Ingeniero Mecánico

ARTICLE IN PRESS

Atmospheric Environment 41 (2007) 6964–6973
www.elsevier.com/locate/atmosenv

Source estimation methods for atmospheric dispersion
K. Shankar RaoÃ
Atmospheric Turbulence and Diffusion Division, Air Resources Laboratory, NOAA, Oak Ridge, TN 37831-2456, USA
Received 22 December 2006; received in revised form 14 April 2007; accepted 25 April 2007

Abstract
Both forwardand backward transport modeling methods are being developed for characterization of sources in
atmospheric releases of toxic agents. Forward modeling methods, which describe the atmospheric transport from sources
to receptors, use forward-running transport and dispersion models or computational fluid dynamics models which are run
many times, and the resulting dispersion field is compared toobservations from multiple sensors. Forward modeling
methods include Bayesian updating and inference schemes using stochastic Monte Carlo or Markov Chain Monte Carlo
sampling techniques. Backward or inverse modeling methods use only one model run in the reverse direction from the
receptors to estimate the upwind sources. Inverse modeling methods include adjoint and tangent linear models, Kalmanfilters, and variational data assimilation, among others.
This survey paper discusses these source estimation methods and lists the key references. The need for assessing
uncertainties in the characterization of sources using atmospheric transport and dispersion models is emphasized.
Published by Elsevier Ltd.
Keywords: Atmospheric transport and dispersion models; Bayesian updating and inferencemethods; Inverse modeling; Adjoint and
tangent linear models; Kalman filtering; Variational data assimilation

1. Introduction
Atmospheric transport and dispersion (ATD)
models are routinely used to assess the impact of
emission sources on air quality for varying meteorological conditions. These models are also used at
nuclear and chemical plants for emergency response
and impact assessmentsfor hazardous substances
accidentally released into the atmosphere.
Trajectory models are widely used to establish
source–receptor relationships for estimation of
atmospheric concentrations or for interpretation
of measurements. While forward trajectories describe the paths of the released particles from the
ÃTel.: +1 865 576 1238; fax: +1 865 576 1327.

E-mail address: Shankar.Rao@noaa.gov1352-2310/$ - see front matter Published by Elsevier Ltd.
doi:10.1016/j.atmosenv.2007.04.064

source to receptors, backward trajectories from
receptors attempt to approximate the particles’
paths to the source. Trajectories are less accurate
when particles travel in the atmospheric boundary
layer because of the effects of turbulence, or when
they encounter convective up- and down-drafts(e.g., Stohl et al., 2002).
Both forward and backward modeling methods
are being developed to estimate the release rate and
duration of chemical, biological, and radiological
(CBR) agents in terrorism-related events. Given the
meteorology and concentrations observed at several
detectors, these methods characterize the source
type, and estimate the source strength and location.
This surveypaper discusses the leading source
estimation methods, and lists the relevant key
references. These methods are broadly classified as

ARTICLE IN PRESS
K. Shankar Rao / Atmospheric Environment 41 (2007) 6964–6973

forward modeling methods, which model the atmospheric transport from sources to receptors, and
backward modeling methods which use one model
run in the reverse direction fromthe receptors to
estimate the upwind sources. The importance of
assessing various uncertainties in the estimation of
sources using ATD models is emphasized.
2. Forward modeling methods
Forward modeling source estimation methods
utilize forward-running dispersion models or computational fluid dynamics algorithms. These models
are typically run many times over the domain of
interest. The...