Bayesian network as a support tool for rapid query of the environmental multimedia distribution of nanomaterials

Muhammad Bilal; Haven Liu; Rong Liu; Yoram Cohen

doi:10.1039/c6nr08583k

Bayesian network as a support tool for rapid query of the environmental multimedia distribution of nanomaterials

Nanoscale. 2017 Mar 23;9(12):4162-4174. doi: 10.1039/c6nr08583k.

Authors

Muhammad Bilal¹, Haven Liu², Rong Liu³, Yoram Cohen⁴

Affiliations

¹ University of California's Center for Environmental Implications of Nanotechnology (UC-CEIN), University of California, Los Angeles, Los Angeles, CA 90095, USA. yoram@ucla.edu m.bilal@ucla.edu and Institute of the Environment and Sustainability (IoES), University of California, Los Angeles, Los Angeles, CA 90095, USA.
² Department of Chemical & Biomolecular Engineering, University of California, Los Angeles, Los Angeles, CA 90095, USA.
³ Institute of the Environment and Sustainability (IoES), University of California, Los Angeles, Los Angeles, CA 90095, USA.
⁴ University of California's Center for Environmental Implications of Nanotechnology (UC-CEIN), University of California, Los Angeles, Los Angeles, CA 90095, USA. yoram@ucla.edu m.bilal@ucla.edu and Department of Chemical & Biomolecular Engineering, University of California, Los Angeles, Los Angeles, CA 90095, USA.

PMID: 28282098
DOI: 10.1039/c6nr08583k

Abstract

An approach is presented describing the development of a Bayesian network (BN) based tool for rapid assessment of the distribution of engineered nanomaterials (ENMs). The methodology was demonstrated via the construction of a BN model for estimating the exposure concentrations of nanomaterials (BN-nanoExpo) based on simulation data derived from a mechanistic multimedia compartmental fate and transport model. The results of simulations of the distribution of six ENMs in eight different regions were generated for a broad range of geographical and meteorological parameters as well as ENM release rates to the air, water and soil major compartments. Test cases with the constructed BN-nanoExpo demonstrated the capability of the BN based model to portray a wide range of simulation results that can be obtained with the mechanistic fate and transport model. Moreover, BN-nanoExpo is shown to be a suitable tool for estimating both ENM concentrations and release rates given partial information, while also enabling assessment of the impact of uncertainties in input data on the predicted outcomes. The results of the current study suggest that there is merit in exploring the utility of the approach to more complex models, which would provide decision makers with powerful tools for rapid assessment of the behavior of nanomaterials in the environment.