Authors: Hannah Burdett*, , Jake Lehner, University of Windsor, Chris Houser, University of Windsor
Topics: Geographic Information Science and Systems, Hazards and Vulnerability, Environmental Perception
Keywords: Machine Learning, Environmental Science, Fuzzy Data, Read Across
Session Type: Poster
Start / End Time: 1:20 PM / 3:00 PM
Room: Napoleon Foyer/Common St. Corridor, Sheraton, 3rd Floor
Presentation File: No File Uploaded
Environmental assessments tend to be limited by incomplete or unobtainable information. This is a particular problem for environmental problems where data inputs are often imprecise or inadequate due to the complexity and scale of natural and anthropogenic processes. To facilitate a transparent and systematic approach to aid in reducing uncertainty read across methods may be applicable to fill data gaps for topology assessment and regulatory decisions in environmental assessments. Read-across is a technique for predicting endpoint information for one substance, the target substance, by using data from the same endpoint from another substance or substances. Currently read across methods have only been applied in toxicological assessments for chemical safety assessments. The focus of this study is to assess the general methodology of read-across for toxicological assessment, and determine whether it is applicable to environmental assessment. To begin generating read-across predictions for environmental processes, existing read-across strategies must be re-evaluated primarily assessing: (1) the similarity between the target(s), and (2) the uncertainties in the read-across process and prediction. Templates from toxicological assessments will act as proposals to assist in assessing similarity. As well, guide the systematic characterization of uncertainty both in the context of the similarity rationale, the read across data and overall approach and conclusion for potential environmental assessments.