You are here:
20190825 - Development and evaluation of consensus meta-model for estimating national concentrations of organic chemicals in surface water (ACS Fall 2019)
Citation:
Sayre, R., P. Saranjampour, K. Isaacs, AND J. Wambaugh. 20190825 - Development and evaluation of consensus meta-model for estimating national concentrations of organic chemicals in surface water (ACS Fall 2019). American Chemical Society, San Diego, CA, August 25 - 29, 2019. https://doi.org/10.23645/epacomptox.9772604
Impact/Purpose:
The current project, EcoSEEM, assesses the correlation between measured values from surface water monitoring sites and results from openly-available mechanistic models that provide nationwide surface water concentration estimatesbased on physical chemical properties to create a consensus regression model for screening-level concentration estimates (and estimate uncertainties) for the thousands of chemicals which may be present in surface water, with or without monitoring data, which can contribute to the prioritization of safety evaluations for a broad range of chemicals.
Description:
U.S. EPA’s Center for Computational Toxicology and Exposure provides tools to rapidly generate quantitative toxicity, human exposure, and internal dose estimates. The SEEM (Systematic Empirical Evaluation of Models) meta-modeling approach uses forward-backward inference to identify when complex models add more signal than noise. SEEM has been used to successfully develop exposure predictions built from different data streams and model results that correspond to human biomonitoring data. The current project, EcoSEEM, extends this approach to surface water concentrations. We assess the correlation between measured values from surface water monitoring sites and results from openly-available mechanistic models that provide nationwide surface water concentration estimates for many chemicals. Seasonal concentrations for 91 chemicals from 1984 to 2014 inform estimates of likely loading values and the predictivity of each model. The three high-throughput models evaluated cover both far- and near-field chemical releases, and predict chemical fate based on various physical chemical properties. The overall association between model results and observations is used to create a consensus regression model for screening-level concentration estimates (and estimate uncertainties) of human and ecological contact with the thousands of chemicals which may be present in surface water, with or without monitoring data, which can contribute to the prioritization of safety evaluations for a broad range of chemicals. This abstract does not necessarily represent the views or policies of the U.S. Environmental Protection Agency.
URLs/Downloads:
DOI: 20190825 - Development and evaluation of consensus meta-model for estimating national concentrations of organic chemicals in surface water (ACS Fall 2019)
RRSAYRE_ECOSEEM_ACS2019_21AUG2019.PDF (PDF, NA pp, 723.339 KB, about PDF)