Environmental Engineering Reference
In-Depth Information
1998); (3) the beta version of the MM5-FLEXPART model (Stohl et al., 2006);
and (4) the NOAA HYSPLIT model (Draxler, 1997). CALPUFF is a puff model
that is currently recommended by the U.S. Environmental Protection Agency
(EPA) to satisfy regulatory permitting requirements involving long range transport
under EPA's
Guideline on Air Quality Models
, and SCIPUFF is a puff model that
serves as the dispersion model core of the Hazard Prediction Assessment Capability
(HPAC) developed for the Defense Threat Reduction Agency. HYSPLIT is a
particle model developed by the National Oceanic and Atmospheric Administration's
Air Research Laboratory (NOAA/ARL), and FLEXPART is a particle dispersion
model developed and maintained by the Norwegian Institute for Air Research
(NILU).
Meteorological fields were developed using MM5-v3.7.2 for CAPTEX releases
1-5, and 7 and ETEX release 1. MM5 was initialized with NCEP/NCAR Reanalysis
data available on 2.5º × 2.5º grid every 6 h. For the CAPTEX releases, 3 nested
domains of 108, 36, and 12 km were used. A single domain of 36 km was used for
the ETEX simulations. Output from the finest resolution nest was used for each
simulation.
2.3. Model experiment design
Intermodel comparisons can be challenging due to the inherent differences in
model technical formulations. Chang et al. (2003) conducted an evaluation of
three Lagrangian puff models. While based upon Gaussian puff formulation, these
models varied significantly in terms of the level of sophistication of their technical
formulation. Cheng et al. (2003) correctly opined that it is necessary to set up a
proper framework to perform an objective and meaningful evaluation when such
models vary significantly in their formulation. A primary focus of their model
evaluation framework centered upon the use of the same observed meteorological
data and similar modeling domains. To the extent practical, default model options
were selected for all models in their evaluation.
Reflecting this evaluation paradigm, a major focus of this study was to provide
a common source of meteorological fields to each of the dispersion models
evaluated. Each of the four models in this exercise requires three-dimensional
meteorological fields as input to the model. For the majority of these models,
meteorological fields from prognostic meteorological models are the primary
source. The MM5 simulations discussed in Section 2.2 were used to supply three-
dimensional prognostic fields for each of the models. However, both CALPUFF and
SCIPUFF typically rely upon its own diagnostic meteorological model (CALMET
and MC-SCIPUFF respectively) to provide three-dimensional meteorological
fields to the dispersion model. Even though MM5 data is ingested as the initial
guess field into CALMET, much of the original MM5 data is not preserved, and
key boundary layer parameters are recomputed. This compromises the evaluation
paradigm - a common meteorological database. In 2008-2009, EPA developed
software to facilitate direct ingestion of prognostic meteorological model data by
the dispersion model with no changes to many of the boundary layer parameters
