222
Air Pollution and Turbulence: Modeling and Applications
8.8 EXAMPLE OF LSDM APPLICATIONS
In order to highlight the possible LSDM applications in different topography and
stability conditions and with reference to different typology of applications (validation studies, impact analysis, scenarios, etc.), in this section we briefl y introduce the
SPRAY model, developed by the team that the authors belong to, and we present a
few examples of its application. All these materials were already published in international journals; thus it is not a newly published work and will be referred to the
original journals.
8.8.1 SPRAY MODEL
SPRAY (Tinarelli et al., 1994, 2000; Ferrero et al., 2001, 2003; Trini Castelli et al.,
2003) is a 3-D model designed to deal with the simulation of passive airborne pollutant dispersion in complex terrain. Basically, it integrates three Langevin equations,
one for each Cartesian component of the velocity fl uctuations (see Equations 8.2,
8.3, and 8.15) according to the Thomson (1987) scheme. The algorithm to account
for the rise of buoyant emissions is the one presented in Section 7.3 (Equations 8.49
and 8.50). Regarding the Eulerian PDF of the vertical turbulent velocities, that is
generally skewed, one can choose between bi-Gaussian PDF (Equation 8.16) and the
Gram-Charlier PDF (Equation 8.26). The model makes use of the inhomogeneous
Gaussian PDF in the horizontal directions (Equation 8.32).
SPRAY also enters an integrated modeling system: RMS, acronym for RAMS,
the atmospheric circulation model (Pielke et al. 1992); MIRS, the parameterization
interface code (Trini Castelli and Anfossi, 1997; Trini Castelli, 2000), calculating the PBL parameters and Lagrangian turbulence fi elds not directly supplied by
RAMS and processing its outputs for the input to SPRAY, which is the last module
of the system.
8.8.2 VALIDATION VERSUS EXPERIMENTS
Model validation is the scientifi c basis for the development and improvement of the
numerical models to make them usable tools both for theoretical studies and for
applications in environmental frameworks. SPRAY model was used and tested in
several case studies and experiments since 1986. Here we report as examples two
works dealing with different approaches, the fi rst considering a comparison with
observations collected from physical modeling in a wind tunnel, the second with a
real fi eld experiment.
The RUSVAL tracer experiment (Khurshudyan et al., 1990) permits to evaluate
the model performances in controlled condition. These type of experiments allow
performing sensitivity analyses on specifi c aspects and parameters in the physics
described by the model. In RUSVAL, the fl ow over a schematic two-dimensional
valley was reproduced in a wind tunnel. RUSVAL data were used by Trini Castelli
et al. (2001) and Ferrero et al. (2003). The main aim of their work was to suggest
proper methods for predicting turbulence fi eld for dispersion models over complex
terrain and, more generally, in horizontally nonhomogeneous conditions. In fact,
© 2010 by Taylor and Francis Group, LLC
Air Pollution and Turbulence: Modeling and Applications
8.8 EXAMPLE OF LSDM APPLICATIONS
In order to highlight the possible LSDM applications in different topography and
stability conditions and with reference to different typology of applications (validation studies, impact analysis, scenarios, etc.), in this section we briefl y introduce the
SPRAY model, developed by the team that the authors belong to, and we present a
few examples of its application. All these materials were already published in international journals; thus it is not a newly published work and will be referred to the
original journals.
8.8.1 SPRAY MODEL
SPRAY (Tinarelli et al., 1994, 2000; Ferrero et al., 2001, 2003; Trini Castelli et al.,
2003) is a 3-D model designed to deal with the simulation of passive airborne pollutant dispersion in complex terrain. Basically, it integrates three Langevin equations,
one for each Cartesian component of the velocity fl uctuations (see Equations 8.2,
8.3, and 8.15) according to the Thomson (1987) scheme. The algorithm to account
for the rise of buoyant emissions is the one presented in Section 7.3 (Equations 8.49
and 8.50). Regarding the Eulerian PDF of the vertical turbulent velocities, that is
generally skewed, one can choose between bi-Gaussian PDF (Equation 8.16) and the
Gram-Charlier PDF (Equation 8.26). The model makes use of the inhomogeneous
Gaussian PDF in the horizontal directions (Equation 8.32).
SPRAY also enters an integrated modeling system: RMS, acronym for RAMS,
the atmospheric circulation model (Pielke et al. 1992); MIRS, the parameterization
interface code (Trini Castelli and Anfossi, 1997; Trini Castelli, 2000), calculating the PBL parameters and Lagrangian turbulence fi elds not directly supplied by
RAMS and processing its outputs for the input to SPRAY, which is the last module
of the system.
8.8.2 VALIDATION VERSUS EXPERIMENTS
Model validation is the scientifi c basis for the development and improvement of the
numerical models to make them usable tools both for theoretical studies and for
applications in environmental frameworks. SPRAY model was used and tested in
several case studies and experiments since 1986. Here we report as examples two
works dealing with different approaches, the fi rst considering a comparison with
observations collected from physical modeling in a wind tunnel, the second with a
real fi eld experiment.
The RUSVAL tracer experiment (Khurshudyan et al., 1990) permits to evaluate
the model performances in controlled condition. These type of experiments allow
performing sensitivity analyses on specifi c aspects and parameters in the physics
described by the model. In RUSVAL, the fl ow over a schematic two-dimensional
valley was reproduced in a wind tunnel. RUSVAL data were used by Trini Castelli
et al. (2001) and Ferrero et al. (2003). The main aim of their work was to suggest
proper methods for predicting turbulence fi eld for dispersion models over complex
terrain and, more generally, in horizontally nonhomogeneous conditions. In fact,
© 2010 by Taylor and Francis Group, LLC