Author
YAKIREVICH, ALEXANDER - Ben Gurion University Of Negev | |
Gish, Timothy | |
Pachepsky, Yakov | |
Guber, Andrey | |
CADY, RALPH - Us Nuclear Regulatory Commission | |
NICHOLSON, THOMAS - Us Nuclear Regulatory Commission |
Submitted to: BARC Poster Day
Publication Type: Abstract Only Publication Acceptance Date: 3/27/2012 Publication Date: 4/19/2012 Citation: Yakirevich, A., Gish, T.J., Pachepsky, Y.A., Guber, A.K., Cady, R., Nicholson, T. 2012. An information theory application to improve understanding of subsurface flow and transport conditions at the BARC OPE3 site. BARC Poster Day. Interpretive Summary: Technical Abstract: Improving understanding of subsurface conditions includes comparison and discrimination of concurrent models. Additional observations can be useful for that purpose. The objective of this work was to implement and test a novel method for optimization of selecting locations for additional observation well. The method is based on the information theory concept of “weights of evidence”, and ensemble modeling with pedotransfer functions to estimate variance of the predicted values. The method was applied to the data from the tracer experiment at the USDA-ARS OPE3 research site. A pulse of KCL solution was applied to an irrigation plot, and chloride concentrations were measured in shallow groundwater at three sampling depths in 12 observations wells. The spatial distribution of soil materials was obtained from cores taken from depths of 0-200 cm with 20 cm increment during installation of observation wells. A three-dimension flow and transport model was developed to simulate the flow and chloride transport for the tracer experiment at the OPE3 site. The search of the optimal location of the augmentation wells was done on a 2D grid. Models of subsurface were compared that included and did not include the subsurface lenses. The method provided explicable and stable solution for the augmentation well location problem. Future research includes the generalizations of the method for multiple responses and for Bayesian augmentation of observation networks. |