Location and Yield of Nuclear Events Using Radionuclide Monitoring Data and Inverse Plume Modeling

A worldwide network of radionuclide monitoring stations has been established to monitor radioactive fall-out from nuclear tests. These stations sample the air for radioactive particulates and Xenon on a daily basis. This project is to develop a softwaresystem to determine the location, timing, and yield of nuclear events and their corresponding estimation probabilities using the sampled data from these monitoring stations. The system combines the capability of an EPA recommended model for atmospherictransport and dispersion with a robust automatic differentiation software package to develop an adjoint model for inverse plume modeling. The measurements obtained from the monitoring network and the adjoint calculations are assimilated based on an optimalscheme utilizing the uncertainty values. Incorporated with near real-time observational meteorological and radionuclide data, the developed system will be an operational tool for the detection of nuclear test events and accidents. This developed softwarehas a very high commercialization potential. The developed software system can be used to support the Comprehensive Test Ban Treaty (CTBT). It can also be used for emergency planning and post-accident investigation for both governmental and privateindustrial applications. In the government markets, there is currently a large need for a product with the capability proposed in this project. Potential application areas include the identification and quantification of sources for nuclear, biological,chemical, and other events. The federal customers include almost all the major agencies such as DOD, DOE, EPA, NASA, and DOC. We estimate the annual market size for the government is about $50 million/year. The future sales revenue can be in the form ofsales and maintenance of the software system with some modifications for the specific needs, and contract work for research and development for new applications.