried volcanic centers, 20-25 km south of Yucca Mountain, has an age of about 3.9 Ma. These results indicate that post-Miocene basaltic volcanism only occurred to the south and west of Yucca Mountain and not to the east, an important constraint in models that forecast the potential location of future volcanism. Other new data considered included the characteristics of dike swarms based on analog volcanoes, variations in crustal extension across the region, mantle tomography, and differences in calculated lithostatic pressure between basins and ranges. The new distribution for the annual frequency of intersection (5th and 95th percentiles of ~6e -10 and 1e -7) overlaps the 1996 distribution but is broader with more weight at both higher and lower values. This results in a slightly higher mean value and a slightly lower median value. The results are consistent with consideration of a broader range of conceptual models for the spatial and temporal behavior of volcanism, as well as more complex models of the geometry of volcanic events, all influenced in part by availability of new data.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2008
- Bibcode:
- 2008AGUFM.V11A2009P
- Keywords:
-
- 8488 Volcanic hazards and risks