Utility of Atmospheric Transport Runs Done Backwards in Time for Source Term Estimation
Abstract
In a general sense, the equations describing transport of an atmospheric contaminant are reversible in time. However, the time reversibility is not perfect. For a source term estimation problem, one atmospheric transport model (ATM) run done backwards in time to represent the influence of a single sample value can yield as much information as hundreds or thousands of forwards ATM runs. Thus, backwards ATM are often used to limit the computational burden of an analysis. A large number of paired ATM runs at 14 release locations were used to explore the agreement between dilution factors for the forward and backwards time directions. Four of the release locations have correlation coefficients between forward and backwards dilution factors greater than 0.9. None of the four have large elevation changes within 100 km of the release location. The two release locations with the lowest correlation coefficients are in mountainous terrain with large elevation differences. Elevation changes have a greater negative influence on the correlations than distance to the sampler or plume transit times. The locations of the releases were estimated using dilution factors from backwards ATM runs. Many of the location estimates are close to the release point, vindicating the general approach ofmore »
 Authors:

 BATTELLE (PACIFIC NW LAB)
 Publication Date:
 Research Org.:
 Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
 Sponsoring Org.:
 USDOE
 OSTI Identifier:
 1530555
 Report Number(s):
 PNNLSA136065
 DOE Contract Number:
 AC0576RL01830
 Resource Type:
 Journal Article
 Journal Name:
 Journal of Environmental Radioactivity
 Additional Journal Information:
 Journal Volume: 203
 Country of Publication:
 United States
 Language:
 English
Citation Formats
Eslinger, Paul W., and Schrom, Brian T. Utility of Atmospheric Transport Runs Done Backwards in Time for Source Term Estimation. United States: N. p., 2019.
Web. doi:10.1016/j.jenvrad.2019.03.006.
Eslinger, Paul W., & Schrom, Brian T. Utility of Atmospheric Transport Runs Done Backwards in Time for Source Term Estimation. United States. doi:10.1016/j.jenvrad.2019.03.006.
Eslinger, Paul W., and Schrom, Brian T. Mon .
"Utility of Atmospheric Transport Runs Done Backwards in Time for Source Term Estimation". United States. doi:10.1016/j.jenvrad.2019.03.006.
@article{osti_1530555,
title = {Utility of Atmospheric Transport Runs Done Backwards in Time for Source Term Estimation},
author = {Eslinger, Paul W. and Schrom, Brian T.},
abstractNote = {In a general sense, the equations describing transport of an atmospheric contaminant are reversible in time. However, the time reversibility is not perfect. For a source term estimation problem, one atmospheric transport model (ATM) run done backwards in time to represent the influence of a single sample value can yield as much information as hundreds or thousands of forwards ATM runs. Thus, backwards ATM are often used to limit the computational burden of an analysis. A large number of paired ATM runs at 14 release locations were used to explore the agreement between dilution factors for the forward and backwards time directions. Four of the release locations have correlation coefficients between forward and backwards dilution factors greater than 0.9. None of the four have large elevation changes within 100 km of the release location. The two release locations with the lowest correlation coefficients are in mountainous terrain with large elevation differences. Elevation changes have a greater negative influence on the correlations than distance to the sampler or plume transit times. The locations of the releases were estimated using dilution factors from backwards ATM runs. Many of the location estimates are close to the release point, vindicating the general approach of using ATM runs backwards in time. Aggregated over the 14 release points, 26% of the location estimates were within 10 km of the release point, 61% were within 25 km, 80% were within 50 km, and 87% were within 75 km. Most of the remaining 13% (poor matches) occur for two release locations in mountainous terrain. Good timereversibility cannot be guaranteed for a new analysis, so we recommend that any source term solution using backwards ATM runs should include comparisons based on forward ATM runs. As an example, calculations show good agreement between dilutions in different timedirections for samples taken after the announced nuclear test by the Democratic People’s Republic of Korea in 2013.},
doi = {10.1016/j.jenvrad.2019.03.006},
journal = {Journal of Environmental Radioactivity},
number = ,
volume = 203,
place = {United States},
year = {2019},
month = {7}
}