Title:
Sensitivity analysis and parameterization of passively measured prompt-diagnostic signatures from a nuclear detonation and the effects on early-time attribution
Sensitivity analysis and parameterization of passively measured prompt-diagnostic signatures from a nuclear detonation and the effects on early-time attribution
Author(s)
Redd, Evan M.
Advisor(s)
Erickson, Anna S.
Editor(s)
Collections
Supplementary to
Permanent Link
Abstract
An orphaned nuclear detonation in an urban environment would create enormous pressure on policy makers to act quickly requiring strong attribution evidence. To answer these needs, this research examines prompt emissions within 60 seconds after detonation. This time period is defined as peri-detonation nuclear forensics and supports traditional post-detonation nuclear forensics which focus on sample/particulate collection and analysis. The competing physical effects and number of unknowns that affect prompt emissions, global and local sensitivity methods must be employed for determining which device characteristics can be determined at early-times. The results of these sensitivities and trends aid a Monte Carlo Filtering matching algorithm that accurately determines yield, fuel-type, mass, and height-of-burst from hypothetical photon emissions from simplified Fat Man and Little Boy design derivatives. Additionally, signatures emanating from nuclear cloud rise models between approximately 3 and 60 seconds refined the results from the Monte Carlo Filtering algorithm, reducing uncertainty and improving accuracy. Superior Monte Carlo Filtering accuracy and lowered uncertainty indicates promise of using signatures from nuclear cloud rise as diagnostic indicators, if earlier data is destroyed or unavailable. Additionally, experimental photon emission sensitivities were shown with varying masses of Pu-239 within 0.01 seconds using select energy ranges from the 47-group photon energy structure. The derived early signatures may provide policy makers with empirical attribution evidence when combined with intelligence from other collection disciplines.
Sponsor
Date Issued
2017-11-21
Extent
Resource Type
Text
Resource Subtype
Dissertation