Paradigms and commonalities in atmospheric source term estimation methods

Modeling the downwind hazard area resulting from the unknown release of an atmospheric contaminant requires estimation of the source characteristics of a localized source from concentration or dosage observations and use of this information to model the subsequent transport and dispersion of the contaminant. This source term estimation problem is mathematically challenging because airborne material concentration observations and wind data are typically sparse and the turbulent wind field chaotic. Methods for addressing this problem fall into three general categories: forward modeling, inverse modeling, and nonlinear optimization. Because numerous methods have been developed on various foundations, they often have a disparate nomenclature. This situation poses challenges to those facing a new source term estimation problem, particularly when selecting the best method for the problem at hand. There is, however, much commonality between many of these methods, especially within each category. Here we seek to address the difficulties encountered when selecting an STE method by providing a synthesis of the various methods that highlights commonalities, potential opportunities for component exchange, and lessons learned that can be applied across methods.