Identification of multiple contamination sources using variational continuous assimilation

摘要

A variational continuous assimilation (VCA) method was proposed as an estimation method for multiple unknown contaminant sources in outdoor conditions. The VCA method is a data assimilation method that minimizes the difference between calculated and observed concentrations by adding a correction term to the governing equation of the computational fluid dynamics (CFD) model. The method was validated via a set of numerical experiments, in which single and multiple unknown source locations and their emission rates were estimated using the VCA method with limited observation data. The results showed that the VCA method could be used to roughly estimate source locations and emission rates, and could reproduce the correct concentration fields. The estimated concentration fields are, however, slightly more widely distributed than the actual concentration fields and are distributed slightly windward of the actual source locations. The estimation accuracy in the case with two sources was equivalent to the case with a single source.