Environmental Engineering Reference
In-Depth Information
1.15 Evaluation of Four Lagrangian Models Against
the Cross-Appalachian and European Tracer
Experiments
B.A. Anderson
1
and R.W. Brode
2
1
U S Environmental Protection Agency, Kansas City, KS, USA
2
U S Environmental Protection Agency, Research Triangle Park, NC, USA
1. Introduction
In this study, a comparison of model performance of four state-of-the practice
Lagrangian dispersion models: CALPUFF, SCIPUFF, HYSPLIT, and FLEXPART
is presented. These models were used to simulate the dispersion of the tracer cloud
for the European Tracer Experiment (ETEX) and the Cross-Appalachian Tracer
Experiment (CAPTEX). Dispersion simulations for each of the modeling systems
were done using a common meteorological dataset derived from the NCAR/PSU
MM5 model. Statistical evaluation procedures used for model verification follow
a hybrid approach based upon the procedures used in the second Atmospheric
Transport Model Evaluation Study (ATMES-II) and the NOAA Data Archive of
Tracer Experiments and Meteorology (DATEM) project. Verification scores show
that the NOAA HYSPLIT model performed best overall, followed by the
SCIPUFF and FLEXPART models. CALPUFF performance was significantly
poorer than the other three models in the ETEX experiment and improved in
CAPTEX, but will require additional analysis to diagnose possible causes of
predicted tracer advection errors noted in the ETEX experiment.
2. Models and Methods
2.1. Mesoscale tracer experiments
This study focuses upon the evaluation of models against two major mesoscale
tracer experiments conducted during the 1980s and 1990s widely used for long
range transport model design and evaluation efforts. The Cross-Appalachian
Tracer Experiment (CAPTEX) consisted of seven releases of a perflourocarbon
tracer called perfluromonomethylcyclohexane (PMCH) in September and October
1983 from Dayton, Ohio and Sudbury, Ontario. Air sampling was conducted at 84
monitoring sites in the United States and Canada, with sampling frequencies of
